Genome-Wide Association Studies of Metabolites in Patients with CKD Identify Multiple Loci and Illuminate Tubular Transport Mechanisms

CLINICAL RESEARCH www.jasn.org

Yong Li,1 Peggy Sekula ,1 Matthias Wuttke,1 Judith Wahrheit,2 Birgit Hausknecht,3 Ulla T. Schultheiss,1 Wolfram Gronwald,4 Pascal Schlosser ,1 Sara Tucci,5 Arif B. Ekici,6 Ute Spiekerkoetter,5 Florian Kronenberg ,7 Kai-Uwe Eckardt,3 Peter J. Oefner,4 and Anna Köttgen ,1 the GCKD Investigators

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

ABSTRACT Background The kidneys have a central role in the generation, turnover, transport, and excretion of metabolites, and these functions can be altered in CKD. Genetic studies of metabolite concentrations can identify proteins performing these functions. Methods We conducted genome-wide association studies and aggregate rare variant tests of the concentrations of 139 serum metabolites and 41 urine metabolites, as well as their pairwise ratios and fractional excretions in up to 1168 patients with CKD. Results After correction for multiple testing, genome-wide significant associations were detected for 25 serum metabolites, two urine metabolites, and 259 serum and 14 urinary metabolite ratios. These included associations already known from population-based studies. Additional findings included an association for the uremic toxin putrescine and variants upstream of an enzyme catalyzing the oxidative deamination of 2 polyamines (AOC1, P-min=2.4310 12), a relatively high carrier frequency (2%) for rare deleterious mis- sense variants in ACADM that are collectively associated with serum ratios of medium-chain acylcarnitines 2 (P-burden=6.6310 16), and associations of a common variant in SLC7A9 with several ratios of lysine to 2 neutral amino acids in urine, including the lysine/glutamine ratio (P=2.2310 23). The associations of this SLC7A9 variant with ratios of lysine to specific neutral amino acids were much stronger than the association with lysine concentration alone. This finding is consistent with SLC7A9 functioning as an exchanger of urinary cationic amino acids against specific intracellular neutral amino acids at the apical membrane of proximal tubular cells. Conclusions Metabolomic indices of specific kidney functions in genetic studies may provide insight into human renal physiology.

J Am Soc Nephrol 29: 1513–1524, 2018. doi: https://doi.org/10.1681/ASN.2017101099 CLINICAL RESEARCH Metabolites are small molecules that represent in- termediates or products of metabolic processes. Received October 16, 2017. Accepted February 9, 2018. They play important roles in energy generation, Y.L. and P.S. contributed equally to this work. signaling, and the regulation of enzymatic reactions. Published online ahead of print. Publication date available at The concentrations of these small molecules in cells www.jasn.org. and body ﬂuids result from a balance of their intake Correspondence: Dr. Anna Köttgen, Institute of Genetic Epi- and generation, their transport across compart- demiology, Medical Center – University of Freiburg, Hugstetter ments, and their breakdown and excretion. The Straße 49, 79106 Freiburg, Germany. Email: anna.koettgen@ kidneys play a central role in all of these processes, uniklinik-freiburg.de providing a rationale for metabolomics research in Copyright © 2018 by the American Society of Nephrology

J Am Soc Nephrol 29: 1513–1524, 2018 ISSN : 1046-6673/2905-1513 1513 CLINICAL RESEARCH www.jasn.org nephrology.1 Previous studies have linked blood metabolite Significance Statement concentrations to common genetic markers across the genome and found a strong genetic component to the measured The kidney is a central organ for metabolite handling. Genetic concentrations of many metabolites.2–10 The implicated genes studies of metabolite concentrations in blood and urine from pa- are often involved in balancing blood metabolite concentrations tients with CKD may reveal aspects of metabolite handling. The authors carried out genome-wide association studies of 139 serum through metabolite generation (e.g., encoding the rate-limiting and 41 urine metabolites and their pairwise ratios among 1168 enzyme), turnover, or excretion (e.g., encoding metabolite patients with CKD. Of particular interest was an association between transporters). In CKD, reduced GFR leads to elevated concen- genetic variants in SLC7A9 and several urinary lysine–to–neutral trations of many metabolites in blood. We therefore hypothe- amino acid ratios. The associations match the biologic function of SLC7A9 sized that the presence of CKD represents a “challenge” model as a renal exchanger of cationic against neutral amino acids, and provide a direct human readout of its substrates in vivo. for metabolite handling. Under such a model, metabolites may The study highlights the potential of linking genomics to metab- be quantifiable that are usually below the limit of detection, and olomics to generate insights into human renal physiology. renal mechanisms that facilitate active metabolite reabsorption or excretion may be altered and could inform about tubular 2 functions. Studying metabolite concentrations in patients with had an eGFR between 30 and 60 ml/min per 1.73 m or an . 2 CKD may therefore allow for the detection of genetic loci influ- eGFR 60 ml/min per 1.73 m and a urinary albumin-to-cre- . . encing such processes. atinine ratio (UACR) 300 mg/g, albuminuria 300 mg/d, a . Only a few previous studies have investigated genetic influ- urinary protein-to-creatinine ratio 500 mg/g, or proteinuria . 15 ences on metabolite concentrations in urine,11–13 which are of 500 mg/d. Trained personnel obtained information on particular interest for the field of nephrology. Not only does clinical data, socio-demographic factors, medical and family the urine contain metabolites that are exclusively or predom- history, medications, and health-related quality of life. The inantly generated in the kidneys and secreted into urine, but leading cause of CKD was ascertained from the treating ne- urinary metabolite concentrations also allow for the modeling phrologist. Moreover, biospecimens (plasma, serum, whole of specific renal functions that may be affected in the presence blood, spot urine) were collected in a standardized way at of CKD. For example, genetic screens for the metabolite’s frac- the enrollment visit, processed, and shipped frozen to a central tional excretion (FE) could identify tubular transport proteins laboratory for routine clinical chemistry and to a central for this metabolite, or genetic variants associated with the biobank for future analyses following standard operating pro- 16 ratio of metabolites in the urine could identify substrates cedures. A complete description of study design and the 14,15 that are counter-transported across the apical tubular cell recruited study population can be found elsewhere. membrane. For the current analysis, serum and urine specimens col- We therefore set out to test the following hypotheses: first, lected at baseline were selected for metabolite measurements genetic investigations of metabolite concentrations in serum from a subset of GCKD participants: all participants re- and urine of patients with CKD can replicate findings from cruited from the Freiburg study center and additionally previous population-based studies and identify additional ge- all GCKD patients from other study centers with autoso- netic loci that may reflect the metabolic challenge posed by mal-dominant polycystic kidney disease, focal segmental CKD. Second, modeling of kidney-specific functions on the glomerulosclerosis, membranous nephropathy, membrano- basis of metabolite concentrations can be used to gain insights proliferative GN, rapid progressive GN, pauci-immune, and – into tubular transport mechanisms and metabolic reactions of anti glomerular basement membrane GN as the leading importance in CKD or detectable in its presence. To test these cause of CKD (Supplemental Material). hypotheses in a proof-of-principle study, we carried out genome-wide association studies (GWAS) of metabolite concen- Genotyping and Imputation trations in serum and urine as well as their FEs and pairwise Genomic DNA was extracted from whole blood using an au- ratios in up to 1168 patients with CKD participating in the tomated magnetic bead–based technology, quantified and German Chronic Kidney Disease (GCKD) study.14 normalized on a pipetting robot platform, and available for 5123 GCKD participants. Genotyping was conducted for 2,612,357 markers at the Helmholtz Center Munich using METHODS Omni2.5Exome BeadChip arrays (Illumina, GenomeStudio, Genotyping Module Version 1.9.4). Data cleaning was carried Study Design and Participants out separately for the Omni2.5 content and the Exome Chip The GCKD study is a prospective cohort study of patients with content of the array. For the Omni2.5 content, data QC and CKD treated by nephrologists. It was approved by the local cleaning followed the protocol of Anderson et al.17 Per-indi- ethics committees and registered in the national registry for vidual QC steps included evaluation of call rate, sex check, clinical studies (DRKS 00003971). Between 2010 and 2012, heterozygosity, cryptic relatedness, and genetic ancestry. Alto- 5217 eligible adult patients provided written consent and gether, 89 individuals were removed during the per-individual were enrolled into the study.14 Patients were included if they QC. In the per–single nucleotide polymorphism (SNP) QC

