[go: up one dir, main page]
More Web Proxy on the site http://driver.im/ Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 May 1.
Published in final edited form as: Kidney Int. 2019 Mar 12;95(5):1197–1208. doi: 10.1016/j.kint.2018.12.017

Identification of 22 novel loci associated with urinary biomarkers of albumin, sodium, and potassium excretion

Daniela Zanetti 1,2, Abhiram Rao 3, Stefan Gustafsson 4, Themistocles L Assimes 1, Stephen B Montgomery 5,6, Erik Ingelsson 1,2,4
PMCID: PMC6535090  NIHMSID: NIHMS1023301  PMID: 30910378

Abstract

Urine biomarkers reflecting kidney function and handling of dietary sodium and potassium are strongly associated with several common diseases including chronic kidney disease, cardiovascular disease, and diabetes mellitus. Knowledge about the genetic determinants of these biomarkers may shed light on pathophysiological mechanisms underlying the development of these diseases. We performed genome-wide association studies of urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr) and urinary sodium: potassium ratio (UNa/UK) in up to 218,450 (discovery) and 109,166 (replication) unrelated individuals of European ancestry from the UK Biobank. Further, we explored genetic correlations, tissue-specific gene expression, and possible genes implicated in the regulation of these biomarkers. After replication, we identified 19 genome-wide significant independent loci associated with UACR, 6 each with UK/UCr and UNa/UCr, and 4 with UNa/UK. In addition to 22 novel associations, we confirmed several established associations, including between the CUBN locus and microalbuminuria. We detected high pairwise genetic correlation across the urinary biomarkers, and between their levels and several physiological measurements. We highlight GIPR, a potential diabetes drug target, as possibly implicated in the genetic control of urinary potassium excretion, and NRBP1, a locus associated with gout, as plausibly involved in sodium and albumin excretion. Overall, we identified 22 novel genome-wide significant associations with urinary biomarkers and confirmed several previously established associations, providing new insights into the genetic basis of these traits and their connection to chronic diseases.

Keywords: chronic kidney disease, genetics, microalbuminuria, urinary biomarkers

Urinary biomarkers are used clinically to assess an individual’s renal function as well as to diagnose and predict the onset of related chronic diseases.1 Such biomarkers include creatinine, albumin, potassium, and sodium, which have been associated with chronic kidney disease (CKD),2,3 cardiovascular disease (CVD),4–6 and type 2 diabetes (T2D).7,8 Compared to blood, biomarkers in urine are less affected by homeostatic mechanisms. This allows greater fluctuations in biomarker levels, which may provide signals that more reliably reflect dynamic changes in biological and pathophysiological processes.9

Little progress has been made in disentangling the genetic determinants of urinary biomarkers despite many observational association studies showing the relevance of these biomarkers.6,10,11 Previous genetic studies of kidney function have been focused on albuminuria8,12 and on estimated glomerular filtration rate based on serum creatinine and/or cystatin C.13,14 To increase the understanding of mechanisms underlying kidney function and disease, we analyzed the genetic underpinnings of 4 urinary biomarkers, specifically of urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK). These biomarkers have been selected for their strong connection with CKD as well as for their association with CVD and T2D. Previous studies have found that microalbuminuria is predictive, independently of traditional risk factors, of CVD within groups of patients with diabetes or hypertension as well as in the general population.15 Likewise, the association of high sodium and low potassium levels with higher blood pressure (BP) is supported by a large body of evidence that includes randomized clinical trials16 and observational studies.6

We performed genome-wide association studies (GWASs) of UACR, UK/UCr, UNa/UCr, and UNa/UK in up to 218,450 (discovery) and 109,166 (replication) participants from the UK Biobank (UKB). Furthermore, for each urinary biomarker, we estimated its heritability; identified genetic associations that were likely mediated by expression or methylation quantitative trait loci (eQTLs or mQTLs); evaluated genetic correlations with several anthropometric, cardiovascular, glycemic, lipid, hematological, and kidney traits; and used a bioinformatic approach to pinpoint possible candidate genes.

Discovering genetic associations and identifying whether they are genetically correlated with other common traits can provide important etiological insights into the control of these biomarkers and ultimately renal function. Such knowledge may in turn improve the predictive use of the respective biomarker and point to new therapeutic strategies to prevent common diseases.

RESULTS

Association analyses

In our discovery GWASs, we identified a total of 52 genome-wide significant independent variants associated with at least 1 of the 4 urinary biomarker ratios: 21 for UACR (n = 218,759), 11 for UK/UCr (n = 218,435), 12 for UNa/UCr (n = 218,450), and 8 for UNa/UK (n = 217,996) (Table 1). Regional plots for the 52 discovered variants are shown in Supplementary Figures S1 to S4. We observed minor inflation in test statistics (λ = 1.08 for UACR; λ = 1.10 for UK/UCr; λ = 1.067 for UNa/UCr; and λ = 1.092 for UNa/UK), which is expected17 under polygenic inheritance in large samples (Supplementary Figure S5), also supported by the observation that all linkage disequilibrium score regression intercepts were close to 1 (Table 2; also see below).

Table 1 |.

Genetic loci associated with the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK) in the discovery and replication genome-wide association studiesa

