Identification of Potential Oxidative Stress-Related Causal Genes for Calculus of Kidney and Ureter: A Multi-Omics Mendelian Randomization Analysis

Abstract

Introduction

Oxidative stress contributes to kidney and ureteral calculi, but the specific genes and mechanisms remain unclear. This study integrates methylation quantitative trait loci (mQTL), expression QTL (eQTL), and protein QTL (pQTL) data with genome-wide association study (GWAS) data to identify oxidative stress-related genes linked to calculi.

Methods

Summary data-based mendelian randomization (SMR) and colocalization analyses were performed using GWAS data from FinnGen (discovery) and UK Biobank (UKB; validation) to identify oxidative stress-related genes associated with calculus risk. External validation employed a genetic hypercalciuric stone-forming (GHS) rat transcriptomic dataset. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses, together with protein-protein interaction (PPI) network, were used to characterize functional pathways and central regulators.

Results

SMR and colocalization analyses in FinnGen identified 172 mQTLs, 27 eQTLs, and 5 pQTLs associated with calculus risk, with subsets validated in UKB. Multi-omics integration highlighted CREM (cg26679713) and MAPT (cg21705961) as key candidates, showing hypermethylation-associated upregulation. CREM was consistently associated with increased risk in both SMR analysis and the GHS model, whereas MAPT showed a protective association in SMR but was upregulated in GHS kidneys. GO and KEGG enrichment of 75 nonredundant genes from eQTL- and pQTL-level associations indicated predominant roles in oxidative stress and inflammation, and enrichment in FoxO, TNF, apoptosis, and IL-17/Th17 pathways. PPI analysis identified 11 hub genes linked to oxidative stress and inflammation.

Conclusion

Oxidative stress-related genes, particularly CREM and MAPT, are potential modulators of kidney and ureteral calculi risk, warranting further mechanistic and translational investigation.

Introduction

Kidney and ureteral calculi, also known as renal stones, nephrolithiasis, or urolithiasis, result from the crystallization of supersaturated minerals in the urine [1]. The prevalence ranges from 5% to 21.1% globally, with particularly high rates in southern Iran and Saudi Arabia [2]. Symptoms include severe flank pain, hematuria, and urinary obstruction, often requiring medical intervention [3]. Current therapeutic strategies focus on pain management, extracorporeal shock wave lithotripsy, and surgical removal, which are often invasive, costly, and may not prevent recurrence [4]. Studies suggest that over 45% of nephrolithiasis cases have a genetic component, involving both monogenic and polygenic factors [5]. However, the underlying genetic mechanisms of stone formation are not fully understood, limiting the development of targeted therapies.

Oxidative stress, caused by the imbalance between reactive oxygen species (ROS) and antioxidant defenses [6], contributes to the pathogenesis of kidney and ureteral calculi by inducing renal cellular damage and apoptosis. Excess ROS disrupts cellular homeostasis by oxidizing lipids, proteins, and DNA, leading to mitochondrial dysfunction [7]. Additionally, ROS stimulates pro-inflammatory signaling pathways like NF-κB and MAPK, exacerbating inflammation [8]. This inflammatory environment not only damages renal cells but also promotes the adhesion of calcium oxalate crystals to the damaged epithelium, facilitating crystal aggregation and stone formation [9]. Given the critical role of oxidative stress in stone formation, identifying oxidative stress-related genes involved in these pathways is critical for understanding the molecular mechanisms underlying kidney and ureteral calculi.

To investigate the causal roles of oxidative stress genes in calculi development, we employed summary data-based mendelian randomization (SMR) analysis [10]. This approach integrates cis-quantitative trait loci (cis-QTL) for DNA methylation (mQTL), gene expression (eQTL), and protein expression (pQTL) with genome-wide association study (GWAS) data. By identifying genetic variants associated with changes in DNA methylation, gene expression, or protein levels and linking them to kidney and ureteral calculi, SMR helps distinguish causal relationships from correlations, providing more precise insights into the genetic mechanisms underlying the disease [11]. Previous studies have demonstrated the power of SMR analysis in accurately identifying and validating pathogenic genes in Crohn’s disease [12], as well as its effectiveness in identifying biomarkers and drug targets in colorectal cancer [13]. These examples highlight the value of SMR in inferring causative genes or molecules and identifying potential therapeutic targets.

