VDR inhibits NLRP3 signal in lupus nephritis by competitively binding with importin 4 to suppress NF-κB nuclear translocation

Rationale: Lupus nephritis (LN) is a major risk factor for morbidity and mortality in systemic lupus erythematosus patients, and lupus nephritis treatment is limited to immunosuppressive therapy with many problems. Vitamin D receptor (VDR) can regulate NLRP3 inammasome which plays critical roles in LN pathogenesis. Objectives: This study was designed to explore the therapeutic effect of VDR agonist on LN and its potential mechanisms, aiming to elucidatethe optimal therapy for LN. Findings: In vivo, treatment of MRL/lpr mice since 8 weeks of age with VDR agonist paricalcitol for 8 weeks decreased disease pathogenesis of LN with markedly improved renal pathological changes, decreased urine protein and serum anti-ds-DNA antibody level in a time-depended manner. In MRL/lpr mice of 16 weeks of age with LN, the expression of NLRP3/caspase-1/IL-1β/IL-18 axis was upregulated detecteded by ELISA, RT-PCR, western blot and immunohistochemistry, while when treated with VDR agonist paricalcitol, expression of this axis was decreased signicantly. Further, it is proved that VDR agonist paricalcitol modulated NLRP3/caspase-1/IL-1β/IL-18 axis via inhibiting NF-κB, in addition, co-immunoprecipitation results showed that VDR agonist suppressed NF-κB nuclear translocation by competitively binding with importin 4. In vitro, anti-dsDNA antibody induced apoptosis and upregulation of NF-κB/NLRP3/caspase-1/IL-1β/IL-18 axis in mRTECs, which could be reversed by VDR agonist paricalcitol. Conclusions: Vitamin D receptor agonist may be a promising novel therapeutic strategy for patients with lupus nephritis, which paves the way for future preclinical/clinical studies.


Introduction
Lupus nephritis (LN) is a major risk factor for morbidity and mortality in systemic lupus erythematosus (SLE) patients, renal-limited lupus nephritis has been reported as well, and 10% of patients with LN will develop to end stage renal disease (ESRD) [1]. Importantly, 10-year survival improves from 46-95% if disease remission can be achieved, lupus nephritis treatment is limited to immunosuppressive therapy, and existing problems include inadequate therapeutic response, medication related side effects, relapses of lupus nephritis, the optimal therapy for LN remains to be elucidated, so it is important to nd potential effective disease remission strategies [2].
LN is immune complex-mediated glomerulonephritis with renal tubular dysfunction and tubulointerstitial in ammation [3,4], recently, it is reported that tubulointerstitial in ammation may be the initiator of chronic renal injury and may predict response to therapy in LN [5]. Immunity/in ammation mechanism is a key component during the pathogenesis of LN [6], it is reported that over 110 immune-genes were differentially expressed between LN and healthy control kidney biopsies [7]. Immune cell in ltration participates in the pathogenesis of LN and associates with clinical features. Immune cells, including monocytes, B cells, and T cells, are recruited to kidney tissue and produce cytokines and chemokines resulting tissue damage [8,9] . Besides, plenty of in ammatory mediators have been implicated in the development and pathogenesis of lupus nephritis, such as IL-6, TNF-α, IFNs, and hyaluronan, which further drive the in ammatory processes [10]. A variety of immune mechanisms are involved in the onset and ampli cation of the in ammatory response in LN. Among them, in ammasome machinery plays an important role in promotion of renal damage during LN development, most in ammasomes involve NOD-, LRR-and pyrin domain-containing protein 3 (NLRP3), in ammasome NLRP3-generated cytokines aggravate nephritis in various murine models of lupus via NF-κB/NLRP3/caspase-1/IL-1β/IL-18 [11].
Vitamin D receptor (VDR) is an ancient nuclear receptor, vitamin D (VD) de ciency is implicated in various diseases, including SLE/LN [12]. VDR plays critical roles in transcriptional regulation, immunity, in ammation and proliferation, it relates with chronic tubulointerstitial changes, such as, cortical interstitial expansion, in ammation and brosis [13]. VDR signaling reduces in ammation by suppressing in ammatory signaling pathway, promoting anti-in ammatory cytokines secretion and regulating T-cell different [14,15]. Recently, it is reported that VDR expression was downregulated in renal tissues of LN patients and was negatively correlated with disease activity and severity [16]. Combined with the pathogenesis mechanisms of LN, researchers proposed that VDR has the potential to be a novel therapeutic target for LN [17].
Untill now, the therapeutic effect of VDR agonist on LN has not been investigated, and the potential mechanisms are also unclear. Recently, it is reported that VDR inhibits NLRP3 activation which participated in LN development as described above in other diseases [18][19][20], at present the following issues need to clarify: 1) whether VDR agonist alleviates LN though inhibiting NLRP3 signal; 2) the molecular mechanism through which VDR inhibits NLRP3 signal.
Here, in this study, we try to test if an orally active VDR agonist would decrease disease pathogenesis in lupus-prone MRL/lpr mice. Additionally, we seek to delineate the cellular and molecular mechanism of action of VDR agonist in LN. We hope to provide the theoretical basis for potential clinical applications of vitamin D/VDR in LN.