1514 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1513–1524, 2018 www.jasn.org CLINICAL RESEARCH steps, SNPs with ,96% call rate, deviating from Hardy–Wein- Analytic performance was assessed by measuring replicates berg equilibrium (P value ,1E25), and those with duplicate of human plasma samples spiked at three different concentra- positions were removed. Detailed steps of data cleaning are tions as quality controls on each measurement plate along with available in the Supplemental Material. The final dataset for the GCKD samples. Reference samples were within the pre- the Omni2.5 genotypes contained 5034 individuals and defined tolerances of the method. Accuracy of the measure- 2,337,794 SNPs. The scripts from Anderson et al.17 which ments, determined by the accuracy of external calibrators, was are on the basis of Plink v1.90 and R programming language in the normal range of the method (deviations from target were used in the genotyping data cleaning steps. #20%). Concentrations were normalized to the median of Cleaned Omni2.5 genotype data were then imputed using spiked reference samples to correct for possible batch effects. IMPUTE2 (v2.3.1)18 following the Best Practices for Imputa- Overall, serum measurements from 1200 individuals and tion on the IMPUTE2 website (https://mathgen.stats.ox.ac. urinary measurements from 1187 individuals could be uk/impute/impute_v2.html#best_practices). SHAPEIT v219 obtained. All concentration values given are in micromolar was used for strand alignment as well as for prephasing. The (micromoles per liter). A detailed overview of the measured 1000 Genomes Project ALL haplotypes – Phase 3 integrated metabolites is provided in Supplemental Table 1. Measure- variantsetwasusedasthereference.Imputationwasper- ments were further subjected to detailed quality control and formed chunk-wise with 5-Mbp sequences in each chunk. data cleaning procedures (Supplemental Material). Initial imputation yielded genotypes for 81,698,995 variants. Filtering for imputation quality (info) value of .0.8 and mi- Additional Variables nor allele frequency (MAF) $1% led to 9,281,895 high-quality Serum creatinine was measured in the GCKD study in a central imputed markers that were subsequently used for GWAS. laboratory using an IDMS traceable enzymatic assay (Creati- Similarly, per-individual QC and per-SNP QC steps were nine plus, Roche). eGFR values were calculated using the cre- performed for the Exome Chip content, with additional checks atinine-based Chronic Kidney Disease Epidemiology Collab- specific to exome chip data as in Guo et al.20 In the per-indi- oration formula.24 The UACR was calculated from urinary vidual QC steps, 96 individuals were removed, and in the per- albumin and creatinine measurements, where creatinine was SNP QC, 3818 SNPs were removed on the basis of filtering for measured using the same assay as in serum and albumin with call rate .95% and Hardy–Weinberg equilibrium P value the ALBU-XS assay (Roche/Hitachi Diagnostics GmbH, Man- $1E25. More detailed information of the Exome Chip QC nheim, Germany). is found in the Supplemental Material. The cleaned Exome Chip dataset contained genotypes of 226,233 variants from GWAS and Exome Chip Analyses 5027 individuals. These cleaned data were postprocessed by Cleaned metabolite data were merged with genotype data zCall21 with a z-score threshold of six and then used in the yielding a dataset for 139 serum metabolites from 1168 indi- Exome Chip association analyses. viduals and 41 urinary metabolites from 1155 individuals for subsequent genetic analyses (Supplemental Material). The Metabolite Measurements concentrations of the urinary metabolites were normalized Samples were shipped frozen to BIOCRATES Life Sciences AG using the probabilistic quotient25 to account for dilution. in Innsbruck (Austria) and stored at 280°C until the time of Within each matrix, pairwise ratios were calculated: Met A/ analysis in 2014/2015 when aliquots were thawed, centrifuged, Met B. For 37 of the metabolites, both serum and urinary and the supernatant was used for analysis. Following the man- concentrations were available, and the FE for each of these ufacturer’s protocol of the commercially available Absolu- metabolites was calculated as (1003Metu3Creas)/(Mets3 teIDQ p180 Kit, serum and urine samples were handled in a Creau) where Metu and Creau are the urinary concentrations uniform and standard way that includes a precipitation/ex- of the metabolite and of creatinine, and Mets and Creas their traction step. Using that kit, .180 metabolites of different serum concentrations. FEs and ratios were subjected to analyte metabolite classes (amino acids, biogenic amines, acylcarni- level QC on the basis of evaluations of completeness, variabil- tines, [lyso-]phosphatidylcholines, sphingomyelins, and hex- ity, and outliers, resulting in 34 FEs, 9591 serum metabolite oses) were quantified from samples (input: 10 ml). ratios, and 820 urinary metabolite ratios for analyses. Before The quantitative assay was on the basis of PITC (phenyl- GWAS, all values (single metabolites, FEs, ratios) were trans- isothiocyanate)-derivatization in the presence of stable iso- formed using rank-based inverse normal transformation. tope-labeled internal standards, followed by flow injection An in-house GWAS analysis pipeline was set up using analysis of acylcarnitines, lipids, and hexose and liquid chro- SNPTEST (v2.5.2; https://mathgen.stats.ox.ac.uk/genetics_soft- matography of amino acids and biogenic amines, with detec- ware/snptest/snptest.html) for association testing and GWA- tionaccomplishedby multiplereaction monitoring on a SCIEX toolbox (V2.2.4)26 for QC of the GWAS results. This pipeline 4000 QTrap mass spectrometer (SCIEX, Darmstadt, Ger- was used for the GWAS of single serum and urinary metabolite many). The experimental metabolomics measurement tech- concentrations and the FEs, using filtered imputed data of ge- nique is described in detail in patents EP 1 897 014 B1 and EP 1 notype dosages and assuming an additive genetic model. As in 875 401 B1.22,23 previous GWAS of metabolite concentrations,4,7 age and sex