This study aims to investigate the potential pathogenic mechanisms of oxidative stress-related genes in kidney and ureteral calculi using multi-omics mendelian randomization methods. By integrating blood-based eQTL, mQTL, and pQTL data with GWAS data, we performed SMR and colocalization analyses to identify genetic variants that influence both QTLs and calculi risk. Through this approach, we seek to uncover key molecular mechanisms and identify novel therapeutic targets for the diagnosis and treatment of kidney and ureteral calculi.

Methods

Study Design

The oxidative stress gene-related mQTL, eQTL, and pQTL datasets were subjected to SMR analysis to evaluate their associations with kidney and ureteral calculi GWAS data sourced from public repositories. Colocalization analysis was used to identify shared causal variants influencing both QTLs and disease risk. Multi-omics integration was conducted to identify candidate genes supported at multiple regulatory levels. These genes were assessed in the GSE75542 transcriptomic dataset from a genetic hypercalciuric stone-forming (GHS) rat model and analyzed by Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and protein-protein interaction (PPI) network methods to identify relevant pathways and regulatory nodes. The flowchart of the study is presented in Figure 1. The protocol design and reporting of this study followed the “STROBE-MR checklist” (STROBE-MR) guidelines.

Fig. 1.

Study design. The flowchart illustrates the integration of oxidative stress-related gene QTL datasets (mQTL, eQTL, and pQTL) with GWAS data from the UK Biobank and FinnGen database. The mQTL dataset includes DNA methylation level-associated SNPs, the eQTL dataset includes DNA expression-associated SNPs, and the pQTL dataset includes protein expression-associated SNPs. These datasets underwent summary data-based mendelian randomization (SMR) analysis. The SMR results were then subjected to colocalization analysis to identify shared causal variants.

Data Source

A total of 2,550 oxidative stress-related protein-coding genes were obtained from the GeneCards database [14] with a relevance score greater than 2.511. The discovery dataset of kidney and ureter calculi was obtained from the FinnGen project (R10 release), including 10,556 cases and 40,068,141 controls. The validation datasets were sourced from UK Biobank (UKB), comprising 3,191 kidney and 2,417 ureteral calculi cases, along with 401,005 controls. The details are summarized in Table S1 (for all online suppl. material, see https://doi.org/10.1159/000551263). Summary data for blood mQTLs were derived from a meta-analysis of two European cohorts [15]: the Brisbane Systems Genetics Study (n = 614) and the Lothian Birth Cohorts (n = 1,366). eQTL data were obtained from a published article [16], comprising genetic data on blood gene expression from 31,684 European individuals. pQTL data were sourced from a previous study [17], including 54,219 UKB participants. The included studies were approved by the respective ethical review committees, and informed consent was obtained from all participants.

SMR Analysis

To identify oxidative stress-related genes associated with calculi, SMR analyses and heterogeneity in dependent instruments (HEIDI) tests were performed using the SMR software (v1.3.1) [18]. The SMR approach integrated summary-level GWAS with mQTL, eQTL, and pQTL data to detect pleiotropic associations that suggest causal relationships. The HEIDI test identifies whether shared genetic variation influences both exposure and outcome by analyzing associations across multiple SNPs within a region. Top associated cis-QTLs were selected within a ± 1,000-kb window around the gene to identify genetic variants (p value of <5.0 × 10⁻⁸), following well-established QTL mapping criteria [19, 20]. SNPs with allele frequency differences greater than 0.2 between any two datasets, including linkage disequilibrium reference samples, QTL summary data, and outcome summary data, were excluded. The maximum allowable proportion of SNPs with allele frequency differences was 0.05 (--diff-freq-prop 0.05). To enhance the validity of the findings, SMR_multi was used to incorporate multiple genetic variants simultaneously. p_SMR <0.05, p_SMR_multi <0.05, and p_HEIDI >0.05 were applied to determine significant associations, following well-accepted SMR analysis thresholds [15, 21]. False discovery rate values were also calculated using the Benjamini-Hochberg procedure and are reported in online supplementary Tables S2–S11 for reference.

