Identification of end stage renal disease associated loci in X chromosome: an X chromosome-wide association study

doi:10.21203/rs.3.rs-5294228/v1

Download PDF

Research Article

Identification of end stage renal disease associated loci in X chromosome: an X chromosome-wide association study

https://doi.org/10.21203/rs.3.rs-5294228/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

X-chromosomal genetic variants have been understudied in end stage renal disease (ESRD), which hold the promise to provide valuable insights into sexually dimorphic traits and diseases.

Methods

We performed a X chromosome-wide association study (XWAS) in a Chinese cohort (N = 2750), comprising 1489 cases with ESRD and 1261 controls, to identify locus associated with ESRD risk.

Results

One sex-shared loci, rs3138874 located in the promoter of COL4A5 were identified in the meta-analysis on the summary statistics from the sex-stratified XWAS. Additionally, 2 male-specific loci, comprising rs142591759 near MIR3202-2 and rs73250616 near SYTL4 were identified in the sex-stratified XWAS in males. Of the 3 ESRD associated loci, 2 were previously found to be associated with estimated Glomerular filtration rate (eGFR) in different populations. Finally, by integrating expression quantitative trait loci (eQTL) statistics from multiple tissues and conducting eQTL colocalization analysis, we found that SYTL4, TSPAN6, NOX1, CSTF2, PCDH19 and RPL36A are the target genes of ESRD associated locus Xq22.1.

Conclusion

Our finding revealed three X-chromosome loci linked to ESRD risk, which provided foundational knowledge for genetic risk prediction and advanced our understanding of the molecular underpinnings of ESRD.

End stage renal disease

X chromosome-wide association study

single nucleotide polymorphism

sexually dimorphic traits

• Prevalence of end stage renal disease (ESRD) was higher in males than in females. To this end, we studied the associations of over 120,000 chromosome X variants with ESRD in 2750 Chinese individuals and further identified three new genomic loci where the variants are associated with ESRD.

• In two of these three associated loci, loci near (Xq22.1) and (Xq28), we observed that there are sex-specific genetic effects on ESRD across various kidney diseases, with significant associations specifically observed in males. Further supporting this observation, Xq22.1 was previously found to be associated with estimated Glomerular filtration rate (eGFR) in other population including European and African ancestry.

• Loci located in the promoter of showed gender-shared effects on ESRD in diverse kidney diseases, and association with eGFR in other population.

• Six plausible prioritized genes, including SYTL4, TSPAN6, NOX1, CSTF2, PCDH19 and RPL36A at male-specific loci Xq22.1, were revealed by using published eQTL database and conducting eQTL colocalization analysis, suggesting the potential mechanism of ESRD pathology.

End-stage renal disease (ESRD) is an irreversible outcome of chronic kidney disease (CKD), leading to diminished personal longevity and quality of life, as well as imposing substantial medical and economical burdens globally[1, 2]. The worldwide morbidity of CKD was estimated to range from 8.5%-9.8%, with approximately 2% of cases progressing to ESRD[3, 4]. Research has identified significant familial clustering of common causes of ESRD, as evidenced by a study involving over 5 million residents[5]. Over the past decades, our insight of genetic susceptibility to renal dysfunction was greatly extended and refined through large-scale population genetic studies[6]. Hundreds of loci and genes have been identified by genome-wide association studies (GWASs) of ESRD, CKD, or kidney function–related traits[7–9], helping in the risk explanation and therapies development for kidney diseases. Particularly, variants in APOL1 have been associated with increased risk of developing ESRD in patients with hypertension and diabetes[10–13]. The sex chromosomes were, however, excluded in most of the GWAS analyses, although the sex disparities in CKD risk have been observed in clinical characteristics and progression [14–16], highlighting the potential genetic differences between genders.

The exclusion is largely because the difference in X chromosome counts between male and females and X-inactivation in females contribute to the complexity of assessing the genotype-phenotype relationship for the sex chromosomes. To date, only a few genetic studies for kidney related traits investigated the role of sex chromosome variants in kidney traits. Notably, the high-profile X-linked Alport syndrome (AS), characterized by progressive kidney disease and ESRD, is caused by pathogenic variants in the COL4A5 (coding the collagen type IV alpha 5 chain) on the X-chromosome. In addition, the female-specific effect underlying ESRD was detected by another GWAS[17], linking variants rs12696304 and rs2736100 in the TERC (coding the telomerase RNA component) with an increased risk of ESRD in females. Prevalence of ESRD was higher in males than in females, and the current findings on the genetic susceptibility of ESRD remain insufficient in elucidating the underlying mechanisms contributing to sex differences in CKD progression. Therefore, it is crucial to reveal both shared and sex-specific effects of the X chromosome genetic variants on ESRD risk to advance the understanding of disease biology and develop new therapeutic strategies.

Here we performed a X chromosome-wide association study (XWAS) to identify common X chromosome risk variants in a Chinese cohort of ESRD. In this study, we first investigated the shared X chromosome-linked genetic risk effect between males and females (sex-shared effect) through a meta-analysis of the genetic risk effects obtained from sex-stratified GWAS analyses in 1623 males and 1127 females. Given the sex-bias of prevalence in ESRD, we also analyzed the sex-specific effects of X-chromosome variants on ESRD. Finally, we performed the colocalization analysis by integrating eQTL statistics of X chromosome from multiple tissue in GTEx to prioritize the most likely candidate genes within each locus.

Study subjects and grouping

This study enrolled 1489 patients diagnosed with ESRD. For the control group, we included 1161 subjects with blood samples for genotyping, who exhibited no substantial evidence for CKD. This study was approved by the ethics committee of Guangdong provincial people's hospital in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants.