J Am Soc Nephrol 29: 1513–1524, 2018 Genetics of Metabolism in CKD 1515 CLINICAL RESEARCH www.jasn.org were included as covariates in all analyses; analyses of serum Power Considerations metabolite concentrations additionally included adjustment The power to detect associations between metabolites and ge- for log(eGFR). After correcting for multiple testing using a Bon- netic variants in a study of 1150 unrelated individuals such as ferroni procedure, statistical significance was defined as ours is provided across a range of MAFs and effect sizes in P,3.6E210 for 139 serum metabolites, P,1.2E209 for 41 uri- Supplemental Table 10. For example, the power to detect an nary metabolites, and P,1.5E209 for 34 FEs; suggestive signif- association at the genome-wide significance level of 5.0E208 icance was defined as P value ,5.0E208. Regional association for genetic markers with MAF.0.1 and an absolute effect size plots of loci with significant associations were generated by Lo- .0.5 is .80%. cusZoom (v1.3).27 The proportion of variance in inverse-normal transformed metabolite trait (y) explained by an index SNP was calculated RESULTS as R2=b23var(SNP)/var(y), where b is the estimated effect of the SNP on y, var(SNP)=23MAFSNP3(12MAFSNP) accord- Among the 1143 patients with CKD for whom both genetic ing to the binomial distribution of the alleles.28 information as well as serum and urinary metabolite measure- Because of the large number of evaluated ratios, a two-step ments were available, the mean eGFR and the median UACR analytic approach was used, with the first step consisting of were 51.6 ml/min per 1.73 m2 and 160.8 mg/g, respectively GWAS of the 9591 serum and 820 urinary metabolite ratios (Table 1). The mean age was 56 years, and 63% of the patients using the software OmicABEL29 on the basis of cleaned, gen- were male. An overview of all metabolites studied, including otyped markers. Adjustment variables are the same as in the their abbreviations, limits of detection, and coefficients of var- single variant analysis by SNPTEST. The OmicABEL results iation, is provided in Supplemental Table 1. were filtered for associations with chi square–association statistics .30, which is equivalent to a P value ,4.3E28. Serum Genetic Association Studies of Metabolites in Patients and urinary metabolite ratios with at least one SNP that passed with CKD: Model Selection this filter were then subjected in a second step to the As a first step, we tested whether known genetic associations SNPTEST-based GWAS pipeline using imputed genotype with metabolites could be replicated in a CKD population, and data, as for single metabolites and FEs. The Bonferroni-cor- evaluated the potential influence of covariate selection and, rected significance threshold was set at 5E28/9591=5.2E212 specifically for urinary metabolites, the method used to ac- for serum metabolite ratios to account for the conduct of 9591 count for dilution. We systematically selected variants with GWAS, and 5E28/820=6.1E211 for urinary metabolite ra- large genetic effects on metabolites from early publications tios. Additionally, the GWAS results of ratios were filtered on of metabolite GWAS3,12 where both variant and metabolite the “P-gain,” defined as the minimum of the two P values of were available in our data. These known genetic associations the single metabolites composing the ratio divided by the P also achieved genome-wide significance among patients with value of the ratio.30 Only associations with a P-gain.9591 in CKD, as illustrated by associations between serum nonanoyl- serum and P-gain.820 in urine were further considered. carnitine concentrations and an SNP in ACADL or serum bu- For the analysis of the exome chip data, two kinds of gene- tyrylcarnitine concentrations and an SNP in ACADS,aswellas based tests of single metabolites, FEs, and metabolite ratios urinary lysine concentrations and an SNP in SLC7A9 (Table were carried out using the R package seqMeta (v1.6.5) and 2). Adjustment of serum metabolite associations for eGFR led using the same phenotypes and adjustment variables as for to increased statistical significance, mainly due to smaller the single variant associations: a burden test and the sequence SEM. Thus, genetic screens with metabolite concentrations kernel association test. For both tests, only variants with in serum were adjusted for eGFR in subsequent analyses. In MAF,1% and likely to be functional (splicing, nonsynony- urine, the association between rs7247977 in SLC7A9 and ly- mous, stopgain, and stoploss variants on the basis of the sine concentrations was best detected when probabilistic quo- SNPInfo annotation file used in the CHARGE Consortium) tient–normalization25 was used to account for urine dilution, were included.31 The test results were filtered and only genes rather than normalization by urinary creatinine concentra- with cumulative minor allele count $10 and with $2 variants tions. The association changed little upon further adjustment per gene were retained. In total, 6924 genes met these filter for UACR. Hence, subsequent GWAS for urinary metabolite criteria. To account for multiple testing, the overall signifi- concentrations were adjusted for age and sex. cance threshold (0.05) was adjusted for the number of respective outcomes (per matrix), the number of genes (n=6924), GWAS and Burden Tests Identify Significant and the number of performed tests (n=2). The significance Associations with 25 Serum Metabolites thresholds were thus P,2.6E208 for 139 serum metabolites, We carried out GWAS of common SNPs (MAF.0.01) with P,8.8E208 for 41 urinary metabolites, P,3.8E210 for 9591 concentrations of 139 metabolites in serum. The genomic in- serum metabolite ratios, P,4.4E209 for 820 urinary metab- flation factors ranged from 0.99 to 1.03, consistent with the olite ratios, and P,1.1E207 for 34 FEs. Suggestive statistical absence of inflation from systematic errors. Multiple previ- significance was defined as P,1E206. ously reported associations were identified at genome-wide

1516 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1513–1524, 2018 www.jasn.org CLINICAL RESEARCH

Table 1. Characteristics of the study sample All Patients in the GCKD Study GCKD Patients with Genetic and Metabolite Characteristic (n=5217) Information (n=1143) Male sex, % (n) 60 (3132) 62.9 (719) Mean age (SD), yr 60.1 (11.96) 56 (13.1) Mean BMI (SD), kg/m2a 29.8 (5.97) 28.6 (5.64) Mean systolic BP (SD), mm Hga 139.5 (20.36) 138.5 (19.11) Diabetes, % (n) 35.5 (1853) 23 (263) Mean serum creatinine (SD), mg/dla 1.5 (0.48) 1.5 (0.52) Mean eGFR (SD), ml/min per 1.73 m2a 49.5 (18.22) 51.6 (21.12) Mean urinary creatinine (SD), mg/dla 82.1 (55.58) 79.4 (53.72) Median urinary albumin-to-creatinine ratio (IQR), 50.9 (9.66–391.79) 160.8 (24.35–872.27) mg/ga CKD cause: primary glomerular disease, % (n) 18.7 (978) 52.4 (599) The number of GCKD patients with genetic data and either serum measurements or urinary measurements is slightly higher: n=1168 and n=1155, respectively. BMI, body mass index; IQR, interquartile range. aThese variables were not available for some patients (proportion,2%). significance (P,3.6E210, see Methods), such as genetic var- concentrations. No gene passed the significance threshold for iants in NAT8 and N-acetyl-ornithine (P=1.0E262), CPS1 multiple testing (P,2.6E208, see Methods), but the burden of and glycine (P=2.4E220), PRODH and proline rare deleterious variants in four genes showed suggestive evi- (P=1.8E217), ACADL and nonanoylcarnitine (P=1.3E235), dence for association (P,1E206) with at least one metabolite and FADS1 and phosphatidylcholines such as PC aa C36:5 (Supplemental Table 3). Among these was an association be- (P=4.7E214; Table 3). The 25 metabolites for which signifi- tween serum concentrations of histidine and rare variants in the cant associations were found mapped into 13 loci. HAL gene (P-burden=7.1E208), encoding histidine ammonia- One locus, intergenic to TMEM176A and AOC1, showed a lyase.34 Rare mutations in HAL can cause autosomal-recessive novel genome-wide significant association with concentration histidinemia (MIM #235800), and associations between rare of serum putrescine (P=2.4E212), a known uremic toxin genetic variants in HAL and blood histidine concentrations (Figure 1).32 The effect of the G allele at the index SNP in have previously been reported in a population-based this locus, rs4725366, on serum putrescine concentration study.9 In that study, 14 rare HAL variants were underlying was twice as large in individuals with eGFR,30 ml/min per the association signal with blood histidine concentrations, 1.73 m2 compared with those with eGFR$60 (20.28 versus whereas in our study 17 such variants were identified (nine 20.60, Table 4). The relationship between the genetic effect of which overlapped). These findings indicate that rare del- size and eGFR, however, was not linear, and a statistical in- eterious variants also influence metabolite concentrations in teraction term between SNP and eGFR category was not sta- patients with CKD. Because additional and unique poten- tistically significant (P=0.3). SNP rs4725366 is significantly tially deleterious variants are identified with each new study associated with the expression of the neighboring AOC1 population examined, results in the suggestive significance gene (P=1.5e222 in testis in the GTEx resource33 that does range identified in our study likely contain other true posi- not contain eQTL information from kidney tissue). AOC1 tive findings that await discovery in future experimental or encodes amine oxidase copper-containing 1, an enzyme cata- larger genetic association studies. lyzing the oxidative deamination of polyamines such as hista- mine, putrescine, and cadaverine. GWAS and Burden Tests of Urinary Metabolites A full table listing the index variant for metabolites with at Confirm Population-Based Findings in CKD least one suggestive association signal (P,5E208, see Meth- Next, we carried out GWAS of common SNPs and metabolite ods) is shown in Supplemental Table 2. These include loci not concentrations in urine. Among the 41 metabolites that passed significantly associated with blood metabolite concentrations quality control, significant associations were observed for two in previous studies, such as the PTPRN2 locus and concentra- previously reported loci, SLC7A9 and urinary lysine concen- tions of lysophosphatidylcholine with acyl residue C18:1 trations (P=4.7E212) and NAT8 and urinary N-acetylorni- (P=1.7E208). Because this locus also showed suggestive evi- thine concentrations (P=9.6E212, Supplemental Table 4). dence of association with the blood concentrations of another These results were stable to additional adjustment for lipid in a previous metabolite GWAS,7 such associations are UACR. Suggestive associations were detected for several excellent candidates to achieve genome-wide significance in loci that were previously described as associated with the me- future, larger association studies. tabolite’s blood concentrations, such as SLC22A1 and butyr- We additionally examined the aggregate effect of rare del- ylcarnitine.10 The rare variant analyses did not lead to the eterious variants within the same gene on serum metabolite identification of an aggregate effect of deleterious variants in