Mice and paricalcitol treatment
The experimental animals were approved by the Institutional Animal Care and Use Committee of Xian jiaotong University. Female MRL/lpr mice (7 weeks of age) and age-and sex-matched C57BL/6 mice were purchased from Model Animal Research Center Of Nanjing University (Nanjing, China). C57BL/6 mice were used as control group, lupus-prone MRL/lpr mice were used as mouse model of LN [21]. The experimental animals were placed in a special pathogen-free environment (24 ± 1 °C, 50 ± 5% relative humidity and normal 12-h light/12-h dark cycle) in the Animal Experiment Center of Xian jiaotong University. All the mice were kept in controlled conditions for one week before the experiment. After 1 week acclimatization, all the MRL/lpr mice were continuously administrated intraperitoneally for 8 weeks with VDR agonist paricalcitol (19-nor-1,25-dihydroxyvitamin D2, PAL, Abbott Laboratories, CA, USA, 300 ng/kg/mouse per dose, 5 times a week), the control groups were given equal dosage of saline. The amount of 24 h proteinuria were respectively obtained from mice using metabolic cages, and was assessed weekly, quantitative analysis of mouse urine protein was performed by bicinchoninic acid (BCA) protein assay kit (Thermo Scienti c, MA, USA). All mice survived to the end of treatment, the mice were sacri ced at 0 week and 8 weeks after intervention (8 weeks old or 16 weeks old), respectively. Serum collected from mouse orbit was obtained by the centrifuge at 3000 rpm for 10 min at 4 °C, and stored at -20 °C for future ELISA use. Fresh Renal tissues were frozen at -80 °C for Western blot and PCR analysis.
Renal tissues were xed in 4% neutral-buffered formalin and embedded in para n for histopathological and immunohistochemistry analysis. Cells Culture Mouse renal tubular epithelial cells (mRTECs) were obtained from Jennio-bio (Guangzhou, China). Cells were cultured in Dulbecco's Modi ed Eagle Medium (DMEM) (Gibco, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, CA, USA), 100 U/mL penicillin, and 100 mg/mL streptomycin at 37 °C and 5% CO 2 in a humidi ed incubator after resuscitation. After starvation for 24 h, cells were stimulated with vehicle or PAL (0.2 ng/ml) for 24 h. In anti-dsDNA antibody stimulation experiment, mRTECs were treated with serum (the concentration of anti-dsDNA antibody was adjusted to 10 ng/ml) obtained from MRL/lpr mice with LN, for control group, mRTECs were treated with serum of same dilution obtained from normal C57BL/6 mice.

Plasmid Construction
VDR overexpression plasmid was constructed. In brief, speci c primers used to amplify the coding region of mouse VDR (NM_009504.04) were Mus-VDR-Forward: 5'-CCCAAGCTTATGGAGGCAATGGCAGCCAG-3' and Mus-VDR-Reverse: 5'-CCGCTCGAGTCAGGAGATCTCATTGCCGA-3'. The above PCR products were TA subcloned into pcDNA3.1-Hygro(+) vector. The inserts were inserted into the XhoI and Hind III sites of the vector, plasmids containing the insert were isolated for sequencing.

Cell Transfection
Cells were plated onto 6-well or 96-well plates at 50% con uence. Transient transfection of siRNA or plasmids of VDR was carried out using Lipofectamine 2000 reagent following the manufacturer's procedure. Brie y, mRTEC cells were washed with serum-free medium and cultured in serum-free medium without antibiotics. The transfection complex (siRNA/plasmids and the transfection reagent mixture) were added to the medium in a drop-wise manner and mixed gently by rocking the media back and forth. After 4-6 h, the cell culture medium was changed back to DMEM containing serum and antibiotics and incubated at 37 °C for 48 h before proliferation assay, Western blot analysis, or PCR experiments.