End stage renal disease

For the study participants, we included individuals diagnosed with ESRD based on International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9‐CM), code 585 or International Classification of Diseases, Tenth Revision, Clinical Modification (ICD‐10‐CM), code N18.5 or N18.6 as well as those with a kidney transplant. Briefly, ESRD is characterized by a irreversible decline in kidney function, typically indicated by a glomerular filtration rate (GFR) below 15 ml/(min 1.73m²), the presence of uremic toxicity, and the need for renal replacement therapy. This includes: 1) Patients receiving renal replacement therapy, such as dialysis, or those who have undergone a kidney transplant 2) Individuals with Stage 5 CKD who may not yet be on dialysis but exhibit significant uremic toxicity and are at a stage where dialysis or transplantation is imminent or being considered

Genotyping, quality control and imputation

Genomic DNA was extracted from blood sample using AxyPrep™ Blood Genomic DNA Purification Miniprep Kit (Axygen, AP-MN-BL-GDNA-250) and quantified using ELIASA(pectra max plus 384, Molecular Devices). Qualified DNA sample of each subject was genotyped by the BeadChip Array Asian Screening Array (Illumina ASAMD-24v1-0) following the standard protocol for an Illumina Infinium™ assay. The GenomeStudio software and the calling algorithm from Illumina were used for normalized intensity data analyses.

The genotypes of 25606 single nucleotide polymorphism (SNPs) in the X chromosome were extracted for subsequent analysis. Prior to genotype imputation, we conducted quality control (QC) procedures on the genotype data following the criteria: (i) SNPs with a maximum per-person call rate lower than 95% were excluded; (ii) SNPs violating the Hardy-Weinberg equilibrium (HWE) assumption, defined as P-values < 1e-6, were excluded. (iii) Low frequency loci with a minor allele frequency (MAF) less than 1% were removed. Samples with missing SNP genotypes greater than 5% were removed. Note there are no samples filtered in our study.

We first pre-phased the genotypes using Eagle (version 2.4.1). Subsequently, we imputed genotype dosages using Beagle(version 5) with the 1000 Genomes Project phase 3 (version 3) reference haplotypes. The reference panels consist of 3202 unrelated samples from 26 populations. Only non-pseudoautosomal regions were phased and imputed. Pre-phasing and imputation were conducted separately for females and males. Imputed SNPs with were excluded based on the following criteria: (i) P value of HWE < 10e-6. (ii) SNP with low imputation quality (imputation quality score DR² < 0.3 and MAF < 0.01). The genetic analyses in this study were conducted using PLINK (version 1.9). The X-chromosome genotype matrix was established by coding females as diploid (0, 1, or 2) and males as homozygous diploid (either 0 or 2).

X chromosome-wide association analysis on ESRD

We first conducted a sex-stratified XWAS on ESRD risk in females and males separately, which used an age-adjusted model of logistic regression under the assumption of additive allelic effects of the SNP dosages. The significance threshold for the X chromosome-wide association was set at P = 10e^− 5 based on the number of independent linkage disequilibrium (LD) blocks. We also used non-additive XWAS models to assess recessive and dominance effects to identify variations associated with ESRD, applying a more stringent threshold at P < 10e^− 6. This threshold was set because, compared to additive model, dominant and recessive disease models generate weaker LD levels in patients[18].

To identify cross-sex ESRD variants, a fixed effect inverse-variance weighted meta-analysis was performed by combining female and male summary statistics using ‘metafor’ R packages (version 3.4.0). To identify sex-shared variants, we retained only the summary statistics of SNPs with low heterogeneity (I² < 20%) between females and males for final meta-analysis. Additionally, a Z score was computed to quantify the sex differences in effect sizes, aiding in the identification sex specific variants.

Fine mapping and functional annotations on ESRD loci

Lead index variants were determined as the most significant variant within ± 5 kb windows. We defined independent associated loci on the basis of genomic positions at least 1Mb apart from each other. Conditional and joint analysis (COJO) and stepwise conditional and joint analysis (SLCT) using GCTA[19] tool were further conducted to identify independent locus in a step-wise forward selection process.

Independent ESRD loci were mapped to target genes based on the index SNP by (i) identifying the nearest gene; (ii) associating with a reported eQTL gene from the GTex database; and (iii) identifying colocalized genes. The colocalization analysis of each locus, including variants within the ± 1 Mb region surrounding the index SNP, was conducted using the R package ‘coloc’. Within this region, the presence of at least one genome-wide significant eQTL variant (P-values < 10e − 5) was required before testing for colocalization. A posterior probability (PP) > 70% for a shared variant between eQTL and ESRD associated locus was defined colocalization.

Association between ESRD loci and kidney traits

A cross-sectional lookup analysis was conducted by referring to two published GWAS in European and Japanese populations, which encompassed the associations between candidate loci and kidney traits, including creatinine, estimate glomerular filtration rate (eGFR), uric acid, urea nitrogen, and albuminuria. ESRD loci identified in our analysis were mapped to the full summary statistics of these two published GWASs to identify replicated loci. Overlapping locus with a P-value < 0.05 in previous studies were considered replicated.

Identification of cross-sex loci associated with ESRD in X chromosome

A total of 2750 subjects (1489 case and 1161 controls) were included in this study, comprising 1127 females and 1623 males. The distribution of primary kidney diseases of ESRD cases in the GDPH cohort were illustrated in Table 1. By analyzing the ESRD cases caused by diverse background chronic kidney diseases, we aimed to identify common genetic variations for kidney progression to ESRD regardless of background chronic kidney disease. Study design is shown in Fig. 1. For the XWAS, 155,759 SNPs were available for the female cohort (comprising 683 cases and 444 controls), and 127,855 SNPs were available for the male cohort (996 cases and 627 controls) after QC procedures and genotype imputation. Detailed demographic characteristics and primary diseases of subjects were illustrated in Table 1.

