Part 2:Clinical And Genetic Spectra Of Autosomal Dominant Tubulointerstitial Kidney Disease Due To Mutations in UMOD And MUC1

Contact: Audrey Hu Whatsapp/hp: 0086 13880143964 Email: audrey.hu@wecistanche.com

Pls click here to Part 1

UMOD-score and urine UMOD levels to guide genetic testing in ADTKD

The score was validated in patients that were UMOD-positive (n ¼ 124) and UMOD-negative (n ¼ 183) from the US ADTKD registry, yielding similarly high sensitivity and low specificity for UMOD mutations using a cut-off of $5 (sensitivity: 97.6%, specificity: 16.4%, NPV: 91.0%, PPV: 44.2%, data not shown), altogether making ADTKD-UMOD very unlikely for score results <5. We tested how the clinical score separated the 2 most common etiologies of ADTKD in a subset of patients with ADTKD-UMOD (n ¼ 125) and ADTKD-MUC1 (n ¼ 80) from the US registry for which at least 5 of the 8 clinical items and/or urinary UMOD levels were available. The clinical UMOD-score alone separated the 2 entities with an AUC of 0.69 (95% CI: 0.62–0.77, P ¼ 0.037) (Figure 7a, left panel). However, the specificity for UMOD increased considerably with higher UMOD-score values (for instance, a score of $8 had a sensitivity of 48.8%, a specificity of 83.7%, and NPV of 50.8% and a PPV of 81.3% for UMOD mutation) (Supplementary Table S5). Only a few patients, mostly those with ADTKD-MUC1, had score results of <5 (Figure 7a, right panel).

We next investigated whether the addition of urinary UMOD levels to the clinical score improved its ability to discriminate ADTKD-UMOD from ADTKD-MUC1. Based on the normalized urinary UMOD values in the reference population [(mg/g creatinine)/eGFR] (Figure 5a, right panel), we assigned, respectively, þ1 and þ3 points for urinary UMOD values between the median and 25th percentile [(0.14–0.23 mg/g creatinine)/eGFR] and below the 25th percentile. Similarly, we assigned, respectively, 1 and 3 points for urinary UMOD values between the median and 75th percentile [(0.23–0.35 mg/g creatinine)/eGFR] and above the 75th percentile. Applied to a cohort of 51 patients with ADTKD-UMOD and 35 patients with ADTKD-MUC1 for which urinary UMOD data were available, this combined clinical and biochemical score separated ADTKD-UMOD from ADTKD-MUC1 with an improved AUC of 0.89.

Cistanche can help with kidney function

Figure 5| UMOD processing in autosomal dominant tubulointerstitial kidney disease (ADTKD)–UMOD and ADTKD-MUC1. (a) Urinary UMOD excretion normalized to urinary creatinine (creat) (left panel) and normalized to urinary creatinine and estimated glomerular filtration rate (eGFR) (right panel) in patients with ADTKD-MUC1and ADTKD-UMOD and a reference population (controls). Median, 25th percentile, and 75th percentile values in the reference population are indicated (right panel). Numerical values (medians and interquartile ranges) for urinary UMOD (uUMOD), eGFR, and sample size are below the graph. Outlier removed with GraphPad (ROUT Q ¼ 1%), 1-way analysis of variance P < 0.0001 for both graphs, and Tukey’s multiple comparison test was applied with no significant (NS) and ***P < 0.0001. (b) Immunofluorescence staining for UMOD (green) and glucose-regulated protein 78 (GRP78; red) in ADTKD-MUC1, ADTKD-UMOD, and normal human kidney (NHK) biopsy. Bars ¼ 50 mm. DAPI, 40,6-diamidino-2-phenylindole. To optimize viewing of this image, please see the online version of this article at www.kidney-international.org. CI: 0.82–0.96, P < 0.001). The cut-off value of $5 still appears as the optimal cut-off value to discriminate ADTKD- UMOD from ADTKD-MUC1 (Youden index: 0.684) with a sensitivity of 94.1%, a specificity of 74.3%, an NPV of 89.7%, and a PPV of 84.2% for a UMOD mutation (Figure 7b; Supplementary Table S5). Based on the clinical and biochemical UMOD-score, we suggest a diagnostic algorithm to guide genetic testing in ADTKD (Figure 8).1,23,24