Colocalization Analysis

Bayesian colocalization analysis was performed to determine whether the observed associations in both traits are driven by the same genetic variants, as it accounts for various sources of variation [22]. Significant associations (p_SMR_multi <0.05, p_SMR <0.05, and p_HEIDI >0.05) were examined using the R package “coloc” to identify shared genetic variants between cis-QTLs and the diseases [23]. SNPs within ± 1,000 kb of the top cis-QTL colocalization region windows were analyzed. Five mutually exclusive hypotheses (H0–H4) were tested: no association with either trait (H0), association with only the first trait (H1), association with only the second trait (H2), association with both traits but with different causal variants (H3), and association with both traits sharing the same causal variant (H4). A posterior probability (PPH4) greater than 0.8 indicates a significant colocalization [22]. In the “coloc” analysis, the prior probability P12 was set to 5 × 10⁻⁵.

Validation in Kidney Tissue Transcriptomes

To validate the expression patterns of SMR-prioritized genes, the GEO bulk transcriptomic dataset GSE75542 was analyzed, comprising kidney samples from GHS Sprague-Dawley rats (n = 3) and controls (n = 3) [24]. Raw cell intensity files were processed using the oligo (v1.66.0) package, and robust multi-array average normalization was applied. Probe annotation was performed with the ragene10sttranscriptcluster.db (v8.8.0) package, focusing on well-annotated core transcripts. Differential expression was assessed using a two-tailed t-test (ggsignif v0.6.4). Human-rat ortholog conversion was conducted with the homologene package (v1.4.68.19.3.27).

GO and KEGG Enrichment Analyses

Functional enrichment of nonredundant oxidative stress-related genes derived from eQTL- and pQTL-based SMR analyses was performed using the clusterProfiler package. GO enrichment was conducted for biological process, molecular function, and cellular component categories. KEGG pathway analysis excluded disease-related pathways. Significant terms were defined by p < 0.05. The top 15 pathways were visualized.

PPI Network Analysis

The STRING database (https://cn.string-db.org/) was used to construct a PPI network for oxidative stress-related genes derived from eQTL- and pQTL-based SMR analyses. A medium confidence threshold (interaction score of ≥0.400) was applied. The network was visualized in Cytoscape, and topological analysis was performed using the cytoHubba plugin (v0.1) with seven algorithms (Degree, Stress, Radiality, Closeness, MCC, MNC, and EPC). The intersection of the top 10-ranked genes across all algorithms defined the hub nodes.

Transcriptome-Wide Association Study

To investigate the association between genes encoding pathogenic proteins and kidney and ureteral stones, we performed a transcriptome-wide association study (TWAS) using the FUSION platform. Kidney tissue gene-level eQTL data were obtained from the GTEx Consortium V8 database (https://gtexportal.org/). The FUSION platform computes a linear sum of locus-specific, independent SNP Z-score weights. We then integrated the genetic effects on kidney and ureteral stones (GWAS Z-scores for kidney and ureteral stones) with mRNA expression weights. Using the FUSION platform, we calculated TWAS expression weights (i.e., SNP-gene expression associations) based on a reference transcriptome [25]. In our analysis, we employed multiple prediction models, including top1, BLUP, LASSO, elastic net (enet), and BSLMM. The model with the best predictive performance was selected to compute mRNA expression weights. Subsequently, we used the estimated gene expression levels to identify susceptibility genes associated with kidney and ureteral stones. Consistency in the direction of effect between TWAS Z-scores and those from the genome-wide proteome-wide mendelian randomization analysis, along with a TWAS p value of <0.05 (p-TWAS <0.05), was considered as evidence that the identified pathogenic genes passed the transcriptome-level TWAS validation.

Statistical Analysis

Statistical analysis was performed using R (v4.3.0). Manhattan plots and forest plots were generated using the “ggplot2” and “forestplot” packages, respectively. The locus plots and effect plots were generated as previously described [18].