Table 1

Overview of female and male cohort studies that contributed to the GWAS meta-analysis
Cohort	Genotyping Platform	Prephasing software	Imputation		SNP quality control			Association analysis		Lambda	No. of SNPs Pre-imputation	No. of SNPs Pre-imputation	Participants(N)	Cases, %	µ (σ) –age cases	µ (σ) –age controls	Propotion of Primary disease in cases
Cohort	Genotyping Platform	Prephasing software	Software	Reference panel	HWE P-value	MAF	ER	Software	Covariates	Lambda	No. of SNPs Pre-imputation	No. of SNPs Pre-imputation	Participants(N)	Cases, %	µ (σ) –age cases	µ (σ) –age controls	Propotion of Primary disease in cases
Females	Illumina Asian Screening Array	eagle	beagle	1000 genomes phase 3	0.000001	> 1%	> 0.3	PLINK	PC1-10	1	25606	155759	1,127	50.66%	52.77 (23.86)	59.09 (14.42)	DN(64.21%); HN(4.63%); IgAN (9.44%); LN (11.11%); PKD (12.41%)
Males	Illumina Asian Screening Array	eagle	beagle	1000 genomes phase 3	/	> 1%	> 0.3	PLINK	PC1-10	1	25606	127855	1,623	56.38%	55.21 (20.60)	55.93 (13.05)	DN(72.81%); HN(5.9%); IgAN (9.8%); LN (1.05%); PKD (10.43%)
Tests for deviation from Hardy Weinberg are 2-tailed. DN:diabetic nephropathy, IgAN: IgA nephropathy, HN: hypertensive nephropathy, LN: lupus nephropathy, PKD: polycystic kidney disease

We first performed the GWAS analysis in the female and male samples separately and did not observe any evidence for genomic inflation in either the female- or male-specific analysis (λ_female = 1.00, λ_male = 1.06)(Fig. 2d and 2f). Given that most SNPs of were in linkage disequilibrium (LD) to each other, the common SNPs (MAF ≥ 0.01) were pruned into 500 independent index SNPs (LD r² < 0.01 within 1,500kb). Therefore, the X chromosome wide significance threshold was set at P < 10e-5 (0.05/500). To uncover the sex-shared locus associated with ESRD on the X chromosome, a meta-analysis on a total of 111,436 common SNPs (shared between the female and male specific analyses) was performed by combining the female- and male-specific summary statistics from the gender-specific XWAS (λ = 1.03). The meta-analysis identified one sex-shared loci at the X chromosome-wide significance with consistent effects across genders (Fig. 2a and 2b), that is rs3138874 located in the promoter of COL4A5 (β = -0.34, MAF = 0.1, P = 2.43e-6). Details of this loci were provided in Table 2, and the regional plots of the association results for this loci were presented in Fig. 3. Importantly, we found that the additive effect model has better performance, with no significant sex-shared loci associated with ESRD identified in either the recessive or dominant models (Fig. S1a and b, Fig. S2a and b). This result is not surprising. A recent study suggested that in most human genomic studies, complex traits exhibit additive effects based on gene dosage reliable, and detection of loci showing non-additive effects requires larger sample sizes (minimum sample sizes of millions) and statistical power [20].

Table 2

Independent SNPs in X chromosome-Wide Significant Loci Identified in the Combined Sex Meta-Analysis
rsID	Pos (chrX/hg38)	Alleles	MAF	Nearest Gene	Group	Sex	β	se	p	bJ	bJ_se	pJ	het (I square)	p-het
rs3138874	108694045	A > G	0.1	COL4A5	Sex-shared	Combined sex	-0.34	0.07	2.43E-06	-	-	-	0	0.73
						Females	-0.36	0.15	6.80E-02	-	-	-
						Males	-0.27	0.08	1.20E-05	-	-	-
rs73250616	100738288	T > C	0.36	SYTL4	Male-specific	Combined sex	-0.14	0.05	2.31E-03	-	-	-	90.8	9.80E-04
						Females	0.11	0.09	1.97E-01	-	-	-
						Males	-0.22	0.05	1.71E-05	-0.23	0.05	1.08E-05
rs142591759	153994161	A > C	0.1	MIR3202-2	Male-specific	Combined sex	-0.23	0.07	1.72E-03	-	-	-	85.02	9.78E-03
						Females	0.13	0.16	4.17E-01	-	-	-
						Males	-0.33	0.08	6.85E-05	-3.30E-01	0.08	7.41E-05
β values are calculated with respect to the minor allele. These values are provided for the overall meta-analysis;
se = standard error; MAF = minor allele frequency; β = effect size; p-het = p value from the sex-heterogeneity test; SNP = single nucleotide polymorphism.
bJ, bJ_se and pJ: effect size, standard error and p-value from a joint analysis of all the selected SNPs.

Identification of sex-specific loci associated with ESRD in X chromosome

In the male-specific XWAS, we identified two independent loci (Fig. 2e) significantly associated with ESRD, including rs142591759 in the intergenic region close to MIR3202-2 and rs73250616 in the intergenic region close to SYTL4 (Table 2). Figure 4 presents the region plot of each male-specific locus. Furthermore, we tested the significance of the difference in effect sizes between genders and males confirmed the male-specific effects of these loci on ESRD (rs142591759: β._female = 0.13, β._male = -0.33, P._zscore = 9.8e-3; rs73250616: β._female = 0.11, β._male = -0.22, P_zscore = 9.8e-4). In contrast, no female-specific locus was identified in the female-specific XWAS (Fig. 2c). Unsurprisingly, we observed no significant associations between ESRD and X chromosome variations in recessive and dominant model in female cohorts (Fig. S1. c and Fig. S2. c).

Association analysis conditioning on IgAN and PKD

To determine whether the ESRD signals were driven by confounding effects, we adjusted for polycystic kidney disease (PKD) status due to its strong genetic basis and potential to confound associations between identified SNPs and ESRD. Additionally, we accounted for IgA nephropathy (IgAN), a common cause of ESRD in East Asian, by including IgAN polygenic risk scores and status as covariates in our GWAS logistic regression models. We also adjusted for IgAN polygenic risk scores, which were calculated using a previously established risk model[21]. Remarkably, the analysis indicated that the association result of identified SNPs were robust to the potential impacts of these etiology (Table 3).