 Discovery
 Replication
CHR POS SNP Function Nearest gene (bp) EA OA EAF β SE P β SE P
UACR
1 47961691 rs10157710 Intergenic FOXD2 (55,328) T C 0.8009 0.026 0.004 1.11 × 10−13 0.023 0.005 6.77 × 10−6
1 171435542 rs16864515 Intergenic PRRC2C (19,148) A C 0.0963 −0.029 0.005 1.37 × 10−9 −0.004 0.007 5.17 × 10−1
2 27598097 rs4665972 Intronic SNX17 C T 0.6071 −0.016 0.003 3.66 × 10−8 −0.020 0.004 2.01 × 10−6
2 203714973 rs10207567 Intronic ICA1L C G 0.8130 0.021 0.004 6.63 × 10−9 0.017 0.005 1.09 × 10−3
2 211540507 rs1047891 Missense CPS1 A C 0.3160 −0.019 0.003 3.23 × 10−10 −0.015 0.004 4.25 × 10−4
2 226985504 rs185291443 Intergenic NYAP2 (466,770) A C 0.0009 0.539 0.051 6.71 × 10−26 0.509 0.073 2.34 × 10−12
2 227459951 rs71431010 Intergenic MIR5702 (63,475) A G 0.0005 0.749 0.068 1.70 × 10−28 0.762 0.097 2.97 × 10−15
2 228511926 rs34823645 Intronic C2orf83 C T 0.0007 0.530 0.061 3.30 × 10−18 0.459 0.086 9.84 × 10−8
2 229160200 rs35311980 Intergenic PID1 (555,042) T C 0.0009 0.423 0.048 2.14 × 10−18 0.288 0.070 4.39 × 10−5
4 77019581 rs6829592 Intronic ART3 G A 0.2676 −0.018 0.003 3.88 × 10−8 −0.002 0.005 6.06 × 10−1
4 190769223 rs4109437 Intronic FRG1-DT A G 0.0382 0.060 0.007 8.55 × 10−16 0.053 0.011 4.90 × 10−7
7 17284577 rs4410790 Intronic AHR C T 0.6348 0.023 0.003 1.77 × 10−14 0.026 0.004 8.17 × 10−10
8 126500031 rs28601761 Intronic TRIB1 G C 0.4203 −0.017 0.003 9.58 × 10−9 −0.018 0.004 8.32 × 10−6
10 16932384 rs45551835 Missense CUBN A G 0.0144 0.198 0.012 4.47 × 10−62 0.170 0.017 2.49 × 10−24
10 16940846 rs45619139 Intronic CUBN G C 0.1016 0.063 0.005 1.09 × 10−40 0.050 0.007 5.51 × 10−14
10 16992011 rs141640975 Missense CUBN A G 0.0026 0.424 0.028 1.26 × 10−51 0.475 0.039 4.13 × 10−34
10 17005743 rs539606836 Intronic CUBN A G 0.0003 0.627 0.099 1.95 × 10−10 0.507 0.132 1.26 × 10−4
10 17025127 rs572663329 Intronic CUBN G C 0.0003 0.633 0.095 2.90 × 10−11 0.467 0.130 3.39 × 10−4
10 77893686 rs67339103 Intronic LRMDA A G 0.2131 0.021 0.003 2.31 × 10−9 0.017 0.005 6.96 × 10−4
15 45713801 rs60476496 Intergenic GATM (19,276) T C 0.2549 −0.021 0.003 3.01 × 10−10 −0.014 0.005 2.02 × 10−3
15 75027880 rs2472297 Intergenic CYP1A1 (9928) T C 0.2663 0.026 0.003 2.87 × 10−16 0.022 0.005 1.28 × 10−6
UK/UCr
1 72748567 rs66495454 Upstream NEGR1 (149) GTCCT G 0.3781 0.016 0.003 3.96 × 10−8 0.009 0.004 3.01 × 10−2
2 211540507 rs1047891 Exonic CPS1 A C 0.3151 −0.033 0.003 1.38 × 10−27 −0.029 0.004 1.20 × 10−11
3 27413566 rs4973766 Downstream SLC4A7 (647) T C 0.4558 −0.021 0.003 5.92 × 10−14 −0.023 0.004 1.06 × 10−8
10 75640062 10:75640062_GTTCG Intergenic CAMK2G (5718) G GTTCA 0.7645 −0.020 0.003 3.41 × 10−9 −0.016 0.005 6.45 × 10−4
14 33367658 rs10134619 Intergenic NPAS3 (36,481) C T 0.1425 −0.022 0.004 3.22 × 10−8 0.002 0.006 7.31 × 10−1
15 45693164 rs35335867 Intronic GATM CT C 0.2485 −0.024 0.003 9.39 × 10−14 −0.022 0.005 1.52 × 10−6
15 75027880 rs2472297 Intergenic CYP1A1 (9928) T C 0.2665 0.022 0.003 1.92 × 10−12 0.018 0.004 3.76 × 10−5
16 68301985 rs2863980 Intronic SLC7A6 C A 0.6261 0.016 0.003 4.22 × 10−8 0.007 0.004 1.05 × 10−1
17 35806016 rs562250921 Intronic TADA2A G C 0.0044 −0.122 0.022 1.71 × 10−8 −0.042 0.030 1.64 × 10−1
19 41352257 19:41352257_GTG Intronic CYP2A6 G GT 0.0850 −0.028 0.005 3.90 × 10−8 −0.014 0.007 5.28 × 10−2
19 46180414 rs34783010 Intronic GIPR T G 0.1933 0.026 0.004 1.26 × 10−13 0.025 0.005 4.74 × 10−7
UNa/UCr
2 27730940 rs1260326 Exonic GCKR C T 0.6082 −0.029 0.003 2.11 × 10−21 −0.0255 0.0043 2.80 × 10−9
2 211540507 rs1047891 Exonic CPS1 A C 0.3151 −0.028 0.003 3.21 × 10−18 −0.0245 0.0045 5.86 × 10−8
2 217666174 rs2541387 ncRNA_intronic FABP5P14 (4285) C A 0.5702 −0.019 0.003 1.05 × 10−10 −0.0092 0.0043 2.98 × 10−2
3 141154542 rs6440008 Intronic ZBTB38 C T 0.3870 −0.017 0.003 1.61 × 10−8 −0.0173 0.0043 6.67 × 10−5
3 168824936 rs545727790 Intronic MECOM A T 0.0003 −0.518 0.095 4.64 × 10−8 0.0815 0.1332 5.41 × 10−1
4 9948870 rs4276278 Intronic SLC2A9 C T 0.4327 0.017 0.003 9.33 × 10−9 0.0065 0.0042 1.22 × 10−1
5 52786075 rs12054747 Intergenic FST (3111) T C 0.2151 −0.020 0.004 4.07 × 10−8 −0.0030 0.0051 5.57 × 10−1
5 84251574 rs144795378 Intergenic C G 0.0079 0.104 0.017 2.13 × 10−9 −0.0252 0.0254 3.20 × 10−1
7 73034929 rs71556736 Intronic MLXIPL T C 0.1305 −0.024 0.004 2.44 × 10−8 −0.0197 0.0062 1.51 × 10−3
8 51947549 rs4873492 Intergenic SNTG1 (240,870) T C 0.1704 0.023 0.004 2.83 × 10−9 0.0189 0.0056 7.80 × 10−4
8 126482077 rs2954021 ncRNA_ntronic TRIB1 (31,429) G A 0.5058 −0.017 0.003 1.56 × 10−8 −0.0132 0.0042 1.75 × 10−3
15 45647522 rs113769380 Upstream RNU6–953P (438) GGGT G 0.2379 −0.023 0.004 1.32 × 10−10 −0.0135 0.0051 7.60 × 10−3
Una/UK
1 72825144 rs6699744 Intergenic RPL31P12
(57,631)		 T A 0.6118 −0.018 0.003 7.72 × 10−9 −0.0077 0.0043 7.15 × 10−2
2 27598097 rs4665972 Intronic SNX17 C T 0.6082 −0.026 0.003 1.40 × 10−17 −0.0203 0.0043 2.24 × 10−6
2 171958363 rs10208341 Intronic TLK1 A G 0.3597 0.019 0.003 6.78 × 10−10 0.0102 0.0043 1.88 × 10−2
3 35683104 rs9860326 Intronic ARPP21 G C 0.3287 0.018 0.003 1.36 × 10−8 0.0051 0.0044 2.53 × 10−1
3 141150026 rs7625643 Intronic ZBTB38 G A 0.4488 −0.019 0.003 6.62 × 10−10 −0.0136 0.0042 1.32 × 10−3
5 87712913 rs6452788 ncRNA_ntronic TMEM161B-AS1 A G 0.2361 −0.021 0.003 7.50 × 10−10 −0.0200 0.0049 4.92 × 10−5
14 69663339 14:69663339_CTTC ncRNA_ntronic EXD2 C CTT 0.4830 0.017 0.003 1.47 × 10−8 0.0081 0.0042 5.70 × 10−2
19 46182304 rs10423928 Intronic GIPR A T 0.1934 −0.023 0.004 2.93 × 10−10 −0.0183 0.0053 5.12 × 10−4
Open in a new tab

CHR, chromosome; EA, effect allele; EAF, effect allele frequency; OA, other allele; POS, position; SNP, single-nucleotide polymorphism.

a

Bonferroni significant threshold for replication: UACR, P = 3.57 × 10−3; UK/UCr, P = 4.55 × 10−3; UNa/UCr, P = 4.17 × 10−3; UNa/UK, P = 6.25 × 10−3.

Associations significant after replication are highlighted in bold.

Table 2 |.

Genome-wide heritability of the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK)

Variable h2g_obs h2g_obs_se h2g_int h2g_int_se
UACR 0.053 0.003 1.022 0.008
UK/UCr 0.053 0.003 1.021 0.008
UNa/UCr 0.042 0.003 1.024 0.009
UNa/UK 0.051 0.003 1.024 0.009
Open in a new tab

h2g, genome-wide heritability; int, intercept; obs, observed.

Of these 52 variants, 35 were significantly associated with a consistent direction in the replication data set. A conservative Bonferroni-corrected threshold of 2.38 × 10−3 for UACR, 4.55 × 10−3 for UK/UCr, 4.17 × 10−3 for UNa/UCr, and 6.25 × 10−3 for UNa/UK (adjusting for the different numbers of significant independent hits in the discovery GWASs) was used to identify significant replications. We identified 19 replicated variants for UACR (n = 109,530), 6 for UK/UCr (n = 109,178), 6 for UNa/UCr (n = 109,166), and 4 for UNa/UK (n = 108,942) (Table 1 and Figure 1).

Figure 1 |. Manhattan plot for the discovery genome-wide association study of the urinary albumin: creatinine ratio (a), urinary potassium: creatinine ratio (b), urinary sodium: creatinine ratio (c), and urinary sodium: potassium ratio (d).