Results

Multi-Omics SMR Screening of Oxidative Stress-Related Genes in Kidney and Ureteral Calculi

To comprehensively identify oxidative stress-related loci associated with kidney and ureteral calculi, we integrated mQTL-, eQTL-, and pQTL-based SMR analyses with GWAS data from FinnGen and UKB cohorts. At the methylation level, 528 CpG sites spanning 260 genes were associated with stone risk, with 76 loci from 37 genes showing significant colocalization evidence (PPH4 > 0.8). Ninety-one of these loci, corresponding to 21 genes, overlapped with subsequent eQTL-GWAS analysis (Fig. 2, online suppl. Table S2). Validation in UKB datasets confirmed 68 loci from 36 genes in the kidney calculi cohort and 22 loci from 14 genes in the ureteral calculi cohort. Of note, 6 loci from 4 genes, including SLC30A10 (cg19313402, cg21648401), DDR1 (cg14447703, cg06200824), MICB (cg13531176), and SREBF2 (cg16000331), were consistently validated in both cohorts (online suppl. Tables S3, S4).

Fig. 2.

SMR analysis of oxidative stress gene methylation associated with kidney and ureteral calculi risk. The forest plot demonstrates the results of SMR analysis for 91 loci from 21 genes, which overlapped with findings from subsequent eQTL-GWAS analysis. Each row represents a specific gene and its corresponding probe ID, along with the p value from SMR analysis (p_SMR), odds ratio (OR) with 95% confidence intervals (95% CI), multi-SMR p value (p_SMR_multi), and posterior probability of co-localization (PPH4). The plot highlights sites with significant colocalization evidence (PPH4 > 0.8), indicating shared genetic variants influencing both methylation levels and calculi risk.

At the transcript level, 61 oxidative stress-related genes were significantly associated with calculi (Fig. 3, online suppl. Table S5), including 29 with positive and the remainder with negative correlation with risk. Thirteen genes showed colocalization evidence. SF3B1 (OR = 1.39, 95% CI: 1.12–1.73) and ABL1 (OR = 1.72, 95% CI: 1.10–2.68) were associated with increased risk and replicated in the UKB ureteral calculi cohort (online suppl. Tables S6, S7).

Fig. 3.

SMR analysis of oxidative stress gene expression associated with calculi risk. The forest plot shows the results of eQTL-GWAS SMR analysis identifying 61 oxidative stress genes significantly associated with calculi. *PPH4 > 0.8.

At the protein level, 15 oxidative stress-related proteins were significantly associated with calculi risk, with CAVYBP and TNFRSF1B showing colocalization evidence (PPH4 > 0.8) (Fig. 4, online suppl. Table S8). Validation in UKB cohorts confirmed associations of MICB with kidney calculi (OR = 0.92, 95% CI = 0.87–0.97), HLA-DRA (OR = 0.89, 95% CI = 0.8–1), and HSPB6 (OR = 0.17, 95% CI = 0.03–0.89) with ureteral calculi (online suppl. Tables S9, S10). Together, this multi-omics screening identified oxidative stress-related genes and proteins consistently linked to stone risk, providing a prioritized pool for further integrative and mechanistic studies.

Fig. 4.

SMR analysis of oxidative stress-related proteins associated with calculi risk. The forest plot shows the SMR analysis results for 15 proteins significantly associated with calculi. *PPH4 > 0.8.

Integration of Multi-Omics Analyses Highlights CREM and MAPT

To better understand the molecular mechanisms underlying calculi risk, we performed an integrative analysis to explore potential causal relationships between gene methylation, expression, and protein levels. The integration of mQTL and eQTL summary data revealed significant associations between specific methylation sites and gene expression (online suppl. Table S11). Negative associations between methylation and gene expression were found for XPO1 (cg15711740) (OR = 0.91, 95% CI = 0.86–0.97) and RPS2 (cg02871659) (OR = 0.89, 95% CI = 0.87–0.92), indicating that higher methylation at these sites may reduce gene expression. On the other hand, positive associations between methylation and gene expression were observed for CREM (cg26679713) (OR = 1.49, 95% CI = 1.33–1.68), RBL2 (cg01016459) (OR = 12.69, 95% CI = 7.24–22.23), and MAPT (cg21705961) (OR = 1.04, 95% CI = 1.01–1.07), suggesting that increased methylation at these sites may enhance gene expression (Table 1). No significant associations were observed in the integration of eQTL and pQTL SMR analysis, possibly due to the limited pQTL coverage.

Table 1.