Table 3

Association between identified SNPs and ESRD from three models adjusted for PKD, IgAN and IgAN PRS
SNP	BP	Group	Base Model		Base + PKD Model		Base + IgAN Model		Base + IgAN PRS ^a		Heterogeneity I²	Q statistic for heterogeneity	p value for heterogeneity
SNP	BP	Group	β^b	Pvalue	β	Pvalue	β	Pvalue	β	Pvalue	Heterogeneity I²	Q statistic for heterogeneity	p value for heterogeneity
rs3138874	108694045	cross-sex	-0.34	2.43E-06	-0.35	3.33E-06	-0.36	2.07E-06	-0.36	1E-06	0	0.02	0.99
rs73250616	100738288	Male	-0.22	1.71E-05	-0.21	7.46E-05	-0.22	5.52E-05	-0.22	2.1E-05	0	0.01	0.99
rs142591759	153994161	Male	-0.33	6.85E-05	-0.33	9.45E-05	-0.34	7.75E-05	-0.33	7.2E-05	0	0.04	0.99
a. IgAN PRS: IgAN polygenic risk scores, which were calculated using a previously established risk model from a recent GWAS, namely Kiryluk, K.et al (PMID: 37337107)
b. β = effect size;
P values were derived from logistic model including the covariates age, sex and genomic PCs (base model), or PKD phenotype (Base + PKD model), or IgAN phenotype (Base + IgAN model), or IgAN PRS (Base + IgAN PRS model)

Gene prioritization via eQTL colocalization analysis

We identified the most likely candidate causal genes within each association locus through eQTL-based colocalization analysis. Among the male-specific loci, we found rs73250616 at locus Xq22.1 was associated with decreased expressions of SYTL4, TSPAN6, NOX1, CSTF2 and PCDH19 (Table 4, at the end of the manuscript). Colocalization analysis further revealed that this locus colocalized with the eQTL SYTL4 ((PP(H4) = 0.998) and RPL36A (PP(H4) = 0.71) in visceral adipose tissue of the omentum and cultured fibroblasts (Table 5, Fig. 5a and Fig. 5b), respectively. Details were illustrated in Table 5. Notably, the renal fibrosis is a crucial mechanism contributing to the progression of kidney disease. This suggests that ESRD-related genetic risk variants in male might contribute to kidney fibrosis by regulating the expression of RPL36A.

Table 4

Annotation of ESRD associated SNP rs73250616 identified in male cohort with respect to expression quantitative trait loci (eQTL) based on GTEx/v8 databases
eQTL Gene	A1	A2	ma_samples	maf	pval_nominal	slope	slope_se	pval_nominal_threshold	tissue
TSPAN6	C	T	188	0.49	4.10E-09	-0.20	0.49	3.23E-02	Esophagus_Gastroesophageal_Junction
TSPAN6	C	T	259	0.48	7.77E-10	-0.17	0.48	2.66E-02	Esophagus_Muscularis
TSPAN6	C	T	293	0.50	1.14E-04	-0.10	0.50	2.52E-02	Lung
NOX1	C	T	88	0.45	4.08E-05	0.18	0.45	4.25E-02	Brain_Cerebellar_Hemisphere
NOX1	C	T	294	0.47	1.40E-06	0.21	0.47	4.27E-02	Nerve_Tibial
CSTF2	C	T	280	0.47	2.34E-12	-0.15	0.47	2.02E-02	Esophagus_Mucosa
SYTL4	C	T	312	0.46	1.06E-20	0.19	0.46	1.90E-02	Adipose_Subcutaneous
SYTL4	C	T	260	0.48	2.38E-31	0.29	0.48	2.28E-02	Adipose_Visceral_Omentum
SYTL4	C	T	221	0.48	2.11E-39	0.47	0.48	3.07E-02	Artery_Aorta
SYTL4	C	T	121	0.47	5.43E-05	0.14	0.47	3.35E-02	Artery_Coronary
SYTL4	C	T	317	0.47	8.67E-13	0.15	0.47	2.04E-02	Artery_Tibial
SYTL4	C	T	67	0.48	2.02E-11	0.47	0.48	6.20E-02	Brain_Amygdala
SYTL4	C	T	75	0.47	3.77E-18	0.44	0.47	4.27E-02	Brain_Anterior_cingulate_cortex_BA24
SYTL4	C	T	101	0.48	5.87E-19	0.38	0.48	3.76E-02	Brain_Caudate_basal_ganglia
SYTL4	C	T	88	0.45	4.63E-10	0.25	0.45	3.71E-02	Brain_Cerebellar_Hemisphere
SYTL4	C	T	103	0.44	2.00E-11	0.27	0.44	3.72E-02	Brain_Cerebellum
SYTL4	C	T	105	0.46	6.08E-25	0.43	0.46	3.53E-02	Brain_Cortex
SYTL4	C	T	88	0.45	3.24E-23	0.46	0.45	3.78E-02	Brain_Frontal_Cortex_BA9
SYTL4	C	T	82	0.43	3.12E-10	0.34	0.43	4.99E-02	Brain_Hippocampus
SYTL4	C	T	84	0.44	1.65E-06	0.24	0.44	4.68E-02	Brain_Hypothalamus
SYTL4	C	T	103	0.46	1.11E-15	0.36	0.46	4.04E-02	Brain_Nucleus_accumbens_basal_ganglia
SYTL4	C	T	89	0.48	3.56E-17	0.33	0.48	3.36E-02	Brain_Putamen_basal_ganglia
SYTL4	C	T	228	0.49	7.38E-11	0.21	0.49	3.19E-02	Breast_Mammary_Tissue
SYTL4	C	T	174	0.47	4.81E-07	0.10	0.47	2.00E-02	Colon_Sigmoid
SYTL4	C	T	208	0.48	1.11E-12	0.18	0.48	2.41E-02	Colon_Transverse
SYTL4	C	T	259	0.48	1.01E-04	0.07	0.48	1.79E-02	Esophagus_Muscularis
SYTL4	C	T	115	0.49	1.30E-22	0.45	0.49	3.92E-02	Liver
SYTL4	C	T	78	0.48	9.51E-08	0.34	0.48	6.04E-02	Minor_Salivary_Gland
SYTL4	C	T	401	0.49	2.79E-08	0.11	0.49	1.94E-02	Muscle_Skeletal
SYTL4	C	T	294	0.47	2.41E-13	0.11	0.47	1.43E-02	Nerve_Tibial
SYTL4	C	T	113	0.44	5.75E-06	0.25	0.44	5.35E-02	Ovary
SYTL4	C	T	163	0.45	2.50E-38	0.57	0.45	3.68E-02	Pancreas
SYTL4	C	T	98	0.44	8.19E-06	0.19	0.44	4.12E-02	Prostate
SYTL4	C	T	296	0.50	1.25E-07	0.09	0.50	1.69E-02	Skin_Not_Sun_Exposed_Suprapubic
SYTL4	C	T	343	0.48	1.80E-07	0.08	0.48	1.45E-02	Skin_Sun_Exposed_Lower_leg
SYTL4	C	T	127	0.46	5.09E-08	0.16	0.46	2.76E-02	Spleen
SYTL4	C	T	189	0.48	7.93E-15	0.27	0.48	3.30E-02	Stomach
SYTL4	C	T	158	0.49	7.37E-10	0.12	0.49	1.90E-02	Testis
SYTL4	C	T	325	0.48	1.16E-97	0.59	0.48	2.23E-02	Thyroid
PCDH19	C	T	293	0.50	1.17E-04	-0.15	0.50	3.76E-02	Lung
These values are provided by eQTL GTEx/v8 eQTL databases: ma_samples Number of samples with minor allele, pval_nominal Nominal p-value, slope Coefficient for the SNP genotype in the linear model, slope_se Standard error of the slope,pval_nominal_threshold Gene-tissue-specific nominal p-value significance threshold as determined by permutations and transcriptome-wide FDR;