Figure 1 |

Open in a new tab

The nearest genes for each locus that showed a significant P value in the replication stage were labeled. Negative log10-transformed P values for each single nucleotide polymorphism (y axis) are plotted by chromosomal position (x axis). The gray line represents the threshold for genome-wide statistically significant associations (P = 5 × 10−8). Red points represent significant hits.

When we performed the GWAS (discovery, n = 217,634; replication, n = 108,800) using albumin as a binary trait, we identified 10 variants to be robustly associated (3.33 × 10−3 in the replication stage after multiple testing correction) (Supplementary Table S1 and Supplementary Figure S6).

Of the 35 replicated lead variants, 22 are novel in relation to kidney function while 11 are located in loci previously associated with kidney function, including cubilin (CUBN),18 carbamoyl phosphate synthase 1 (CPS1),13 glycine amidino-transferase (GATM),19 and glucokinase regulator (GCKR).20 Of the 35 significantly replicated variants, there was some overlap between several of the 4 biomarkers. Specifically, rs34783010 (UK/UCr) and rs10423928 (UNa/UK) are located in the gastric inhibitory polypeptide receptor (GIPR) locus, rs1047891 (UACR, UK/UCr, and UNa/UCr) is located in the CPS1 locus, rs4665972 (UACR and UNa/UK) in the sorting nexin-17 (SNX17) locus, rs42472297 (UACR and UNa/UK) close to the cytochrome P450 family 1 subfamily A member 1 (CYP1A1) locus, and rs6440008 (UNa/UCr) and rs7625643 (UNa/UK) in the gene zinc finger and BTB domain containing 38 (ZBTB38) locus.

Phenotypic and genetic correlations

We detected positive phenotypic correlation between all pairs of urinary biomarkers (0.076–0.719), except between UK/UCr and UNa/UK (−0.402; P < 2.2 × 10−16) (Supplementary Table S2). We found evidence of strong direct genetic correlations between all pairs of urinary biomarkers (0.201–0.728), except between UACR and UNa/UK (−0.0491; P = 1.17 × 10−1). We also detected inverse genetic correlation between UK/UCr and UNa/UK (−0.523; P = 1.65 × 10−96) (Supplementary Table S3). UACR and binary albumin showed positive and significant genetic (0.335 ± 0.044; P = 3.21 × 10−14) and phenotypic (0.384 ± 0.001; P < 2.20 × 10−16) correlations. In general, we observed high consistency among the phenotypic and genetic correlations between all pairs of urinary biomarkers (Supplementary Figure S7).

Furthermore, we observed significant genetic correlations between the urinary biomarkers and several anthropometric, cardiovascular, glycemic, lipid, and kidney traits (Figure 2; Supplementary Table S3). Specifically, we identified a significant and positive genetic correlation of UNa/UK and several traits related to cardiometabolic disease, including body mass index, body fat, obesity, waist-to-hip ratio, coronary artery disease, BP, T2D, levels of fasting glucose and fasting insulin, as well as of UACR with systolic BP. In addition, we observed positive correlations of potassium and high-density lipoprotein cholesterol with estimated glomerular filtration rate based on serum creatinine as well as of sodium and UACR with estimated glomerular filtration rate based on serum creatinine. We detected a negative correlation between potassium and several traits related to cardiometabolic disease, including body mass index, body fat, obesity, waist-to-hip ratio, coronary artery disease, BP, levels of fasting glucose and fasting insulin, triglycerides, and CKD. We confirmed the negative correlation between sodium, obesity, and fasting insulin levels. In addition, we detected a negative correlation of height with UACR, UNa/UCr, and UNa/UK. Details on the source of data for the 29 traits analyzed are given in Supplementary Table S3. The heritability of the 4 urinary biomarkers was in the range of 0.042 ≤ h2g ≤ 0.053 (Table 2). The heritability of binary albumin was 0.017 ± 0.002. The heritability estimates for albumin are likely to be conservative, as a high proportion of the samples was below the lower limit of detection (6.7 mg/l; n = 228,262) and truncated to this level.

Figure 2 |. Genetic correlations of the urinary albumin: creatinine ratio (UACR) and urinary potassium: creatinine ratio (UK/UCr) (a) and of the urinary sodium: creatinine ratio (UNa/UCr) and urinary sodium: potassium ratio (UNa/UK) (b) with each other and other clinical traits.

Figure 2 |

Open in a new tab

Significant correlations after Bonferroni correction (P < 4.46 × 10−4) are highlighted with a red triangle. 2hrGluAdjBMI, 2-h glucose adjusted for body mass index; AF, atrial fibrillation; BMI, body mass index; CAD, coronary artery disease; CKD, chronic kidney disease; DBP, diastolic blood pressure; eGFRcrea, estimated glomerular filtration rate based on serum creatinine; eGFRcys, estimated glomerular filtration rate based on serum cystatin C; FG, fasting glucose; FI, fasting insulin; HbA1C, glycated hemoglobin; HDL, high-density lipoprotein; HR, heart rate; IS, ischemic stroke; LDL, low-density lipoprotein; SBP, systolic blood pressure; T2D, type 2 diabetes; TC, total cholesterol; TG, triglyceride; WHR, waist-to-hip ratio.

eQTLs

Of the 35 replicated lead variants, 10 showed evidence of significant eQTLs (Supplementary Table S4). The largest number of significant eQTLs was associated with gene expression in brain (52), artery (19), and skin (19). The most significant eQTLs in loci associated with UNa/UCr (rs1260326; P = 3.70 × 10−19) and UNa/UK (rs4665972; P = 6.60 × 10−22) were in relation to SNX17 in skeletal muscle. SNX17 has previously been implicated in myocardial infarction.21 The strongest eQTL detected in the GWASs of UACR and UK/UCr were for spermatogenesis associated 5 like 1 (SPATA5L1; rs60474696) and secretory carrier membrane protein 2 (SCAMP2; rs2472297) in cells transformed fibroblasts (P = 8.90 × 10−49) and skin (P = 2.90 × 10−9), respectively. The SPATA5L1 gene has previously been associated with CKD.14

Colocalization with eQTL and mQTL summary statistics

We tested the colocalization of GWAS lead variants with 92 and 1636 genome-wide significant cis-eQTLs and cis-mQTLs, respectively (P < 5 × 10−8), located within 1 Mb of the top replicated variants and controlled false discovery rate using the Benjamini-Yekutieli method (5% false discovery rate) in both analyses (Supplementary Table S5). To exclude summary data–based Mendelian randomization (SMR) results that may reflect pleiotropy, we performed the heterogeneity in dependent instruments (HEIDI) test, which considers the pattern of associations using all the single-nucleotide polymorphisms (SNPs) that are significantly associated with gene expression (eQTLs) in the cis region. After excluding heterogeneous variants (Pheterogeneity in dependent instrument < 0.05) and SNP in linkage disequilibrium (R2 > 0.8), we detected significant eQTL colocalizations for UACR, UNa/UCr, and UNa/UK in the nuclear receptor binding protein 1 (NRBP1) locus (rs11684134; SMR, P = 2.20 × 10−4, P = 5.69 × 10−5, and P = 8.24 × 10−5, respectively) (Table 3). No homogeneous and significant eQTL colocalization signals were detected for urinary potassium. We observed 19 significant independent mQTL probe colocalizations for UACR, 15 for UK/UCr, 7 for UNa/UCr, and 8 for UNa/UK (Table 3).

Table 3|.