DISCUSSION

This international cohort study represents the largest dataset of patients with ADTKD-UMOD and ADTKD- MUC1 reported to date, providing new insights into the phenotype and disease progression of the main subtypes of ADTKD. Because of the autosomal dominant inheritance and regional familial clustering, considerable differences in the prevalence of ADTKD subgroups are mentioned in national cohorts.2,15,20 In the International ADTKD Cohort, ADTKD-UMOD represents the most

frequent subtype of ADTKD with an estimated prevalence of 37.1%, followed by ADTKD-MUC1 in 35.1% of families that are UMOD-negative, and an estimated overall prevalence of 21.0%. Of note, a systematic effort to screen for mutations in HNF1B, REN, DNAJB11, and SEC61A1 is ongoing in the 133 families that are UMOD- negative and MUC1-negative and for mutations in MUC1 in the 141 families that are UMOD-negative in the registry.

Based on the large sample size, we observed distinct features in the clinical presentation of ADTKD-UMOD and ADTKD-MUC1, with relevance for clinical practice and patient counseling. Kidney disease appears more severe in patients with ADTKD-MUC1, with a higher prevalence of ESKD (58% vs. 44% in ADTKD-UMOD, P ¼ 0.04), an earlier onset of ESKD (36 years vs. 46 years in ADTKD-UMOD, P < 0.001), and a shorter median renal survival (46 years vs. 54 years in ADTKD-UMOD, P ¼ 0.013). Previous studies re- ported an older age at ESKD (mean: 44.9 years) in patients with ADTKD-MUC1,8 which could be explained by the inclusion

Figure 6| Clinical UMOD-score and performance in the Belgo-Swiss ADTKD registry. (a) Clinical UMOD-score based on clinical, biochemical, histological, and imaging data. Attributed points for specific characteristics are shown on the right. after routine work-up including urinary sediment and urinalysis and kidney imaging. interstitial fibrosis, tubular atrophy, thickening and lamellation of tubular basement membranes, tubular dilatation (microcysts), negative immunofluorescence for the compliment, and Igs. proteinuria >300 mg/dl, persistent hematuria (both eumorphic and dysmorphic) in repeated urinalysis. hemoglobin A1c >10% or repeated blood pressure measurements > 160/100 mm Hg and/or corresponding clinical findings of hypertensive cardiopathy and/or nephropathy. eAt least 1 cyst at any location is diagnosed by ultrasonography, computed tomography scan, or magnetic resonance imaging. Example: 35-year-old patient, gout onset at 32 years (þ1); serum uric acid 550 mmol/l (þ3); estimated glomerular filtration rate 55 ml/min per 1.73 m2, bland urine analysis and sediment, kidneys without cysts and normal size on magnetic resonance imaging, no diabetes or hypertension (þ2 for chronic kidney disease [CKD] of unknown origin); and family history of CKD documented on 3 generations (þ2) yields a total clinical UMOD-score of 8 points. (b) Receiver-operating characteristics curve of the clinical UMOD-score in the Belgo-Swiss registry (n ¼ 211 patients with autosomal dominant tubulointerstitial kidney disease [ADTKD] with available data) are as follows: area under the curve (AUC): 0.72; 95% confidence interval (CI): 0.66–0.79; P < 0.001; the cut-off value of $5 has a sensitivity of 98.1% and specificity of 41.4% for UMOD mutation; negative predictive value: 94.3%; positive predictive value: 59.1%. (c) Histogram of clinical UMOD-score results in patients who are UMOD-positive (n ¼ 106) and UMOD- negative (n ¼ 105). The red horizontal line indicates the cut-off value of 5.of historically affected patients (clinically affected relatives of genetically diagnosed patients), whereas we only included individuals with an established genetic diagnosis. The heterogeneity of ADTKD-MUC1 in terms of CKD and/or renal disease progression is intriguing and suggests considerable modifier effects.

Gout has been classically described in patients with UMOD mutations. Indeed, our data suggest that gout is strikingly more prevalent and of significantly earlier onset in ADTKD-UMOD than in ADTKD-MUC1. Defective urinary concentration resulting in polydipsia and polyuria has been described in patients with ADTKD-UMOD, most likely because of impaired activity of TAL-based Naþ-Kþ-2Cl - cotransporter.15,17 Plasma volume contraction and compensatory higher reabsorption activity of the proximal tubule including upregulation of Naþ-coupled urate transporters most likely explain the hyperuricemia phenotype in ADTKD-UMOD.25,26 A similar mechanism was shown in

aged Umod knockout mice that displayed the reduced activity of the Naþ-Kþ-2Cl -cotransporter.26 Even though ADTKD- MUC1 presumably originates from the distal tubule, gout was considerably less prevalent in this disorder.

