Novel Mechanism for the Anti-Inflammatory Effects Of

BASIC RESEARCH www.jasn.org

TRAF3 Modulation: Novel Mechanism for the Anti-inﬂammatory Effects of the Vitamin D Receptor Agonist Paricalcitol in Renal Disease

Sandra Rayego-Mateos,1,2 Jose Luis Morgado-Pascual,1,3 José Manuel Valdivielso,2,3 Ana Belén Sanz,3,4 Enrique Bosch-Panadero,4 Raúl R. Rodrigues-Díez ,1 Jesús Egido,5 Alberto Ortiz,3,4 Emilio González-Parra,4 and Marta Ruiz-Ortega 1,3

Due to the number of contributing authors, the afﬁliations are listed at the end of this article.

ABSTRACT Background CKD leads to vitamin D deficiency. Treatment with vitamin D receptor agonists (VDRAs) may have nephroprotective and anti-inflammatory actions, but their mechanisms of action are poorly understood. Methods Modulation of the noncanonical NF-kB2 pathway and its component TNF receptor–associated factor3(TRAF3)bytheVDRAparicalcitolwasstudied in PBMCs from patients with ESKD, cytokine- stimulated cells, and preclinical kidney injury models. Results In PBMCs isolated from patients with ESKD, TRAF3 protein levels were lower than in healthy controls. This finding was associated with evidence of noncanonical NF-kB2 activation and a proinflammatory state. However, PBMCs from patients with ESKD treated with paricalcitol did not exhibit these features. Experiments in cultured cells confirmed the link between TRAF3 and NF-kB2/inflammation. Decreased TRAF3 ubiquitination in K48-linked chains and cIAP1-TRAF3 interaction mediated the mechanisms of paricalcitol action.TRAF3 overexpression by CRISPR/Cas9 technology mimicked VDRA’s effects. In a preclinical model of kidney injury, paricalcitol inhibited renal NF-kB2 activation and decreased renal inflammation. In VDR knockout mice with renal injury, paricalcitol prevented TRAF3 downregulation and NF-kB2–dependent gene upregulation, suggesting a VDR-independent anti-inflammatory effect of paricalcitol. Conclusions These data suggest the anti-inflammatory actions of paricalcitol depend on TRAF3 modulation and subsequent inhibition of the noncanonical NF-kB2 pathway, identifying a novel mechanism for VDRA’s effects. Circulating TRAF3 levels could be a biomarker of renal damage associated with the inflammatory state.

JASN 31: 2026–2042, 2020. doi: https://doi.org/10.1681/ASN.2019111206

CKD is a devastating condition that affects 5%–7% thought to contribute to both CKD progression of the population worldwide. The number of pa- and cardiovascular mortality.1,6 Deficiency in tients with CKD and ESKD) is increasing due to the growing incidence of diabetes mellitus type 2, hy- 1 pertension, obesity, and aging. Patients with CKD Received November 23, 2019. Accepted April 7, 2020. frequently show vitamin D deficiency due to insuf- fi Published online ahead of print. Publication date available at cient generation of the active metabolite 1,25- www.jasn.org. dihydroxyvitamin D by injured kidneys. CKD is Correspondence: Dr. Marta Ruiz-Ortega, Molecular and Cellular associated with impaired immune system responses Biology in Renal and Vascular Pathology, IIS-Fundación Jiménez and chronic systemic inflammation, characterized Díaz, Avenida Reyes Católicos, 2, 28040 Madrid, Spain. Email: by elevated circulating proinflammatory cytokines [email protected] – and activated circulating cells.2 5 Inflammation is Copyright © 2020 by the American Society of Nephrology

2026 ISSN : 1046-6673/3109-2026 JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH vitamin D and its active metabolites has been associated with Significance Statement adverse outcomes in patients with ESKD and in the general population.7 However, despite evidence that vitamin D supple- TNF receptor–associated factor 3 (TRAF3) downregulation is a key mentation or vitamin D receptor (VDR) agonists (VDRAs), feature promoting inflammation in CKD, and noncanonical NF-kB2 fl such as paricalcitol, may have antiproteinuric and anti- activation is a key driver of in ammation in this context. TRAF3 fl levels in PBMCs are decreased in patients on hemodialysis and in ammatory properties in preclinical and clinical kidney dis- could be a biomarker for the inflammatory state. Paricalcitol may – ease, there is no agreement on the overall clinical effect.7 21 reverse TRAF3 downregulation in a vitamin D receptor– Thus, in clinical trials, vitamin D improved endothelial func- independent manner, suggesting novel signaling pathways behind tion in patients with CKD but did not preserve kidney function the anti-inflammatory effect of paricalcitol. in patients with type 2 diabetes.22–24 Paricalcitol had an erratic effect on albuminuria but did not alter left ventricular mass with ESKD on hemodialysis, treated or not with paricalcitol, 21,25,26 index. Systematic reviews and meta-analyses have also as well as from healthy donors (used as controls). Inclusion been inconsistent on the effect of vitamin D or its analogues on criteria were as follows: age .45 years; male sex; informed 27–31 kidney or cardiovascular outcomes. These differences re- consent; and absence of active inflammatory, infectious, or fi main unexplained and may depend on speci c patient charac- malignant diseases at the initiation or during the study. fi teristics such as the prior existence of vitamin D de ciency, TheIIS-Fundación Jiménez Díaz (IIS-FJD) Ethics Committee 32,33 dietary salt intake, or other features. An improved under- approved this study. The primary aim was to assess the expres- standing of the underlying molecular mechanisms activated by sion of inflammation-related genes in PBMCs. Patients were fi paricalcitol may shed light on the contradictory clinical nd- enrolled after providing written informed consent. PBMCs ings and facilitate the design of novel therapeutic strategies. were separated from whole blood samples (30 ml into EDTA k fl 34,35 The NF- B pathway is a key regulator of in ammation. tubes) using Ficoll density gradient centrifugation. In total, k TheroleofthecanonicalNF- B1 pathway in kidney disease 503106 PBMCs were obtained from 30 ml of whole blood has been well characterized. The canonical pathway proceeds and 63106 PBMCs were used to isolate protein, RNA, and k a through phosphorylation of I B and p65, and the subse- nuclear/cytosolic protein levels using the NE-PER Reagent – k quent translocation of the active p65 NF- B complex to the (Pierce) following the manufacturer’s instructions. nucleus, where it binds to specific promoters and regulates gene transcription.34–37 In particular, p65 phosphorylation Cultured Cells on Ser536 has been linked to proinflammatory gene regula- Human kidney proximal tubule epithelial cells (HK2 cell line, tion.38 Activation of the noncanonical NF-kB2 pathway is ATCC CRL-2190) were grown in RPMI 1640 with 10% FBS, regulated by NF-kB–inducing kinase (NIK), which collabo- 1% glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin, rates with IkB kinase a (IKKa) to induce phosphorylation- 5 mg/ml insulin-transferrin-selenite (ITS), and 36 ng/ml hy- dependent ubiquitination and processing of p100 NF-kB2 to drocortisone in 5% carbon dioxide at 37°C. When cells p52 NF-kB2 and nuclear translocation of p52/RelB.34 In hu- reached 60%–70% confluence, they were serum depleted for man and experimental kidney diseases, elevated renal NF-kB 24 hours before the experiment. activity correlates with upregulation of proinflammatory Murine proximal tubular epithelial cells (MCT cell line) parameters.34–37,39,40 However, most studies focused on the were originally obtained from Dr. Eric Neilson (Vanderbilt canonical NF-kB1 pathway, and data about the noncanonical University) and used for gene expression studies. These cells NF-kB2 pathway are scarce. TNF receptor–associated factors were grown in RPMI 1640 with 10% ineFBS, 1% glutamine, (TRAFs) are critical signaling adaptors downstream of proin- 100 U/ml penicillin, and 100 mg/ml streptomycin in 5% car- flammatory receptors. Whereas TRAF2, TRAF5, and TRAF6 bon dioxide at 37°C. When cells reached 60%–70% conflu- activate the canonical NF-kB1 signaling, TRAF3 inhibits the ence, they were maintained in RPMI with 1% FBS for noncanonical NF-kB2 pathway by promoting NIK ubiquitin– 24 hours. dependent degradation.41 In this paper we described for the Cells were stimulated with 100 ng/ml recombinant human first time that circulating TRAF3 levels are decreased in CKD soluble TNF-like weak inducer of apoptosis (TWEAK; Milli- on hemodialysis and could be a biomarker of renal damage pore) or 200 mg/ml p-cresyl sulfate (APExBIO). Concentra- associated with the inflammatory state. Paricalcitol restored tions of TWEAK and p-cresyl sulfatep-CS) were based on pre- TRAF3 levels inhibiting the noncanonical NF-kB2 pathway viously published dose-response experiments.42,43 In some and decreased renal inflammation. experiments, cells were preincubated for 48 hours with 12 mmol/L paricalcitol (Abbott) before stimulation. DMSO, used as a solvent in some cases, had no effect on cell viability or METHODS on gene expression levels (not shown).

PBMC Isolation Experimental Animal Models Intracellular protein and mRNA levels were measured in All procedures on animals were performed according to the eripheral blood mononuclear cells (PBMCs) from patients recommendations issued by the European Community and