J Am Soc Nephrol 29: 1513–1524, 2018 Genetics of Metabolism in CKD 1517 CLINICAL RESEARCH www.jasn.org

, 2 29 35 22 25 07 08 12 12 any gene with P 1E 06. The number of loci associated with 2 2 2 2 2 2 2 2 urinary metabolite concentrations among patients with CKD Value

P in this study needs to be interpreted in light of the assay used to quantify metabolites: the BIOCRATES AbsoluteIDQ p180 Kit is a targeted assay originally developed to quantify blood metabolite concentrations, only a subset of which are detected in urine (Supplemental Table 1). 0.46 0.036 3.1E 0.41 0.038 1.4E 0.29 0.042 7.3E 0.23 0.041 1.5E 0.29 0.042 9.7E Effect SEM GWAS and Burden Tests of Serum Metabolite Ratios Highlight Variants Central to Lipid and Acylcarnitine Metabolism Metabolite concentrations can be used to model physiologic Allele functions of the kidney. Speciﬁcally, ratios of metabolites can Freq Effect reﬂect not only the activity of enzymatic reactions, but also co- or counter-transport at the apical or basolateral membrane of tubular cells, resulting in synchronized changes of these me-

Allele tabolites in urine and blood, respectively. We therefore first Ref/Effect evaluated the association between common genetic variants and metabolite ratios in serum. Genome-wide significant associations were identified between genetic variants and 259 metabolite ratios mapping into 11 loci (Supplemental Table 5), many of which implicated known genes central to lipid biosynthesis. These analyses confirmed FADS1 as a major player in the metabolism of phosphatidylcholines, ACADM to evaluate correction of urinary metabolite concentrations for urine dilution.

25 and ACADS in metabolism of medium-chain acylcarnitines, SPTLC3 in the synthesis of sphingomyelins, as well as 3,35 ed for nonanoylcarnitine (serum) and lysine (urine) PDXDC1 in phospholipid metabolism. fi SNP Chr Position In the gene-based tests of rare variants on serum metabolite rs2014355 12 121175524 T/C 0.28 0.43 0.043 1.6E rs2286963 2 211060050 T/G 0.34 0.46 0.041 7.0E rs7247977 19 33358355 T/C 0.39ratios, 0.22 we 0.042 2.6E identified an exome-wide significant association (P,2.4E210) for one gene, ACADM. Significant associations were observed for several metabolite ratios, with the lowest P value observed for the decanoylcarnitine/dodecanoylcarnitine ratio from the burden test (P=6.6E216, Figure 2). The low burden test P value, despite the modest sample size and low frequency of the evaluated deleterious variants (MAF,0.01), suggested large and Model concordant effects of the contributing variants affecting the metabolite concentrations. Indeed, the four variants in ACADM that contributed to the test were all associated with increases in the adjustment for age, sex, eGFR adjustment for age, sex, eGFR adjustment for age, sex adjustment for age, sex decanoylcarnitine/dodecanoylcarnitine ratio (Figure 2). ACADM encodes medium-chain acyl-CoA dehydrogenase, which catalyzes the initial reaction in the b-oxidation of C4 to C12 straight-chain b acyl-CoAs. Mutations in the ACADM gene cause MCAD deficiency (MIM #201450), a condition that is part of newborn screening on the basis of acylcarnitine analysis in blood. Three of the four variants detected in our screen have been reported as cause of classic

Normalization MCAD deﬁciency if present in the homozygous state (p.E43K, p. NA NA creatinine-normalized adjustment for age, sex creatinine-normalized adjustment for age, sex, lnUACR pq-normalized adjustment for age, sex pq-normalized adjustment for age, sex, lnUACR Y67H, p.K329E), conﬁrming the functional and clinical relevance of these mutations. The mutation p.K329E gives rise to a severe disease phenotype, whereas the clinical picture is milder for the

a mutations p.E43K and p.Y67H, suggesting residual enzyme function. These observations from monogenic manifestations of the Evaluation of known genetic associations in a CKD population, exempli disease are consistent with the effect sizes observed for heterozy- gous carriers in our study, with p.K329E showing the largest effect Metabolite on the decanoylcarnitine/dodecanoylcarnitine ratio. The last var- Nonanoylcarnitine, C9 NA Butyrylcarnitine, C4 NA Lysine We compared the state-of-the-art normalization with urinary creatinine and the probabilistic quotient normalization Metabolite measurements were transformed before analysis using rank-based inverse normal transformation. Small differences in comparison to results from subsequent GWAS originate from differences in ﬁ Table 2. Chr, chromosome; Ref,a reference; Freq, frequency; NA, not available; pq, probabilistic quotient.b software and covariate adjustment. Serum Urine iant identi ed in our study, p.I364T, is yet of uncertain clinical

1518 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1513–1524, 2018 mScNephrol Soc Am J 29: 1513