Figure 7| UMOD-score comparing autosomal dominant tubulointerstitial kidney disease (ADTKD)–UMOD versus ADTKD-MUC1 in the US ADTKD registry. (a, left panel) The receiver-operating characteristics curve of the clinical UMOD-score in the US registry (n ¼ 205 patients with ADTKD-UMOD and ADTKD-MUC1 with available data) are as follows: area under the curve (AUC): 0.69; 95% confidence interval (CI): 0.62–0.77; P < 0.037. A cut-off value of $8 has a sensitivity of 48.8% and specificity of 83.7% for UMOD mutations, while a cut-off value of $5 has a sensitivity of 97.6% and specificity of 15.0% for UMOD mutations. (a, right panel) Histogram of clinical UMOD-score results in patients with ADTKD-UMOD (n ¼ 125) and ADTKD-MUC1 (n ¼ 80). (b, left panel) Receiver-operating characteristics curve of the clinical UMOD-score including urine UMOD levels in the US registry (n ¼ 86 patients with ADTKD-UMOD and ADTKD-MUC1 with available urinary UMOD data) are as follows: AUC: 0.89; 95% CI: 0.82–0.96; P < 0.001. The cut-off value of $5 has the highest Youden index for discrimination (0.684) and has a sensitivity of 94.1% and specificity of 74.3% for UMOD mutation; negative predictive value: 89.7%; positive predictive value: 84.2%. (b, right panel) Histogram of clinical þ urinary UMOD-score results in patients with ADTKD-UMOD (n ¼ 51) and ADTKD-MUC1 (n ¼ 35). The red horizontal line indicates the cut-off value of 5. Q, quartile.

We investigated 2 cardinal biological features described in ADTKD-UMOD with likely pathophysiological relevance: aberration in UMOD export mechanisms and induction of ER stress. Based on the observation that MUC1 is expressed in the distal kidney tubule including the TAL where it colocalizes with UMOD6 and on the observation that MUC1fs is accumulating in other MUC1-expressing tissues (skin, breast, lung, colon) without causing extrarenal manifestations,7 one could hypothesize that MUC1fs might interact with UMOD in TAL. Yet, in contrast to ADTKD- UMOD, we found no difference in the urinary level of UMOD between patients with ADTKD-MUC1 and the normal population. Furthermore, analysis of MUC1-mutant kidney biopsies revealed a normal distribution of UMOD in TAL cells, without evidence for ER stress (GRP78 expression), which is a hallmark of ADTKD-UMOD. These novel findings suggest that the processing of UMOD is not altered in ADTKD-MUC1 and that ER stress is not the main finding in ADTKD-MUC1. Along the same line, a recent study found entrapment of MUC1fs in vesicles of the early secretory pathway in models of ADTKD-MUC1.19

Previous reports described the intracellular accumulation of UMOD in kidney biopsies from patients with ADTKD- UMOD.1,2 However, such staining is not available in a large number of patients, preventing us from speculating on its value in clinical decision-making. In our experience, the

Figure 8| Diagnostic algorithm for suspected autosomal dominant tubulointerstitial kidney disease (ADTKD) based on clinical

UMOD-score and urinary UMOD levels. progressive loss of renal function, bland urinary sediment, normal-to-mild albuminuria and/or proteinuria, normal-sized kidneys on ultrasound, and no consumption of drugs linked to tubulointerstitial nephritis. assessed by validated details. cFor diagnostic algorithms including other ADTKD genes, refer to Devuyst et al.1 Alternative diagnoses include nephronophthisis (autosomal recessive), autosomal dominant polycystic kidney disease (large cystic kidneys), autosomal dominant glomerulopathies(proteinuria and/or hematuria), other causes of tubulointerstitial kidney disease (autoimmune, tubulointerstitial nephritis, and uveitis syndrome) including drugs and toxins (nonsteroidal anti-inflammatory drugs, aristolochic acid, calcineurin inhibitors, lithium). CKD, chronic kidney disease.

UMOD staining is operator-dependent, requiring rigorous positive and negative controls, and it might depend on the underlying UMOD mutation. Furthermore, the availability of kidney biopsies is restricted. The assessment of urinary UMOD levels in patients at the time of diagnosis and during disease progression might offer a noninvasive diagnostic tool and biomarker in ADTKD-UMOD. Because urinary UMOD levels show a positive correlation with eGFR (for eGFR <90 ml/min per 1.73 m2) and tubular mass,23,24 they need to be normalized for residual eGFR and interpreted against matched controls. Based on data from a large control cohort, we show here that urinary UMOD (in mg/g creatinine to account for urine concentration) normalized for eGFR can be applied in the clinical setting of ADTKD.