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inflammation 2027 BASIC RESEARCH www.jasn.org the protocol was approved by IIS-FJD Animal Research Ethical staining in a Dako Autostainer. The steps were as follows: Committee and by the Community of Madrid All animal work (1) endogenous peroxidase blockade; (2) incubation with pri- took place in the Laboratory of Molecular and Cellular Biology mary antibodies anti-CD3 (1:300; Dako) or anti-F4/80 in Renal and Vascular Pathology of the IIS-FJD at the Auton- (1:5000; Serotec), anti-p52 NF-kB2 (1:50; Cell Signaling), oma University of Madrid. anti-RelB (1:50; Santa Cruz Biotechnology), and anti-CCL- At the time of euthanasia, animals were anesthetized with 21 (1:90; Santa Cruz Biotechnology); (3) washing; and (4) 5 mg/kg xylazine (Bayer AG) and 35 mg/kg ketamine (Pfizer), DUOFLEX Doublestain EnVision treatment, using 3,39- and kidneys were perfused in situ with cold saline before re- diaminobenzidine as chromogen. For F4/80 staining, a rabbit moval. Then, kidney portions were fixed in buffered formalin anti-rat linker was used before EnVision. Sections were coun- for immunohistochemistry studies or immediately frozen in terstained with Carazzi hematoxylin. The intensity of the re- liquid nitrogen for gene and protein studies. active mark was obtained using Image-Pro Plus software. For each sample (processed by duplicate in a blinded manner), the TWEAK Administration average value was obtained from the analysis of four fields C57BL/6 female mice (9–12 weeks, weight 20 g, seven to eight (203 objective) as density/mm2 or percentage stained area animals per group) received a single intraperitoneal (i.p.) in- versus total analyzed area. Data are expressed as fold increase jection of 0.5 mg TWEAK dissolved in saline and were eutha- over control mice, as mean6SEM of 8–10 animals per group. nized 24 hours later, as previously described.42 Controls were Negative controls include nonspecific Ig and no primary an- injected with the saline vehicle (n58–10 mice per group). tibody (not shown). TWEAK endotoxin levels were ,0.1 ng/mg, confirmed by matrix-assisted laser desorption/ionization–time of flight Protein Studies (MALDI-TOF) (not shown). Some animals received parical- Proteins were obtained from treated cells or mouse kidneys citol (750 ng/kg per day, i.p.; Abbot) or the NF-kB2 inhibitor using lysis buffer (50 mmol/L Tris-hydrochloride, 150 mol/L SN52 (0.7 mg/mouse per day), starting 48 hours before sodium chloride, 2mmol/L EDTA, 2 mmol/L EGTA, 0.2% TWEAK administration. Animals were euthanized 24 hours Triton X-100, 0.3% IGEPAL, 10 ml/ml proteinase inhibitor later. The dose of paricalcitol was chosen based on prior ex- cocktail, 0.2 mmol/L PMSF, and 0.2 mmol/L orthovanadate). perience in mild inflammatory kidney conditions. To determine protein content, the bicinchoninic acid method was used. Unilateral Ureteral Obstruction For Western blot, cell (25 mg/lane) and kidney Unilateral ureteral obstruction (UUO) is used as a model of (100–150 mg/lane) protein extracts were separated on accelerated CKD. It was established in male C57BL/6 mice, 6%–12% polyacrylamide-SDS gels under reducing condi- under isoflurane-induced anesthesia. The left ureter was li- tions. Samples were then transferred onto nitrocellulose gated with silk (5/0) at two locations and cut between ligatures membranes (Bio-Rad), blocked with Tris-buffered saline/5% to prevent urinary tract infection (obstructed kidney).44,45 defatted milk/0.05% Tween 20, and incubated overnight at Two groups were studied: the untreated group or the group 4°C with the following antibodies (dilution): anti–NF-kB2 treated with paricalcitol (750 ng/kg per day), with n56–8mice p52 (1:500), phosphorylated-p65 NF-kB (1:500), and anti– per group. Studies compared both kidneys (contralateral ver- p65 NF-kB (1:500; Cell Signaling); and anti–CCL-21A sus obstructed) in each mouse. (sc-25445, 1:500), anti–RelB (sc-226, 1:500), anti–p-IKKa (sc-101706, 1:500), anti–p-IkBa (sc-8404, 1:500), anti–IkBa Folic Acid Nephropathy (sc-371, 1:500), anti–TRAF3 (sc-6933, 1:250), anti–cIAP1 (sc- Folic acid (FA) nephropathy is a classic model of kidney tu- 271419, 1:500), anti–NIK (sc-7211, 1:500 Santa Cruz Biotech- bulointerstitial injury, inflammation, and AKI that has been nology). Membranes were subsequently incubated with reported in humans. Mice received a single i.p. injection of FA peroxidase-conjugated IgG secondary antibody and devel- (300 mg/kg; Sigma) in 0.3 mol/L sodium bicarbonate or ve- oped using an ECL Chemiluminescence Kit (Amersham). hicle, and were euthanized 48 hours later. Two groups were Loading controls included anti–glyceraldehyde-3-phosphate studied: the untreated group or the group treated with pari- dehydrogenase (GAPDH; 1:10000; Chemicon), anti–a tubu- calcitol (25 mg/kg per day), starting 24 hours before FA, lin (1:5000; Sigma-Aldrich), or anti–histone H1 (sc-8030, n57–8 per group. The dose of paricalcitol was chosen based 1:250; Santa Cruz Biotechnology) for nuclear proteins or total on dose-finding preliminary experiments. protein levels in phosphorylation studies. Autoradiographs were scanned using the Gel Doc EZ imager and analyzed Renal Histology and Immunohistochemistry with the Image Lab 3.0 software (Bio-Rad). Paraffin-embedded kidney sections (3 mm) were stained using ELISA was used to evaluate levels of the chemokines CCL-2 conventional methods. Antigen retrieval was performed by PT and CCL-5 (eBioscience). In renal samples, total protein con- Link system (Dako Diagnostics) with sodium citrate buffer tent was determined by the bicinchoninic acid method, and (10 mmol/L) adjusted to pH 6–9, depending on the immuno- equal amounts of protein were analyzed. Data are expressed as histochemical marker, followed by immunohistochemical n-fold increase over the mean of control levels.

2028 JASN JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH

Gene Expression Studies instructions. Cells were then incubated with 10% heat- Total RNA was isolated from cells and mouse kidney samples inactivated FBS for 24 hours, followed by 24 hours in serum- with Trizol (Invitrogen). The cDNA was synthesized using the free medium. High-Capacity cDNA Archive Kit (Applied Biosystems) using 2 mg total RNA primed with random hexamer primers. Mul- Coimmunoprecipitation Assays tiplex real-time PCR was performed using the following Ap- Cells growing in tissue culture dish plates were lysed in plied Biosystems expression assays for murine and human 300–500 mlTriton–NP-40 lysis buffer (50 mmol/L Tris- samples: ccl-2 Mm00441242_m1, ccl-5 Mm_ 01302428_m1, hydrochloride pH 8, 150 mmol/L sodium chloride, 1 mmol/ il-6 Mm_00446190_m1, ccl-19 Mm_00839967_g1, ccl- L PMSF, 1% NP-40/IGEPAL, and a phosphatase-inhibitor 21a Mm_036466971_gH, TRAF3 Hs_00936781_m1, cocktail [Set II; Calbiochem]), scraped off the dish, and CCL-2 Hs00234140_m1, CCL-5 Hs00982282_m1, IL-6 incubated for 1 hour at 4°C with shaking. Cell lysates were Hs00174131_m1, CCL-19 Hs_00171149_m1, and CCL-21A precleared by incubating with 10 ml protein A agarose bead Hs_00989654_m1. Data were normalized to 18S; 4210893E slurries (0.5 ml agarose/2 ml PBS) for 30 minutes at 4°C and (VIC) and gapdh Mm99999915_g1. The mRNA copy num- then centrifuged three times for 5 minutes at 2500 rpm to wash bers were calculated for each sample by the instrument soft- supernatants. Precleared lysates were incubated with 2.5–5 mg ware using Ct value. Results are expressed in copy numbers, mouse monoclonal anti-TRAF3 antibody (sc-6933; Santa Cruz calculated relative to unstimulated cells or control mice, after Biotechnology) overnight at 4°C. The immune complexes were normalization against 18S or gapdh. Spike-in was not used as captured by the addition of protein A/G PLUS-agarose (20 ml) an internal control. bead slurries for 1 hour at 4°C. The agarose beads were collected by centrifugation, washed three times with lysis buffer, resuspen- ELISA-Based NF-kB2 and RelB Assay ded in 23 Laemmli sample buffer, boiled for 5 minutes, and Nuclear and cytoplasmic fractions were separated from renal subjected to SDS-PAGE. Then Western blot was performed tissues using the NE-PER Reagent following the manufac- as described above, using 1:250 anti-ubiquitin Lys48 specific turer’s instructions. In renal nuclear extracts, NF-kB2 and (05-1307; Millipore) or 1:250 anti-TRAF3 (sc-6933) antibodies. RelB DNA binding activity was measured as binding to an oligonucleotide containing the NF-kB consensus site using a Statistical Analyses TransAM NF-kB Family kit (Active Motif) with an antibody All results are expressed as mean6SEM. Differences between in- that only recognizes active NF-kB2 p52 and RelB. tervention groups and controls were assessed by Mann–Whitney test. P,0.05 was considered significant. Statistical analysis was Activation of Gene Expression by Magnetofection conducted using the SPSS statistical software (version 11.0). Gene overexpression was achieved in cultured cells using the activation TRAF3 CRISPR/Cas9 DNA plasmid (Santa Cruz Biotechnology). Subconfluent cells were transfected with RESULTS magnetofection (OZ biosciences) mixture (1 mlPolyMag CRISPR magnetofection reagent plus 1 mg DNA plasmid) Paricalcitol Inhibits NF-kB2 Activation without for 30 minutes above the magnetic plate according to the man- Modulating NF-kB1 Pathway in PBMCs from Patients ufacturer’s instructions. Then, cells were incubated with 10% with CKD heat-inactivated FBS for 24 hours and incubated in serum-free We used PBMCs to study paricalcitol anti-inflammatory medium for 24 hours before experiments. Cells were stimu- mechanisms PBMCs. In PBMCs from patients with ESKD, lated with recombinant human soluble TWEAK (Millipore). both the canonical and noncanonical NF-kB pathways were In some experiments, cells were preincubated for 48 hours activated, as demonstrated by increased p65/NF-kB1 and RelB with 12 mmol/L paricalcitol (Abbott) before stimulation. Ac- or p52/NF-kB2 DNA binding activity (Figure 1, A and B). tivation of the NF-kB2 pathway was assessed by gene expres- Interestingly, in PBMCs from patients with ESKD who were sion and Western blot analysis using 1:200 anti-p52/NF-kB2 treated with paricalcitol, p65/NF-kB1 activation was similar to and 1:200 anti–CCL-21A. Anti-TRAF3 (1:250) was used for patients with ESKD who were not treated with paricalcitol, overexpression specificity and efficiency assessment, and anti- where the activation of the NF-kB2 pathway was inhibited GAPDH (1:10,000) as loading control. (Figure 1, A and B). In PBMCs from patients with ESKD, proinflammatory gene expression levels were higher than in Gene Silencing controls, but were significantly lower in patients with ESKD Gene silencing in cultured cells was performed using a prede- treated with paricalcitol (Figure 1C). signed small interfering RNA (siRNA) corresponding to membrane associated, rapid response steroid-binding Paricalcitol Inhibits NF-kB2 Activation in (MARRS; Ambion). Subconfluent cells were transfected for Cytokine-Stimulated Cultured Cells 24 hours with 25 nmol/L siRNA using 50 nmol/L Lipofectamine TWEAK is one of the few cytokines that activates both NF-kB RNAiMAX (Invitrogen) according to the manufacturer’s pathways.34 In PBMCs from healthy donors, paricalcitol