– $ ﬁ P, 2 54 2018 1524, Table 3. Common variants (MAF 0.05) associated with serum metabolites at genome-wide signi cance ( 3.6E 10) Proportion of Variance Metabolite SNP Chr Position NEA EA Effect SEM P Value AF_EA Genes Explained C10 rs12134854 1 76137606 T C 20.30 0.04 1.43E212 0.31 0.04 SLC44A5(dist=60807),ACADM(dist=52437) Ac-Orn rs375811360 2 73870324 C G 20.84 0.05 1.04E262 0.21 0.24 NAT8 C9 rs3764913 2 211074909 T C 0.46 0.04 1.34E235 0.34 0.09 ACADL Gly rs1047891 2 211540507 C A 0.39 0.04 2.43E220 0.31 0.07 CPS1 C18:2 rs12194000 6 110776012 A G 20.31 0.05 3.44E211 0.24 0.03 SLC22A16 C4 rs146203232 6 160521885 C T 20.54 0.06 1.57E218 0.09 0.05 IGF2R Putrescine rs4725366 7 150522054 A G 20.36 0.05 2.42E212 0.81 0.04 TMEM176A(dist=19846),AOC1(dist=27511) C0 rs1171616 10 61468589 G T 0.38 0.05 2.84E215 0.77 0.05 SLC16A9 PC ae C38:4 rs11320420 11 61542006 GA G 20.28 0.04 2.35E210 0.33 0.04 MYRF PC ae C38:5 rs11320420 11 61542006 GA G 20.33 0.04 2.79E213 0.33 0.05 MYRF PC ae C36:5 rs11320420 11 61542006 GA G 20.34 0.04 1.47E214 0.33 0.05 MYRF PC ae C38:5 rs174537 11 61552680 G T 20.32 0.04 9.45E213 0.33 0.04 MYRF PC ae C36:5 rs174537 11 61552680 G T 20.34 0.04 2.47E214 0.33 0.05 MYRF PC aa C38:5 rs174537 11 61552680 G T 20.36 0.04 1.37E216 0.33 0.06 MYRF PC aa C38:4 rs102275 11 61557803 T C 20.40 0.04 9.76E221 0.33 0.07 TMEM258 lysoPC a C20:4 rs102275 11 61557803 T C 20.50 0.04 2.63E232 0.33 0.11 TMEM258 PC aa C36:4 rs102274 11 61557826 T C 20.41 0.04 8.07E221 0.33 0.07 TMEM258 PC aa C36:3 rs174560 11 61581764 T C 0.28 0.04 2.52E210 0.29 0.03 FADS1 PC aa C36:5 rs174566 11 61592362 A G 20.32 0.04 4.66E214 0.33 0.05 FADS1(dist=7833),FADS2(dist=3351) C4 rs2014355 12 121175524 T C 0.42 0.04 3.90E227 0.28 0.07 ACADS PC aa C28:1 rs7160525 14 64232220 G A 0.42 0.05 6.57E215 0.16 0.05 SGPP1(dist=37464),SYNE2(dist=87463) 2 SGPP1 SYNE2 eeiso eaoimi CKD in Metabolism of Genetics

SM (OH) C14:1 rs112329286 14 64239877 T TTTTA 0.36 0.05 5.44E 12 0.16 0.04 (dist=45121), (dist=79806) www.jasn.org PC ae C32:1 rs1077989 14 67975822 A C 20.28 0.04 1.97E212 0.48 0.04 TMEM229B Asn rs2282377 14 104571820 C G 0.43 0.06 1.23E211 0.13 0.04 ASPG Pro rs117935223 22 18911333 C A 0.62 0.07 1.79E217 0.09 0.06 PRODH Sample size: n=1168. Chr, chromosome; NEA, noneffect allele; EA, effect allele; AF_EA, allele frequency of effect allele; Genes, gene the SNP maps into or genes neighboring the SNP and the corresponding distances in base pairs. LNCLRESEARCH CLINICAL 1519 CLINICAL RESEARCH www.jasn.org

Table 4. Effect of G allele of rs4725366 (AOC1 locus) on serum putrescine concentrations stratiﬁed by eGFR eGFR Category n Effect SEM P Value (ml/min per 1.73 m2) $60 293 20.28 0.10 6.37E203 45–59 343 20.41 0.09 4.16E206 30–44 391 20.30 0.09 1.34E203 ,30 117 20.60 0.15 1.12E204 n,samplesize.

intronic index SNP rs12460876 show higher urinary lysine/AA0 ratios than those homozygous for the major T allele, indicating less efficient function of the encoded SLC7A9 transport protein. Figure 1. Regional association plot illustrates the presence of Modeling of the association between rs12460876 and all pairwise serum putrescine-associated SNPs upstream of AOC1 and amino acid ratios in urine revealed specificlysine/AA0 ratios with downstream of TMEM176A on chromosome 7. y axis, negative strong associations (Figure 3B), representing a readout of human log10(P value) from GWAS analysis; x axis, genomic position. Linkage disequilibrium with the index SNP rs4725949 (purple) in SLC7A9 function in vivo. Other amino acid ratios showed little the 1000 Genomes Project EUR reference population is color- (mostly ornithine-containing ratios) or no association, such as 2 coded as detailed in the legend. lysine/AA ratios (aspartate, glutamate). The evaluation of the aggregate effect of rare variants on urinary metabolite ratios did not yield findings that achieved exome-wide significance (Figure 2). Our findings reveal a relatively high carrier significance (P,4.4E209). There were, however, biologically frequency of potentially deleterious ACADM mutations with large plausible findings of suggestive significance (P,1E206, Supple- effects on metabolite concentrations in an adult population of mental Table 8). For example, rare variants in SLC36A2 European ancestry (23 of 1157=2.0%; 1.9% without p.I364T). were associated with the urinary glycine-to-serine ratio This relatively high frequency is also observed among European (P=4.0E207). This matches the known transport capacity of ancestry individuals in the gnomAD database, suggesting that this the encoded transporter for glycine and its localization to the S1 finding is more widely applicable. Full results of all serum metab- segment of the proximal tubule.13,38 Transport capacity for serine olite ratios associated at suggestive significancewithrarevariants has only been described for murine SLC36A2 so far39 and could in one or more genes are listed in Supplemental Table 6. now be tested experimentally for human SLC36A2. Another example is the association between rare variants in PAH and the GWAS of Urine Metabolite Ratios Highlight Amino phenylalanine/tyrosine ratio (P=5.9E207), which matches the Acids Counter-Transported by SLC7A9 function of the encoded phenylalanine hydroxylase that catalyzes There were several genetic loci with significant SNP associa- the rate-limiting step in phenylalanine catabolism, the hydroxyl- tions with urinary metabolite ratios (Supplemental Table 7). ation of phenylalanine to tyrosine. Although an earlier study had Interestingly, these associations delivered valuable insights reported the association between rare variants in PAH and blood into renal physiology: genetic variants in SLC7A9 showed sig- phenylalanine concentrations,9 our findings highlight that mod- nificant associations with lysine-containing ratios that were eling of enzyme function as the ratio of the enzyme’ssubstrateand many orders of magnitude stronger than the association of product adds additional information. In addition, the examina- the respective SNP with urinary lysine concentrations alone tion of urine adds information, because the gene is exclusively (Figure 3). This is exemplified by the lysine/glutamine ratio expressed in liver and kidney, and urinary concentrations inte- (P=2.2E223, P-gain=2.1E+11; see Methods), the lysine/tyro- grate the information of reactions occurring in the kidney. More sine ratio (P=3.1E222, P-gain=1.5E+10), and the lysine/thre- generally, these findings suggest that the study of metabolite ratios onine ratio (P=5.8E222, P-gain=1.5E+10). Other significant in urine of patients with CKD in a larger study sample is a prom- lysine-containing ratios contained alanine, glycine, serine, as- ising avenue for future research. paragine, phenylalanine, and tryptophan. Of note, all of these amino acids carry no net electric charge at physiologic pH, con- GWAS and Burden Tests of Metabolite FEs sistent with the known role of SCL7A9 as an exchanger of cationic Lastly,GWASofcommonSNPswiththeFEsof34metabolites amino acids (lysine, arginine, and ornithine) and cystine in yielded one finding of genome-wide significance, the associa- the renal tubular lumen against intracellular neutral amino acids tion between variants in the MLEC gene at the ACADS locus (AA0) described in in vitro studies.36,37 This finding illustrates the and the FE of butyrylcarnitine (C4, P=2.8E214; Supplemental potential of modeling metabolite concentrations to uncover spe- Table 9). However, these associations were not as significant as cific physiologic functions of the kidney. The concept is illustrated the ones with serum butyrylcarnitine concentration alone in Figure 3A: individuals homozygous for the minor C allele at the (P=3.9E227, Table 3). No significant or suggestive associations

