Inhibition of Interleukin-33 to Reduce Glomerular Endothelial Inflammation in Diabetic Kidney Disease

Introduction Inflammation is a significant contributor to cardiorenal morbidity and mortality in diabetic kidney disease (DKD). The pathophysiological mechanisms linking systemic, subacute inflammation and local, kidney injury-initiated immune maladaptation is partially understood. Methods Here, we explored the expression of proinflammatory cytokines in patients with DKD; investigated mouse models of type 1 and type 2 diabetes (T2D); evaluated glomerular signaling in vitro; performed post hoc analyses of systemic and urinary markers of inflammation; and initiated a phase 2b clinical study (FRONTIER-1; NCT04170543). Results Transcriptomic profiling of kidney biopsies from patients with DKD revealed significant glomerular upregulation of interleukin-33 (IL-33). Inhibition of IL-33 signaling reduced glomerular damage and albuminuria in the uninephrectomized db/db mouse model (T2D/DKD). On a cellular level, inhibiting IL-33 improved glomerular endothelial health by decreasing cellular inflammation and reducing release of proinflammatory cytokines. Therefore, FRONTIER-1 was designed to test the safety and efficacy of the IL-33–targeted monoclonal antibody tozorakimab in patients with DKD. So far, 578 patients are enrolled in FRONTIER-1. The baseline inflammation status of participants (N > 146) was assessed in blood and urine. Comparison to independent reference cohorts (N > 200) validated the distribution of urinary tumor necrosis factor receptor 1 (TNFR1) and C-C motif chemokine ligand 2 (CCL2). Treatment with dapagliflozin for 6 weeks did not alter these biomarkers significantly. Conclusion We show that blocking the IL-33 pathway may mitigate glomerular endothelial inflammation in DKD. The findings from the FRONTIER-1 study will provide valuable insights into the therapeutic potential of IL-33 inhibition in DKD.


Targeted Analysis of Inflammatory Gene Expression
Data from the Karolinska Institute/Sahlgrenska University Hospital (KI/SU) human kidney cohort were used to analyze inflammatory marker expression in patients with DKD.Briefly, RNA sequencing was performed on paired, microdissected glomerular and tubulointerstitial tissue from patients diagnosed with DKD (n=19; 15 males; median age, 61 [range: 30-85] years; CKD stages 1-4) and living kidney donors (n=20; 12 males; median age: 56 [30-70] years), as previously described. 21The P values were adjusted for multiple testing using the Benjamini-Hochberg method to identify proinflammatory cytokines that were significantly differentially expressed in each tissue compartment.
The threshold for significance was set at an adjusted P value of <0.05.

Kidney Expression of IL-33 in Patients with CKD
Human kidney samples were processed according to the European Renal cDNA Bank (ERCB) protocol as previously described. 22,23IL-33 expression in biopsies from living kidney donors was compared with IL-33 expression in biopsies from patients with various CKD etiologies.

Immunohistochemistry
Fixed 3-4 μm human kidney sections were stained on an automated immunohistochemistry robot (Flatbed Autostainer Plus, Dako; Santa Clara, CA, USA).The EnVision FLEX High pH (Link) detection kit (Dako K8000) was used.Paraffin sections underwent heat-induced epitope retrieval in Tris/EDTA buffer, pH 9 (K8004, Dako) in a microwave oven.Endogenous peroxidase activity was blocked with EnVision FLEX Peroxidase-Blocking Reagent for 5 min at room temperature.For immunostaining, a goat anti-human IL-33 antibody (AF3625, R&D Systems; Minneapolis, MN, USA) was incubated for 30 (Dako) for a further 30 min and transferred to DAB solution for 4 min.Sections were stained with Mayer's hematoxylin, dehydrated with ethanol, and mounted under cover slips.Analysis was done by using the Aperio ePathology digital image analysis software.