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inflammation 2029 BASIC RESEARCH www.jasn.org inhibited TWEAK-induced NF-kB2 activation without mod- (Supplemental Figure 1), thus displaying a clear-cut anti- ulating the canonical NF-kB1 pathway (Figure 2A). Similar inflammatory effect. Interestingly, in TWEAK-injected mice, effects were observed in cells stimulated with p-cresyl sulfate, a SN52 prevented the upregulation of specificNF-kB2 target representative uremic toxin that is not adequately cleared by genes, but also NF-kB1–regulated genes (il-6, ccl-2,andccl-5) dialysis (Figure 2B).43,46 Additionally, paricalcitol inhibited (Supplemental Figure 1). TWEAK-induced upregulation of several proinflammatory In UUO, paricalcitol was previously shown to decrease kid- genes, such as ccl-2, ccl-5,andil-6, as well as the specificNF- ney inflammatory cell infiltration and expression of ccl-2 and kB2–regulated genes ccl-21a and ccl-19(Figure 2C). ccl-5, without blocking nuclear p65/NF-kB1 translocation.49 Next, we performed studies in cultured human tubular Therefore, we investigated whether paricalcitol modulated the epithelial HK2 cells to represent renal cells expressing com- NF-kB2 pathway in UUO. Paricalcitol prevented the increased ponents of the noncanonical NF-kB2 pathway that promote RelB levels and nuclear localization observed in obstructed kidney injury.47,48 In cultured tubular cells, TWEAK increased kidneys (Figure 5A). Moreover, it blocked p100/p52 process- p65 and IkBa phosphorylation and downregulated cytosolic ing as well as p52 nuclear translocation and DNA binding IkBa levels (Figure 3, A and B). Preincubation with paricalci- activity (Figure 5, A–C). By contrast, activation of NF-kB1, eval- tol did not modulate TWEAK-induced NF-kB1 activation uated by p65 nuclear translocation and IkBa phosphorylation, (Figure 3, A and B) but inhibited TWEAK-induced IKKa was not modified in paricalcitol-treated obstructed kidneys phosphorylation, the increase in NIK, and nuclear p52 and (Figure 5, D and E), confirming prior reports.49 Importantly, RelB(Figure3,A,CandD).Paricalcitolalsoinhibited in obstructed kidneys, paricalcitol prevented inflammatory cell TWEAK-induced cytokine gene expression (Figure 3, E and infiltration and the increased expression of proinflammatory F). Moreover, NF-kB2 inhibition with SN52 prevented the factors (such as il-6, ccl-2,andccl-5) and specificNF-kB2– TWEAK-induced upregulation of proinflammatory genes regulated chemokines (Supplemental Figure 4). (Figure 3, E and F), mimicking the effects of paricalcitol. Next, we explored paricalcitol effects in FA-induced renal damage in mice. Paricalcitol only blocked NF-kB2 activation Paricalcitol Inhibits NF-kB2 Activation In Vivo (as assessed by RelB and p52 nuclear translocation, p52 protein We next explored the effect of paricalcitol on NF-kB2 activa- levels, DNA binding activity of RelB, and NF-kB2–dependent tion in three different preclinical models: TWEAK-induced cytokines such as ccl-21a), whereas NF-kB1 activation was not renal inflammation, UUO, and FA-induced AKI. modified. Paricalcitol also prevented the decrease in renal func- TWEAK administration in mice causes an acute renal tion as assessed by BUN levels (Supplemental Figure 5). inflammatory response and activates both canonical and noncanonical NF-kB pathways.34 In TWEAK-injected mice, par- Paricalcitol Increases TRAF3 Levels icalcitol significantly decreased the number of infiltrating TRAF3 plays an essential role in regulating the activation of the monocytes/macrophages and T lymphocytes (Supplemental noncanonical NF-kB2 signaling pathway by the TNF receptor Figure 1, A and B) and upregulated proinflammatory genes, superfamily.41,50,51 such as ccl-2, ccl-5,andil-6 (Supplemental Figure 1, C and D). First, we investigated TRAF3 levels in patients with ESKD In kidneys from TWEAK-injected mice, IkBa phosphoryla- and their modulation by paricalcitol. In PBMCs from patients tion and nuclear p65 NF-kB levels were higher than in with ESKD, TRAF3 protein was downregulated compared controls, but were not modified by paricalcitol. However, with healthy controls, without a decrease in TRAF3 mRNA paricalcitol inhibited TWEAK-induced activation of kidney levels (Figure 6, A and B), suggesting that TRAF3 protein levels NF-kB2, as demonstrated by reduced nuclear p52 and RelB were regulated post-translationally. Moreover, in patients with accumulation in tubular cells. Moreover, paricalcitol inhibi- ESKD treated with paricalcitol, PBMC TRAF3 levels were sim- ted TWEAK-induced p52 and RelB DNA binding activity ilar to the range of healthy controls, without differences in (Figure 4). TRAF3 mRNA levels (Figure 6, A and B), suggesting that par- TWEAK via the NF-kB2 pathway drives the synthesis of icalcitol prevents TRAF3 protein degradation. Next, we evalu- specific chemokines, such as CCL-21A and CCL-19 in tubular ated TRAF3 levels in experimental kidney injury. Kidney TRAF3 cells.34 In TWEAK-injected mice, paricalcitol reduced renal levels were downregulated in the models of TWEAK adminis- ccl-21a and ccl-19 gene expression and CCL-21A protein levels tration, FA, and UUO, and this effect was prevented by parical- (Supplemental Figures 1E and 2A). Interestingly, CCL-21A citol (Figure 6, C–E). In cultured tubular cells and in PBMCs was located in renal tubules in TWEAK-injected mice from healthy donors, TWEAK downregulated TRAF3 levels and (Supplemental Figure 2, C and D), that is, in the same renal this was again prevented by paricalcitol (Figure 6, F and G). structures where active NF-kB2 components RelB and p52 were located (Figure 4C). Mechanisms of TRAF3 Modulation by Paricalcitol In TWEAK-injected mice, inhibition of the NF-kB2 pathway In cultured cells, TRAF3 overexpression mimicked paricalci- with SN52, a synthetic peptide that blocks the nuclear translo- tol actions. In HK2 cells, transfection with a CRISPR/Cas9 cation of the p52/NF-kB2 heterodimer (Supplemental Figure 3), TRAF3 activation plasmid blocked TWEAK-induced p52 ac- significantly decreased the inflammatory cell infiltration tivation, and the increased expression of proinflammatory

2030 JASN JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH

A B 8 * p52 7 6 Rel B 5 3 4 * 4

3 2 * * 2 1 # NFkB2 subunits DNA binding activity (n-fold) 1 binding activity (n-fold) p65 NF-kB1 subunit DNA 0 0 Control +VDRA VDRA VDRA

ESKD Control Control ESKD ESKD

C CCL-2 CCL-5 IL-6 8

* 6 *

mRNA levels (n-fold) 2 # #

ESKD ESKD ESKD Control Control Control

ESKD+VDRA ESKD+VDRA ESKD+VDRA

Figure 1. Paricalcitol (VDRA) inhibits noncanonical but not canonical activation of NF-kB, and decreases proinflammatory genes in PBMCs from patients with ESKD on hemodialysis. Activation of canonical and noncanonical NF-kB pathways in PBMCs was assessed by (A) RelA/p65 or (B) p52 and RelB DNA binding activity, respectively. (C) Gene expression was assessed by real-time PCR. Number of patients: five to eight per group. Differences between intervention and control groups were assessed by Mann–Whitney test. *P,0.05 versus healthy control; #P,0.05 versus patients with ESKD not treated with paricalcitol. genes such as CCL-2, CCL-5, and IL-6,aswellasNF-kB2– lowered to values in the range of healthy controls (Figure 8A). regulated genes (Figure 7). Moreover, cIAP1 levels were higher in whole protein extracts Next, we investigated upstream mechanisms involved in ofPBMCsfrompatientswithESKDthanfromhealthycon- TRAF3 degradation triggered by paricalcitol. A previous study trols, whereas in patients with ESKD treated with paricalci- demonstrated that TNF superfamily receptor–induced activa- toltheywereclosetocontrolvalues(Figure8A).Similar tion of NF-kB2 signaling is coupled with TRAF3 degradation results were observed in cell culture experiments. In HK2 and concomitant accumulation of NIK.52 The E3 ubiquitin cells, TWEAK stimulation induced TRAF3 ubiquitination ligases cIAPs promote TRAF3 K48-linked ubiquitination, and CIAP1-TRAF3 complex formation and this was preven- thus tagging it for proteasomal degradation.53–59 In PBMCs ted by paricalcitol (Figure 8B). from patients with ESKD, TRAF3 immunoprecipitation studies showed increased TRAF3 ubiquitination in the K48-linked Paricalcitol Decreases NF-kB2 Activation chains, leading to the formation of cIAP1-TRAF3 complexes Independently of VDR and MARRS (Figure 8A). In patients treated with paricalcitol, TRAF3 ubiq- To evaluate the role of VDR in paricalcitol actions, we used uitination and cIAP1-TRAF3 interaction were significantly VDR knockout mice. Surprisingly, paricalcitol decreased

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inﬂammation 2031 BASIC RESEARCH www.jasn.org

A TWEAK B p-CS Vehicle +VDRA Vehicle +VDRA

p-p65 p-p65

p52 p52

GAPDH GAPDH

4 p-p65 * * * 5 p52 3 4

* # 3 2 * *

(n-fold) 2 (n-fold) 1 #

protein/ GAPDH 1 protein/ GAPDH

0 0

TW TW TW TW Vehicle Vehicle Vehicle Vehicle TW+VDRA TW+VDRA TW+VDRA TW+VDRA

C 6 * ccl-2 ccl-5 * il-6 ccl-19 4 * ccl-21a

* # * 2 # # #

mRNA levels (n-fold) #

TW TW TW TW TW VDRA VDRA VDRA VDRA VDRA Vehicle Vehicle Vehicle Vehicle Vehicle TW+VDRA TW+VDRA TW+VDRA TW+VDRA TW+VDRA

Figure 2. Paricalcitol decreased noncanonical NF-kB2 activation but not canonical NF-kB activation in PBMCs from healthy donors. Human PBMCs were pretreated with 12 mM paricalcitol for 24 hours, before stimulation with (A) 100 ng/ml TWEAK (TW) or (B) 200 mg/ml p-cresyl sulfate (p-CS) for 24 hours. (C) Activation of canonical NF-kB1 was assessed as phosphorylated p65 in cytosolic fractions and noncanonical activation of NF-kB2 as p52 levels. PBMCs from control individuals were pretreated with 12 mmol/L paricalcitol or vehicle for 24 hours, and then stimulated with TWEAK for 6 hours. Gene expression was assessed by real-time PCR. Differences between intervention and control groups were assessed by Mann–Whitney test. *P,0.05 versus control; #P,0.05 versus TWEAK or p-CS–stimulated samples. Data are expressed as mean6SEM of four experiments for protein levels and ﬁve experiments for gene expression levels.

TWEAK-induced renal inflammation and prevented TRAF3 with TWEAK. In MARRS silenced cells, paricalcitol dimin- downregulation and NF-kB2–dependent gene upregulation, ished TWEAK-induced NF-kB2 activation, suggesting a as seen by the increase in ccl-21a expression in VDR knockout MARRS-independent effect of paricalcitol (Supplemental mice (Supplemental Figure 6), suggesting a VDR-independent Figure 7). anti-inflammatory effect of paricalcitol. MARRS has been suggested as a VDRA receptor. To test whether MARRS may mediate paricalcitol effects on NF-kB2 DISCUSSION activation, MARRS was blocked by gene silencing. HK2 cells were transfected with a MARRS siRNA or a control siRNA and Several lines of evidence suggest an anti-inflammatory activity then pretreated or not with paricalcitol before stimulation of vitamin D and VDRAs in CKD.60 However, clinical

2032 JASN JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH

A B p-p65 C p-IKKα TWEAK 4 p-IkBα NIK α * * IkB 4 * 3 * * Vehicle +VDRA * 3 65 kDa 2 p-p65 2 #

(n-fold) #

p-IKBα 36kDa 1 * * (n-fold) 1 36 kDa α protein levels / GAPDH IKB 0

protein levels / GAPDH 0 α 88 kDa Cytosolic p-IKK TW TW TW TW Vehicle Vehicle Vehicle TW NIK 125 kDa Vehicle Vehicle TW+VDRA TW+VDRA TW+VDRA TW+VDRA TW+VDRA GAPDH 35-38 kDa D p65 65 kDa * 2.5 * * p65 * p52 52 kDa 2.0 p52 Rel B Rel B 62 kDa 1.5 #

Nuclear # Histone 21 kDa 1.0 H1 (n-fold) 0.5

0.0 protein levels/ Histone H1

TW TW TW Vehicle Vehicle Vehicle TW+VDRA TW+VDRA TW+VDRA E F 20 ccl-2 5 ccl-5 * * ccl-19 il-6 ccl-21a 15 * 4 # * * # 3 # 10 # # # # 2 # 5 #

mRNA levels (n-fold) #

mRNA levels (n-fold) 1

0 0

TW TW TW TW TW Vehicle Vehicle Vehicle Vehicle Vehicle TW+VDRATW+SN52 TW+VDRATW+SN52 TW+VDRATW+SN52 TW+VDRATW+SN52 TW+VDRATW+SN52

Figure 3. Paricalcitol decreased noncanonical NF-kB2 activation but not canonical NF-kB activation induced by TWEAK in renal tubular epithelial cells. (A–D) Human tubular cells were pretreated with 12 mM paricalcitol for 48 hours before stimulation with 100 ng/ml TWEAK (TW) for 15 minutes (NF-kB1 pathway) or 6 hours (NF-kB2 pathway). (A and B) Activation of canonical NF-kB1 was assessed by IkBa or p65 phosphorylation levels in total protein extracts, and p65 levels in cytosolic or nuclear fractions. Activation of the noncanonical NF-kB2 pathway was assessed by (A and C) evaluating NIK and phosphorylated IKKa levels in total protein extracts and (A and D) the subcellular location of p100/p52 and Rel B subunits. GAPDH and histone H1 were used as total and nuclear protein loading controls, respectively. (A) Representative Western blots. (B–D) Quantiﬁcation of protein levels expressed as mean6SEM of four experiments. *P,0.05 versus control; #P,0.05 versus TWEAK. Paricalcitol decreased proinﬂammatory genes in tubular cells. Cultured murine tubular epithelial cells were stimulated with 100 ng/ml TWEAK for 6 hours. Cells were preincubated for 48 hours with 12 mmol/L paricalcitol or 60 ng/ml SN52. (E and F) Gene expression was assessed by real-time PCR. Data expressed as mean6SEM of three to six experiments. Differences between intervention and control groups were assessed by Mann–Whitney test. *P,0.05 versus control; #P,0.05 versus TWEAK-treated cells.