Table 5

Colocalizatio analysis of ESRD associated locus and eQTLs.
locus	nsnps	PP.H0.abf	PP.H1.abf	PP.H2.abf	PP.H3.abf	PP.H4.abf	Gene.Symbol	Tissue	Group
Xq22.1	1352	0	5.71E-22	0	0.0025	0.9975	SYTL4	Adipose_Visceral_Omentum	Male
Xq22.1	1005	0	1.96E-01	0	0.0944	0.7099	RPL36A	Cells_Cultured_fibroblasts	Male
For all independent signals, we retrieved the nearby genes and performed colocalization analyses of association statistics and eQTLs retrieved from GTEx v8. Association statistics of all analysis groups were considered.
Co-localization of association and eQTL signals was assumed if PP(H4) ≥ 70%.

Exploring association between ESRD-associated loci and renal traits

Existing evidences have highlighted the significant roles of decreased eGFR, elevated gout[22], blood urea nitrogen[23] and uric acid[24] in predicting the risk of ESRD[25]. Therefore, we further explored the relationships between ESRD loci and kidney-related traits by looking up the published summary statistics of two GWASs, which comprised cohorts from BioBank Japan, the Chronic Kidney Disease Genetics Consortium, UK Biobank, and FinnGen. Out of ESRD associated loci, two loci, including rs3138874 in COL4A5, and rs73250616 near SYTL4, were identified with nominally significant association with eGFR. Specifically, male-specific SNP rs73250616 also showed association with creatine. (Details were presented in Table 6)

Table 6

Look-up analysis of ESRD associated SNPs in previously GWAS summary data for kidney traits.
rsID	Nearest gene	Position	E/A	EAF	Trait	Group	Beta	Se	P	Cohort	Ancestry
rs3138874	COL4A5	108694045	A/G	0.23	eGFR	cross-sex	-0.00142	0.00056	9.19E-03	the Chronic Kidney Disease Genetics Consortium	European; Asian; African
rs73250616	SYTL4	100738288	T/C	0.36	eGFR	Male	-0.00094	0.00038	1.12E-02	the Chronic Kidney Disease Genetics Consortium	European; Asian; African
rs73250616	SYTL4	100738288	T/C	0.36	creatinine	Male	0.0035	0.0013	7.28E-03	UK Biobank and FinnGen	European
We obtained the summary statistics of associations between X-chromosomal variants and kidney traits from two recent GWAS,
namely Sakaue et al. (PMID 34594039), and Scholz. (PMID 38233393).
We tested if ESRD SNPs were also associated with kidney related traits based on these two GWAS summary data, which were at least nominal significance in one-sided testing of the β coefficients of additive genetic effects considering expected effect directions.

In this study, we performed association analyses between the variations of X chromosome and ESRD in a Chinese cohort (1489 cases and 1161 controls), and identified three independent loci in total significantly associated with ESRD risk. One of these loci showed gender-shared effects, while two loci showed male-specific effects with significant larger effects on ESRD compared to that in females. Two of the three loci were previously found to be associated with eGFR in different populations. Additionally, a colocalization analysis revealed two plausible prioritized genes, SYTL4 and RPL36A at Xq22.1, suggesting the potential mechanism of ESRD pathology. In summary, our findings highlighted the importance of X chromosome variation on genetic susceptibility of kidney disease and proposed a set of ESRD-associated loci with potential in biomarker or therapy development.