Significant independent expression quantitative trait locus (eQTL) and methylation quantitative trait locus (mQTL) probe colocalizations for the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK)

ProbeID ProbeChr Gene Probe_bp topSNP A1 A2 Freq b_SMR se_SMR P_SMR P_HEIDI
eQTL
UACR
ILMN_1670096   2 NRBP1 27664880 rs11684134 G A 0.417 0.110 0.030 2.20 × 10−4 0.311
UNa/UCr
ILMN_1670096   2 NRBP1 27664880 rs11684134 G A 0.417 0.136 0.034 5.69 × 10−5 1.40 × 10−1
UNa/UK
ILMN_1670096
mQTL 		  2 NRBP1 27664880 rs11684134 G A 0.417 0.129 0.033 8.24 × 10−5 1.64 × 10−1
UACR
cg14242246   2 CAD 27434262 rs62130680 C T 0.145 −0.012 0.003 1.19 × 10−5 2.41 × 10−1
cg13521797   2 ICA1L 203737787 rs74675536 A G 0.129 −0.054 0.012 4.66 × 10−6 7.47 × 10−2
cg05158606   2 ICA1L 204194156 rs1971819 G C 0.180 0.079 0.017 6.68 × 10−6 1.24 × 10−1
cg20080878   8 126479070 rs6999569 G A 0.453 −0.044 0.009 3.65 × 10−7 6.35 × 10−2
cg20381115 15 GATM 45671148 rs2172874 G T 0.276 −0.027 0.004 8.15 × 10−11 1.44 × 10−1
cg05926586 15 CCDC33 74592786 rs12904134 G A 0.529 0.019 0.003 2.85 × 10−10 5.20 × 10−2
cg06369322 15 CCDC33 74607544 rs2415239 C T 0.353 0.072 0.018 5.47 × 10−5 4.04 × 10−1
cg05454635 15 74657911 rs8032336 T A 0.366 −0.054 0.013 1.65 × 10−5 2.42 × 10−1
cg24482024 15 74660566 rs35807116 C T 0.431 −0.045 0.008 4.79 × 10−8 6.64 × 10−2
cg19144497 15 74836094 rs9672559 A G 0.116 −0.056 0.016 5.99 × 10−4 1.64 × 10−1
cg20004910 15 EDC3 75018852 rs7161903 G A 0.150 0.045 0.010 1.37 × 10−5 3.97 × 10−1
cg01359532 15 75045126 rs11072508 C T 0.395 −0.053 0.008 1.18 × 10−10 1.28 × 10−1
cg14664628 15 SCAMP2 75095509 rs3765066 G A 0.379 −0.055 0.011 2.25 × 10−7 7.03 × 10−2
cg10253484 15 75165896 rs11857376 A G 0.546 −0.035 0.005 1.54 × 10−13 6.58 × 10−2
cg27066162 15 75199472 rs7497026 T C 0.487 −0.078 0.018 2.06 × 10−5 7.93 × 10−2
cg15611336 15 SCAMP5 75248496 rs35103257 G T 0.545 0.074 0.016 2.44 × 10−6 3.10 × 10−1
cg21236593 15 75287447 rs62027288 C T 0.420 0.067 0.016 1.92 × 10−5 6.42 × 10−2
cg01701819 15 SCAMP5 75287824 rs12911421 C A 0.239 0.013 0.004 3.68 × 10−4 2.58 × 10−1
cg24222504 15 SNUPN 75918608 rs56178382 T A 0.226 −0.076 0.018 2.14 × 10−5 9.52 × 10−2
UK/UCr
cg06730238   3 NEK10 27131936 rs7618713 C T 0.209 0.056 0.016 3.48 × 10−4 2.10 × 10−1
cg16540259 10 NDST2 75572220 rs13013 C A 0.376 −0.077 0.019 3.70 × 10−5 7.41 × 10−1
cg00564723 10 CAMK2G 75632066 rs2688626 G A 0.225 0.035 0.006 1.03 × 10−9 2.82 × 10−1
cg06521280 10 PLAU 75671378 rs2227560 A G 0.228 0.062 0.015 2.75 × 10−5 1.44 × 10−1
cg18435832 15 GATM 45671155 rs2172874 G T 0.276 −0.032 0.004 1.90 × 10−15 5.48 × 10−2
cg05019826 15 74891207 rs2472297 T C 0.215 0.079 0.015 8.12 × 10−8 2.34 × 10−1
cg05549655 15 75019143 rs12903896 C T 0.405 −0.057 0.015 1.81 × 10−4 7.67 × 10−1
cg12101586 15 75019203 rs11072508 C T 0.395 −0.062 0.017 2.45 × 10−4 7.52 × 10−1
cg14664628 15 SCAMP2 75095509 rs3765066 G A 0.379 −0.035 0.009 1.92 × 10−4 2.49 × 10−1
cg04877966 15 ULK3 75135169 rs9210 T C 0.310 0.019 0.005 4.93 × 10−5 9.10 × 10−2
cg10253484 15 75165896 rs11857376 A G 0.546 −0.020 0.004 7.28 × 10−6 3.88 × 10−1
cg15611336 15 SCAMP5 75248496 rs35103257 G T 0.545 0.051 0.013 1.27 × 10−4 3.05 × 10−1
cg17716663 19 ERCC1 45978434 rs8106229 G T 0.406 0.017 0.005 4.21 × 10−4 1.29 × 10−1
cg19822309 19 GIPR 46181546 rs10423928 A T 0.214 −0.032 0.004 3.34 × 10−16 2.87 × 10−1
cg15455864 19 SYMPK 46319398 rs1132645 C G 0.318 −0.040 0.011 1.58 × 10−4 4.78 × 10−1
UNa/UCr
cg25309888   2 27988724 rs12477908 A C 0.282 −0.032 0.008 9.01 × 10−5 1.23 × 10−1
cg01879723   3 ZBTB38 141119262 rs34169305 T C 0.307 0.068 0.017 6.76 × 10−5 5.14 × 10−1
cg13411656   3 ZBTB38 141121016 rs4683605 A C 0.425 0.054 0.010 1.01 × 10−7 2.39 × 10−1
cg03060555   3 ZBTB38 141131189 rs13069734 A G 0.310 0.109 0.023 1.61 × 10−6 8.60 × 10−2
cg08082507   7 GTF2IRD2 72685715 rs707395 T C 0.185 0.051 0.013 4.35 × 10−5 7.12 × 10−2
cg14087351   7 MLXIPL 73037990 rs3812316 G C 0.116 0.024 0.004 3.83 × 10−9 2.96 × 10−1
cg20080878   8 126479070 rs6999569 G A 0.453 −0.047 0.009 1.54 × 10−7 2.13 × 10−1
UNa/UK
cg12404044   2 26911734 rs1275985 C T 0.407 −0.012 0.003 3.07 × 10−4 1.84 × 10−1
cg27084654   3 ZBTB38 141118615 rs13098180 A T 0.383 0.050 0.011 6.14 × 10−6 1.85 × 10−1
cg01879723   3 ZBTB38 141119262 rs34169305 T C 0.307 0.069 0.017 5.99 × 10−5 3.18 × 10−1
cg13411656   3 ZBTB38 141121016 rs4683605 A C 0.425 0.055 0.010 4.87 × 10−8 2.88 × 10−1
Cg03060555   3 ZBTB3S 141131189 rs13069734 A G 0.310 0.109 0.023 1.43 × 10−6 7.46 × 10−2
cg10604528   5 87437959 rs7448105 C T 0.197 −0.056 0.012 3.01 × 10−6 1.82 × 10−1
Cg09277709 19 SIX5 46224285 rs2341097 T C 0.323 −0.054 0.016 6.90 × 10−4 8.33 × 10−2
cg24278423 19 SYMPK 46367240 rs11083783 C T 0.083 −0.040 0.010 1.18 × 10−4 5.75 × 10−1
Open in a new tab

A1, effect allele; A2, other allele; b, beta; bp, base position; Chr, chromosome; Freq, minor allele frequency; HEIDI, heterogeneity in dependent instruments; SMR, summary data-based Mendelian randomization; SNP, single-nucleotide polymorphism.

The strongest mQTL colocalization detected for UACR was for a probe on chromosome 15 (lead SNP rs11857376; SMR, P = 1.54 × 10−13). We detected consistent signals in GWAS and mQTL colocalization analyses for islet cell autoantigen 1 like (ICA1L) on chromosome 2 (rs74675536 and rs1971819; SMR, P ≤ 6.68 × 10−6) and for the GATM locus on chromosome 15 (rs2172874; SMR, P = 8.15 × 10−11) (Table 3).