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inﬂammation 2033 BASIC RESEARCH www.jasn.org

A TWEAK + B TWEAK + Vehicle TWEAK VDRA Vehicle TWEAK VDRA

Cytosolic Nuclear p-lkBa p65 GAPDH Red ponceau

6 4 * *

3 * 4

2 * (n-fold) (n-fold) 2 1 p-lkB α /GAPDH p65 / Red ponceau 0 0 Vehicle TW TW+ Vehicle TW TW+ VDRA VDRA

TWEAK + C D Vehicle TWEAK VDRA REL B p52 p52

GAPDH 4

Vehicle * 3

2 #

p52 / GAPDH 1 protein levels (n-fold) 0

TWEAK VehicleTW TW+ VDRA

E 2.0 * p52 * Rel B 1.5 # 1.0 (n-fold) TWEAK + VDRA 0.5 NFkB2 subunits/ DNA binding activity

0.0

TW TW Vehicle Vehicle TW+VDRA TW+VDRA

Figure 4. Paricalcitol only inhibits TWEAK-induced noncanonical NF-kB activation in vivo. Mice were treated with paricalcitol 750 ng/kg per day starting 48 hours before 0.5 mg TWEAK (TW), and euthanized 24 hours after TWEAK administration. (A and B) In TWEAK-injected mice, increased (A) cytosolic phosphorylated IkBa levels and (B) nuclear p65 levels were found, but they were not modulated by paricalcitol treatment, suggesting that paricalcitol did not modulate NF-kB1 activation. GAPDH and Red Ponceau were used as loading controls. Data expressed as mean6SEM of five to eight mice per group. *P,0.05 versus control. (C) Nuclear location of p52 and RelB in renal tubules of TWEAK-injected mice assessed by immunohistochemistry. (D) Western blot assessment of NF-kB2 activation (p52 levels) in whole-kidney protein extracts. (E) In isolated renal nuclear protein extracts, p52 and RelB NF-kB2 DNA binding activity was measured by ELISA. All data expressed as mean6SEM of five to eight animals per group. Differences between intervention and control groups were assessed by Mann–Whitney test. *P,0.05 versus control; #P,0.05 versus TWEAK. outcomes have been inconsistent, suggesting an incomplete involves TRAF3 modulation and subsequent inhibition of the understanding of the pathways involved. This work, studies noncanonical NF-kB2 pathway. in vitro, in experimental renal damage, and in patients with Vitamin D or VDRAs reduce renal inflammatory cell in- ESKD on hemodialysis demonstrate a novel mechanism im- filtration in experimental renal damage.14,60–63 In the UUO plicated in the anti-inflammatory actions of paricalcitol that model, paricalcitol was previously shown to decrease

2034 JASN JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH

A REL B p52 B Ob+ NOb+ NOb Ob VDRA VDRA

p52

GAPDH NOb 8 * 6 # 4 (n-fold) 2 Ob p52 / GAPDH

0 NOb Ob Ob+ NOb+ VDRA VDRA C 6 * p52 Rel B Ob+VDRA

4 * # # (n-fold) 2 NF-kB subunits activity 0 NOb+VDRA

Ob Ob NOb NOb

Ob+VDRA Ob+VDRA NOb+VDRA NOb+VDRA

D Ob+ NOb+ E Ob+ NOb+ NOb Ob VDRA VDRA NOb Ob VDRA VDRA

Nuclear Cytosolic p65 p-IkBα

Red GAPDH Ponceau

60 10 * * 8 40 * 6 * 4 20 2 plkB- α /GAPDH (n-fold)

p65/Red Ponceau (n-fold) 0 0 Nob Ob ob+ Nob+ Nob Ob ob+ Nob+ VDRA VDRA VDRA VDRA

Figure 5. Paricalcitol inhibits NF-kB2 but not NF-kB1 activation in experimental UUO in mice. Mice were treated with paricalcitol 750 ng/kg per day, starting 24 hours before UUO, and studied 5 days after UUO. (A) Immunohistochemistry in parafﬁn-embedded kidney sections. Representative animal from each group (magniﬁcation 2003). (B) NF-kB2 activation was assessed as p52 levels in Western blots of total protein extracts. (C) In isolated renal nuclear proteins, p52 and RelB DNA binding activity was measured by ELISA. (D and E) Western blot assessment of nuclear levels of (D) p65 and (E) cytosolic IkBa phosphorylation were used to evaluate activation of the NF-kB1 pathway. In each mouse, obstructed (Ob) kidneys were compared with the corresponding contralateral nonobstructed (Nob) kidney. Data expressed as mean6SEM of four to eight animals per group. Differences between intervention and control groups were assessed by Mann–Whitney test. *P,0.05 versus contralateral nonobstructed kidneys; #P,0.05 versus obstructed kidneys.

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inﬂammation 2035 BASIC RESEARCH www.jasn.org

A ESKD B control ESKD +VDRA 8 * TRAF3 65-70 kDa 7 6 GAPDH 5 ** 4 4 2.5 # 3 2.0 (n-fold) 1.5 * 2 TRAF3 mRNA levels

(n-fold) 1.0 1

TRAF3/ GAPDH 0.5 0 0.0 Control +VDRA Control +VDRA ESKD ESKD

C TWEAK DEFA +VDRA Control +VDRA Control +VDRA NobOb Ob Nob TRAF3 TRAF3 TRAF3 GAPDH GAPDH GAPDH

3 # 2.0 4 # 1.5 3 2 #

* 1.0 * 2 (n-fold) 1 (n-fold) *

TRAF3/ GAPDH 0.5 TRAF3/ GAPDH 1

0 0.0 0 Control +VDRA Control +VDRA NobOb Ob Nob TWEAK FA +VDRA

FGTWEAK TWEAK Control +VDRA Control +VDRA TRAF3 TRAF3

GAPDH α-Tubulin

1.5 1.4 # # * 1.2 1.0 * 1.0 (n-fold) (n-fold) 0.5 0.8 TRAF3/ GAPDH TRAF3/ α -Tubulin

0.0 0.6

Figure 6. Paricalcitol restored TRAF3 levels in PBMCs from patients with CKD, in experimental kidney injury, and in cultured cells. (A) TRAF3 protein levels in PBMCs from patients with ESKD treated or not with paricalcitol were determined by Western blot. Number of patients: ﬁve to eight per group. *P,0.05 versus control; #P,0.05 versus ESKD cells. (B) TRAF3 mRNA levels in PBMCs from patients with ESKD treated or not with paricalcitol were determined by real-time PCR. Number of patients: ﬁve to eight per group. *P,0.05

2036 JASN JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH interstitial fibrosis64 and renal inflammation, without block- inflammatory cells. SN52, a specific blocker p52/NF-kB2 nu- ing nuclear p65/NF-kB1 translocation.49 We have now ob- clear translocation, inhibited TWEAK-induced renal inflam- served that paricalcitol inhibited activation of the NF-kB2 mation and upregulation of proinflammatory genes both pathway without interfering with NF-kB1 activation, both in in vitro and in vivo. Interestingly, SN52 inhibited genes not injured murine kidneys and in cultured renal cells. TWEAK is regulated by NF-kB1, such as ccl-21a and ccl-19, supporting one of the few cytokines that activates both NF-kB pathways. the existence of a crosstalk between NF-kB pathways.41 Experimental studies, using pharmacologic or genetic ap- The mechanisms by which TNF superfamily receptors such proaches, have clearly demonstrated that TWEAK activates as Fn14 activate the NF-kB2 pathway remain controversial. its receptor Fn14 to induce renal inflammation,65–69 suggest- NIK is continuously synthesized and degraded in resting cells ing this cytokine is relevant in renal diseases. In cultured tu- but becomes stabilized and accumulates in response to TNF bular epithelial cells and in the kidneys in vivo, paricalcitol receptor superfamily members.78 TRAF3 is a key regulator of inhibited the successive steps of noncanonical NF-kB activa- NIK ubiquitin-dependent degradation and protein levels by tion: p-IKKa;NIK,NF-kB2/p52, and RelB protein expression, forming a complex with the E3 ubiquitin ligase cIAP (cIAP1 or and ccl-21a and ccl-19 gene expression, suggesting anti- cIAP2), in which TRAF3 serves as the NIK-binding inflammatory effects mediated by the modulation of NF-kB2– adapter.41,77,79,80 Activation of TNF superfamily receptors in- controlled genes. Prior in vitro studies support our findings. In duces TRAF3 degradation, thereby promoting NIK accumula- human proximal tubular cells, paricalcitol did not affect TNF- tion and NF-kB2 p100 processing to p52.41,52,55,81,82 Indeed, a–mediated IkBa phosphorylation and degradation, nor p65 TRAF3 knockout cells exhibit elevated NIK expression and activation and its nuclear translocation.49 Despite this, we constitutive NF-kB2 p100 processing to p52. More importantly, have found that paricalcitol downregulated many canonical deletion of the NF-kB2 p100 gene rescued the TRAF3 null phe- NF-kB1 genes, such as ccl-2 and ccl-5, in experimental renal notype,51 showing that TRAF3 regulates NF-kB2 p100 process- damage and in cytokine-stimulated cells. Recent studies sug- ing. We found that TRAF3 overexpression blocked TWEAK/ gest a crosstalk between both NF-kB pathways in which the Fn14-induced NF-kB2 p100 processing to p52 and upregulation recruitment of the noncanonical pathway, such as NIK, boosts of NF-kB2 target cytokines, suggesting a link between TRAF3 the consequences of NF-kB activation.48,49,70 Moreover, parical- and the NF-kB2 pathway. Thus, TRAF3 overexpression in vitro citol diminished TWEAK-induced CCL-21A upregulation and mimics the effects of paricalcitol, suggesting a role for TRAF3 blocked NF-kB2 activation in tubular epithelial cells in vivo,sug- modulation in the anti-inflammatory action of paricalcitol. gesting these cells are likely the primary target of vitamin D anti- TRAF2/3 and cIAP1/2 form a cytoplasmic ubiquitin complex inflammatory actions. Accordingly, in cultured human tubular that induces Lys-48–linked ubiquitination and degradation of epithelial cells, NIK overexpression increased nuclear RelB/p52 TRAF3.82 In PBMCs from patients with ESKD, cIAP1-TRAF3 levels and DNA binding activity and expression of proinflamma- complex formation and Lys-48–linked ubiquitination were in- tory cytokines, including CCL-2,CCL-5, IL-6 and IL-8.71 creased in association to NF-kB2 activation and a proinflamma- Blockade of the canonical NF-kB activation, using different tory phenotype. Importantly, paricalcitol reduced TRAF3 K48- inhibitors, such as parthenolide or SN50 peptides, inhibited linked chain ubiquitination and prevented cIAP1-TRAF3 com- experimental inflammation, including kidney damage.37,72,73 plex formation in all of the experimental systems studied. There is recent evidence of noncanonical NF-kB2 activation in Paricalcitol and calcitriol bind to and activate the VDR to reg- human and experimental kidney disease, including increased ulate gene expression.83 As an example, paricalcitol promotes IKKa phosphorylation, and NIK and RelB levels, as well as VDR/p65 complex formation to inhibit p65 transactivation of protection from AKI by RelB downregulation or NIK defi- CCL-5/RANTES gene transcription. However, this mechanism ciency.47,48,74,75 Additionally, NIK or RelB silencing protected did not occur for CCL-2/MCP1.49 To evaluate the role of VDR cultured tubular cells from cell death or inflammatory re- in the action of paricalcitol, we studied VDR knockout mice.84–86 sponses to advanced glycation end products or inflammatory Surprisingly, paricalcitol decreased TWEAK-induced renal in- cytokines.48,53,75,76 However, there are no studies specifically flammation and normalized TRAF3 levels in VDR knockout linking NF-kB2 to renal inflammation. Now, we have found mice. This indicates that paricalcitol has VDR-independent renal activation of the NF-kB2 pathway in several models of anti-inflammatory effects. The VDR-independent effects of renal damage, which is associated with kidney infiltration by VDRAs have been known for some time. Calcitriol also exerts

versus control. (C–E) Evaluation of TRAF3 levels in experimental models of renal damage, assessed by Western blot. Mice were treated with paricalcitol starting before induction of kidney injury. (C) TWEAK injection (0.5 mg/mouse), (D) 300 mg/kg FA, or (E) UUO. (F and G) Paricalcitol restored TRAF3 levels in cytokine-stimulated cells. (F) Human tubular cells or (G) PBMCs from healthy donors were pretreated with paricalcitol 48 hours before TWEAK stimulation (100 ng/ml) for 24 hours. TRAF3 protein levels by Western blot. Data expressed as mean6SEM of four to eight animals per group or of three in vitro experiments. Differences between intervention and control groups were assessed by Mann–Whitney test. *P,0.05 versus control; #P,0.05 versus cells from patients with ESKD, injured- kidney cells, or TWEAK-treated cells. Nob, contralateral nonobstructed kidneys; Ob, obstructed kidneys.