A recent study based on exome sequencing reported mutations in UMOD accounting for w3% of all patients with a genetic finding in this cohort.13 However, considerable hurdles in the diagnostic approach of ADTKD subtypes persist. These include but are not limited to (i) limited availability of MUC1 testing due to technical challenges, (ii) lack of

validated diagnostic or genetic algorithm due to unappreciated clinical differences between ADTKD subtypes, and (iii) missing disease biomarkers due to small and scattered disease cohorts. For everyday practice and cost-effectiveness, practical tools such as scoring systems are very useful to guide genetic testing.1 The Belgo-Swiss registry was instrumental in delineating a clinical UMOD-score because it revealed key discriminatory clinical features, including a positive family history of CKD and/or gout, age at presentation, the prevalence of kidney disease and progression to ESKD, and history of gout. Of interest, renal cysts are less common in patients with ADTKD-UMOD, which is in line with previous studies.8,15,20 The delineated clinical UMOD-score showed an excellent NPV for UMOD mutations (cut-off $5) in the Belgo-Swiss (NPV: 94.3%) and US (NPV: 91.0%) registries. As ADTKD- UMOD and ADTKD-MUC1 present considerable clinical overlap, we were not surprised that the clinical UMOD-score separated modestly between these 2 entities (AUC: 0.69). Yet, higher UMOD-score values showed a solid specificity for UMOD mutations (e.g., cut-off $8: specificity of 83.7% and PPV of 81.3% for a UMOD mutation). Adding urinary

UMOD measurements, a pathophysiological biomarker for ADTKD-UMOD, considerably increased the discriminating power of the score (AUC: 0.89) with a PPV of 84.2% for a UMOD mutation (cut-off $5 points). Because the progression of kidney disease and the prevalence and onset of gout seems dependent on the underlying genetic diagnosis, a genetic diagnosis is recommended as it might impact the management of patients with ADTKD (e.g., follow-up, scheduling of renal transplantation, and gout-preventive strategies). Furthermore, targeted therapies might be in reach at least for ADTKD-MUC1.

The limits of this study include the retrospective “real-life” cohort design of consecutively recruited patients, with inherent difficulties such as limited access to full clinical information, missing DNA samples for further genetic testing, and lack of strict inclusion and exclusion criteria. We

included all genetically resolved cases of a given family, potentially introducing the risk for selection bias. However, we estimate that this represents a negligible risk as in general only 1 to 2 patients were included per family and considerable intrafamilial clinical variability exists in ADTKD.8,20 Because kidney biopsies are rarely performed in these diseases and yield nonspecific findings (e.g., interstitial fibrosis, tubular atrophy), we did not include histopathology information in the analysis. A survey of histopathology results from the Belgo-Swiss registry showed that interstitial fibrosis with tubular atrophy (in w60% of available pathology reports) and interstitial nephritis (in w40% of available pathology reports) were the preponderant histological findings in patients with ADTKD-UMOD and those who were UMOD- negative. A more detailed histological description of biopsies performed in ADTKD-UMOD and ADTKD-MUC1 warrants a dedicated analysis.

It should be pointed out that systematic screening for UMOD mutations in all 10 coding exons has only been performed in a subset of patients with ADTKD. Based on pre- vious screens and whole-exome sequencing, we estimate that very few UMOD mutations outside exons 3 and 4 might have been missed in ADTKD-UMOD.13,15 Furthermore, large deletions or insertions in UMOD are not detected by direct sequencing methods. With the availability of gene panel testing and next-generation sequencing approaches, the utility of a clinical score in directing targeted gene testing will probably decrease. However, at the current stage, MUC1 mutations are missed by next-generation sequencing and the availability of specialized testing is limited. To the best of our knowledge, clinical-grade genetic testing for MUC1 is only performed by the Broad Institute. For these reasons, we estimate that simple clinical and biochemical tools to estimate pretest probability impacts on diagnostic workup and potentially reduce the costs associated with unjustified genotyping.

In conclusion, this large international retrospective cohort study provides a detailed phenotype analysis of patients with ADTKD-UMOD and ADTKD-MUC1. The clinical hallmarks of the 2 most common ADTKD subtypes are hyperuricemia and early gout in ADTKD-UMOD and a heterogeneous, but generally more severe kidney disease in ADTKD-MUC1. The clinical UMOD-score is a sensitive and, coupled to urinary UMOD levels, potentially specific tool to select patients for genetic UMOD testing. These results should help clinicians to improve diagnostic rates, clinical management, and patient counseling in ADTKD.