1520 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1513–1524, 2018 www.jasn.org CLINICAL RESEARCH

adjustments across these studies unfortu- nately preclude a systematic comparison of genetic effect sizes between population-based individuals and patients with CKD to answer the question of whether CKD as a metabolic challengeisreflected in different geneticeffect sizes. A systematic comparison of genetic effect sizes between population-based individuals and patients with CKD should therefore be the focus of future joint studies that har- monize the methods for metabolite quantification and data analysis across studies apriori. There was one novel association between serum putrescine concentrations and genetic ACADM Figure 2. Rare coding variants in are cumulatively associated with the serum variants at the AOC1 locus, encoding an en- concentrations of acylcarnitine ratios. The upper panel contains a schematic repre- zyme catalyzing the oxidative deamination of ACADM fi sentation of the gene and the mapping of the identi ed deleterious variants polyamines such as putrescine.43 The doubling to exons. The table contains P values for the cumulative effect of the variants on the P in genetic effect size in patients with CKD with concentrations of the listed acylcarnitine ratios, as well as the association values for , the single variants that contribute to the gene-based tests (burden test and sequence eGFR 30 compared with those with $ 2 kernel association test, see Methods). The allele frequencies of these variants in Eu- eGFR 60 ml/min per 1.73 m may hint at ropean individuals in gnomAD database are: 0.0032 for rs147559466, 0.00099 for the presence of dose-response effects with rs121434280, 0.0061 for rs77931234, and 0.00049 for rs150710061. C10, dec- worsening kidney function and illustrate the anoylcarnitine; C12, dodecanoylcarnitine; C14:1, Tetradecenoylcarnitine; cMAF, cu- value of studying “challenged” populations mulative MAF; MAC, minor allele count; SKAT, Sequence Kernel Association Test. (such as disease cohorts) in general. Again, however, differences across studies complicate (P,1E206) were identified for the aggregate effect of rare comparisons: the lack of reported associations with putrescine in deleterious variants in any gene and the FEs of the 34 evaluated previous population-based studies may be because it was not in- metabolites. Thus, investigation of the FEs of metabolites did cluded on targeted assays or removed during quality control, be- not provide additional insights. cause it was concentrated too lowly for quantification, because it was measured but not identified as putrescine, or because it showed no genetic associations and was therefore not reported. Although DISCUSSION transcriptional activation of AOC1 has recently been linked to kidney morphogenesis in mice by affecting polyamine breakdown,44 it Our study has several main findings: first, genetic influences on cannot be inferred from our study whether elevated putrescine metabolite concentrations in serum and urine from previous pop- concentrations occur as a consequence of CKD or contribute caus- ulation-based studies2–4,7–9,11,40,41 translate to patients with CKD. ally to CKD etiology and/or progression. Second, the study of patients with CKD revealed a novel association In terms of biologic insights, our study offers snapshots of renal between concentrations of a uremic toxin, putrescine,42 and varia- physiology in humans and generates novel hypotheses for experi- tion near the gene encoding amine oxidase copper-containing 1. mental confirmation. This former is exemplified by the identifica- Third, metabolite ratios are useful composite indices that reflect tion of the neutral amino acids transported in exchange for urinary underlying physiologic mechanisms such as enzymatic reactions lysine in humans in vivo, and the latter by the potential transport and counter-transport mechanisms of importance in nephrology, capacity of human SLC36A2 for serine. Several previous studies as exemplified by variants in SLC7A9 and the urinary ratios of lysine reported associations between the same index or a proxy variant at and several neutral amino acids. Our study generates multiple new the SLC7A9 locus with urinary lysine concentrations6,11–13 or some hypotheses about the role of identified genes in human metabolism lysine-containing ratios,11 but without biologic interpretations in that can now be tested experimentally. light of tubular counter-transport mechanisms. Interestingly, The great majority of genomic loci identified in our study have GWAS of eGFR in the general population could show that the previously been detected in equally sized or larger GWAS of me- rs12460876 C allele is associated not only with higher urinary lysine tabolite concentrations in population-based studies. Our study re- concentrations but also with higher eGFR and lower CKD preva- veals that these genetic determinants of metabolite concentrations lence and incidence,6,13,45 consistent with the association between are also important in the setting of CKD. Thus, patients with both higher urinary lysine concentrations and decreased risk for CKD.13 CKD and metabolite-increasing genotypes may represent a partic- The molecular mechanisms connecting reduced SLC7A9-medi- ularly high-risk group for adverse events related to such metabolites. ated transport, resulting in higher urinary lysine and cystine and The different techniques and assays used for metabolite quanti- higher intracellular AA0 concentrations, to better eGFR/lower CKD fication as well as the different transformations and covariate risk are presently unclear.

J Am Soc Nephrol 29: 1513–1524, 2018 Genetics of Metabolism in CKD 1521 CLINICAL RESEARCH www.jasn.org

Studies of genetic influences on metabolite concentrations in patients with CKD have the potential to ultimately deliver insights of clinical importance; for example, by identifying genetic predispositions to unfavorable profiles of CKD risk factors or to CKD comorbidities. This may be especially relevant for rare variants with large effects on CKD risk factors, the presence and prevalence of which is illustrated by the discovery of rare coding ACADM variants in our study. Illuminating pathways that underlie processes such as hyperlipidemia or stone formation may eventually help to better treat or prevent these common CKD-related conditions. In addition, genetic discoveries related to metabolites that are considered uremic toxins such as putrescine can be used to test their causal involvement in disease progression once large-scale genetic studies of CKD progression become available. Some limitations of our study warrant mention: the use of the BIOCRATES p180 assay for metabolite quantification results in a limited number of metabolites to study. Although nontargeted approaches to quantify metabolites result in a much greater number of metabolites for study and cover classic uremic toxins that are not part of the BIOCRATES assay, the latter was chosen in this study because it includes isotope-labeled standards for many metabolites that allow for comparison and integration of their concentrations across urine and blood. We could not comprehensively com- pare genetic effect sizes on metabolite concentrations in patients with CKD to those from previous population-based studies because of differences in methods to quantify metabolites and statistical data analysis. Our study was limited to individuals of Eu- ropean ancestry, and can therefore not nec- essarily be generalized. Twenty-four-hour urine collections were not available in the Figure 3. Counter-transport of amino acids by SLC7A9 in the renal proximal tubule GCKD study, which may have led to in- cell is revealed by the analysis of urinary metabolite ratios. The upper part (A) contains creased variation of urinary metabolite a schematic representation of SLC7A9 function in the proximal tubule, reabsorbing concentrationsbecause of differences in cir- dibasic urinary amino acids such as lysine in exchange for intracellular neutral amino cadian rhythms resulting in reduced power acids. Individuals homozygous for the minor C allele at the index SNP rs12460876 0 to detect associations. Urate is one of few show higher lysine/neutral amino acid (AA ) ratios in the urine, indicating less efficient lysine reuptake at the brush border. Listed amino acids are restricted to the ones metabolites whose FE has been assessed measured in this study. The lower panel (B) represents a heatmap of the strength of the association between genotype at intronic rs12460876 (scaled chi square statistics) and the evaluated pairwise metabolite ratios in urine. It highlights the AA0 counter- isoleucine; lys, lysine; orn, ornithine; phe, transported against lysine in vivo.Forcomparison,lysinebyitselfhasascaledchi phenylalanine; pro, proline; ser, serine; thr, square statistic of 0.39, indicated by light orange color. ala, alanine; asn, asparagine; threonine; trp, tryptophan; tyr, tyrosine; val, asp, aspartate; gln, glutamine; glu, glutamate; gly, glycine; his, histidine; ile, valine.