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inﬂammation 2037 BASIC RESEARCH www.jasn.org

B CRISPR ControlTWEAKTWEAK + TRAF3TRAF3 CRISPRTWEAKTWEAK +VDRA

CCL-21A 21 kDa

p52 A GAPDH 8 *

6 p52 CCL-21A 4 4 *

(n-fold) * **# 2 3 TRAF3 mRNA levels # # # 0 2

Control TRAF3 CRISPR (n-fold) 1 Protein levels/ GADPH 0

ControlTWEAK TWEAK ControlTWEAK TWEAK

TRAF3 CRISPRTWEAK +VDRA TRAF3 CRISPRTWEAK +VDRA

TWEAK + TRAF3 CRISPR TWEAK + TRAF3 CRISPR C CCL-2 D CCL-21A 6 CCL-5 5 CCL-19 IL-6 * 4 * # 4 * 3 * * # 2 2 # # # mRNA levels (n-fold) # mRNA levels (n-fold) 1 # 0 0

ControlTWEAK ControlTWEAK ControlTWEAK Control TWEAK ControlTWEAK

TWEAK+VDRA TWEAK+VDRA TWEAK+VDRA TWEAK+VDRA TWEAK+VDRA

TWEAK+TRAF3 CRISPR TWEAK+TRAF3 CRISPRTWEAK+TRAF3 CRISPR TWEAK+TRAF3 CRISPR TWEAK+TRAF3 CRISPR

Figure 7. TRAF3 overexpression mimics the actions of paricalcitol. TRAF3 overexpression was achieved in cultured cells using activation TRAF3/CRISPR/Cas9 DNA plasmid. Cells were stimulated with recombinant human soluble TWEAK (100 ng/ml). In some experiments, cells were preincubated for 48 hours with 15 mmol/L paricalcitol before TWEAK stimulation. (A) TRAF3 mRNA levels are increased in cells transfected with CRISPR/Cas9 TRAF3 activation plasmid as evaluated by real-time PCR. (B) NF-kB2 pathway activation was assessed by Western blot of NF-kB2 p52 and the NF-kB2–regulated cytokine CCL-21A. (C and D) Gene expression of the proinflammatory factors (C)CCL-2, CCL-,andIL-6 or (D) CCL-21A and CCL-19 were evaluated by real-time PCR. Data expressed as mean6SEM of three to five independent experiments. Differences between intervention and control groups were assessed by Mann–Whitney test. *P,0.05 versus control; #P,0.05 versus TWEAK-treated cells. fast nontranscriptional responses involving an increase in As indicated in the introduction, interventional studies on intracellular calcium and activation of intracellular ki- vitamin D or VDRAs in patients with CKD have been inconclu- nases.87,88 A novel calcitriol receptor, MARRS, was proposed sive, with more recent meta-analyses being disappointing.27–31 to mediate nongenomic actions of calcitriol in chick intesti- In some instances, potential confounders have been identified, nal basolateral membranes.89 However, in MARRS-silenced such as salt intake or baseline vitamin D status.32,33 Some of these cells, paricalcitol still prevented TWEAK-induced NF-kB2/p52 confounders, such as salt intake, are known to modulate macro- activation, indicating that paricalcitol modulates NF-kB2 acti- phage activation, but whether they modulate TRAF3 is currently vation in a VDR- and MARRS-independent manner. unknown.90,91 Interestingly, sodium retention has been linked to

2038 JASN JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH

ABESKD TWEAK I.P:TRAF3 I.P:TRAF3 Control +VDRA Control +VDRA I.B:Ub lys-48 I.B:Ub lys-48

Polyubiquitin chains linked through the Lys48 residue

Heavy chain* Heavy chain*

TRAF3 TRAF3

cIAP-1 70 kDa cIAP-1

Figure 8. Modulation of TRAF3 ubiquitination and cIAP1-TRAF3 complex formation was involved in paricalcitol restoration of TRAF3 levels. (A) In PBMCs from patients with ESKD on hemodialysis treated or not with paricalcitol (VDRA), TRAF3 was immunoprecipitated (I.P.) by an anti-TRAF3 antibody followed by SDS-PAGE and Immunoblotting (I.B.) using an anti-ubiquitin antibody. Representative experiment. TRAF3 antibody was used as loading control. cIAP1 levels were also evaluated. Number of patients five to eight per group. (B) TWEAK-induced TRAF3 ubiquitination was prevented by paricalcitol in HK2 cells stimulated with TWEAK and treated or not with paricalcitol. Figures show a representative I.P. experiment out of three performed.* IgG heavy chain. modulation of a TRAF3-interacting protein (TRAF3 interact- patent holder Autonomous University of Madrid and IIS-FJD) pending. All ing protein 2, TRAF3IP2), resulting in cardiovascular in- remaining authors have nothing to disclose. jury.92 The present findings of an anti-inflammatory action of the VDRA paricalcitol mediated by the modulation of TRAF3 levels and NF-kB2 activation open new avenues to FUNDING further optimize therapeutic approaches and clinical trial design. Thus, paricalcitol analogues selected for TRAF3 modu- This work was supported by Instituto de Salud Carlos III (ISCIII; Carlos III lation and baseline risk stratification by TRAF3 expression in Health Institute), Centro de Investigación Biomédica en Red Diabetes y patient PBMCs represent novel strategies for improving drug Enfermedades Metabólicas Asociadas (CIBERDEM; Biomedical Research selection and clinical trial design. Networking Center in Diabetes and Associated Metabolic Disorder), Ministerio de Economía y Competitividad, and European Union European Currently, there is no drug in clinical use targeting NF-kB2. fi Regional Development Fund grants PI17/00119, PI19/00588, PI19/00815, and However, information from patients de cient in NFKB2 is in PI17/01495. This work was also supported by Red de Investigación Renal line with our observations because it is associated with im- (REDinREN) grant RD16/0009; Sociedad Española de Nefrologia; NOVEL- mune deficiency, implying a role for NF-kB2 in inflammation REN-CM: Enfermedad renal crónica: nuevas Estrategias para la prevención, and immune defenses (https://omim.org/entry/615577). Diagnóstico y tratamiento grants B2017/BMD-3751, B2017/BMD-3686, and CIFRA2-CM; and the European Union’s Horizon 2020 research and innova- In conclusion, our data clearly demonstrate that, in pa- “ fi fl tion program for the IMPROVE-PD ( Identi cation and Management of tients with ESKD, paricalcitol decreases in ammatory factors Patients at Risk–Outcome and Vascular Events in Peritoneal Dialysis”) proj- in PBMCs cells by preserving TRAF3 expression and prevent- ect under the Marie Skłodowska-Curie grant agreement number 812699. ing activation of the noncanonical NF-kB2 pathway, thus A.B. Sanz was supported by an ISCIII Miguel Servet grant. S. Rayego- ’ identifying TRAF3/NF-kB2 as an important target for VDRAs. Mateos s salary was supported by the Ministerio de Economía, Industria y Competitividad, Gobierno de España “Juan de la Cierva de Formacion” train- Paricalcitol upregulation of TRAF3 was independent from the ing program, grant FJCI-2016-29050. presence of VDR and MARRS, thus facilitating the design of novel molecules with enhanced anti-inflammatory activity that may be used for kidney protection. Further studies should characterize the effect of earlier CKD stages on TRAF3 levels ACKNOWLEDGMENTS and their modulation by paricalcitol or other VDRAs. We want to thank Susana Carrasco at the IIS-FJD for help with immunohistochemical procedures and María Soledad Sanchez DISCLOSURES Fernández for help with the acquisition of healthy donor samples. Dr. Sandra Rayego-Mateos contributed to the design of the ex- J. Egido, J.L. Morgado-Pascual, A. Ortiz, E. Gonzalez-Parra, S. Rayego- periments, data acquisition, analysis, interpretation of the data, and Mateos, and M. Ruiz-Ortega have a patent entitled “In Vitro Method for De- drafted the manuscript; Dr. José Luis Morgado-Pascual contributed tecting Renal Disease” (European patent application number, EP19382470.3; to data acquisition, mouse model development, and data analysis;

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inﬂammation 2039 BASIC RESEARCH www.jasn.org