METHODS

International ADTKD Cohort

The International ADTKD Cohort consists of patients from the Belgo-Swiss ADTKD registry and the US ADTKD registry. The inclusion criteria were those defined by the Kidney Disease: Improving Global Outcomes consensus2 and included the following in any combination: a family history compatible with the autosomal dominant inheritance of CKD with features of ADTKD, including progressive loss of kidney function, bland urinary sediment, absent-to-mild albuminuria and/or proteinuria, normal-sized or small-sized kidneys on ultrasound; and/or (in absence of a positive family history of CKD) a history of early-onset hyperuricemia and/or gout and/or the presence of interstitial fibrosis and/or tubular atrophy on kidney biopsy. Exclusion criteria included the following: a different genetic diagnosis (non-ADTKD), the presence of enlarged cystic kidneys, proteinuria (>1 g/24 h) and/or consistent hematuria, long-standing or uncontrolled diabetes mellitus, or arterial hypertension, and the consumption of drugs linked to tubulointerstitial nephritis. Only patients screened for mutations of UMOD and/or MUC1 were included in the cohort. Anonymized demographics and clinical and genetic information were recorded in a database. This study was approved by the institutional review board of Wake Forest School of Medicine), the Université Catholique de Louvain (UCL) Medical School, and the European Community’s Seventh Framework Programme “European Consortium for High-Throughput Research in Rare Kidney Diseases (EURenOmics).

Belgo-Swiss ADTKD registry. The Belgo-Swiss ADTKD registry includes patients referred to UCLouvain and the University of Zurich (UZH) by clinical partners mostly from Europe (Supplementary Appendix S1). In 2019, the registry included 275 patients who had been enrolled since 2003. The clinical data included a family pedigree, onset and evolution of kidney function decline, the onset of hyperuricemia and/or gout (age of gout onset was defined as the patient’s age at the first episode of gouty arthritis) and fractional excretion of uric acid, imaging and histopathology data (where available), and information on potential extrarenal manifestations (e.g., pancreatic enzymes, liver function tests). ESKD was defined as eGFR<10 ml/min or the initiation of renal replacement therapy (dialysis or kidney transplantation).

US ADTKD registry. The US ADTKD registry includes families with tubulointerstitial kidney disease referred to Wake Forest School of Medicine in 1999 (Supplementary Appendix S1). Information collected included demographics, pedigree, age of ESKD, laboratory values, and ultrasound results.

Genetic testing

Informed written consent was obtained from all patients. Genomic DNA was isolated from peripheral blood leukocytes using standard procedures and DNA was stored at 4 C.

Direct sequencing of UMOD exons was initially performed by Sanger sequencing, as previously described.27 More recently, the UMOD gene is analyzed by massively parallel sequencing using a tubulopathy gene panel designed by the work package tauopathies of the European Consortium EURenOmics.28,29 Mutational analysis was carried out in exons 3 and 4 for all enrolled patients and in all 10 coding exons for a subset of patients.

MUC1 genotyping was performed using a MUC1 VNTR sequencing approach coupled with a spectrometry-based probe extension assay as previously described.7,30 MUC1 testing was provided by the Broad Institute of MIT and Harvard 30 and the First Faculty of Medicine, Charles University.7 Nucleotide numbering reflects cDNA numbering with þ1 corresponding to the A of the ATG translation initiation codon in the reference sequence (NM_003361.3). Alamut Visual software (Interactive Biosoftware, Rouen, France; www.interactivebiosoftware.com) was used to assist in determining variant pathogenicity. Identified variants were successively checked against relevant databases, such as Clinvar (National Center for Biotechnology Information, Bethesda, MD; HTTPS:// www.ncbi.nlm.nih.gov/clinvar/), Human Gene Mutation Database (Institute of Medical Genetics in Cardiff, Cardiff, UK; http://www. hgmd.cf.ac.uk/ac/index.php), VarSome (Saphetor, Lausanne, Switzerland; https://varsome.com/), and local databases to assess for previous publication.

Variants were considered disease-causing based on previous re- ports, family segregation analysis, or prediction algorithms (Sorting Intolerant from Tolerant [SIFT], Align GVD, mutation taster, and Polymorphism Phenotyping v2 [PolyPhen-2]) for pathogenicity.