Integration of mQTL-GWAS and eQTL-GWAS analyses

Expo_ID	Outco_Gene	p_SMR	p_SMR_multi	p_HEIDI	OR_SMR (95% CI)
cg15711740	XPO1	0.005	0.003	0.259	0.91 (0.86–0.97)
cg26679713	CREM	3.30E-11	2.02E-12	0.127	1.49 (1.33–1.68)
cg02871659	RPS2	7.30E-15	4.89E-11	0.403	0.89 (0.87–0.92)
cg01016459	RBL2	7.01E-19	7.01E-19	0.352	12.69 (7.24–22.23)
cg21705961	MAPT	0.004	4.38E-04	0.055	1.04 (1.01–1.07)

Considering significant colocalization evidence for CREM and MAPT in mQTL-GWAS (Fig. 2, online suppl. S1A, S2A, S2B) and eQTL-GWAS (Fig. 3, online suppl. S1B, S2C, S2D) analyses, we further examined their roles in kidney and ureteral calculi risk. In the mQTL-GWAS analysis, CREM (cg26679713) methylation was positively associated with increased risk (OR = 1.19, 95% CI = 1.06–1.33, PPH4 = 0.64; Fig. 2, online suppl. S3A), while MAPT (cg21705961) methylation was negatively associated with the risk (OR = 0.90, 95% CI = 0.84–0.96, PPH4 = 0.84; Fig. 2 and online suppl. S3B). Similarly, in eQTL-GWAS analysis, CREM expression was positively linked to calculi risk (OR = 1.54, 95% CI = 1.18–2.01, PPH4 = 0.64; Fig. 3, online suppl. S3c), while MAPT expression showed a protective effect (OR = 0.48, 95% CI = 0.29–0.80, PPH4 = 0.84; Fig. 3, online suppl. S3D). Furthermore, the integration of methylation-expression relationships suggests that hypermethylation of CREM (cg26679713) and MAPT (cg21705961) may increase their expression. Genomic annotation based on the UCSC Genome Browser indicated that cg21705961 is located within the MAPT gene body, whereas cg26679713 resides in a GeneHancer-predicted distal regulatory region of CREM (online suppl. Fig. S4). In this context, the positive methylation-expression association observed for MAPT reflects gene body methylation, which is associated with active transcription and the regulation of transcriptional elongation and splicing [26]. For CREM, the association at cg26679713 reflects shared genetic regulation acting on distal enhancer activity and CREM transcription [27]. Based on these results, we hypothesize that hypermethylation of CREM (cg26679713) may upregulate its gene expression, thereby increasing the risk of kidney and ureter calculi. Conversely, hypermethylation of MAPT (cg21705961) may also upregulate its gene expression, but this could potentially reduce the risk of kidney and ureter calculi.

External Validation of Crem and Mapt Expression in a GHS Rat Model

Given the multi-omics evidence linking CREM and MAPT to kidney and ureteral calculi, we evaluated their transcriptional profiles in kidney tissue from a GHS rat model (GEO: GSE75542), which closely mimics the pathophysiology of human calcium-based nephrolithiasis [24]. After robust multi-array average normalization and human-rat ortholog mapping, differential expression analysis between GHS rats (n = 3) and wild-type controls (n = 3) revealed significant upregulation of Crem and Mapt in GHS kidneys (Fig. 5). Crem upregulation was consistent with SMR-predicted risk, whereas Mapt upregulation was opposite to the protective effect predicted by SMR.

Fig. 5.

Expression of Crem and Mapt in kidney tissue from a genetic hypercalciuric stone-forming (GHS) rat model. Box plots display normalized transcript levels in kidney samples from wild-type controls (n = 3) and GHS rats (n = 3) based on the GSE75542 dataset. *p < 0.05.

Functional Enrichment and PPI Network Analysis

To elucidate the biological context of the SMR-prioritized oxidative stress-related genes in kidney stone disease, we performed GO and KEGG enrichment analysis on 75 nonredundant candidates derived from 61 eQTL- and 15 pQTL-associated genes/proteins. The most enriched biological processes included “response to peptide hormone” and “response to oxidative stress,” pointing to hormonal and redox regulation in renal pathophysiology. KEGG pathways were dominated by FoxO, TNF, and apoptosis signaling, along with immune-related modules such as IL-17/Th17 differentiation, outlining a mechanistic link between oxidative stress, inflammation, and cell death in stone formation (Fig. 6; online suppl. Table S12). PPI network analysis revealed a densely connected gene set. Eleven hub nodes consistently ranked highest across seven topology algorithms: ABL1, ESR1, GSK3B, MAP2K1, MAPT, PTGS2, STAT5B, TGFB1, TRAF6, XPO1, and XRCC6 (online suppl. Fig. S5; Table S13). These results integrate multi-omics genetic associations into coherent pathways and central network regulators with potential mechanistic relevance to nephrolithiasis.