The strongest mQTL colocalization detected for potassium was for GIPR (lead SNP rs10423928; SMR, P = 3.34 × 10−16) on chromosome 19. Furthermore, we detected consistent signals in GWAS and mQTL colocalization analyses for GATM (lead SNP rs2172874; SMR, P = 1.90 × 10−15) on chromosome 15 and for calcium/calmodulin-dependent protein kinase II gamma (CAMK2G) (lead SNP rs2688626; SMR, P = 1.03 × 10−9) (Table 3).

We detected significant independent mQTL colocalizations for UNa/UCr-related variants and methylation in ZBTB38 on chromosome 3 (lead SNPs rs34169305, rs4683605, and rs13069734; SMR, P ≤ 6.76 × 10−5) and consistent signals in GWAS and mQTL colocalization analyses for protein-coding gene MLX interacting protein like (MLXIPL) on chromosome 7 (rs3812316; SMR, P = 3.83 × 10−9) (Table 3).

We detected consistent signals in GWAS and mQTL colocalizations for UNa/UK in ZBTB38 on chromosome 3 for 4 significant independent probes (lead SNPs rs13098180, rs34169305, rs4683605, and rs13069734; SMR, P ≤ 5.99 × 10−5) (Table 3).

In addition, we detected shared eQTL and mQTL colocalizations for different urinary biomarkers. UACR, UNa/UCr, and UNa/UK shared an eQTL colocalization on chromosome 2 with the same lead variant rs11684134 (NRBP1), indicating that this is a plausible variant implicated in these traits. UACR and UK/UCr shared 3 mQTL colocalizations on chromosome 15 on the GATM, SCAMP5, and SCAMP2 genes with the same lead variants (rs2172874, rs35103257, and rs3765066, respectively). UNa/UCr and UNa/UK also shared 3 mQTL colocalizations on chromosome 3 with the same lead variants (rs34169305, rs4683605, and rs13069734) on the gene ZBTB38, and UK/UCr (rs1132645) and UNa/UK (rs1108378) shared an mQTL colocalization on the symplekin (SYMPK) gene on chromosome 19.

DISCUSSION

Principal findings

We achieved 3 broad goals in this study of up to 337,536 unrelated individuals of European ancestry from the UK Biobank. We (i) established genetic determinants of UACR, UK/UCr, UNa/UCr, and UNa/UK; (ii) evaluated genetic correlation between these biomarkers and a number of clinical traits; and (iii) explored the associated loci using colocalization analyses based on expression and methylation data. Our main findings are severalfold. First, we report a total of 22 novel loci associated with ≥1 of the 4 urinary biomarkers, providing new leads on biological processes involved in regulating UACR, UK/UCr, UNa/UCr, and UNa/UK. Second, our analyses indicate low heritability but high pairwise genetic correlations between the 4 urinary biomarkers as well as significant genetic correlations with several traits related to CKD, CVD, and T2D. Third, we identify 1 and 49 independent colocalization events of GWAS data with blood gene expression and DNA methylation, respectively, providing evidence of potential genes implicated in these loci. As an example of the latter, we identified GIPR, a potential drug target in the treatment of obesity-associated metabolic disorders, as likely involved in mechanisms affecting urinary potassium excretion, and NRBP1, a gene associated with gout, as implicated in sodium and albumin excretion and/or metabolism.

Comparison with prior literature

This study is the first GWAS of UK/UCr, UNa/UCr, and UNa/UK to our knowledge, while genetic determinants of albuminuria have already been explored in previous studies.8,18 In this study, we replicated the well-known association of the CUBN locus with albuminuria and extend prior GWAS of albumin by highlighting several novel pathways influencing this trait. For example, we highlighted 2 variants—rs4410790 and rsrs2472297—located in the aryl hydrocarbon receptor (AHR) and close to the CYP1A1 gene, respectively, previously found associated with coffee consumption.22 When we performed the analyses using albumin as a binary trait, we replicated most of the associations identified in the UACR GWAS, most prominently for the CUBN locus (Supplementary Table S6).

In the UK/UCr GWAS, we replicated the known association of variants in the CPS1 locus (rs1047891; P = 1.38 × 10−27), previously associated with glomerular filtration rate (serum creatinine),13 and with GATM (rs35335867; P = 9.39 × 10−14), a gene that encodes a mitochondrial enzyme previously associated with serum creatinine.14

Several replicated variants for UNa/UCr (rs1260326 on GCKR and rs2954021 near Tribbles Pseudokinase 1 (TRIB1); P = 2.11 × 10−21 and P = 1.56 × 10−8, respectively) and for UACR and UNa/UK (rs4665972 on SNX17; P = 3.60 × 10−8 and P = 1.40 × 10−17, respectively) showed previous association with triglyceride levels,23 while another variant (rs10423928; P = 2.93 × 10−10) is located in the GIPR gene, as discussed in detail below, and has been previously found to be associated with T2D24 and obesity.25 The 2 variants on GCKR and SNX17 genes (rs1260326 and rs4665972, respectively) are not independent and show a strong linkage disequilibrium (R2 = 0.92), so they are likely to represent the same underlying signal. Details of all these previous studies and their comparisons with the UKB associations can be found in Supplementary Table S7.

Previous studies have shown a strong association of sodium and potassium intake with BP16 and CVD,5 but there is a lack of studies of the genetic underpinning of these biomarkers, except for albumin. Indeed, a recent study supported the evidence that albuminuria could increase the risk of CVD with an increase in BP.26

Phenotypic and genetic correlations

We found evidence of high positive correlations (both genetic and phenotypic) between all pairs of urinary biomarkers analyzed, except between UK/UCr and UNa/UK. We confirmed a negative association of BP with potassium and a positive association of BP with UNa/UK. We also confirmed significant positive genetic correlations between UNa/UK and several anthropometric measurements (body mass index, body fat, height, and waist-to-hip ratio), coronary artery disease, T2D, and levels of fasting insulin and fasting glucose. Some of the correlations detected might be caused by associations with creatinine (that was used to normalize the other biomarkers); for example, the direct genetic correlations of UACR, UK/UCr, and UNa/UCr with estimated glomerular filtration rate are likely caused by the correlation of blood and urine creatinine.

We report heritability estimates of the 4 urinary biomarkers in the range of 4% to 5%, which are lower than those of many complex traits, but consistent with prior studies of kidney-related serum biomarkers. Specifically, estimates for serum albumin: creatinine ratio, potassium, and sodium have been reported to be in the range of 13% to 19%, lowest for potassium and highest for creatinine27 when using family-based designs.

Novel biology

We identified NRBP1, previously associated with gout, as significantly associated with UACR, UNa/UCr, and UNa/UK in our eQTL colocalization analyses. Gout is a complex disorder caused by the deposition of monosodium urate crystals within joints and other tissues. Central to the development of gout is elevated serum uric acid concentrations.28 A previous study demonstrated that hypomethylation of its promoter region, B1, is associated with increased gene expression both in vitro and in vivo. Moreover, gout-associated increased NRBP1 expression is regulated through methylation-dependent transcription factor AP-2 alpha binding to the B1 region, which might be involved in the pathogenesis of gout.29 Furthermore, another locus associated with urinary sodium, MLXIPL, has been associated with reduced serum uric acid concentrations and consequently a lower risk of gout, but also with increased coffee consumption.30 Multiple studies have reported that increased coffee consumption is associated with reduced serum uric acid concentrations31 and lower risk of developing gout.32,33 This association has been attributed to several potential mechanisms including improved insulin resistance34 and caffeine-mediated inhibition of xanthine oxidase.35 A recent study showed a potential causal role of uric acid transporters in improving renal function,36 whereas other Mendelian randomization studies have demonstrated that gout, the clinical manifestation of hyperuricemia, is causally associated with body mass index37 and CVD,38 but not with T2D39 and BP.37 Regarding the association with BP, another study reported that sodium intake was significantly associated with uric acid and BP in opposite directions.40 The increase in sodium intake was associated with lower uric acid and higher BP. These findings suggest that uric acid is unlikely to mediate changes in BP in the context of dietary sodium intake. This may be explained by the effects of sodium intake on glomerular filtration rate and excretion or absorption of urate. Previous studies have shown that reabsorption of sodium and urate accompany one another41 at different sites in the nephron.42 Thus, it is possible that decreased renal reabsorption of sodium from excess sodium intake contributes to a decrease in urate reabsorption.40