Immunohistochemistry
Protein expression levels were detected in para n-embedded lung sections using the Importin4 (1:50  antibodies, which were visualized using HRP-conjugated goat anti-rabbit/anti-mouse secondary antibody. Slides were counterstained with hematoxylin, dehydrated and mounted.

Histopathologic Analysis
Renal tissues were collected and xed in 10% neutral buffered formalin usually for 48-96 hours. Tissues were then routinely processed, embedded, sectioned (∼4 µm), and stained with routine H&E stains for general examination. Periodic acid Schiff (PAS) stains were selected for observing renal glomerular basement membrane and the mesangial area.

Flow Cytometry
Apoptosis of cells was determined by ow cytometric analysis using the AnnexinV-APC/7-AAD Assay kit (BD Biosciences, CA, USA) according to the manufacturer's instructions. Brie y treated mRTEC cells were harvested with trypsin, washed in cold PBS twice, and resuspended in binding buffer. A volume of 100 µL of the solution was transferred to a 5 mL culture tube and 5 µL AnnexinV-APC and 5 µL 7-AAD were added and incubated at room temperature for 15 minutes in the dark. Then, 400 µL of binding buffer was added to the culture tube. Samples were analyzed using CytoFLEX (Beckman Coulter, CA, USA) within 1hour after staining. Apoptotic cells were de ned as AnnexinV-APC-positive, 7-AAD-negative cells.
Co-Immunoprecipitation (Co-IP) Cells were collected and lysed with lysis buffer on ice for 10-15 min and centrifuged at 10000 rpm at 4 °C for 10 min. The supernatant fractions were collected and incubated with appropriate antibody at 4 °C overnight and precipitated with protein A/G-agarose beads (Santa Cruz Biotechnology, sc-2003, Texas, USA) for another 3-6 h at 4 °C. The beads were washed with the lysis buffer 3 times by centrifugation at 3000 rpm for 5 min at 4 °C. The immunoprecipitated proteins were separated by SDS-PAGE, and western blotting was performed as previously described. The primary antibodies were as follows: anti-importin 4 (rabbit, abcam, Ab181046, MA, USA), anti-NF-κB (mouse, Santa Cruz, SC-166588, Texas, USA), anti-NLRP3 (mouse, Santa Cruz, SC-13133, Texas, USA). The second antibodies were HRP-goat anti-rabbit secondary antibody (BOSTER biological technology, BA1054, Wuhan, China) and HRP-goat anti-mouse secondary antibody (BOSTER biological technology, BA1051, Wuhan, China).

Statistical analysis
Data were presented as means ± SEM. Comparison between the groups of data was evaluated using the Student's unpaired t-test. For multiple comparisons, one-way ANOVA was used with a Bonferroni post hoc test. A p value < 0.05 was considered statistically signi cant.