Dr. José Manuel Valdivielso and Dr. Ana Belén Sanz contributed data factors in chronic kidney disease: the NEFRONA study. Nephrol Dial acquisition, critical review, and financial support; Dr. Enrique Bosch- Transplant 29: 1415–1422, 2014 Panadero and Dr. Raul R. Rodrigues-Díez contributed to data ac- 6. Vanholder R, Fouque D, Glorieux G, Heine GH, Kanbay M, Mallamaci F, et al.; European Renal Association European Dialysis; Transplant As- quisition and critical review; Dr. Alberto Ortiz, Dr. Jesús Egido, and sociation (ERA-EDTA) European Renal; Cardiovascular Medicine Dr. Emilio González-Parra contributed to the critical review of the (EURECA-m) working group: Clinical management of the uraemic manuscript and financial support; and Dr. Marta Ruiz-Ortega syndrome in chronic kidney disease. Lancet Diabetes Endocrinol 4: contributed to the design of the experiments, analysis and inter- 360–373, 2016 pretation of the data, financial support, and drafted the manuscript. 7. Gonzalez-Parra E, Rojas-Rivera J, Tuñón J, Praga M, Ortiz A, Egido J: fi Vitamin D receptor activation and cardiovascular disease. Nephrol Dial All authors have approved the nal version of this manuscript for Transplant 27[Suppl 4]: iv17-iv21, 2012 publication. 8. Agarwal R, Acharya M, Tian J, Hippensteel RL, Melnick JZ, Qiu P, et al.: Dr. Alberto Ortiz reports personal fees from Amicus, Amgen, Antiproteinuric effect of oral paricalcitol in chronic kidney disease. AstraZeneca, Fresenius Medical Care, Genzyme, Kyowa-Kirin, Kidney Int 68: 2823–2828, 2005 Menarini, Mundipharma, Otsuka, and Shire, as well as grants from 9. Liu L-J, Lv JC, Shi SF, Chen YQ, Zhang H, Wang HY: Oral calcitriol for reduction of proteinuria in patients with IgA nephropathy: a random- Genzyme, outside the submitted work. ized controlled trial. Am J Kidney Dis 59: 67–74, 2012 10. Cheng J, Zhang W, Zhang X, Li X, Chen J: Efficacy and safety of paricalcitol therapy for chronic kidney disease: a meta-analysis. Clin J Am SUPPLEMENTAL MATERIAL Soc Nephrol 7: 391–400, 2012 11. Zhang Z, Sun L, Wang Y, Ning G, Minto AW, Kong J, et al.: Re- noprotective role of the vitamin D receptor in diabetic nephropathy. This article contains the following supplemental material online at Kidney Int 73: 163–171, 2008 http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2019111206/-/ 12. Branisteanu DD, Leenaerts P, van Damme B, Bouillon R: Partial pre- DCSupplemental. vention of active Heymann nephritis by 1 alpha, 25 dihydroxyvitamin Supplemental Figure 1. Paricalcitol or NF-kB2 blockade decreased D3. Clin Exp Immunol 94: 412–417, 1993 TWEAK-induced renal inflammation. 13. Wang Y, Zhou J, Minto AW, Hack BK, Alexander JJ, Haas M, et al.: Altered vitamin D metabolism in type II diabetic mouse glomeruli may Supplemental Figure 2. Paricalcitol inhibits TWEAK-induced provide protection from diabetic nephropathy. Kidney Int 70: 882–891, upregulation of specificNF-kB2 targets in the kidney. 2006 Supplemental Figure 3. SN52 peptide blocks NF-kB2 activation 14. Sanchez-Niño MD, Sanz AB, Carrasco S, Saleem MA, Mathieson PW, and NF-kB2 nuclear translocation in a experimental TWEAK- Valdivielso JM, et al.: Globotriaosylsphingosine actions on human glomerular podocytes: implications for Fabry nephropathy. Nephrol induced renal damage. – fl Dial Transplant 26: 1797 1802, 2011 Supplemental Figure 4. Paricalcitol decreases renal in ammation 15. Freundlich M, Quiroz Y, Zhang Z, Zhang Y, Bravo Y, Weisinger JR, et al.: in experimental Unilateral Ureteral Obstruction (UUO) in mice. Suppression of renin-angiotensin gene expression in the kidney by Supplemental Figure 5. Paricalcitol inhibits NF-kB2, but not paricalcitol. Kidney Int 74: 1394–1402, 2008 NF-kB1 activation in folic acid-induced renal injury. 16. Piecha G, Kokeny G, Nakagawa K, Koleganova N, Geldyyev A, Berger I, fi Supplemental Figure 6. Paricalcitol reduces TWEAK-induced re- et al.: Calcimimetic R-568 or calcitriol: equally bene cial on progression fl of renal damage in subtotally nephrectomized rats. Am J Physiol Renal nal in ammation in VDR KO mice. Physiol 294: F748–F757, 2008 Supplemental Figure 7. Paricalcitol inhibits TWEAK-induced up- 17. Matsui I, Hamano T, Tomida K, Inoue K, Takabatake Y, Nagasawa Y, regulation of specificNF-kB2 targets in MARRS gene silenced cells. et al.: Active vitamin D and its analogue, 22-oxacalcitriol, ameliorate puromycin aminonucleoside-induced nephrosis in rats. Nephrol Dial Transplant 24: 2354–2361, 2009 18. Teng M, Wolf M, Lowrie E, Ofsthun N, Lazarus JM, Thadhani R: Survival REFERENCES of patients undergoing hemodialysis with paricalcitol or calcitriol therapy. NEnglJMed349: 446–456, 2003 1. Ortiz A, Fernandez-Fernandez B: Humble kidneys predict mighty heart 19. Alborzi P, Patel NA, Peterson C, Bills JE, Bekele DM, Bunaye Z, et al.: troubles. Lancet Diabetes Endocrinol 3: 489–491, 2015 Paricalcitol reduces albuminuria and inflammation in chronic kidney 2. Niewczas MA, Pavkov ME, Skupien J, Smiles A, Md Dom ZI, Wilson JM, disease: a randomized double-blind pilot trial. Hypertens 52: et al.: A signature of circulating inflammatory proteins and development of 249–255, 2008 end-stage renal disease in diabetes. Nat Med 25: 805–813, 2019 20. Fishbane S, Chittineni H, Packman M, Dutka P, Ali N, Durie N: Oral 3. Heine GH, Ortiz A, Massy ZA, Lindholm B, Wiecek A, Martínez-Castelao paricalcitol in the treatment of patients with CKD and proteinuria: a A, et al.; European Renal and Cardiovascular Medicine (EURECA-m) randomized trial. AmJKidneyDis54: 647–652, 2009 working group of the European Renal Association-European Dialysis 21. de Zeeuw D, Agarwal R, Amdahl M, Audhya P, Coyne D, Garimella T, and Transplant Association (ERA-EDTA): Monocyte subpopulations et al.: Selective vitamin D receptor activation with paricalcitol for re- and cardiovascular risk in chronic kidney disease. Nat Rev Nephrol 8: duction of albuminuria in patients with type 2 diabetes (VITAL study): a 362–369, 2012 randomised controlled trial. Lancet 376: 1543–1551, 2010 4. Zhang J, Hua G, Zhang X, Tong R, Du X, Li Z: Regulatory T cells/T-helper 22. de Boer IH, Zelnick LR, Ruzinski J, Friedenberg G, Duszlak J, Bubes VY, cell 17 functional imbalance in uraemic patients on maintenance hae- et al.: Effect of vitamin D and omega-3 fatty acid supplementation on modialysis: A pivotal link between microinflammation and adverse kidney function in patients with type 2 diabetes: a randomized clinical cardiovascular events. Nephrology (Carlton) 15: 33–41, 2010 trial. JAMA 322: 1899, 2019 5. Betriu A, Martinez-Alonso M, Arcidiacono MV, Cannata-Andia J, 23. Levin A, Tang M, Perry T, Zalunardo N, Beaulieu M, Dubland JA, et al.: Pascual J, Valdivielso JM, et al.; Investigators from the NEFRONA Randomized controlled trial for the effect of vitamin D supplementation on Study: Prevalence of subclinical atheromatosis and associated risk vascular stiffness in CKD. Clin J Am Soc Nephrol 12: 1447–1460, 2017

2040 JASN JASN 31: 2026–2042, 2020 www.jasn.org BASIC RESEARCH

24. Kumar V, Yadav AK, Lal A, Kumar V, Singhal M, Billot L, et al.: A ran- 45. Rodrigues-Díez R, Rodrigues-Díez RR, Rayego-Mateos S, Suarez- domized trial of vitamin D supplementation on vascular function in Alvarez B, Lavoz C, Stark Aroeira L, et al.: The C-terminal module IV of CKD. J Am Soc Nephrol 28: 3100–3108, 2017 connective tissue growth factor is a novel immune modulator of the 25. Thadhani R, Appelbaum E, Pritchett Y, Chang Y, Wenger J, Tamez H, Th17 response. Lab Invest 93: 812–824, 2013 et al.: Vitamin D therapy and cardiac structure and function in patients 46. Rossi M, Campbell K, Johnson D, Stanton T, Pascoe E, Hawley C, et al.: with chronic kidney disease: the PRIMO randomized controlled trial. Uraemic toxins and cardiovascular disease across the chronic kidney JAMA 307: 674–684, 2012 disease spectrum: an observational study. Nutr Metab Cardiovasc Dis 26. Parvanova A, Trillini M, Podestà MA, Iliev IP, Ruggiero B, Abbate M, 24: 1035–1042, 2014 et al.; PROCEED Study Organization and the Scientific Writing Acad- 47. Starkey JM, Haidacher SJ, LeJeune WS, Zhang X, Tieu BC, Choudhary S, emy (SWA) 2016: Moderate salt restriction with or without paricalcitol in et al.: Diabetes-induced activation of canonical and noncanonical nuclear type 2 diabetes and losartan-resistant macroalbuminuria (PROCEED): a factor-kappaB pathways in renal cortex. Diabetes 55: 1252–1259, 2006 randomised, double-blind, placebo-controlled, crossover trial. Lancet 48. Ortiz A, Husi H, Gonzalez-Lafuente L, Valiño-Rivas L, Fresno M, Sanz Diabetes Endocrinol 6: 27–40, 2018 AB, et al.: Mitogen-activated protein kinase 14 promotes AKI. JAmSoc 27. de Borst MH, Hajhosseiny R, Tamez H, Wenger J, Thadhani R, Goldsmith Nephrol 28: 823–836, 2017 DJ: Active vitamin D treatment for reduction of residual proteinuria: a 49. Tan X, Wen X, Liu Y: Paricalcitol inhibits renal inflammation by pro- systematic review. J Am Soc Nephrol 24: 1863–1871, 2013 moting vitamin D receptor-mediated sequestration of NF-kappaB 28. Hu X, Shang J, Yuan W, Zhang S, Jiang Y, Zhao B, et al.: Effects of signaling. JAmSocNephrol19: 1741–1752, 2008 paricalcitol on cardiovascular outcomes and renal function in patients 50. Yao Z, Xing L, Boyce BF: NF-kappaB p100 limits TNF-induced bone with chronic kidney disease: a meta-analysis. Herz 43: 518–528, 2018 resorption in mice by a TRAF3-dependent mechanism. J Clin Invest 29. Lundwall K, Jacobson SH, Jörneskog G, Spaak J: Treating endothelial 119: 3024–3034, 2009 dysfunction with vitamin D in chronic kidney disease: a meta-analysis. 51. Zarnegar B, Yamazaki S, He JQ, Cheng G: Control of canonical NF- BMC Nephrol 19: 247, 2018 kappaB activation through the NIK-IKK complex pathway. Proc Natl 30. Beveridge LA, Khan F, Struthers AD, Armitage J, Barchetta I, Acad Sci U S A 105: 3503–3508, 2008 Bressendorff I, et al.: Effect of vitamin D supplementation on markers of 52. Liao G, Zhang M, Harhaj EW, Sun S-C: Regulation of the NF-kappaB- vascular function: a systematic review and individual participant meta- inducing kinase by tumor necrosis factor receptor-associated factor 3- analysis. J Am Heart Assoc 7: e008273, 2018 induced degradation. J Biol Chem 279: 26243–26250, 2004 31. Barbarawi M, Kheiri B, Zayed Y, Barbarawi O, Dhillon H, Swaid B, et al.: 53. Hu J, Zhu XH, Zhang XJ, Wang PX, Zhang R, Zhang P, et al.: Targeting Vitamin D supplementation and cardiovascular disease risks in more TRAF3 signaling protects against hepatic ischemia/reperfusions injury. than 83 000 individuals in 21 randomized clinical trials: a meta-analysis. JHepatol64: 146–159, 2016 JAMA Cardiol 4: 765–776, 2019 54. Häcker H, Tseng P-H, Karin M: Expanding TRAF function: TRAF3 as a tri- 32. Fernandez-Fernandez B, Ortiz A: Paricalcitol and albuminuria: tread faced immune regulator. Nat Rev Immunol 11: 457–468, 2011 carefully. Lancet Diabetes Endocrinol 6: 3–5, 2018 55. Yang X-D, Sun S-C: Targeting signaling factors for degradation, an 33. Ortiz A, Sanchez Niño MD, Rojas J, Egido J: Paricalcitol for reduction of emerging mechanism for TRAF functions. Immunol Rev 266: 56–71, 2015 albuminuria in diabetes. Lancet 377: 635–636, author reply 636–637, 56. Hu H, Brittain GC, Chang JH, Puebla-Osorio N, Jin J, Zal A, et al.: 2011 OTUD7B controls non-canonical NF-kB activation through deubiqui- 34. Sanz AB, Sanchez-Niño MD, Ramos AM, Moreno JA, Santamaria B, tination of TRAF3. Nature 494: 371–374, 2013 Ruiz-Ortega M, et al.: NF-kappaB in renal inflammation. JAmSoc 57. Vallabhapurapu S, Matsuzawa A, Zhang W, Tseng PH, Keats JJ, Wang Nephrol 21: 1254–1262, 2010 H, et al.: Nonredundant and complementary functions of TRAF2 and 35. Rangan G, Wang Y, Harris D: NF-kappaB signalling in chronic kidney TRAF3 in a ubiquitination cascade that activates NIK-dependent al- disease. Front Biosci (Landmark Ed) 14: 3496–3522, 2009 ternative NF-kappaB signaling. Nat Immunol 9: 1364–1370, 2008 36. Baud L, Fouqueray B, Bellocq A, Haymann J-P, Peltier J: [Inflammation, 58. Hostager BS, Haxhinasto SA, Rowland SL, Bishop GA: Tumor necrosis prelude to renal sclerosis: The importance of NF-kappa B]. J Soc Biol factor receptor-associated factor 2 (TRAF2)-deficient B lymphocytes 196: 269–273, 2002 reveal novel roles for TRAF2 in CD40 signaling. J Biol Chem 278: 37. Guijarro C, Egido J: Transcription factor-kappa B (NF-kappa B) and 45382–45390, 2003 renal disease. Kidney Int 59: 415–424, 2001 59. Gardam S, Turner VM, Anderton H, Limaye S, Basten A, Koentgen F, 38. Cui R, Tieu B, Recinos A, Tilton RG, Brasier AR: RhoA mediates angiotensin et al.: Deletion of cIAP1 and cIAP2 in murine B lymphocytes constitu- II-induced phospho-Ser536 nuclear factor kappaB/RelA subunit exchange tively activates cell survival pathways and inactivates the germinal on the interleukin-6 promoter in VSMCs. Circ Res 99: 723–730, 2006 center response. Blood 117: 4041–4051, 2011 39. Mezzano S, Aros C, Droguett A, Burgos ME, Ardiles L, Flores C, et al.: 60. Mizobuchi M, Morrissey J, Finch JL, Martin DR, Liapis H, Akizawa T, NF-kappaB activation and overexpression of regulated genes in human et al.: Combination therapy with an angiotensin-converting enzyme diabetic nephropathy. Nephrol Dial Transplant 19: 2505–2512, 2004 inhibitor and a vitamin D analog suppresses the progression of renal 40. Ruiz-Ortega M, Ruperez M, Lorenzo O, Esteban V, Blanco J, Mezzano S, insufficiency in uremic rats. J Am Soc Nephrol 18: 1796–1806, 2007 et al.: Angiotensin II regulates the synthesis of proinflammatory cytokines 61. Kuhlmann A, Haas CS, Gross ML, Reulbach U, Holzinger M, Schwarz U, and chemokines in the kidney. Kidney Int Suppl 62: S12–S22, 2002 et al.: 1,25-Dihydroxyvitamin D3 decreases podocyte loss and podo- 41. Sun S-C: The noncanonical NF-kB pathway. Immunol Rev 246: cyte hypertrophy in the subtotally nephrectomized rat. Am J Physiol 125–140, 2012 Renal Physiol 286: F526–F533, 2004 42. Sanz AB, Justo P, Sanchez-Niño MD, Blanco-Colio LM, Winkles JA, 62. Hirata M, Makibayashi K, Katsumata K, Kusano K, Watanabe T, Kreztler M, et al.: The cytokine TWEAK modulates renal tubulointer- Fukushima N, et al.: 22-Oxacalcitriol prevents progressive glomerulo- stitial inflammation. JAmSocNephrol19: 695–703, 2008 sclerosis without adversely affecting calcium and phosphorus metab- 43. Poveda J, Sanchez-Niño MD, Glorieux G, Sanz AB, Egido J, Vanholder olism in subtotally nephrectomized rats. Nephrol Dial Transplant 17: R, et al.: p-cresyl sulphate has pro-inflammatory and cytotoxic actions 2132–2137, 2002 on human proximal tubular epithelial cells. Nephrol Dial Transplant 29: 63. Panichi V, Migliori M, Taccola D, Filippi C, De Nisco L, Giovannini L, 56–64, 2014 et al.: Effects of 1,25(OH)2D3 in experimental mesangial proliferative 44. Ucero AC, Benito-Martin A, Izquierdo MC, Sanchez-Niño MD, Sanz AB, nephritis in rats. Kidney Int 60: 87–95, 2001 Ramos AM, et al.: Unilateral ureteral obstruction: beyond obstruction. 64. Tan X, Li Y, Liu Y: Paricalcitol attenuates renal interstitial fibrosis in Int Urol Nephrol 46: 765–776, 2014 obstructive nephropathy. JAmSocNephrol17: 3382–3393, 2006