In addition, we observed colocalization of the GWAS signal for UK/UCr with mQTL probes for CAMK2G. The protein encoded by this gene, CAMK2, is a multifunctional serine/threonine kinase with numerous roles in human physiology. Dysfunction in CAMK2-based signaling has been linked with a host of cardiovascular phenotypes including heart failure and arrhythmia. CAMK2 levels are elevated in human and animal disease models of heart disease, and this protein has been suggested as a novel therapeutic target for cardiac arrhythmia.43 In addition, another previous study demonstrated that dysregulated embryonic regulator of G protein signaling expression and oxidative activation of CAMK2G may potentially contribute to congenital heart defects.44 Normal blood levels of potassium are critical for maintaining normal cardiac rhythm. As K+ currents control the repolarization process of the cardiomyocyte action potential, the K+ channel function determines membrane potential and refractoriness of the myocardium. Both low blood potassium levels (hypokalemia) and high blood potassium levels (hyperkalemia) can lead to abnormal heart rhythms.45

We observed colocalization of the GWAS signal for UACR and UK/UCr in the GATM locus with mQTL signals. The GATM locus is not novel in terms of association with renal function, as previous studies have shown its association with CKD and reduced glomerular filtration rate.13,14 GATM, which stands for glycine amidinotransferase, encodes a rate-limiting enzyme involved in creatinine biosynthesis and has been suggested to act as a functional link between statin-mediated lowering of cholesterol and susceptibility to statin-induced myopathy.46 In addition, a previous study showed a significant association of SNPs in the GATM locus with plasma and urine creatinine, but not with cystatin C in plasma, another biomarker of renal function.14 This association may indicate that the regulatory variants in this locus influence creatinine production, rather than creatinine excretion.

Likewise, variants in the GIPR locus significantly associated with UK/UCr in our GWAS and mQTL colocalization have been previously associated with T2D, obesity, and related phenotypes.24,25 The glucose-dependent insulinotropic peptide has a central role in glucose homeostasis through its amplification of insulin secretion. Incretin hormones such as glucose-dependent insulinotropic peptide act to promote efficient uptake and storage of energy after food ingestion and have become important players for glucose homeostasis in pancreatic and extrapancreatic tissue. A recent study demonstrated that mice with selective ablation of GIPR in β cells exhibited lower levels of meal-stimulated insulin secretion, decreased expansion of adipose tissue mass, and preservation of insulin sensitivity as compared with controls.47 Hence, GIPR represents a potential therapeutic in the treatment of diabetes. To test whether potassium was directly related to glucose metabolism, previous hyperglycemic clamp studies were performed in healthy volunteers in whom hypokalemia was induced. One of these studies demonstrated that experimentally induced hypokalemia led to impaired glucose tolerance by reducing insulin secretion in response to glucose load. Potassium depletion was associated with a decrease in pancreatic β-cell sensitivity to hyperglycemia with a reduction in insulin release.48 In another study, hypokalemia was induced through the use of thiazide diuretics. In a first stage, participants demonstrated subsequent impaired glucose tolerance, but after potassium supplementation was given, the defects in insulin release in response to glucose loads were corrected. This further implicates hypokalemia itself as being the causative agent of glucose abnormality.49

Strengths and limitations

Strengths of the present study include the very large sample size with both genetic profiling and phenotypic data, which enabled us to detect a large number of genetic associations, the use of state-of-the-art methods to validate our results including a conservative analytical framework with strict multiple testing correction, and various downstream analyses to highlight important biology and potential genes related to urinary biomarkers. Our study is the most comprehensive study of the genetics of urinary biomarkers to date, combining GWASs, genetic correlation, and eQTL and mQTL colocalization analyses.

We also acknowledge some limitations. First, we did not replicate our findings in an external study sample because of the unavailability of cohorts with these phenotypes of comparable size. For this reason, we used a strict discovery-replication strategy. Second, participants included in our analyses were restricted to middle-aged and elderly individuals of European ancestry. Hence, the generalizability of our results to other age groups and ethnicities is unknown. Third, we analyzed a spot measurement of the electrolyte: creatinine ratio and, because of the unavailability in the UKB, we did not analyze other useful measures to provide a complete picture of renal function, such as 24-hour urine collections, food frequency questionnaires, blood biomarker data, urinary glucose, and/or urea.

CONCLUSIONS

We report 22 novel genome-wide significant loci associated with UACR, UK/UCr, UNa/UCr, and UNa/UK based on analyses in up to 337,536 individuals from the general population, confirming several known associations and providing new insights into the genetic basis of renal traits and their connection to chronic diseases. We detected strong genetic correlations between urinary biomarkers and significant genetic correlations of urinary biomarkers with several cardiometabolic traits. We also highlight several most plausible genes associated with these urinary biomarkers, such as GIPR, a potential diabetes drug target, as implicated in the genetic control of urinary potassium, and NRBP1, a locus associated with gout, as implicated in urinary sodium and albumin excretion.

METHODS

Phenotype

We analyzed the urinary biomarkers normalizing for creatinine (UACR, UK/UCr, and UNa/UCr) as well as UNa/UK. In addition, albumin was dichotomized for secondary analyses into a binary trait (≤30 mg/l = 0 and >30 mg/l = 1). The baseline characteristics of UKB participants who were included in the analyses (n = 337,537) are summarized in Table 4. The distribution of the 4 urinary biomarkers before normalization is shown in Supplementary Figure S8.

Table 4 |.

Baseline characteristics of UK Biobank participants included in the present study (discovery, n = 225,024; replication, n = 112,512)

Variable Discovery Replication
Sex: female 120,921 (53.74) 60,347 (53.64)
Age (yr) 56.88 ± 7.99 56.87 ± 7.98
T2D 11,068 (4.92) 5,604 (4.98)
Smoking: current 22,735 (10.10) 11,249 (10.00)
SBP (mm Hg) 140 ± 20 140 ± 20
DBP (mm Hg) 82 ± 11 82 ± 11
BMI (kg/m2) 27.39 ± 4.75 27.42 ± 4.76
WHR 0.87 ± 0.09 0.87 ± 0.09
Body fat (%) 31.34 ± 8.53 31.37 ± 8.51
Weekly alcohol intake (g) 134.32 ± 159.45 134.68 ± 160.51
UACR 17.526 ± 76.969 17.442 ± 74.064
UK/UCr 8.476 ± 3.890 8.444 ± 3.993
UNa/UCr 10.602 ± 6.672 10.619 ± 7.736
UNa/UK 1.400 ± 0.873 1.406 ± 0.874
Open in a new tab

BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure; T2D, type 2 diabetes; UK/UCr, urinary potassium: creatinine ratio; UACR, urinary albumin: creatinine ratio; UNa/UCr, urinary sodium: creatinine ratio; UNa/UK, urinary sodium: potassium ratio; WHR, waist-to-hip ratio.

Data are mean ± SD or n (%).

Genetic association analysis

We performed a discovery GWAS (random two-third sample from eligible individuals; n = 225,024) and a replication GWAS (remaining one-third of eligible individuals; n = 112,512). In addition, to maximize power for genetic correlation and colocalization analyses, we performed a meta-analysis of the discovery and replication samples.

Colocalization with eQTL and mQTL summary statistics

We performed SMR50 with blood cis-eQTL and cis-mQTL data to evaluate the evidence for colocalization between all genome-wide significant independent variants and white blood cell gene expression or methylation signals.

Supplementary Material

Supplementary Materials. Supplementary Methods and References.