Results
1. Treatment of MRL/lpr mice with VDR agonist decreases disease pathogenesis of LN.
To test if VDR agonist paricalcitol would improve diesease in MRL/lpr mice with LN, MRL/lpr mice received 300 ng/kg/mouse of paricalcitol (Pari) or saline 5 times a week at 8 weeks of age for 8 weeks. Age-and sex-matched C57BL/6 mice were used as non-nephritic controls. Effects of Pari treatment on kidney disease were analyzed by assessing renal pathology and measuring proteinuria, serum anti-ds-DNA antibody. Compared with control group, renal pathology in the model group was deteriorated, as showed in Fig. 1A and 1B, we can see that glomerular mesangial cells and endothelial cells proliferated signi cantly, glomerular volume was increased compensatoryly, glomerular capillary wall damage and swollen renal tubular epithelial cells were observed, in addition, abundant in ammatory cells were in ltrated in renal tissues, while Pari treatment markedly improved the renal pathology in MRL/lpr mice with LN. Compared with controls, urine protein of MRL/lpr mice was increased gradually in a timedepended manner and reached the highest-level at 16 weeks of age, which could be reversed by Pari treatment (Fig. 1C). The same phenomenon was observed in the change of serum anti-ds-DNA antibody, as showed in Fig. 1D, at 8 weeks of age, there was no difference among the three groups, while at 16 weeks of age, the serum anti-ds-DNA antibody concentration was increased signi cantly compared with controls, and Pari treatment could lower the increased level of serum anti-ds-DNA antibody in MRL/lpr mice. Here we conclude that MRL/lpr mice presented with LN at 16 weeks of age, and Pari treatment could reduce disease severity of LN.
Next, we tested the role of NLRP3/caspase-1/IL-1β/IL-18 axis in MRL/lpr mice with LN and the effect of Pari treatment on this pathway axis. The expression of NLRP3/caspase-1/IL-1β/IL-18 axis in renal tissue among the three groups at 8 weeks of age showed no difference ( Fig. 2A). Figure 2B showed that NLRP3/caspase-1/IL-1β/IL-18 axis was upregulated over time when the LN presented, compared with control mice, the expression of NLRP3/caspase-1/IL-1β/IL-18 axis was increased abviously in MRL/lpr mice at 16 weeks of age, and after treated with Pari, the expression of the axis was decreased signi cantly. Immunohistochemistry showed that the immunoreactivity of NLRP3 and caspase-1 signi cantly increased in the renal tissue from MRL/lpr mice at 16 weeks of age compared with those from control mice, and decreased after Pari treatment (Fig. 2C). NLRP3 and caspase-1-positive signals were predominantly found along the renal tubule epitheliums (Fig. 2C, e-f). These results suggest that NLRP3/caspase-1/IL-1β/IL-18 axis participated in the pathogenesis of LN, and it is a potential target for VDR agonist to prevent LN.
Then we investigated the upstream molecule through which VDR agonist Pari inhibited NLRP3/caspase-1/IL-1β/IL-18 axis in MRL/lpr mice with LN, and NF-κB that plays an important role in immunity and in ammation process was tested. Similar trend was observed on the expression of NF-κB as that of NLRP3/caspase-1/IL-1β/IL-18 axis. As showed in Fig. 2A and 2B, among the three groups at 8 weeks of age, the expression of NF-κB showed no difference. Compared with control mice, the expression of NF-κB was increased in MRL/lpr mice at 16 weeks of age signi cantly, and could be downregulated after treated with Pari. In addition, immunohistochemistry (Fig. 3C) showed that the immunoreactivity of NF-κB signi cantly increased in the renal tissue from MRL/lpr mice at 16 weeks of age compared with those from control mice, and decreased after Pari treatment. Together, these ndings suggest that VDR agonist improved MRL/lpr mice with LN through NF-κB/NLRP3/caspase-1/IL-1β/IL-18 pathway.

VDR agonist suppresses NF-κB nuclear translocation by competitively binding with importin4.
As we know, tubulointerstitial in ammation may be the initiator of chronic renal injury, and VDR relates with chronic tubulointerstitial changes, our results also showed that NLRP3 were predominantly found along the renal tubule epitheliums, then we further explored the regulation mechanism of NF-κB inactivation by VDR agonist in mRTECs. From the Fig. 4A, we can see that both the total protein and nuclear protein of NF-κB were upregulated in mRTECs when treated with anti-dsDNA (10 ng/ml) that extracted from serum of MRL/lpr mice with LN, and Pari treatment could offset this effect, while the opposite trend was observed in the case of VDR expression. The obove ndings suggest that expression of VDR and NF-κB had a "trade-off" relationship, including their nuclear translocation. Co-Immunoprecipitation results (Fig. 4B) showed that both VDR and NF-κB could combine with importin-4 to form a complex, and after overexpressing VDR, the binding of NF-κB and importin-4 was signi cantly reduced, indicating that VDR and NF-κB competitively binded with importin 4 to translocate into nucleus. 5. VDR agonist reverses anti-dsDNA-induced upregulation of NF-κB/NLRP3/caspase-1/IL-1β/IL-18 axis in mRTECs.
It is reported that anti-dsDNA antibody could induce apoptosis of resident renal cells [22,23]. In our study, the percentage of early apoptotic cells (APC+/7-AAD−, the lower right quadrant of density plot) of mRTECs treated with anti-dsDNA antibody was clearly higher than that of control group (Fig. 6A), while after Pari treatment, the increased percentage of early apoptotic cells induced by anti-dsDNA antibody could be reversed. Knockdown of NF-κB/NLRP3 decreased the percentage of early apoptotic cells further, indicating that NF-κB/NLRP3/caspase-1/IL-1β/IL-18 axis, as a target of VDR, involved in mRTECs apoptosis. Moreover, in vivo experiments, the expression of Bax that functions as an apoptotic activator was increased in MRL/lpr mice with LN, which could be reversed by Pari treatment, and further decreased when knockdown NF-κB/NLRP3. For the expression of Bcl-2, which is an antiapoptotic protein and a member of the Bcl-2 family, opposite trend was observe compared with Bax (Fig. 6B). This indicates that VDR agonist inhibits anti-dsDNA antibody-induced mRTECs apoptosis via NF-κB/NLRP3/caspase-1/IL-1β/IL-18 axis.