Fig. 6.

GO and KEGG enrichment analysis of SMR-prioritized oxidative stress-related genes in kidney stone disease. The top enriched 15 terms from GO biological process (BP), cellular component (CC), molecular function (MF), and KEGG pathway analyses are shown. Dot size indicates the number of genes in each term; color scales represent adjusted p values for BP, CC, MF, and KEGG categories.

Investigation of Kidney-Specific Effects via TWAS

To assess whether the candidate genes identified by SMR also show functional relevance in kidney tissue, we performed a TWAS using GTEx kidney expression data. However, no significant associations were observed between the genetically predicted expression of any SMR-prioritized gene and kidney stone risk (all p-TWAS >0.05; online suppl. Table S14). This finding suggests that the genetic effects identified by SMR may primarily reflect systemic regulatory mechanisms rather than kidney-specific transcriptional changes.

Discussion

In this study, we conducted SMR analysis on mQTL, eQTL, and pQTL data, identifying oxidative stress-related genes potentially involved in kidney and ureteral calculi. Integration of multi-omics evidence prioritized CREM and MAPT, with methylation-expression patterns suggesting divergent effects on calculi risk. Transcriptomic assessment in the GHS rat model supported differential expression of these genes in stone-forming kidneys. Functional characterization through GO and KEGG enrichment, together with PPI network analysis, indicated their involvement in oxidative stress regulation, TNF signaling, and apoptosis. These convergent findings provide mechanistic insights and potential molecular targets for future therapeutic intervention.

GO and KEGG enrichment highlighted apoptosis, oxidative stress response, and inflammatory pathways such as TNF, IL-17, and Th17 cell differentiation as central processes linking CREM, MAPT, and other prioritized genes to kidney stone pathogenesis. For example, TGFB1, an important PPI hub, mediates CaOx crystal-induced tubular injury via activation of the TGF-β/Smad pathway, promoting apoptosis, oxidative stress, epithelial-mesenchymal transition, and fibrosis through regulatory interactions such as lncRNA-ATB-miR-200 signaling, which contribute to the renal injury environment underlying stone development [28]. Likewise, RBL2, which encodes retinoblastoma-like protein 2 (p130) that regulates cell cycle progression and apoptosis, is targeted by miRNAs such as rno-miR-672-5p and rno-miR-674-5p that are upregulated in hypercalciuric stone-forming kidneys, influencing inflammation, oxidative stress, and calcium metabolism [24]. In addition, ESR1, another PPI hub that modulates estrogen-mediated signaling and exerts antioxidant and anti-apoptotic effects, was upregulated in a calcium oxalate rat model following treatment with Bushen Huashi decoction, suggesting its involvement in protective pathways against oxidative stress and tubular epithelial cell injury in kidney stone disease [29]. Collectively, these findings suggest that kidney stone pathogenesis involves a complex interplay of pro-injury and protective signaling networks, and that nodal regulators may represent mechanistic targets for modifying disease risk.

CREM (cAMP response element modulator) hypermethylation was associated with increased expression and elevated kidney stone risk, supported by multi-omics integration and upregulation in the GHS model. As a member of the CREB transcription factor family, CREM orchestrates transcriptional programs that modulate oxidative stress and inflammatory responses [30]. Under acute stress, CREM induces antioxidant enzymes such as superoxide dismutase and catalase to neutralize ROS [31]. However, persistent activation shifts CREM function toward upregulating pro-inflammatory cytokines, including TNF-α and IL-6 [32]. In renal epithelial cells, CREM has been implicated in cystogenesis and inflammation, processes that share mechanistic overlap with crystal-induced tubular injury [33]. Our enrichment analysis additionally linked CREM to the “membrane lipid metabolic process” (with CYP1B1 and GLB1), suggesting that CREM-mediated lipid remodeling may influence membrane microdomain organization and receptor localization, thereby potentiating cytokine signaling and ROS-lipid peroxidation feedback loops in stone-forming kidneys. Association with “adrenergic signaling in cardiomyocytes” reflects its integration into cAMP-PKA pathways [34]. Sustained activation of cAMP-PKA signaling in renal epithelial cells can promote stone formation by driving downstream pro-lithogenic pathways [35]. Our study showed that CREM (cg26679713) hypermethylation may increase its expression. CREM hypermethylation, induced by silica nanoparticles, can trigger mitochondrial-dependent apoptosis via caspase-9/3 activation and Bax upregulation [36], exacerbating tubular injury and facilitating crystal deposition [37]. Taken together, our findings suggest that CREM is a potential upstream regulator linking oxidative stress, membrane lipid remodeling, inflammatory signaling, and apoptosis in the pathogenesis of kidney and ureteral calculi.