Figure S1. Regional association and linkage disequilibrium plots for 21 genome-wide significant loci for the urinary albumin: creatinine ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S2. Regional association and linkage disequilibrium plots for 11 genome-wide significant loci for the urinary potassium: creatinine ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S3. Regional association and linkage disequilibrium plots for 12 genome-wide significant loci for the urinary sodium: creatinine ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S4. Regional association and linkage disequilibrium plots for 8 genome-wide significant loci for the urinary sodium: potassium ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S5. Q–Q plots for genetic associations for the urinary albumin: creatinine ratio (A), urinary potassium: creatinine ratio (B), urinary sodium: creatinine ratio (C), and urinary sodium: potassium ratio (D) in the discovery analysis.

Figure S6. Manhattan plot for the discovery genome-wide association study of binary albuminuria. The nearest genes for each locus that showed a significant P value in the replication were labeled. Negative log10-transformed P values for each single-nucleotide polymorphism (y axis) are plotted by chromosomal position (x axis). The gray line represents the threshold for genome-wide statistically significant associations (P = 5 × 10−8). Red points represent significant hits.

Figure S7. Scatter plot of the genetic and phenotypic correlations of the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK).

Figure S8. Histogram of the frequencies of the urinary albumin: creatinine ratio (UACR) (A), urinary potassium: creatinine ratio (UK/UCr) (B), urinary sodium: creatinine ratio (UNa/UCr) (C), and urinary sodium: potassium ratio (UNa/UK) (D) before normalization.

Table S1. Genetic loci associated with urinary albuminuria in the discovery and replication genome-wide association studies.

Table S2. Phenotypic correlation between urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK).

Table S3. Genetic correlation of the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK) with other traits.

Table S4. Significant expression quantitative trait loci for replicated variants in urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK) genome-wide association studies.

Table S5. Significant expression quantitative trait locus (eQTL) and methylation quantitative trait locus (mQTL) probe colocalizations for the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK).

Table S6. Genetic loci associated with the urinary albumin: creatinine ratio (UACR) in relation with the genome-wide association study of binary albuminuria.

Table S7. Previous studies and their comparisons with the UK Biobank associations.

ACKNOWLEDGMENTS

This research has been conducted using the UK Biobank resource under application number 13721. The research was performed with support from the National Institutes of Health (grant nos. 1R01HL135313–01 and 1R01DK106236–01A1) and the Stanford Diabetes Research Center (award no. P30DK116074). DZ was supported by the AHA Postdoctoral Fellowship (19POST34370115).

Footnotes

Data availability

Data sets related to this article are available at UKB resource (https://www.ukbiobank.ac.uk/). GWAS summary statistics for the 4 urinary biomarkers performed using all the European unrelated samples in the UKB are available at GRASP resource (https://grasp.nhlbi.nih.gov/FullResults.aspx).

DISCLOSURE

EI is a scientific advisor for Precision Wellness and has received consulting fees from Olink Proteomics for work unrelated to the present project. All the other authors declared no competing interests.

Supplementary material is linked to the online version of the paper at www.kidney-international.org.

REFERENCES

  • 1.Pisitkun T, Johnstone R, Knepper MA. Discovery of urinary biomarkers. Mol Cell Proteomics. 2006;5:1760–1771. [DOI] [PubMed] [Google Scholar]
  • 2.Rossing K, Mischak H, Dakna M, et al. Urinary proteomics in diabetes and CKD. J Am Soc Nephrol. 2008;19:1283–1290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Tesch GH. Review: Serum and urine biomarkers of kidney disease: a pathophysiological perspective. Nephrology (Carlton). 2010;15:609–616. [DOI] [PubMed] [Google Scholar]
  • 4.Ingelsson E, Sundstrom J, Lind L, et al. Low-grade albuminuria and the incidence of heart failure in a community-based cohort of elderly men. Eur Heart J. 2007;28:1739–1745. [DOI] [PubMed] [Google Scholar]
  • 5.Cook NR, Obarzanek E, Cutler JA, et al. Joint effects of sodium and potassium intake on subsequent cardiovascular disease: the Trials of Hypertension Prevention follow-up study. Arch Intern Med. 2009;169:32–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Jackson SL, Cogswell ME, Zhao L, et al. Association between urinary sodium and potassium excretion and blood pressure among adults in the United States: National Health and Nutrition Examination Survey, 2014. Circulation. 2018;137:237–246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Ekinci EI, Cheong KY, Dobson M, et al. High sodium and low potassium intake in patients with type 2 diabetes. Diabet Med. 2010;27:1401–1408. [DOI] [PubMed] [Google Scholar]
  • 8.Teumer A, Tin A, Sorice R, et al. Genome-wide association studies identify genetic loci associated with albuminuria in diabetes. Diabetes. 2016;65: 803–817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Wu J, Gao Y. Physiological conditions can be reflected in human urine proteome and metabolome. Expert Rev Proteomics. 2015;12:623–636. [DOI] [PubMed] [Google Scholar]
  • 10.Mente A, O’Donnell MJ, Rangarajan S, et al. Association of urinary sodium and potassium excretion with blood pressure. N Engl J Med. 2014;371:601–611. [DOI] [PubMed] [Google Scholar]
  • 11.Rein P, Saely CH, Vonbank A, et al. Impact of the albumin to creatinine ratio and the coronary artery state on vascular events. Am J Cardiol. 2014;113:1616–1620. [DOI] [PubMed] [Google Scholar]
  • 12.Pattaro C Genome-wide association studies of albuminuria: towards genetic stratification in diabetes? J Nephrol. 2018;31:475–487. [DOI] [PubMed] [Google Scholar]
  • 13.Pattaro C, Teumer A, Gorski M, et al. Genetic associations at 53 loci highlight cell types and biological pathways relevant for kidney function. Nat Commun. 2016;7:10023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kottgen A, Pattaro C, Boger CA, et al. New loci associated with kidney function and chronic kidney disease. Nat Genet. 2010;42:376–384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Weir MR. Microalbuminuria and cardiovascular disease. Clin J Am Soc Nephrol. 2007;2:581–590. [DOI] [PubMed] [Google Scholar]
  • 16.Binia A, Jaeger J, Hu Y, et al. Daily potassium intake and sodium-to-potassium ratio in the reduction of blood pressure: a meta-analysis of randomized controlled trials. J Hypertens. 2015;33:1509–1520. [DOI] [PubMed] [Google Scholar]
  • 17.Wood AR, Esko T, Yang J, et al. Defining the role of common variation in the genomic and biological architecture of adult human height. Nat Genet. 2014;46:1173–1186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Boger CA, Chen MH, Tin A, et al. CUBN is a gene locus for albuminuria. J Am Soc Nephrol. 2011;22:555–570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Reichold M, Klootwijk ED, Reinders J, et al. Glycine amidinotransferase (GATM), renal Fanconi syndrome, and kidney failure. J Am Soc Nephrol. 2018;29:1849–1858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Gorski M, van der Most PJ, Teumer A, et al. 1000 Genomes-based meta-analysis identifies 10 novel loci for kidney function. Sci Rep. 2017;7:45040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Zhao D, Li X, Liang H, et al. SNX17 produces anti-arrhythmic effects by preserving functional SERCA2a protein in myocardial infarction. Int J Cardiol. 2018;272:298–305. [DOI] [PubMed] [Google Scholar]
  • 22.Josse AR, Da Costa LA, Campos H, El-Sohemy A. Associations between polymorphisms in the AHR and CYP1A1-CYP1A2 gene regions and habitual caffeine consumption. Am J Clin Nutr. 2012;96:665–671. [DOI] [PubMed] [Google Scholar]
  • 23.Coram MA, Duan Q, Hoffmann TJ, et al. Genome-wide characterization of shared and distinct genetic components that influence blood lipid levels in ethnically diverse human populations. Am J Hum Genet. 2013;92:904–916. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Saxena R, Hivert MF, Langenberg C, et al. Genetic variation in GIPR influences the glucose and insulin responses to an oral glucose challenge. Nat Genet. 2010;42:142–148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Berndt SI, Gustafsson S, Magi R, et al. Genome-wide meta-analysis identifies 11 new loci for anthropometric traits and provides insights into genetic architecture. Nat Genet. 2013;45:501–512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Haas ME, Aragam KG, Emdin CA, et al. Genetic association of albuminuria with cardiometabolic disease and blood pressure. Am J Hum Genet. 2018;103:461–473. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Moulin F, Ponte B, Pruijm M, et al. 5C.09: heritability of renal function parameters and electrolyte levels in the Swiss population. J Hypertens. 2015;33(suppl 1):e70. [Google Scholar]
  • 28.Dalbeth N, Merriman TR, Stamp LK. Gout. Lancet. 2016;388:2039–2052. [DOI] [PubMed] [Google Scholar]
  • 29.Zhu Z, Meng W, Liu P, et al. DNA hypomethylation of a transcription factor binding site within the promoter of a gout risk gene NRBP1 upregulates its expression by inhibition of TFAP2A binding. Clin Epigenetics. 2017;9:99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Hutton J, Fatima T, Major TJ, et al. Mediation analysis to understand genetic relationships between habitual coffee intake and gout. Arthritis Res Ther. 2018;20:135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Park KY, Kim HJ, Ahn HS, et al. Effects of coffee consumption on serum uric acid: systematic review and meta-analysis. Semin Arthritis Rheum. 2016;45:580–586. [DOI] [PubMed] [Google Scholar]
  • 32.Choi HK, Willett W, Curhan G. Coffee consumption and risk of incident gout in men: a prospective study. Arthritis Rheum. 2007;56:2049–2055. [DOI] [PubMed] [Google Scholar]
  • 33.Choi HK, Curhan G. Coffee consumption and risk of incident gout in women: the Nurses’ Health Study. Am J Clin Nutr. 2010;92:922–927. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Carlstrom M, Larsson SC. Coffee consumption and reduced risk of developing type 2 diabetes: a systematic review with meta-analysis. Nutr Rev. 2018;76:395–417. [DOI] [PubMed] [Google Scholar]
  • 35.Kela U, Vijayvargiya R, Trivedi CP. Inhibitory effects of methylxanthines on the activity of xanthine oxidase. Life Sci. 1980;27:2109–2119. [DOI] [PubMed] [Google Scholar]
  • 36.Hughes K, Flynn T, de Zoysa J, et al. Mendelian randomization analysis associates increased serum urate, due to genetic variation in uric acid transporters, with improved renal function. Kidney Int. 2014;85:344–351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Palmer TM, Nordestgaard BG, Benn M, et al. Association of plasma uric acid with ischaemic heart disease and blood pressure: Mendelian randomisation analysis of two large cohorts. BMJ. 2013;347:f4262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Kleber ME, Delgado G, Grammer TB, et al. Uric acid and cardiovascular events: a mendelian randomization study. J Am Soc Nephrol. 2015;26:2831–2838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Sluijs I, Holmes MV, van der Schouw YT, et al. A Mendelian randomization study of circulating uric acid and type 2 diabetes. Diabetes. 2015;64:3028–3036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Juraschek SP, Choi HK, Tang O, et al. Opposing effects of sodium intake on uric acid and blood pressure and their causal implication. J Am Soc Hypertens. 2016;10:939–946.e942 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Ter Maaten JC, Voorburg A, Heine RJ, et al. Renal handling of urate and sodium during acute physiological hyperinsulinaemia in healthy subjects. Clin Sci (Lond). 1997;92:51–58. [DOI] [PubMed] [Google Scholar]
  • 42.Weinman EJ, Knight TF, McKenzie R, Eknoyan G. Dissociation of urate from sodium transport in the rat proximal tubule. Kidney Int. 1976;10:295–300. [DOI] [PubMed] [Google Scholar]
  • 43.Hund TJ, Mohler PJ. Role of CaMKII in cardiac arrhythmias. Trends Cardiovasc Med 2015;25:392–397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Chakravarti B, Yang J, Ahlers-Dannen KE, et al. Essentiality of regulator of G protein signaling 6 and oxidized Ca2+/calmodulin-dependent protein kinase II in Notch signaling and cardiovascular development. J Am Heart Assoc. 2017;6(11). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Abuelo JG. Low dialysate potassium concentration: an overrated risk factor for cardiac arrhythmia? Semin Dial. 2015;28:266–275. [DOI] [PubMed] [Google Scholar]
  • 46.Mangravite LM, Engelhardt BE, Medina MW, et al. A statin-dependent QTL for GATM expression is associated with statin-induced myopathy. Nature. 2013;502:377–380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Campbell JE, Ussher JR, Mulvihill EE, et al. TCF1 links GIPR signaling to the control of beta cell function and survival. Nat Med. 2016;22:84–90. [DOI] [PubMed] [Google Scholar]
  • 48.Rowe JW, Tobin JD, Rosa RM, Andres R. Effect of experimental potassium deficiency on glucose and insulin metabolism. Metabolism. 1980;29:498–502. [DOI] [PubMed] [Google Scholar]
  • 49.Helderman JH, Elahi D, Andersen DK, et al. Prevention of the glucose intolerance of thiazide diuretics by maintenance of body potassium. Diabetes. 1983;32:106–111. [DOI] [PubMed] [Google Scholar]
  • 50.Zhu Z, Zhang F, Hu H, et al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. 2016;48:481487. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Materials. Supplementary Methods and References.