The variants were classified according to the guidelines published by the American College of Medical Genetics in 2015.31 Variants of interest were verified by Sanger sequencing.

Measurements of urinary levels of UMOD

A validated enzyme-linked immunosorbent assay method was used to measure urinary UMOD levels (second-morning urine sample) from 86 patients with ADTKD.21 Urinary creatinine was measured using a Synchron DXC800 analyzer (Beckman Coulter, Fullerton, CA). The reference samples (n ¼ 2717) were obtained from the Cohort Lausannoise, a population-based study including 6000 people 35 to 75 years of age from the city of Lausanne, Switzerland.22 eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation. Informed consent was obtained from all participating individuals.

UMOD expression constructs

cDNA of human wild-type UMOD was cloned in pcDNA 3.1(þ) (Thermo Fisher Scientific, Waltham, MA) and a hemagglutinin tag was inserted after the leader peptide in between T26 and S27 in the protein sequence.32 The C150S and L284P mutant isoforms were obtained by mutagenesis using the QuikChange Lightning mutagenesis kit (Agilent, Santa Clara, CA) following the manufacturer’s instructions. Primers were designed using the software QuikChange Primer Design Program (Agilent). Primers used for mutation C150S: forward (50 /30 ) gatggcactgtgagtcctccccgggctcctg, reverse (50 /30 ) caggagcccggggaggactcacagtgccac and for mutation L248P: forward (50 /30 ) cccgagtgtcacccggcgtactgcaca, reverse (50 /30 ) tgtgcagtacgccgggtgacactcggg.

Cell culture conditions

HEK293 cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, 200 U/ml penicillin, 200 mg/ml streptomycin, and 2 mmol/l glutamine at 37 C, 5% CO2.

HEK293 cells were transfected using Lipofectamine 2000 (Thermo Fisher Scientific) following the manufacturer’s protocol and analyzed 24 hours after transfection.

Western blot

Cells were lysed in octylglucoside lysis buffer (50 mmol/l Tris-HCl, pH 7.4, 150 mmol/l NaCl, 60 mmol/l octyl b-D-glucopyranoside, 10 mmol/l NaF, 0.5 mmol/l sodium orthovanadate, 1 mmol/l glycerophosphate and protease inhibitor cocktail [Sigma-Aldrich, St. Louis, MO]) for 1 hour at 4 C followed by 10 minutes of centrifugation at 17,000g. Soluble fractions were quantified by the Bio-Rad Protein Assay (Bio-Rad Laboratories, Hercules, CA). Western blot experiments were performed as described in Schaeffer et al.32 Antibodies used were mouse purified anti-HA.11 Epitope Tag antibody (dilution 1:1000; 901502; BioLegend, San Diego, CA) and mouse monoclonal anti-b-actin (dilution 1:20,000; A2228; Sigma-Aldrich).

Immunofluorescence

Kidney biopsies. Immunodetection of UMOD and GRP78 was performed on 5-mm-thick kidney sections obtained from nephrectomy samples of patients with ADTKD-UMOD (female, 41 years old, ESKD; male, 42 years old, ESKD) and ADTKD-MUC1 (female, 60 years old, ESKD; male, 47 years old, ESKD). Slides were deparaffinized in xylene and rehydrated in a graded ethanol series. Antigen retrieval was carried out for 10 minutes with citrate buffer (pH 6.0) at 98 C. After 20 minutes in blocking solution, slides were incubated overnight with GRP78 primary antibody (1:300; ab21685; Abcam, Cambridge, UK), followed by incubation with Alexa- Fluor555-conjugated goat anti-rabbit antibody for 45 minutes (1:200; Invitrogen, Thermo Fisher Scientific). The slides were probed with sheep anti- UMOD primary antibody (1:800; K90071C; Meridian Life Science Inc., Memphis, TN), followed by AlexaFluor488- conjugated donkey anti-sheep (1:200; Invitrogen). Coverslips were mounted with Prolong gold antifade reagent with 40,6-diamidino-2- phenylindole (Invitrogen) and analyzed under a Zeiss LSM 510 Meta Confocal microscope (Carl Zeiss, Jena, Germany) with high numerical aperture lenses (Plan-Neofluar 20/0.5). The use of these samples has been approved by the Université Catholique de Louvain Ethical Review Board.33