1522 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1513–1524, 2018 www.jasn.org CLINICAL RESEARCH from different urine collection methods in studies of healthy 2. Gieger C, Geistlinger L, Altmaier E, Hrabé de Angelis M, Kronenberg F, individuals. In these studies, the FEs for urate quantified from Meitinger T, et al.: Genetics meets metabolomics: A genome-wide association fi PLoS Genet spot urine were similar to those from 24-hour urines,46–48 but study of metabolite pro les in human serum. 4: e1000282, 2008 3. Illig T, Gieger C, Zhai G, Römisch-Margl W, Wang-Sattler R, Prehn C, this may not pertain to all metabolites. et al.: A genome-wide perspective of genetic variation in human me- Our study represents the first study of genetic influences on tabolism. Nat Genet 42: 137–141, 2010 metabolite profiles in a large population of patients with CKD. 4. Suhre K, Shin SY, Petersen AK, Mohney RP, Meredith D, Wägele B, et al; We were able to examine up to 1168 patients with CKD and CARDIoGRAM: Human metabolic individuality in biomedical and Nature – benefitted from the standardized data collection, sample pre- pharmaceutical research. 477: 54 60, 2011 5. Kettunen J, Tukiainen T, Sarin AP, Ortega-Alonso A, Tikkanen E, Lyytikäinen processing and storage procedures, DNA extraction, and gen- LP, et al.: Genome-wide association study identifies multiple loci influencing otyping in the GCKD study. Our study shows that genetic human serum metabolite levels. Nat Genet 44: 269–276, 2012 influences on metabolite profiles in the general population 6. Rueedi R, Ledda M, Nicholls AW, Salek RM, Marques-Vidal P, Morya E, translate to patients with CKD and that novel genetic associ- et al.: Genome-wide association study of metabolic traits reveals novel PLoS Genet ations can be identified in this study population. Modeling gene-metabolite-disease links. 10: e1004132, 2014 7. Shin SY, Fauman EB, Petersen AK, Krumsiek J, Santos R, Huang J, et al.; of renal metabolite handling such as counter-transport mech- Multiple Tissue Human Expression Resource (MuTHER) Consortium: An anisms delivered novel insights into human renal (patho-) atlas of genetic influences on human blood metabolites. Nat Genet 46: physiology and generated new hypotheses for experimental 543–550, 2014 confirmation. Future extensions of sample size and metabolite 8. DraismaHHM,PoolR,KoblM,JansenR,PetersenAK,VaarhorstAAM,etal.: fi coverage hold great promise to generate additional insights Genome-wide association study identi es novel genetic variants contributing to variation in blood metabolite levels. Nat Commun 6: 7208, 2015 into human renal physiology in vivo. 9.RheeEP,YangQ,YuB,LiuX,ChengS,DeikA,etal.:Anexomearray study of the plasma metabolome. Nat Commun 7: 12360, 2016 10. Long T, Hicks M, Yu HC, Biggs WH, Kirkness EF, Menni C, et al.: Whole- genome sequencing identifies common-to-rare variants associated ACKNOWLEDGMENTS with human blood metabolites. Nat Genet 49: 568–578, 2017 11. Suhre K, Wallaschofski H, Raffler J, Friedrich N, Haring R, Michael K, We are grateful for the willingness of the patients to participate in the et al.: A genome-wide association study of metabolic traits in human urine. Nat Genet 43: 565–569, 2011 German Chronic Kidney Disease (GCKD) study. The enormous effort 12. Raffler J, Friedrich N, Arnold M, Kacprowski T, Rueedi R, Altmaier E, of the study personnel of the various regional centers is highly ap- et al.: Genome-wide association Study with targeted and non-targeted preciated. We thank the large number of nephrologists who provide NMR metabolomics identifies 15 novel loci of urinary human metabolic routine care for the patients and collaborate with the GCKD study. The individuality. PLoS Genet 11: e1005487, 2015 GCKD Investigators are listed in the Supplementary Information. 13. McMahon GM, Hwang SJ, Clish CB, Tin A, Yang Q, Larson MG, et al.; CKDGen Consortium: Urinary metabolites along with common and The work of P.S., Y.L., and A.K. and metabolite quantification was rare genetic variations are associated with incident chronic kidney supported by grants KO 3598/2-1 and KO 3598/4-1 of the German Re- disease. Kidney Int 91: 1426–1435, 2017 search Foundation (DFG). The work of M.W., U.T.S., and A.K. was 14. Eckardt KU, Bärthlein B, Baid-Agrawal S, Beck A, Busch M, Eitner F, supported by DFG CRC 1140. P. Schlosser was supported by DFG CRC et al.: The German Chronic Kidney Disease (GCKD) study: Design and 992. A.K. was additionally supportedbyaDFGHeisenbergProfessorship methods. Nephrol Dial Transplant 27: 1454–1460, 2012 15. Titze S, Schmid M, Köttgen A, Busch M, Floege J, Wanner C, et al.; GCKD (KO 3598/3-1); U.T.S. was additionally supported by the Else Kroener study investigators: Disease burden and risk profile in referred patients with Fresenius Kolleg Nierenfunktionsstörungen als Komplikation von Sys- moderate chronic kidney disease: Composition of the German Chronic temerkrankungen (NAKSYS). Genotyping was supported by Bayer Kidney Disease (GCKD) cohort. Nephrol Dial Transplant 30: 441–451, 2015 Pharma Aktiengesellschaft (AG). Metabolite quantification was addi- 16. Prokosch HU, Mate S, Christoph J, Beck A, Köpcke F, Stephan S, et al.: tionally supported by the Else Kroener Fresenius Foundation and the Designing and implementing a biobanking IT framework for multiple research scenarios. Stud Health Technol Inform 180: 559–563, 2012 German Ministry of Education and Research (BMBF) Gerontosys II 17. Anderson CA, Pettersson FH, Clarke GM, Cardon LR, Morris AP, NephAge Project, 031 5896A. The GCKD study was/is funded by grants Zondervan KT: Data quality control in genetic case-control association from the BMBF (grant number 01ER0804) and the Kuratorium fuer studies. Nat Protoc 5: 1564–1573, 2010 Heimdialyse (KfH) Foundation for Preventive Medicine. 18. Howie BN, Donnelly P, Marchini J: A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet 5: e1000529, 2009 DISCLOSURES 19. Delaneau O, Zagury JF, Marchini J: Improved whole-chromosome phasing for disease and population genetic studies. Nat Methods 10: 5–6, 2013 J.W. is employed by BIOCRATES Life Sciences Aktiengesellschaft (AG) 20. Guo Y, He J, Zhao S, Wu H, Zhong X, Sheng Q, et al.: Illumina human (Innsbruck, Austria). exome genotyping array clustering and quality control. Nat Protoc 9: 2643–2662, 2014 21. Goldstein JI, Crenshaw A, Carey J, Grant GB, Maguire J, Fromer M, REFERENCES et al.; Swedish Schizophrenia Consortium; ARRA Autism Sequencing Consortium: zCall: A rare variant caller for array-based genotyping: Genetics and population analysis. Bioinformatics 28: 2543–2545, 2012 1. Kalim S, Rhee EP: An overview of renal metabolomics. Kidney Int 91: 22. Ramsay, S.L.B.D. (Heidelberg 3084, AU), Guggenbichler, W. (Austrasse 61–69, 2017 26A, A-6063 Rum, AT), Weinberger, K.M. (Weidach 82, A-6414

J Am Soc Nephrol 29: 1513–1524, 2018 Genetics of Metabolism in CKD 1523 CLINICAL RESEARCH www.jasn.org