In contrast to CREM, MAPT (cg21705961) hypermethylation in blood QTL data was associated with increased gene expression and a reduced risk of kidney and ureteral calculi, suggesting a potential protective role. MAPT encodes the tau protein, which stabilizes microtubules and supports cytoskeletal integrity [38]. Proper microtubule function is essential for maintaining cellular homeostasis, facilitating autophagic clearance of damaged organelles and crystal fragments, and preventing the accumulation of injury signals that promote nephrolithiasis [39]. Enhanced MAPT expression may therefore preserve tubular epithelial integrity, inhibit crystal adhesion, and reduce calcium oxalate-induced cell injury [40]. GO enrichment linked MAPT to immune and inflammatory regulation, including neuroinflammatory response, myeloid leukocyte activation, and macrophage activation, while KEGG analysis placed MAPT in MAPK signaling together with MAP2K1, MAPKAPK3, TGFB1, TRAF2, and TRAF6, of which MAP2K1, TGFB1, and TRAF6 are PPI hub nodes and established mediators of oxidative stress, inflammation, and epithelial-mesenchymal transition. This co-enrichment suggests that increased MAPT expression may stabilize cytoskeletal and vesicular systems under oxidative stress, attenuate pro-inflammatory signaling, and preserve epithelial integrity, collectively mitigating pathways that induce kidney stone formation.

However, transcriptomic profiling of kidneys from the GHS rat model revealed MAPT upregulation in stone-forming tissue, opposite to the protective association predicted by SMR. This apparent contradiction likely reflects regulatory heterogeneity between blood and kidney tissue, as well as disease-stage-specific effects. Kidney and ureteral calculi are not purely local disorders but are associated with systemic metabolic and inflammatory dysregulation [41]. Blood-derived QTLs primarily capture stable germline regulatory architecture that shapes systemic metabolic and inflammatory set points, whereas renal transcriptomic changes reflect local responses under active injury. In systemic blood-based QTLs, elevated MAPT expression may indicate an inherited background associated with cytoskeletal stability and efficient autophagy, whereas in diseased renal tissue, MAPT upregulation may represent an acquired, stress-responsive transcriptional change. In this scenario, MAPT induction may be insufficient to restore microtubule homeostasis, as crystal deposition, ROS accumulation, and pro-inflammatory signaling could overwhelm cytoprotective pathways. This interpretation is supported by prior findings that MAPT dysfunction promotes mitochondrial injury, ROS overproduction, and calcium dysregulation [42], suggesting that MAPT upregulation in injured kidneys may be a marker of cellular stress rather than an effective defense. These results highlight the importance of integrating tissue-specific transcriptomic data with systemic genetic associations when inferring pathogenic mechanisms.

An important aspect of our study was to bridge the findings from blood-based QTLs to renal pathophysiology. To this end, a kidney-specific TWAS was performed. Interestingly, the SMR findings were not validated, suggesting that the genetically predicted expression of genes like MAPT in healthy kidney tissue is not associated with calculus risk. This discrepancy is an important finding in itself and points to several non-mutually exclusive possibilities. Firstly, it strongly suggests tissue-specific effects, where the genetic associations captured in blood may reflect systemic pathways, such as inflammation or metabolic dysregulation, which are known contributors to stone formation, rather than a direct cis-regulatory mechanism within the kidney. Secondly, there is a substantial difference in statistical power between the blood eQTL cohort (n > 31,000) used for SMR and the GTEx kidney cohort (n ≈ 900). The smaller sample size of GTEx may be insufficient to generate robust expression prediction models for all genes or to detect subtle associations. Finally, the genetic regulation in healthy postmortem tissue from GTEx may not fully represent the dynamic, pathological processes involved in active stone formation. Therefore, the combined SMR and TWAS results suggest that the candidate genes may influence stone risk primarily through systemic mechanisms, which warrants further investigation.