HEK293 cells. Cells grown on a coverslip were fixed in 4% paraformaldehyde for 15 minutes, permeabilized 10 minutes with 0.5% Triton, and blocked for 30 minutes with 10% donkey serum. Cells were labeled for 1 hour, 30 minutes at room temperature with a mouse purified anti-HA.11 Epitope Tag antibody (dilution 1:500; 901502; BioLegend) and a rabbit polyclonal anti-calreticulin (dilution 1:500; C4606; Sigma-Aldrich) followed by incubation for 1 hour with the appropriate AlexaFluor conjugated secondary antibodies (dilution 1:500; Thermo Fisher Scientific). Cells were stained with 40,6-diamidino-2-phenylindole and mounted using fluorescence mounting medium (Dako; Agilent). All pictures were taken with an UltraVIEW ERS spinning disk confocal microscope (UltraVIEW ERS-Imaging Suite Software; Zeiss 63X/1.4; PerkinElmer Life and Analytical Sciences, Boston, MA). All images were imported in Photoshop CS (Adobe Systems, Mountain View, CA) and adjusted for brightness and contrast.

Generation and validation of the ADTKD UMOD-score

The weighted UMOD-score was based on ADTKD criteria, specific clinical characteristics of ADTKD-UMOD (i.e., early gout onset and hyperuricemia), and parameters that are negatively associated with ADTKD (i.e., providing an alternative explanation for CKD: proteinuria and/or hematuria, diabetes and/or uncontrolled hypertension, renal cysts and/or enlarged kidneys).2,15,20 For weighting the items of the score, we used integer values between 1 and þ3. A score of þ2 was given for the general ADTKD criteria,2 þ1 or þ3 for the UMOD- specific clinical and laboratory findings, and 1 for each negatively associated item. The score was first tested in the Belgo-Swiss ADTKD registry and validated in the US ADTKD registry. To discriminate ADTKD-UMOD from ADTKD-MUC1, we defined a normal range of urinary UMOD [(mg/g creatinine)/eGFR] using 2717 urine samples from the general population. Based on the pathophysiology of ADTKD-UMOD, on previous reports,34 as well as on our findings (Figure 5a), we assigned, respectively, þ1 and þ3 points for urinary UMOD values between the median and 25th percentile and below the 25th percentile of normal urinary UMOD levels. Similarly, we assigned, respectively, 1 and 3 points for urinary UMOD values between the median and 75th percentile and above the 75th percentile of normal urinary UMOD levels. Conceptualization of the score was based on the previously published hepatocyte nuclear factor 1b score.35

Statistical analysis

Quantitative parameters are presented as median and IQR (25th, 75th percentiles) (for scale variables) or mean SD (for continuous variables), and qualitative parameters are presented as fractions with percentages. Categorical variables were compared using the c2 test. Continuous variables were compared using the Mann–Whitney U test or unpaired t-test. Analysis of variance testing with Tukey’s multiple comparison test was used to compare urinary UMOD levels. Kaplan- Meier curves were generated to display ESKD-free and gout-free survival. Patients who had not reached ESKD or developed gout at the end of the study (outcome of interest not occurred during follow-up time) were considered censored individuals. Censoring time was defined as age at last follow-up. A log-rank test was used for the comparison of survival curves. The performance of the UMOD-score was assessed by calculating the AUC of the receiver-operating characteristic curve. The Youden index was used to define the optimal discriminatory cut-off point for the UMOD-score. Statistical analysis was performed using SPSS Statistics (IBM Corp., Armonk, NY). P < 0.05 was considered statistically significant, 2-sided tests were used.

DISCLOSURE

All the authors declared no competing interests.