Figure S1. Regional association and linkage disequilibrium plots for 21 genome-wide significant loci for the urinary albumin: creatinine ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S2. Regional association and linkage disequilibrium plots for 11 genome-wide significant loci for the urinary potassium: creatinine ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S3. Regional association and linkage disequilibrium plots for 12 genome-wide significant loci for the urinary sodium: creatinine ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S4. Regional association and linkage disequilibrium plots for 8 genome-wide significant loci for the urinary sodium: potassium ratio in the discovery analysis. The y axis represents the negative log10 of the single-nucleotide polymorphism (SNP) P value, and the x axis represents the position on the chromosome (chr), with the name and location of genes in the UCSC Genome Browser shown in the bottom panel. The SNP with the lowest P value in the region is marked by a purple diamond. The colors of the other SNPs indicate the r2 of these SNPs with the lead SNP. Plots were generated with LocusZoom. GWAS, genome-wide association study.

Figure S5. Q–Q plots for genetic associations for the urinary albumin: creatinine ratio (A), urinary potassium: creatinine ratio (B), urinary sodium: creatinine ratio (C), and urinary sodium: potassium ratio (D) in the discovery analysis.

Figure S6. Manhattan plot for the discovery genome-wide association study of binary albuminuria. The nearest genes for each locus that showed a significant P value in the replication were labeled. Negative log10-transformed P values for each single-nucleotide polymorphism (y axis) are plotted by chromosomal position (x axis). The gray line represents the threshold for genome-wide statistically significant associations (P = 5 × 10−8). Red points represent significant hits.

Figure S7. Scatter plot of the genetic and phenotypic correlations of the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK).

Figure S8. Histogram of the frequencies of the urinary albumin: creatinine ratio (UACR) (A), urinary potassium: creatinine ratio (UK/UCr) (B), urinary sodium: creatinine ratio (UNa/UCr) (C), and urinary sodium: potassium ratio (UNa/UK) (D) before normalization.

Table S1. Genetic loci associated with urinary albuminuria in the discovery and replication genome-wide association studies.

Table S2. Phenotypic correlation between urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK).

Table S3. Genetic correlation of the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK) with other traits.

Table S4. Significant expression quantitative trait loci for replicated variants in urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK) genome-wide association studies.

Table S5. Significant expression quantitative trait locus (eQTL) and methylation quantitative trait locus (mQTL) probe colocalizations for the urinary albumin: creatinine ratio (UACR), urinary potassium: creatinine ratio (UK/UCr), urinary sodium: creatinine ratio (UNa/UCr), and urinary sodium: potassium ratio (UNa/UK).

Table S6. Genetic loci associated with the urinary albumin: creatinine ratio (UACR) in relation with the genome-wide association study of binary albuminuria.

Table S7. Previous studies and their comparisons with the UK Biobank associations.

NIHMS1023301-supplement-Supplementary_Materials.pdf (5.5MB, pdf)

RESOURCES