JASN 31: 2026–2042, 2020 TRAF3/VDRAs and Renal Inﬂammation 2041 BASIC RESEARCH www.jasn.org

65. Wiley SR, Winkles JA: TWEAK, a member of the TNF superfamily, is a 79. Vucic D: The role of ubiquitination in TWEAK-stimulated signaling. multifunctional cytokine that binds the TweakR/Fn14 receptor. Cyto- Front Immunol 4: 472, 2013 kine Growth Factor Rev 14: 241–249, 2003 80. Boutaffala L, Bertrand MJ, Remouchamps C, Seleznik G, Reisinger F, 66. Winkles JA: The TWEAK-fn14 cytokine-receptor axis: discovery, bi- Janas M, et al.: NIK promotes tissue destruction independently of the ology and therapeutic targeting. Nat Rev Drug Discov 7: 411–425, alternative NF-kB pathway through TNFR1/RIP1-induced apoptosis. 2008 Cell Death Differ 22: 2020–2033, 2015 67. Brown SAN, Ghosh A, Winkles JA: Full-length, membrane-anchored 81. He JQ, Zarnegar B, Oganesyan G, Saha SK, Yamazaki S, Doyle SE, et al.: TWEAK can function as a juxtacrine signaling molecule and activate Rescue of TRAF3-null mice by p100 NF-kappa B deficiency. JExpMed the NF-kappaB pathway. J Biol Chem 285: 17432–17441, 2010 203: 2413–2418, 2006 68. Ruiz-Ortega M, Ortiz A, Ramos AM: Tumor necrosis factor-like weak 82. Gardam S, Sierro F, Basten A, Mackay F, Brink R: TRAF2 and TRAF3 signal inducer of apoptosis (TWEAK) and kidney disease. Curr Opin Nephrol adapters act cooperatively to control the maturation and survival signals Hypertens 23: 93–100, 2014 delivered to B cells by the BAFF receptor. Immunity 28: 391–401, 2008 69. Sanz AB, Izquierdo MC, Sanchez-Niño MD, Ucero AC, Egido J, Ruiz- 83. Valdivielso JM: The physiology of vitamin D receptor activation. In: Ortega M, et al.: TWEAK and the progression of renal disease: clinical Peritoneal Dialysis - from Basic Concepts to Clinical Excellence,edited translation. Nephrol Dial Transplant 29[Suppl 1]: i54–i62, 2014 by Ronco C, Crepaldi C, Cruz DN, , Basel, Switzerland, Karger Medical 70. Valiño-Rivas L, Vaquero JJ, Sucunza D, Gutierrez S, Sanz AB, Fresno M, and Scientific Publishers, 2009, pp 206–212 et al.: NIK as a druggable mediator of tissue injury. Trends Mol Med 25: 84. Li YC, Kong J, Wei M, Chen ZF, Liu SQ, Cao LP: 1,25-Dihydroxyvitamin 341–360, 2019 D(3) is a negative endocrine regulator of the renin-angiotensin system. 71. Zhao Y, Banerjee S, LeJeune WS, Choudhary S, Tilton RG: NF-kB-in- JClinInvest110: 229–238, 2002 ducing kinase increases renal tubule epithelial inflammation associated 85. Kato S, Takeyama K, Kitanaka S, Murayama A, Sekine K, Yoshizawa T: In with diabetes. Exp Diabetes Res 2011: 192564, 2011 vivo function of VDR in gene expression-VDR knock-out mice. JSteroid 72. Barnes PJ, Karin M: Nuclear factor-kappaB: a pivotal transcription factor Biochem Mol Biol 69: 247–251, 1999 in chronic inflammatory diseases. NEnglJMed336: 1066–1071, 1997 86. Xiang W, Kong J, Chen S, Cao LP, Qiao G, Zheng W, et al.: Cardiac 73. Tomita N, Morishita R, Lan HY, Yamamoto K, Hashizume M, Notake M, hypertrophy in vitamin D receptor knockout mice: role of the systemic et al.: In vivo administration of a nuclear transcription factor-kappaB and cardiac renin-angiotensin systems. Am J Physiol Endocrinol Metab decoy suppresses experimental crescentic glomerulonephritis. JAm 288: E125–E132, 2005 Soc Nephrol 11: 1244–1252, 2000 87. de Boland AR, Nemere I: Rapid actions of vitamin D compounds. J Cell 74. Loverre A, Ditonno P, Crovace A, Gesualdo L, Ranieri E, Pontrelli P, Biochem 49: 32–36, 1992 et al.: Ischemia-reperfusion induces glomerular and tubular activation 88. Marcinkowska E, Wiedłocha A, Radzikowski C: 1,25-Dihydroxyvitamin of proinflammatory and antiapoptotic pathways: differential modula- D3 induced activation and subsequent nuclear translocation of MAPK is tion by rapamycin. J Am Soc Nephrol 15: 2675–2686, 2004 upstream regulated by PKC in HL-60 cells. Biochem Biophys Res 75. Feng B, Chen G, Zheng X, Sun H, Zhang X, Zhang ZX, et al.: Small in- Commun 241: 419–426, 1997 terfering RNA targeting RelB protects against renal ischemia- 89. Nemere I, Safford SE, Rohe B, DeSouza MM, Farach-Carson MC: reperfusion injury. Transplantation 87: 1283–1289, 2009 Identification and characterization of 1,25D3-membrane-associated 76. Benedetti G, Fokkelman M, Yan K, Fredriksson L, Herpers B, Meerman rapid response, steroid (1,25D3-MARRS) binding protein. J Steroid J, et al.: The nuclear factor kB family member RelB facilitates apoptosis Biochem Mol Biol 89–90: 281–285, 2004 of renal epithelial cells caused by cisplatin/tumor necrosis factor a 90. Jantsch J, Schatz V, Friedrich D, Schröder A, Kopp C, Siegert I, et al.: Cuta- synergy by suppressing an epithelial to mesenchymal transition-like neous Na1 storage strengthens the antimicrobial barrier function of the skin phenotypic switch. Mol Pharmacol 84: 128–138, 2013 and boosts macrophage-driven host defense. Cell Metab 21: 493– 501, 2015 77. Dejardin E, Droin NM, Delhase M, Haas E, Cao Y, Makris C, et al.: The 91. Zhang W-C, Zheng XJ, Du LJ, Sun JY, Shen ZX, Shi C, et al.: High salt lymphotoxin-beta receptor induces different patterns of gene expres- primes a specific activation state of macrophages, M(Na). Cell Res 25: sion via two NF-kappaB pathways. Immunity 17: 525–535, 2002 893–910, 2015 78. Cuarental L, Sucunza-Sáenz D, Valiño-Rivas L, Fernandez-Fernandez B, 92. Sakamuri SSVP, Valente AJ, Siddesha JM, Delafontaine P, Siebenlist U, Sanz AB, Ortiz A, et al.: MAP3K kinases and kidney injury. Nefrologia Gardner JD, et al.: TRAF3IP2 mediates aldosterone/salt-induced car- 39: 568–580, 2019 diac hypertrophy and fibrosis. Mol Cell Endocrinol 429: 84–92, 2016

AFFILIATIONS

1Molecular and Cellular Biology in Renal and Vascular Pathology, Fundación Instituto de Investigación Sanitaria-Fundación Jiménez Díaz,Universidad autonoma de madrid, Madrid, Spain 2Vascular and Renal Translational Research Group. Institut de Receca Biomedica de Lleida (IRBLleida), Lleida, Spain 3REDinREN (Red de Investigación Renal), Madrid, Spain 4Laboratory of Nephrology and Hypertension, Fundación Instituto de Investigación Sanitaria-Fundación Jiménez Díaz-Universidad Autónoma Madrid, 28040 Madrid, Spain. 5Renal, Vascular and Diabetes Research Laboratory, IIS-Fundación Jiménez Díaz.Universidad Autónoma. 28040 Madrid, Spain; Spanish Biomedical Research Centre in Diabetes and Associated Metabolic Disorders (CIBERDEM). 28029 Madrid, Spain

2042 JASN JASN 31: 2026–2042, 2020 Supplementary material table of contents

Supplementary Figure 1. Paricalcitol or NF-κB2 blockade decreased TWEAK-induced renal inflammation. Supplementary Figure 2. Paricalcitol inhibits TWEAK-induced upregulation of specific NF-κB2 targets in the kidney. Supplementary Figure 3. SN52 peptide blocks NF-κB2 activation and NF-κB2 nuclear translocation in a experimental TWEAK-induced renal damage. Supplementary Figure 4. Paricalcitol decreases renal inflammation in experimental Unilateral Ureteral Obstruction (UUO) in mice. Supplementary Figure 5. Paricalcitol inhibits NF-κB2, but not NF-κB1 activation in folic acid- induced renal injury. Supplementary Figure 6. Paricalcitol reduces TWEAK-induced renal inflammation in VDR KO mice. Supplementary Figure 7. Paricalcitol inhibits TWEAK-induced upregulation of specific NF-κB2 targets in MARSS gene silenced cells.

and T lymphocytes,respectively. Representativeanimalfromeach group.Magnification200X. ( sacrificed 24 hours after TWEAK administration. ( with paricalcitol750ng/Kg/dayor theNF- Supplementary Figure 1. Paricalcitol or NF-! were assessed byMann–Whitney test. *p<0.05 determined by Real Time PCR. Data expressed as mean±SEM of 5-8 animals per group. Differences between intervention and control groups CCL-2 and CCL-5protein levelswereevaluated byELISA. ( from totalrenal extracts andproinflammatorygene expressionlevels( B A D F4/80+ cells CD3+ cells % Stained area vs. Renal protein levels Total area