Mieming, AT), Graber, A. (Fishnalerstrasse 14/Top3, A-6020 Innsbruck, identifies novel loci associated with circulating phospho- and sphin- AT), Stöggl, W.M. (Hoher Weg 9/7, A-6020 Innsbruck, AT); Device for golipid concentrations. PLoS Genet 8: e1002490, 2012 quantitative analysis of a metabolite profile. (BIOCRATES Life Sciences 36. Munck BG: Lysine transport across the small intestine. Stimulating and AG (Innrain 66, 6020 Innsbruck, AT), 2014) inhibitory effects of neutral amino acids. J Membr Biol 53: 45–53, 1980 23. Ramsay, S.L.B.D. (Heidelberg 3084, AT), Stöggl, W.M. (Hoher Weg 9/7, 37. Munck BG: Transport of neutral and cationic amino acids across the A-6020 Innsbruck, AT), Weinberger, K.M. (Weidach 82, A-6414 brush-border membrane of the rabbit ileum. J Membr Biol 83: 1–13, Mieming, AT), Graber, A. (Fischnalerstrasse Innsbruck 14/Top 3, A- 1985 6020 Innsbruck, AT), Guggenbichler, W. (Austrasse 26A, A-6063 Rum, 38. Bröer S, Bailey CG, Kowalczuk S, Ng C, Vanslambrouck JM, Rodgers H, AT); Apparatus for analyzing a metabolite profile. (BIOCRATES Life et al.: Iminoglycinuria and hyperglycinuria are discrete human pheno- Sciences AG (Innrain 66, 6020 Innsbruck, AT), 2014) types resulting from complex mutations in proline and glycine trans- 24. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF III, Feldman HI, porters. JClinInvest118: 3881–3892, 2008 et al.; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration): 39. Kennedy DJ, Gatfield KM, Winpenny JP, Ganapathy V, Thwaites DT: A new equation to estimate glomerular filtration rate. Ann Intern Med Substrate specificity and functional characterisation of the H+/amino 150: 604–612, 2009 acid transporter rat PAT2 (Slc36a2). Br J Pharmacol 144: 28–41, 2005 25. Dieterle F, Ross A, Schlotterbeck G, Senn H: Probabilistic quotient 40. Rhee EP, Ho JE, Chen MH, Shen D, Cheng S, Larson MG, et al.: A ge- normalization as robust method to account for dilution of complex bi- nome-wide association study of the human metabolome in a commu- ological mixtures. Application in 1H NMR metabonomics. Anal Chem nity-based cohort. Cell Metab 18: 130–143, 2013 78: 4281–4290, 2006 41. Burkhardt R, Kirsten H, Beutner F, Holdt LM, Gross A, Teren A, et al.: 26. Fuchsberger C, Taliun D, Pramstaller PP, Pattaro C; CKDGen consor- Integration of genome-wide SNP data and gene-expression profiles tium: GWAtoolbox: An R package for fast quality control and handling reveals six novel loci and regulatory mechanisms for amino acids and of genome-wide association studies meta-analysis data. Bioinformatics acylcarnitines in whole blood. PLoS Genet 11: e1005510, 2015 28: 444–445, 2012 42. Igarashi K, Ueda S, Yoshida K, Kashiwagi K: Polyamines in renal failure. 27. Pruim RJ, Welch RP, Sanna S, Teslovich TM, Chines PS, Gliedt TP, et al.: Amino Acids 31: 477–483, 2006 LocusZoom: Regional visualization of genome-wide association scan 43. Chassande O, Renard S, Barbry P, Lazdunski M: The human gene for results. Bioinformatics 26: 2336–2337, 2010 diamine oxidase, an amiloride binding protein. Molecular cloning, se- 28. Pattaro C, Teumer A, Gorski M, Chu AY, Li M, Mijatovic V, et al.; ICBP quencing, and characterization of the promoter. JBiolChem269: Consortium; AGEN Consortium; CARDIOGRAM; CHARGe-Heart Fail- 14484–14489, 1994 ure Group; ECHOGen Consortium: Genetic associations at 53 loci 44. Kirschner KM, Braun JF, Jacobi CL, Rudigier LJ, Persson AB, Scholz H: highlight cell types and biological pathways relevant for kidney func- Amine oxidase copper-containing 1 (AOC1) is a downstream target tion. Nat Commun 7: 10023, 2016 gene of the Wilms tumor protein, WT1, during kidney development. J 29. Fabregat-Traver D, Sharapov SZ, Hayward C, Rudan I, Campbell H, Biol Chem 289: 24452–24462, 2014 Aulchenko Y, et al.: High-performance mixed models based genome-wide 45. Köttgen A, Pattaro C, Böger CA, Fuchsberger C, Olden M, Glazer NL, association analysis with omicABEL software. F1000 Res 3: 200, 2014 et al.: New loci associated with kidney function and chronic kidney 30. Petersen AK, Krumsiek J, Wägele B, Theis FJ, Wichmann HE, Gieger C, et al.: disease. Nat Genet 42: 376–384, 2010 On the hypothesis-free testing of metabolite ratios in genome-wide and 46. Quaratino CP, Di Sciascio N, Rucci C, Ciaglia P, Giacomello A: The metabolome-wide association studies. BMC Bioinformatics 13: 120, 2012 normal range of serum urate levels and of fractional urate excretion. 31. Grove ML, Yu B, Cochran BJ, Haritunians T, Bis JC, Taylor KD, et al.: Adv Exp Med Biol 370: 91–93, 1994 Best practices and joint calling of the HumanExome BeadChip: The 47. Giacomello A, Di Sciascio N, Quaratino CP: Relation between serum CHARGE consortium. PLoS One 8: e68095, 2013 triglyceride level, serum urate concentration, and fractional urate ex- 32. Duranton F, Cohen G, De Smet R, Rodriguez M, Jankowski J, Vanholder R, cretion. Metabolism 46: 1085–1089, 1997 et al.; European Uremic Toxin Work Group: Normal and pathologic con- 48. Perez-Ruiz F, Calabozo M, Erauskin GG, Ruibal A, Herrero-Beites AM: centrations of uremic toxins. J Am Soc Nephrol 23: 1258–1270, 2012 Renal underexcretion of uric acid is present in patients with apparent 33. Consortium GT; GTEx Consortium: Human genomics. The Genotype- high urinary uric acid output. Arthritis Rheum 47: 610–613, 2002 Tissue Expression (GTEx) pilot analysis: Multitissue gene regulation in humans. Science 348: 648–660, 2015 34. Suchi M, Harada N, Wada Y, Takagi Y: Molecular cloning of a cDNA See related article, “Insights into CKD from Metabolite GWAS,” on pages Biochim Biophys Acta – encoding human histidase. 1216: 293 295, 1993 1349–1351. 35. Demirkan A, van Duijn CM, Ugocsai P, Isaacs A, Pramstaller PP, Liebisch G, et al.; DIAGRAM Consortium; CARDIoGRAM Consortium; CHARGE This article contains supplemental material online at http://jasn.asnjournals. Consortium; EUROSPAN consortium: Genome-wide association study org/lookup/suppl/doi:10.1681/ASN.2017101099/-/DCSupplemental.

AFFILIATIONS

1Institute of Genetic Epidemiology, Department of Biometry, Epidemiology, and Medical Bioinformatics, and 5Department of General Pediatrics, Center for Pediatrics and Adolescent Medicine, Faculty of Medicine and Medical Center, University of Freiburg, Freiburg, Germany; 2BIOCRATES Life Sciences Aktiengesellschaft, Innsbruck, Austria; 3Department of Nephrology and Hypertension and 6Insitute of Human Genetics, University of Erlangen-Nürnberg, Erlangen, Germany; 4Institute of Functional Genomics, University of Regensburg, Regensburg, Germany; and 7Division of Genetic Epidemiology, Department of Medical Genetics, Molecular and Clinical Pharmacology, Medical University of Innsbruck, Innsbruck, Austria

1524 Journal of the American Society of Nephrology J Am Soc Nephrol 29: 1513–1524, 2018