Genetic variations in X chromosome have been linked to kidney disease, accounting for part of the pathogenic mechanism[26]. Rs3138874 located in COL4A5 was the top cross-sex locus associated with ESRD in the present study. Notably, since patients with Alport syndrome were not included in the current study, the findings suggest a potential shared causal effect of rs3138874 in COL4A5 on renal progression to kidney failure. Although COL4A5 is typically considered a common pathogenic gene for Alport syndrome (AS) [27], it is important to note that COL4A5-related diseases are not limited to Alport syndrome. Some individuals carrying COL4A5 variants may not have Alport syndrome. Previous evidence further supported the robustness of our findings, indicating that COL4A5 variants are more prevalent than COL4A3 and COL4A4 variants in cohorts with kidney failure[28–32]. Moreover, existing evidence has emphasized the role of COL4A5 gene variants in several kidney diseases, such as focal segmental glomerulosclerosis (FSGS), kidney failure of unknown cause, or familial IgA glomerulonephritis[33]. The COL4A5 is one of the genes coding the alpha chain for type IV collagen, which contributes to the integrality of the glomerular basement membrane (GBM) to ensure the normal function of the glomerulus[34]. Variants in COL4A5 would result in abnormal synthesis or assembly of type IV collagen, which in turn affects the structure and function of the GBM. Besides, variations in COL4A5 were also identified associated with renal functions and ESRD in subjects with polycystic kidney disease, amyloidosis or focal segmental glomerulosclerosis[35], emphasizing the importance of COL4A5 for renal pathology. Overall, our finding emphasizes that identifying COL4A5 variant rs3138874, regardless of the clinical presentation or genders means that affected individuals should be treated precisely for the prevent and management of ESRD.

No X chromosome-wide significant findings were detected for ESRD in male cohort,which is consistent with evidence showed a higher risk of ESRD in males[36]. New risk locus at Xq22.1 in males was identified in sex-specific genome-wide association study. Based on eQTL colocalization, for male-specific risk locus at Xq22.1, we assigned Synaptotagmin-like protein 4 (Slp-4) and RPL36A as plausible candidate genes. Slp-4, also known as granuphilin, the Rab27A-Slp4-a complex involved in controlling the release of vaso-active substances such as inflammatory mediators in the vasculature[37]. Importantly, granuphilin also binds to Rab27 on the insulin granule[38]. The effect of this gene on insulin secretion was impaired by its mutation, and dysfunction of insulin secretion implicated in immune and metabolism kidney disease progression. For the other target gene RPL36A, previous study have indicated that the knockdown of this gene downregulates GLA expression associated with fabry disease in vitro model, which is a rare genetic disorder causing serious kidney problems[39]. Further study is needed to understand the mechanisms by which variants of SYTL4 and RPL36A are associated with ESRD.

Notably, evidence from eQTL association statistics of multiple tissues further revealed the potential of NOX1, CSTF2 and TSPAN6 as the target genes of lead SNP rs73250616 at the locus Xq22.1. NOX1, a member of the NADPH oxidase family, is implicated in microvascular dysfunction in metabolic diseases[40] and has a significant role in renal inflammation and fibrosis[41]. The second gene CSTF2, of which expression was upregulated in tubular epithelial cells (TEC) of fibrotic kidneys[42]. For the TSPAN6 gene, previous research identified an association between its mutation and eGFR[26]. This family of membrane proteins is ubiquitously expressed and involved in a multitude of cellular processes. Although a direct role with respect to ESRD was not yet described, the gene family was shown to be associated with immune function, there is evidence linking this gene family to immune responses, suggesting their possible influence on the chronic inflammation processes associated with CKD[43]. In summary, our findings reinforce and substantiate the link between these gene and kidney disease or kidney function suggested by prior research.

In summary, we conducted a comprehensive genetic association analysis of X chromosome variations with ESRD, and identified novel sex-shared and male-specific ESRD loci, respectively. These findings offer new insights into the genetic predisposition and the sexual dimorphism of ESRD, priority of candidate gene targets for further molecular studies and therapeutic development for ESRD.

Data Availability

Summary statistics of this study are available upon request from the corresponding author. The raw data are not publicly available due to ethical restrictions.

Acknowledgments

We are grateful to the participants and all staff members in this study. This study was supported by the China National GeneBank in computational resources.

Funding Statement

This project was supported by the Guangdong-Hong Kong Joint Laboratory on Immunological and Genetic Kidney Diseases (2019B121205005), National Natural Science Foundation of China (81920108008, 82170714), Guangdong Province High-Level Hospital Construction Project (DFJHBF202101), GDPH Supporting Fund for Talent Program (DFJH2018, KY0120220261), the Guangdong Basic and Applied Basic Research Foundation (No. 2021A1515111054), Guangzhou Municipal Science and Technology Project of China (202201011246), Guangzhou Science and Technology Project (202206080010), Guangdong Province High-Level Hospital Construction Project (DEHBF202101).

Author Contributions

Funding acquisition, project administration, and resources providing: Jianjun Liu, Xueqing Yu; Writing of the Manuscript: Xiaohong Zhou; Statistical methods and analysis: Xiaohong Zhou; Revision of manuscript: Jianjun Liu; Sample ascertainment and data generation: Dianchun Shi; Writing-review and editing: Yibin Liu. Subjects recruiting: Ming Li, Zhiming Ye, Wei Chen, Meng Wang, Dongying Fu, Yanna Wang, Hua Gan, Ping Fu, Xiaojun Tan, Yaozhong Kong, Jihong Chen, Jinghong Zhao

Ethics Declaration

Ethics approval and consent to participate

The studies involving human participants were reviewed and approved by Ethical Committee for Guangdong Provincial People's Provincial Hospital. Written informed consent to participate in this study was provided by the participants.

Consent for publication

Not applicable.

Competing interests

The authors declare they have no competing interests.