Discussion
To our knowledge, this is the rst study to reveal the therapeutic effect of VDR agonist on LN and its potential mechanism. In the present study, we investigated the effects of paricalcitol on LN in MRL/lpr mice and demonstrated that 1) VDR agonist prevents LN in MRL/lpr mice and inhibits anti-dsDNA antibody-induced mRTECs apoptosis signi cantly by targeting NF-κB/NLRP3/caspase-1/IL-1β/IL-18 axis.
Nowdays many clues showed that VDR agonist-vitamin D analogues have the potential to be clinical treatment candidates for many kinds of dieases, such as, cancer and chronic renal diseases. Several studies have suggested that low circulating levels of vitamin D are associated with both an increased risk of developing cancer and poor outcome in patients with cancer, including renal cell carcinoma (RCC), especially colorectal, oral and breast cancer. VDR had anticancer effects, and supplementation with vitamin D analogues also could enhance response to standard therapies [24,25], in addition, VDR activation was reported to ameliorate chemotherapy drug cisplatin-induced acute tubular necrosis [26]. Preclinical experiments suggest that vitamin D has the therapeutic potential for RCC through suppressing renal cancer cell proliferation and enhancing cancer cell death [27,28]. For chronic kidney diseases, VDR agonists play a protective role in several models of renal disease. In dietary fat-induced renal disease mouse model, VDR agonists have been shown to prevent proteinuria, podocyte injury, mesangial expansion, accumulation of extracellular matrix proteins, in ltration with macrophages, and activation of markers of oxidative stress, in ammation, and brosis [29]. Recent clinical trials demonstrated that oral calcitriol can decrease proteinuria in IgA nephropathy patients, in its rats model, use of vitamin D can alleviate renal tissue damage by regulating immune response and NF-κB/TLR4 pathway [30]. A clinical observational study showed that supplementation with oral cholecalciferol for 6 months attenuate hypertension and proteinuria and delay the progression of polycystic kidney disease [31]. Together, VDR would be a novel potential therapeutic target for kidney diseases.
Vitamin D/VDR is involved in SLE pathogenesis, vitamin D3 could reduce the severity of SLE in MRL/l mice [32], it is reported that vitamin D de ciency could be a signi cant predictor of nephritis in SLE, and has a direct relationship with increased disease activity and nephritis [33,34]. Until now, the treatment effect of Vitamin D/VDR on LN has not be researched yet, in our study, we demonstrated that VDR agonist could prevent LN in MRL/lpr mice by improving renal pathological changes and decreasing proteinuria and serum anti-ds-DNA antibody. In vitro, we also found that VDR acitivation could ameliorate anti-dsDNA antibody-induced mRTECs apoptosis. All of these results indicate that VDR agonist could decrease disease pathogenesis of LN, which has the potential for clinical application.
Recent studies indicate that VDR is involved in the immune/in ammation regulation, several clues show that VDR can regulate NLRP3 in ammasome, VDR as a a nuclear receptor could translocate into nucleus to form a complex with NLRP3 and then block the formation of in ammasome [35]. While the mechanisms of VDR inhibiting NLRP3 in ammasome pathway remains unclear, especially in LN. According previous papers in other diseases, BRCC3-mediated deubiquitination of NLRP3 could be inhibited by VDR, and then NLRP3 activation is inhibited [18]. It has been proved that yes-associated protein 1 (YAP1) inhibit NLRP3 activation [36,37], and VDR agonist could negatively regulate NLRP3 in ammation activation through activating YAP1 [20]. NLRP3 in ammasome has a crosstalk with Aryl hydrocarbon receptor (AhR) and NF-κB [38], vitamin D3 can increase the activation of AhR signaling, AhR blocks NF-κB binding sites in the NLRP3 promoter region, which result to suppress NLRP3 in ammasome activation [39]. In this study, we found a novel mechanism through which VDR inhibiting NLRP3 in ammasome. We revealed the crosstalk among VDR, importin 4 and NF-κB, nuclear translocation of VDR or NF-κB was decided by the histone nuclear import protein importin 4 that is known to shuttle between the cytoplasm and nucleus. Importin 4, identi ed as a interactive partner of VDR, is responsible for the ligand-independent nuclear translocation of VDR [40]. NF-κB-importins interaction mediates NF-κB translocation and NF-κB-dependent gene transcription [41,42], while the interaction of NF-κB and importin 4 has not clari ed yet. Our study showed that both VDR and NF-κB interact with importin 4 to complete nuclear transcription, and VDR could bond importin4 competitively to suppress NF-κB nuclear translocation and NF-κB-dependent NLRP3 transcription.