mRNA-based Assessment of IL-33 Pathway Signature
Human renal biopsies from the ERCB cohort were microdissected for glomeruli and tubulointerstitial compartment separation and profiled by microarray analysis as described previously. 22,24Gene Set Variation Analysis (GSVA) was performed in this dataset and a signature score (scaled as -1 ≤× ≤1) was calculated based on expression data from a predefined gene set (IL-33, interleukin 1 receptorlike 1 [ST2], interleukin 1 receptor accessory protein [IL1RAP], TNF receptor-associated factor 6 [TRAF6], myeloid differentiation primary response 88 [MYD88], and interleukin-1 receptor-associated kinase 4 [IRAK-4]). 25use Models and Analysis db/db Uninephrectomy 5-week-old male db/db mice (BKS.Cg-Dock7m +/+ Leprdb/J homozygotes; Charles River, Italy) were housed in the Medimmune Biological Sciences Unit (Cambridge, UK) on a 12 h light/dark cycle, with ad libitum access to water and Purina 5008 chow.All animal protocols and procedures complied with the guidelines and regulations of a UK Home Office Project Licence, which had been reviewed and approved by the local Animal Welfare and Ethical Review Body (AWERB).
Mice were subjected to right nephrectomy under anesthesia, with post-operative analgesia, at 7 weeks to hasten the development of DKD.Mice were randomized by urine albumin-to-creatinine ratio (UACR) at 10 weeks of age.Anti-ST2 (ZY0NP0_E06) mIgG1 (n=13) or an isotype control antibody (n=14) was administered intraperitoneally from 11-16 weeks of age.All treatments were administered at a dose of 10 mg/kg three times per week.Urine was collected at 10, 13, and 15 weeks of age.Terminal tissues were collected at 16 weeks of age.
Body Mass, Treatment Exposure, Creatinine, Albuminuria, and HbA 1c   Body mass was monitored once every 2/3 days and other parameters were measured at 10 (baseline), 13, and/or 15 weeks of age.Blood samples were collected from the tail vein using a K2EDTA microvette.To collect urine samples, mice were housed in metabolic cages with free access to food and water for an 18-h period.To reduce evaporation, the urine collectors were placed in a pre-chilled plastic thermoblock (Tecniplast; Buguggiate VA, Italy).Fed glucose was measured by taking a drop of blood from a tail prick and applying it to a Nova StatStrip glucose meter (Data Sciences International; St. Paul, MN, USA).Antibody plasma exposure was determined by an enzyme-linked immunosorbent assay (ELISA) developed in house that detected anti-mouse ST2 (ZYONP0-F06).Urinary creatinine and albumin concentrations and percentage hemoglobin A1c (HbA1c) were measured using the Cobas C111 analyzer (Roche; Basel, Switzerland).

Histological Analysis
Kidneys were collected at study termination, fixed in 10% neutral buffered formalin for 48 h and paraffin embedded.Fixed 3-to 4-μm sections were stained with periodic acid-Schiff (PAS) stain to examine mesangial expansion and glomerular damage.30-40 glomeruli were scored per kidney from the upper half of the section throughout the full thickness of the cortex.Mice were scored 0-4, based on extent of PAS-positive mesangial expansion, scarcity of nuclei, and decrease in capillary luminal spaces.

Streptozotocin (STZ) High-fat Diet
This study was conducted by RenaSci Limited (Nottingham, UK) in accordance with the project license and Animals (Scientific Procedures) Act 1986.C57bl6/J (4-6 weeks of age, >15 g) mice were sourced from Charles River Laboratories (Wilmington, MA, USA) and RD12492 (a 60% high-fat diet) from Research Diets (New Brunswick, NJ, USA).Upon arrival, mice were introduced to the high-fat diet, which commenced 3 weeks prior to diabetes induction and persisted through the study.
Streptozotocin (STZ), a cytotoxic compound targeting pancreatic islet β-cells, was used to abolish insulin secretion and induce hyperglycemia.For a period of 5 consecutive days, STZ or a vehicle control was administered intraperitoneally at repeat low doses (50 mg/kg).The study was concluded 16 weeks after the initiation of STZ dosing, and the kidneys were collected for RNA and protein analysis of IL-33.