0.0 0.5 1.0 1.5 0.0 0.5 1.0 (n-fold)1.5 2.0 2.5

Control

Control TWEAK+VDRA TWEAK Vehicle

TWEAK TWEAK+VDRA TWEAK TWEAK TWEAK * TWEAK+VDRA TWEAK+SN52 * CD3 # CCL-2 + # # Control !B2 inhibitorSN520.7mg/day(two doses;day-1;0)starting48hoursbefore TWEAK 0.5µg,and Control F4/80 TWEAK TWEAK+VDRA CCL-5 vs control. # p<0.05 vs TWEAK. TWEAK * B2 blockade decreased TWEAK-induced renal inflammation. C57BL/6 mice were treated TWEAK TWEAK TWEAK+VDRA + A) Immunohistochemistry using anti-F4/80 and anti-CD3 identified monocyte/macrophages TWEAK+SN52 * # # # E) NF-!B2-regulated cytokines, C E Renal mRNA levels (n-fold) ccl-2, ccl-5andil-6 ) weredetermined byReal Time PCR. (

0 1 2 3 4 Renal RNA levels Control

(n-fold) TWEAK+VDRATWEAK 0.0 0.5 1.0 1.5 2.0 2.5 TWEAK Control TWEAK+SN52 *

TWEAK+VDRATWEAK # TWEAK ccl-2

Control # * and ccl-21a and

B) TWEAK+SN52 TWEAK+VDRATWEAK Staining quantification (C)RNA was obtained # TWEAK

TWEAK+SN52 * ccl-5 TWEAK +SN52

Control # # ccl-19 geneexpression levels were

TWEAK+VDRATWEAK

Control # TWEAK il-6

* TWEAK+VDRA TWEAK+SN52 TWEAK TWEAK TWEAK+SN52 * # # ccl-21a ccl-19 # # D) Kidney A B TWEAK TWEAK

Vehicle +VDRA Vehicle +SN52

CCL-21A 21 kDa CCL-21A

GAPDH 35-38 kDa GAPDH

4 5 * * 4 3

3 2

(n-fold) 2 # (n-fold) #

1 protein levels 1 CCL-21/GAPDH CCL-21A/GAPDH 0 0 Vehicle +VDRA Vehicle +SN52 TWEAK TWEAK C Control TWEAK Vehicle TWEAK+VDRA TWEAK TWEAKControl TWEAK +VDRA

TWEAK+SN52 CCL21A CCL21A

D 8 * 6

4 # Total area Total 2 % Stained area vs.

Vehicle +VDRA TWEAK

Control TWEAK Supplementary Figure 2. Paricalcitol inhibits TWEAK-induced upregulation of specific NF-!B2 targets in the kidney. Mice were treated with paricalcitol 750 ng/Kg/day or the NF-!B2 inhibitor SN52 0.7 mg/ day (two doses; day -1; day 0) starting 48 hours before TWEAK 0.5 µg, and sacrificed 24 hours after TWEAK administration.TWEAK+VDRA (A and B) CCL21A protein levels were evaluated by western blot in total kidney proteins. GAPDH was used as loading control. (C) CCL21A immunohistochemistry located CCL21A expression to tubular cells. Representative mouse from each group. (D) Quantification of CCL21A stained area vs total area. Data expressed as mean ± SEM of 5-8 mice per group. Differences between intervention and control groups were assessed by Mann– Whitney test. *p<0.05 vs control. # p<0.05 vs TWEAK Control TWEAK TWEAK+SN52 A p52

B C TWEAK TWEAK Control +SN52 Control +SN52 Cytosolic p52 52 kDa 36 kDa p-I!B" GAPDH GAPDH 4 * 3 * 3 * 2 #

(n-fold) 2 (n-fold) p52/ GAPDH

1 p-I ! B " / GAPDH 1

0 0

Supplementary Figure 3. SN52 peptide blocks NF-!B2 activation and NF-!B2 nuclear translocation in a experimental TWEAK-induced renal damage. Mice were pretreated with the NF-!B2 inhibitorControl SN52 (0.7 mg/day;TWEAK two doses; day -1; day 0) starting 48 hoursControl beforeTWEAK administration of TWEAK 0.5 µg, and were sacrificed 24 hours after TWEAK administration. (A) P52 immunohistochemistry in paraffin-embedded kidney sections. Representative animal from each groupTWEAK+SN52 (magnification 200x). (B and C) Western blot for p52 (B)TWEAK+SN52 and I!B" phosphorylation (C). Data expressed as mean ± SEM of 5-8 animals per group. Differences between intervention and control groups were assessed by Mann–Whitney test. *p<0.05 vs control. # p<0.05 vs TWEAK. A NOb Ob Ob+VDRA NOb+VDRA cells + CD3 cells + F 4.80

140 B ccl-2 120 * ccl-5 * il-6 100 100 tnf-! 80 * il-17a 60 # #

50 40 * * 30 # 20 # 10 Renal mRNA Levels (n-fold) Renal mRNA

Ob Ob Ob Ob Ob Nob Nob Nob Nob Nob

Ob+VDRA Ob+VDRA Ob+VDRA Ob+VDRA Ob+VDRA Nob+VDRA Nob+VDRA Nob+VDRA Nob+VDRA Nob+VDRA

C 15 ccl-19 ccl-21a 10 * *

5 # # Renal mRNA Renal mRNA Levels (n-fold)

Ob Ob Nob Nob

Supplementary Figure 4. Paricalcitol decreases renal inflammationOb+VDRA in experimentalOb+VDRA Unilateral Ureteral Obstruction (UUO) in mice. Mice were treated with paricalcitol 750 ng/Kg/day, starting 24 hoursNob+VDRA before UUO, andNob+VDRA studied 5 days after UUO. In paraffin-embedded kidney sections, immunohistochemistry using anti-F4/80 and anti-CD3 identified monocyte/macrophages and T lymphocytes, respectively. A. Representative animal from each group. Magnification 200X. B, C. In RNA obtained from total renal extracts, proinflammatory gene expression (ccl-2, ccl-5, il-6, tnf-! and il-17a) and specific NF!B2-regulated gene expression (ccl-19 and ccl-21) were determined by Real Time PCR. Data expressed as mean±SEM of 4-8 animals per group. Differences between intervention and control groups were assessed by Mann–Whitney test. *p<0.05 vs contralateral non-obstructed (NOb) kidney; #p<0.05 vs obstructed (Ob) kidneys. A B C REL B p52 FA FA Vehicle +VDRA Vehicle +VDRA p52 Cytosolic p-IkB" GAPDH IkB" pIKBalpha wb modelo AF+Z Vehicle Vehicle 2.5 8 * # * 2.0 6 1.5 * 4 1.0 (n-fold) (n-fold)

FA 2 0.5 p52 / GAPDH p-IKB- " / GAPDH 0.0 0

3 FA * D FA Control FA+VDRA Control FA+VDRA 2 # FA +VDRA +VDRA FA 1 FA levels (n-fold)

Vehicle +VDRA 0 F Renal CCL-2 protein Vehicle +VDRA CCL-21A

E FA FA GAPDH 1.4 Control FA+VDRA 4 * 1.2 *

3 1.0 # # (n-fold)

(n-fold) 2 0.8

1 binding activity CCL-21A/ GAPDH Rel B subunits DNA Rel B subunits DNA 0.6 0 Vehicle +VDRA

G FA 15 ccl-2 ccl-5 il-6 ccl21a ccl-19 FA ControlH FA FA+VDRA Control * FA+VDRA 300 * 10

200 #

(n-fold) *

5 * # (mg/dl) # * BUN levels 100 Renal mRNA levels Renal mRNA * # # 0 0 Vehicle +VDRA FA

FA FA FA FA FA

Control Control Control Control Control FA FA+VDRA FA+VDRA FA+VDRA FA+VDRA FA+VDRA Control FA+VDRA Supplementary Figure 5. Paricalcitol inhibits NF-!B2, but not NF-!B1 activation in folic acid-induced renal injury. Mice were treated with paricalcitol 25 #g/Kg/day starting 24 hours before folic acid (FA) 300 mg/kg or vehicle (sodium bicarbonate 0.3 mol/L) administration, and studied 24 hours after FA injection. (A) Immunohistochemistry disclosed nuclear localization of p52 and RelB that was decreased by paricalcitol. Representative animal from each group (magnification 200x). (B and C). Western blotting of p52, as evidence of NF-!B2 activation (B) and I!B" phosphorylation as evidence of NF-!B1 activation (C). (D) CCL2 protein levels evaluated by ELISA. (E). In isolated renal nuclear proteins, RelB DNA binding activity was assessed by ELISA. (F) CCL21 protein levels evaluated in total renal protein extracts by Western Blot. (G) RNA was obtained from total renal extracts, and proinflammatory gene expression levels were determined by Real Time PCR. H. Data of serum BUN levels are shown. Data expressed as mean±SEM of 5-8 animals per group. Differences between intervention and control groups were assessed by Mann–Whitney test.. *p<0.05 vs control; #p<0.05 vs folic acid kidneys.

Control TWEAK KO VDR +TWEAK KO VDR+TWEAK +VDRA CD3+ cells

TWEAK

B C Control KO VDR KO VDR+VDRA

4 ccl-2 TRAF3 65-70 kDa * ccl-21a * GAPDH 3 Red Ponceau

2.0 $ 2

* 1.5 $ $ # #

(n-fold) * * # 1.0 *

Renal mRNA levels Renal mRNA 1 (n-fold)

0.5 TRAF3/ GAPDH

0 0.0 Control TWEAK +VDRA

TWEAK+KO VDR

controlTWEAK controlTWEAK TWEAK TWEAK control TWEAK +KO VDR +KO VDR TWEAK+KO VDR TWEAK+KO VDR TWEAK+KO VDR

TWEAK+KO VDR+VDRA TWEAK+KO VDR+VDRA Supplementary Figure 6. Paricalcitol reduces TWEAK-induced renal inflammation in VDR KOTWEAK+KO mice. VDR+VDRASome animals were pretreated with paricalcitol 750 ng/kg/day, starting 48 hours before a single dose of TWEAK 0.5 µg, and were sacrificed 24 hours after TWEAK administration. (A) Immunohistochemistry using anti-CD3 identified T lymphocytes. Representative animal from each group. Magnification 200X. (B) RNA was obtained from total renal extracts, and proinflammatory gene expression levels were determined by Real Time PCR. *p<0.05 vs WT; #p<0.05 vs TWEAK; $p<0.05 vs TWEAK+KO VDR. (C) Paricalcitol restored TRAF3 levels in experimental renal damage induced by TWEAK in VDR KO mice. TRAF3 protein levels evaluated by Western blot. Data expressed as mean±SEM of 5-8 animals per group. *p<0.05 vs control; #p<0.05 vs injured kidney. Differences between intervention and control groups were assessed by Mann–Whitney test. *p<0.05 vs control; #p<0.05 vs injured-kidney. p52

GAPDH

MARRS 57 kDa

4 * # # 3 (n-fold) 2

1 P52 protein levels/ GAPDH

Supplementary Figure 7. Paricalcitol inhibits TWEAK-induced upregulation of specific NF-!B2 targets in MARRS gene silenced cells. MARRS gene silencing was achieved in cultured cells using a predesigned and validated SiRNA against MARRS. Cells were stimulated with recombinant human soluble TWEAK 100 ng/ml. In some experiments, cells were preincubated for 48 hours withSiRNA paricalcitol Cnt 12 #mol/L prior to TWEAK stimulation. NF-!B2 pathway activation was assessed by western blot of NF!B2 p52. DataSiRNA expressed MARSS as mean±SEM of 3-5 independent experiments. *p<0.05 vs control; #p<0.05 vs TWEAK-treated cells. SiRNA Cnt+TWEAK SiRNA MARSS+TWEAK SiRNA Cnt+TWEAK+PARIC SiRNA MARSS+TWEAK+PARIC