ACKNOWLEDGMENTS

EO is supported by the Fonds National de la Recherche Luxembourg (grant 6903109), the University Research Priority Program “Integrative Human Physiology, ZIHP” of the University of Zurich, and an Early Postdoc Mobility-Stipendium of the Swiss National Science Foundation (P2ZHP3_195181). LR is supported by the Italian Society of Nephrology (SIN) under the “Adotta un Progetto di Ricerca” program, Telethon-Italy (grant GGP14263), and the Italian Ministry of Health (grant RF-2010-2319394). KK and AJB are supported by the National Institutes of Health–National Institute of Diabetes and Digestive and Kidney Diseases (grant R21 DK106584). M, KH, and SK were supported by the Ministry of Health of the Czech Republic (grant NV17-29786A), the Ministry of Education of the Czech Republic (grant LTAUSA19068), and by Charles University in Prague (institutional programs UNCE/MED/007 and PROGRES-Q26/LF1). They thank the National Center for Medical Genomics (grant LM2015091) for help with UMOD and MUC1 sequencing. OD is supported by the European Community’s Seventh Framework Programme (grant 305608), the European Reference Network for Rare Kidney Diseases (project 739532), the Swiss National Science Foundation’s National Center of Competence in Research Kidney Control of Homeostasis program, and the Swiss National Science Foundation (grant 310030-189044). OD and OB were supported by the Gebert-Rüf Foundation for research on ADTKD-UMOD. JAS is supported by Kidney Research UK and the Northern Counties Kidney Research Fund. RT is funded by Instituto de Salud Carlos III: Redes Temáticas de Investigación Cooperativa Red de Investigación Renal (grant RD16/0009) and Fondo de Investigación Sanitarias Fondo Europeo de Desarollo Regional (grant PI15/01824, PI18/00362). JM is supported by the Fonds de la Recherche Scientifique–Communauté Française de Belgique. DGF was supported by the Swiss National Centre of Competence in Research TransCure, the Swiss National Centre of Competence in Research Kidney Control of Homeostasis program, and the Swiss National Science Foundation (grants 31003A_152829 and 33IC30_166785/1). Genetic testing for MUC1 is supported by the Slim Initiative for Genomic Medicine in the Americas, a collaboration of the Broad Institute with the Carlos Slim Foundation.

We thank all participating patients and families. The Cohort Lausannoise is acknowledged for providing reference urine samples and eGFR information. We are grateful to Maegan Harden and her team on the Broad Genomics Platform for expert assistance with genetic testing for MUC1.

Parts of these data were presented as a poster during the 2017 American Society of Nephrology Kidney Week (October 31– November 5, 2017, New Orleans, Louisiana).

SUPPLEMENTARY MATERIAL

Supplementary File (PDF)

Appendix S1. Referring Physicians.

Figure S1. The L284P uromodulin shows a trafficking defect.

Figure S2. Freedom from ESKD and gout in ADTKD-UMOD according to noncysteine versus cysteine mutations.

Figure S3. Association between urinary uromodulin and glomerular filtration in the general population.

Figure S4. Uromodulin processing in ADTKD-UMOD and ADTKD- MUC1.

Figure S5. Design and flowchart of the Belgo-Swiss ADTKD registry. Figure S6. Clinical characteristics of UMOD-positive and UMOD- negative patients in the Belgo-Swiss ADTKD registry.

Table S1. List of the 106 UMOD mutations in the Belgo-Swiss and US ADTKD registries.

Table S2. Allele frequency and pathogenicity scores for novel UMOD variants.

Table S3. Clinical characteristics of ADTKD patients in the US ADTKD registry according to genetic diagnosis.

Table S4. Coordinates of the receiver-operating characteristics (ROC) curve of the clinical UMOD-score in the Belgo-Swiss ADTKD registry (UMOD-positive and UMOD-negative patients) (n ¼ 211 patients with available data).

Table S5. Coordinates of the receiver-operating characteristics (ROC) curve of the UMOD-score in the US ADTKD registry (patients with ADTKD-UMOD and ADTKD-MUC1) without and with urinary uromodulin (n ¼ 205 and n ¼ 86 with urinary uromodulin).

REFERENCES

1. Devuyst O, Olinger E, Weber S, et al.Autosomal dominant tubulointerstitial kidney disease, Nat Rey Dis Primers.2019:5:60.

2. Eckardt KU, Alper SL, Antignac C, et al Autosomal dominant tubulointerstitial kidney disease: diagnosis, classification, and management—a KDIGO consensus report. Kidney Int.2015;88:676-683.

3. Hart TC, Gorry MC, Hart PS, et al. Mutations of the UMOD gene are responsible for medullary cystic kidney disease 2 and familial juvenile hyperuricaemic nephropathy. J Med Genet.2002;39:882-892.

4. Devuyst O. Olinger E, Rampoldi L Uromodulin: from physiology to rare and complex kidney disorders. Nat Rev Nephrol.2017;13:525-544.

5. Kirby A, Gnirke A, Jaffe DB, et al. Mutations causing medullary cystic kidney disease type 1 lie in a large VNTR in MUC1 missed by massively parallel sequencing. Nat Genet.2013;45:299-303.

6. Knaup KX, HackenbeckT, Popp B, et al.Biallelic expression of mucin-1 in 21. Youhanna S, Weber J, Beaujean V, et al.Determination of uromodulin in autosomal dominant tubulointerstitial kidney disease: implications for nongenetic disease recognition.J Am Soc Nephrol.2018292298-2309.