This study identifies significant associations between oxidative stress-related gene methylation, expression, and protein levels in kidney and ureteral calculi, providing insights into potential risk and protective mechanisms. A limitation is the reliance on blood-derived QTL datasets, which may not fully reflect kidney-specific regulatory patterns. The inconsistency of MAPT between SMR and kidney tissue expression highlights the need for direct tissue-based QTL mapping and functional assays to resolve context-dependent effects. Protein-level inference is further constrained by the limited availability of pQTL data. Although the largest publicly available plasma pQTL resource from the UK Biobank (54,219 participants) was used, pQTL information was available for only 521 of the 2,550 oxidative stress-related protein-coding genes, and among the 91 loci corresponding to 21 genes overlapping mQTL- and eQTL-GWAS analyses, only SERPINC1 and FUT8 had available pQTL data. Consequently, protein-level evidence was sparse and served as supplementary support, not a primary criterion for locus prioritization. In addition, plasma pQTLs mainly capture circulating, secreted, or leakage-associated proteins and may underrepresent kidney-intrinsic structural or matrix components relevant to stone formation. Future availability of larger datasets and kidney tissue-specific pQTL resources may expand proteomic coverage and facilitate the identification of additional kidney-relevant proteins. From a pathway perspective, the results suggest that TNF signaling and apoptosis are potentially influenced by genes like MAPT and may contribute to the pathogenesis of nephrolithiasis. Future functional studies, including in vivo experiments using animal models and validation in patient-derived kidney tissue, are warranted to clarify these mechanisms and assess their therapeutic relevance.

In conclusion, this study integrates blood-derived QTL and GWAS data with GO, KEGG, and PPI analyses to identify oxidative stress-related loci linked to kidney and ureteral calculi. CREM emerged as a consistent risk-promoting gene in both SMR analysis and transcriptomic data from a public GHS rat model, enriched in pathways related to oxidative stress, inflammation, and adrenergic signaling. MAPT, despite showing a protective association in SMR, was upregulated in the rat model and mapped to immune activation and MAPK signaling together with network hubs such as MAP2K1 and TGFB1, suggesting tissue-specific or context-dependent regulation. These findings prioritize CREM and MAPT for mechanistic investigation, while also underscoring the need for kidney-specific QTL mapping and functional validation in human renal tissue. Our results highlight the critical need for future functional studies to dissect the roles of genes such as MAPT and the associated TNF signaling and apoptosis pathways in the pathogenesis of kidney stones.

Statement of Ethics

An ethics statement was not required for this study type, as it is based exclusively on data extracted from public repositories.

Conflict of Interest Statement

All authors declare that they have no conflicts of interest.

Funding Sources

This study was supported by the Key Scientific Research Foundation of the Education Department of Province Anhui (2023AH051938), the Health Research Program of Anhui(AHWJ2023A30170), the Key Scientific Research Foundation of Bengbu Medical College(2022byzd028), and the Key Scientific Research Foundation of Bengbu Medical College(2020byzd090).

Author Contributions

Wu Yue Gao and Xin Liang carried out the studies, participated in collecting data, and drafted the manuscript. Yuan Yuan Guo and Bei Bei Liu performed the statistical analysis and participated in its design. Wei Sun and Yi Ming Sun participated in the acquisition, analysis, or interpretation of data and drafted the manuscript. All authors read and approved the final manuscript.

Funding Statement

This study was supported by the Key Scientific Research Foundation of the Education Department of Province Anhui (2023AH051938), the Health Research Program of Anhui(AHWJ2023A30170), the Key Scientific Research Foundation of Bengbu Medical College(2022byzd028), and the Key Scientific Research Foundation of Bengbu Medical College(2020byzd090).

Data Availability Statement

All data generated or analyzed during this study are included in this published article. Further inquiries can be directed to the corresponding author.