Evans RW, Manninen DL, Garrison LP Jr., Hart LG, Blagg CR, Gutman RA, Hull AR, Lowrie EG. The quality of life of patients with end-stage renal disease. N Engl J Med. 1985;312(9):553–9.
Nasser J, Bergman DT, Fulco CP, Guckelberger P, Doughty BR, Patwardhan TA, Jones TR, Nguyen TH, Ulirsch JC, Lekschas F, et al. Genome-wide enhancer maps link risk variants to disease genes. Nature. 2021;593(7858):238–43.
Collaboration GBDCKD. Global, regional, and national burden of chronic kidney disease, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2020;395(10225):709–33.
Wang L, Xu X, Zhang M, Hu C, Zhang X, Li C, Nie S, Huang Z, Zhao Z, Hou FF, et al. Prevalence of Chronic Kidney Disease in China: Results From the Sixth China Chronic Disease and Risk Factor Surveillance. JAMA Intern Med. 2023;183(4):298–310.
Skrunes R, Svarstad E, Reisaeter AV, Vikse BE. Familial clustering of ESRD in the Norwegian population. Clin J Am Soc Nephrol. 2014;9(10):1692–700.
Vivante A. Genetics of Chronic Kidney Disease. N Engl J Med. 2024;391(7):627–39.
Hanson RL, Craig DW, Millis MP, Yeatts KA, Kobes S, Pearson JV, Lee AM, Knowler WC, Nelson RG, Wolford JK. Identification of PVT1 as a candidate gene for end-stage renal disease in type 2 diabetes using a pooling-based genome-wide single nucleotide polymorphism association study. Diabetes. 2007;56(4):975–83.
Kottgen A, Glazer NL, Dehghan A, Hwang SJ, Katz R, Li M, Yang Q, Gudnason V, Launer LJ, Harris TB, et al. Multiple loci associated with indices of renal function and chronic kidney disease. Nat Genet. 2009;41(6):712–7.
Divers J, Ma L, Brown WM, Palmer ND, Choi Y, Israni AK, Pastan SO, Julian BA, Gaston RS, Hicks PJ, et al. Genome-wide association study for time to failure of kidney transplants from African American deceased donors. Clin Transpl. 2020;34(6):e13827.
Friedman DJ, Pollak MR. APOL1 Nephropathy: From Genetics to Clinical Applications. Clin J Am Soc Nephrol. 2021;16(2):294–303.
Beckerman P, Bi-Karchin J, Park AS, Qiu C, Dummer PD, Soomro I, Boustany-Kari CM, Pullen SS, Miner JH, Hu CA, et al. Transgenic expression of human APOL1 risk variants in podocytes induces kidney disease in mice. Nat Med. 2017;23(4):429–38.
Genovese G, Friedman DJ, Ross MD, Lecordier L, Uzureau P, Freedman BI, Bowden DW, Langefeld CD, Oleksyk TK, Uscinski Knob AL, et al. Association of trypanolytic ApoL1 variants with kidney disease in African Americans. Science. 2010;329(5993):841–5.
Tzur S, Rosset S, Shemer R, Yudkovsky G, Selig S, Tarekegn A, Bekele E, Bradman N, Wasser WG, Behar DM, et al. Missense mutations in the APOL1 gene are highly associated with end stage kidney disease risk previously attributed to the MYH9 gene. Hum Genet. 2010;128(3):345–50.
Yu MK, Lyles CR, Bent-Shaw LA, Young BA, Pathways A. Risk factor, age and sex differences in chronic kidney disease prevalence in a diabetic cohort: the pathways study. Am J Nephrol. 2012;36(3):245–51.
Carrero JJ, Hecking M, Chesnaye NC, Jager KJ. Sex and gender disparities in the epidemiology and outcomes of chronic kidney disease. Nat Rev Nephrol. 2018;14(3):151–64.
Neugarten J, Acharya A, Silbiger SR. Effect of gender on the progression of nondiabetic renal disease: a meta-analysis. J Am Soc Nephrol. 2000;11(2):319–29.
Sun Q, Liu J, Cheng G, Dai M, Liu J, Qi Z, Zhao J, Li W, Kong F, Liu G, et al. The telomerase gene polymorphisms, but not telomere length, increase susceptibility to primary glomerulonephritis/end stage renal diseases in females. J Transl Med. 2020;18(1):184.
Schrodi SJ, Garcia VE, Rowland C, Jones HB. Pairwise linkage disequilibrium under disease models. Eur J Hum Genet. 2007;15(2):212–20.
Yang J, Lee SH, Goddard ME, Visscher PM. GCTA: a tool for genome-wide complex trait analysis. Am J Hum Genet. 2011;88(1):76–82.
Palmer DS, Zhou W, Abbott L, Wigdor EM, Baya N, Churchhouse C, Seed C, Poterba T, King D, Kanai M, et al. Analysis of genetic dominance in the UK Biobank. Science. 2023;379(6639):1341–8.
Kiryluk K, Sanchez-Rodriguez E, Zhou XJ, Zanoni F, Liu L, Mladkova N, Khan A, Marasa M, Zhang JY, Balderes O, et al. Genome-wide association analyses define pathogenic signaling pathways and prioritize drug targets for IgA nephropathy. Nat Genet. 2023;55(7):1091–105.
Yu KH, Kuo CF, Luo SF, See LC, Chou IJ, Chang HC, Chiou MJ. Risk of end-stage renal disease associated with gout: a nationwide population study. Arthritis Res Ther. 2012;14(2):R83.
Seki M, Nakayama M, Sakoh T, Yoshitomi R, Fukui A, Katafuchi E, Tsuda S, Nakano T, Tsuruya K, Kitazono T. Blood urea nitrogen is independently associated with renal outcomes in Japanese patients with stage 3–5 chronic kidney disease: a prospective observational study. BMC Nephrol. 2019;20(1):115.
Obermayr RP, Temml C, Gutjahr G, Knechtelsdorfer M, Oberbauer R, Klauser-Braun R. Elevated uric acid increases the risk for kidney disease. J Am Soc Nephrol. 2008;19(12):2407–13.
Kovesdy CP, Coresh J, Ballew SH, Woodward M, Levin A, Naimark DM, Nally J, Rothenbacher D, Stengel B, Iseki K, et al. Past Decline Versus Current eGFR and Subsequent ESRD Risk. J Am Soc Nephrol. 2016;27(8):2447–55.
Scholz M, Horn K, Pott J, Wuttke M, Kuhnapfel A, Nasr MK, Kirsten H, Li Y, Hoppmann A, Gorski M, et al. X-chromosome and kidney function: evidence from a multi-trait genetic analysis of 908,697 individuals reveals sex-specific and sex-differential findings in genes regulated by androgen response elements. Nat Commun. 2024;15(1):586.