IL-33 Detection
Kidney lysates were prepared by homogenization of kidney tissue in 0.05% Triton x-100 in phosphate-buffered saline buffer with added protease inhibitors.Mouse IL-33 protein levels from kidney homogenates were quantified using the MILLIPLEX Mouse TH17 Magnetic Bead Panel (MTH17MAG-47K, Merck Millipore; Burlington, MA, USA), according to the manufacturer's instructions.
Cells were seeded into 6-well plates at 2×10 6 cells/well and were incubated with indicated agonists, or left unstimulated, for 24 h.Human IL-33 protein levels from cell lysates were measured using a U-PLEX Human IL-33 Assay (K151WFK, Meso Scale Discovery; Rockville, MD, USA).

Recombinant IL-33 Protein Production
Cloning, expression, and purification of wild-type (WT) human IL-33 (aa 112-270) and a variant with all four cysteine residues mutated to serine (IL-33C>S) that is resistant to oxidation were synthesized as described previously. 26

MAP Kinase Assay
Cells were seeded in 24-well plates at 5×10 4 cells/well and treated with 30 ng/mL of IL-33 with or without 1 μg/mL of IL-33-neutralizing antibody (tozorakimab) or human IgG1 isotype control, or left untreated, and incubated at 37°C for 30 min.Cells were washed and lysed with corresponding kit buffers.Protein extracts were centrifuged at 14,000 RPM at 4°C and protein concentrations were determined using the BCA Protein Assay Kit (23225, Thermo; Waltham, MA, USA).Phosphorylated mitogen-activated protein kinases (MAPK), p38, and c-Jun N-terminal kinase (JNK) were detected using a Meso Scale diagnostic assay (Phospho-p38 kit, K150CYD, Meso Scale Discovery; Phospho-JNK kit, K150CUD, Meso Scale Discovery, respectively), according to the manufacturer's instructions.

Cytokine Detection
Cells were seeded into 24-well plates at 1×10 5 cells/well.In some assays, cells were stimulated with a full dose range of IL-33 and incubated at 37°C for 24 h.In other assays, cells were treated with 30 ng/mL of IL-33 with or without 1 μg/mL of tozorakimab or human IgG1 isotype control, or left untreated, and incubated at 37°C for 24 h.Supernatants were collected for a detection of proinflammatory cytokines using a Meso Scale diagnostic assay (V-PLEX Proinflammatory Panel kit, K15049D or K15053D, Meso Scale Discovery; R-PLEX Human TNF-RI Assay, K1510VR, Meso Scale Discovery; and V-PLEX Human MCP-1 kit, K151NND, Meso Scale Discovery), according to the manufacturer's instructions.

Proliferation Assays
Cells were seeded in 96-well plates at 5×10 4 cells/well and incubated at 37°C for 16-18 h.Cells were then serum and cell growth supplement starved and incubated at 37°C for a further 24 hours.Cells were treated with increasing concentrations of IL-33 or platelet-derived growth factor-BB (100-14B-2UG, PeproTech) or epidermal growth factor (236-EG-200, R&D Systems) used as positive controls and incubated at 37°C for 18 h.Cells were pulsed with a 10 µM 5-ethynyl-2´-deoxyuridine (EdU) solution for a further 4 h.EdU incorporation was assessed using Amplex UltraRed reagent and the Click-iT EdU Proliferation Assay (C10499, Invitrogen) and fluorescence emission measurement, following the manufacturer's instructions.dapagliflozin on Day 85. Participants and all trial personnel (except the members of the independent data review monitoring committee) will be blinded to treatment assignments throughout the trial.
After randomization, in-person trial visits were performed until Day 169.At each follow-up visit, vital signs were recorded, blood and urine samples were sent for laboratory assessment, and information on potential trial outcomes, adverse events (AEs), concomitant therapies, and adherence to the trial regimen were collected.Before trial completion, each participant underwent a final trial visit.