Gibson JT, Huang M, Shenelli Croos Dabrera M, Shukla K, Rothe H, Hilbert P, Deltas C, Storey H, Lipska-Zietkiewicz BS, Chan MMY, et al. Genotype-phenotype correlations for COL4A3-COL4A5 variants resulting in Gly substitutions in Alport syndrome. Sci Rep. 2022;12(1):2722.
Bullich G, Domingo-Gallego A, Vargas I, Ruiz P, Lorente-Grandoso L, Furlano M, Fraga G, Madrid A, Ariceta G, Borregan M, et al. A kidney-disease gene panel allows a comprehensive genetic diagnosis of cystic and glomerular inherited kidney diseases. Kidney Int. 2018;94(2):363–71.
Groopman EE, Marasa M, Cameron-Christie S, Petrovski S, Aggarwal VS, Milo-Rasouly H, Li Y, Zhang J, Nestor J, Krithivasan P, et al. Diagnostic Utility of Exome Sequencing for Kidney Disease. N Engl J Med. 2019;380(2):142–51.
Ottlewski I, Munch J, Wagner T, Schonauer R, Bachmann A, Weimann A, Hentschel J, Lindner TH, Seehofer D, Bergmann C, et al. Value of renal gene panel diagnostics in adults waiting for kidney transplantation due to undetermined end-stage renal disease. Kidney Int. 2019;96(1):222–30.
Lata S, Marasa M, Li Y, Fasel DA, Groopman E, Jobanputra V, Rasouly H, Mitrotti A, Westland R, Verbitsky M, et al. Whole-Exome Sequencing in Adults With Chronic Kidney Disease: A Pilot Study. Ann Intern Med. 2018;168(2):100–9.
Connaughton DM, Kennedy C, Shril S, Mann N, Murray SL, Williams PA, Conlon E, Nakayama M, van der Ven AT, Ityel H, et al. Monogenic causes of chronic kidney disease in adults. Kidney Int. 2019;95(4):914–28.
Savige J, Harraka P. Pathogenic Variants in the Genes Affected in Alport Syndrome (COL4A3-COL4A5) and Their Association With Other Kidney Conditions: A Review. Am J Kidney Dis. 2021;78(6):857–64.
Naylor RW, Morais M, Lennon R. Complexities of the glomerular basement membrane. Nat Rev Nephrol. 2021;17(2):112–27.
Bleyer AJ, Westemeyer M, Xie J, Bloom MS, Brossart K, Eckel JJ, Jones F, Molnar MZ, Kotzker W, Anand P, et al. Genetic Etiologies for Chronic Kidney Disease Revealed through Next-Generation Renal Gene Panel. Am J Nephrol. 2022;53(4):297–306.
Harris RC, Zhang MZ. The role of gender disparities in kidney injury. Ann Transl Med. 2020;8(7):514.
van Breevoort D, Snijders AP, Hellen N, Weckhuysen S, van Hooren KW, Eikenboom J, Valentijn K, Fernandez-Borja M, Ceulemans B, De Jonghe P, et al. STXBP1 promotes Weibel-Palade body exocytosis through its interaction with the Rab27A effector Slp4-a. Blood. 2014;123(20):3185–94.
Torii S, Takeuchi T, Nagamatsu S, Izumi T. Rab27 effector granuphilin promotes the plasma membrane targeting of insulin granules via interaction with syntaxin 1a. J Biol Chem. 2004;279(21):22532–8.
Al-Obaide MA. The Knockdown of RPL36A Downregulates GLA Expression Associated with Fabry Disease In Vitro Model: PO1600. J Am Soc Nephrol. 2020;31:508.
Xu L, Balzarolo M, Robinson EL, Lorenz V, Della Verde G, Joray L, Mochizuki M, Kaufmann BA, Valstar G, de Jager SCA, et al. NOX1 mediates metabolic heart disease in mice and is upregulated in monocytes of humans with diastolic dysfunction. Cardiovasc Res. 2022;118(14):2973–84.
Gray SP, Jha JC, Kennedy K, van Bommel E, Chew P, Szyndralewiez C, Touyz RM, Schmidt H, Cooper ME, Jandeleit-Dahm KAM. Combined NOX1/4 inhibition with GKT137831 in mice provides dose-dependent reno- and atheroprotection even in established micro- and macrovascular disease. Diabetologia. 2017;60(5):927–37.
Tan Y, Zheng T, Zhang R, Chen S, Cheng Q, Zhang J, Wang R, Chen M, Na N. Alternative polyadenylation writer CSTF2 forms a positive loop with FGF2 to promote tubular epithelial-mesenchymal transition and renal fibrosis. Biochim Biophys Acta Mol Basis Dis. 2022;1868(12):166541.
Florin L, de Winde CM. Recent advancements in the understanding of tetraspanin functions. Med Microbiol Immunol. 2020;209(4):393–5.

No competing interests reported.

SupplementaryFiguresOct18.docx

Download PDF

Reviewers invited by journal
02 Nov, 2024
Editor assigned by journal
25 Oct, 2024
Submission checks completed at journal
25 Oct, 2024
First submitted to journal
19 Oct, 2024

You are reading this latest preprint version

Identification of end stage renal disease associated loci in X chromosome: an X chromosome-wide association study

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Highlights

INTRODUCTION

MATERIALS AND METHODS

Study subjects and grouping

End stage renal disease

Genotyping, quality control and imputation

X chromosome-wide association analysis on ESRD

Fine mapping and functional annotations on ESRD loci

Association between ESRD loci and kidney traits

RESULTS

Identification of cross-sex loci associated with ESRD in X chromosome

Identification of sex-specific loci associated with ESRD in X chromosome

Association analysis conditioning on IgAN and PKD

Gene prioritization via eQTL colocalization analysis

Exploring association between ESRD-associated loci and renal traits

DISCUSSION

Declarations

Data Availability

Acknowledgments

Funding Statement

Author Contributions

Ethics Declaration

Ethics approval and consent to participate

Consent for publication

Competing interests

References

Additional Declarations

Supplementary Files

Status:

Version 1