Outcomes
The primary outcome measure is the change from baseline to Day 169 (Week 24) in UACR compared with placebo to assess the effect of tozorakimab on albuminuria.Secondary outcomes are measures of safety and tolerability such as AEs, vitals, and electrocardiograms, as well as anti-drug antibody incidence throughout the study, and other efficacy measures such as tozorakimab serum concentrations and the proportion of participants with >30%, >40%, and >50% reduction in UACR.
Exploratory objectives aim to assess IL-33 activity by evaluating change and percentage change from baseline for urine and/or serum/plasma biomarker of inflammation and fibrosis relevant to DKD progression.
The primary outcome will be assessed with a mixed model for repeated measures (MMRM) analysis of change from baseline to week 24 in UACR without multiplicity adjustment.The model is adjusted for fixed categorical effects of treatment, visit, and treatment-by-visit interaction, randomization stratification factors, and the continuous covariates of baseline log UACR and baseline log UACR-byvisit interaction.UACR will be log-transformed before entering the data in the MMRM analysis to alleviate the skewness of the data.Randomization stratification factors are with SGLT2i background or not, in Japan or the rest of the world.

Linear Mixed-effect Model with Trend
The effectiveness of the therapy (Tr) with respect to the control treatment performance at inhibiting albuminuria progression will be assessed with a linear mixed-effect model.S1,S2 The Yij, representing ith urine ACR (mg albumin/g creatinine) observed at jth assessment point (T), follows the linear growth model: Yij=a0i+a1i*Tj+eij, where a0i and a1i denote individual intercept and slope parameters, respectively, and eij∼N(0,σ) represents model error.Both intercept and slope are assumed to express random effects: a0i=g00+g01*Tri+u0i, a1i=g10+g11*Tri+u1i, with u0i ∼ N(0, σ0) and u1i ∼ N(0, σ1).The parameters g00, g10, and g01, g11 represent the parameter's fixed effects; σ, and σ0, σ1 correspond to intra-and inter UACR variance, respectively.Treatment was defined as 'effective' if the albuminuria progression inhibition of the treated group compared with the control group was significantly different (P<0.05).Human samples from the ERCB were profiled by microarray analysis.Cohort consists of multiple CKD etiologies collected from both TI and the Glom in the kidney cortex.

Figure S1 .
Figure S1.IL-33 mRNA levels in an independent cohort of patients with advanced DKD and correlation with eGFR.

Figure
Figure S2.IL-33 protein expression in biopsies from healthy donors and patients with DKD.

Figure S8 .
Figure S8.Cell type-specific expression of IL-33 in the kidney by single-nucleus RNA sequencing.

Figure S9 .
Figure S9.Intracellular IL-33 protein expression in response to inflammatory stimuli in (A) mesangial cells and (B) proximal tubule epithelial cells.

Figure
Figure S11.IL-33 induces a modest or no inflammatory response through ST2 signaling in (A-C) mesangial cells and in (D-G) RPTEC.

Figure S14 .
Figure S14.Assessment of urine inflammatory biomarkers in patients with DKD.

Table S1 .
Reference cohorts for exploratory analysis.Two subgroups of participants from the Sun-MACRO study were identified.These were either fast progressors (n=30), defined as having an eGFR decline of >5 mL/min in 12 months, or participants with a stable eGFR (n=30), defined as having an eGFR change of <1 mL/min/year.IMPROVE.Overall, 34 participants with T2D from the IMPROVE study were included in these analyses.Participants had HbA1c levels between 55 and 100 mmoL/moL, and a first morning void UACR of ≥100 mg/g and <3500 mg/g.CKD obs.An observational study to investigate the immunopathophysiology of diabetic kidney disease.CITRINE.The CITRINE study enrolled 60 participants.Participants were adults with T2D, serum urate concentration ≥6.0 mg/dL, eGFR ≥30 mL/min/1.73m 2 , and a UACR of 30-3500 mg/g.

Table S2 .
Number of human samples used for IL-33 gene signature analyses.
Blinding11a If done, who was blinded after assignment to interventions (for example, participants, care providers, those assessing outcomes) and how Supplement 11b If relevant, description of the similarity of interventions Not applicable Statistical methods 12a Statistical methods used to compare groups for primary and secondary outcomes Supplement 12b Methods for additional analyses, such as subgroup analyses and adjusted analyses Supplement For each primary and secondary outcome, results for each group, and the estimated effect size and its precision (such as 95% confidence interval)