The carbonic anhydrase 3 protein in urine: a potential biomarker to monitor atherosclerotic renal artery stenosis

Abstract

Background

Atherosclerotic renal artery stenosis (ARAS) represents the predominant etiology of renal artery stenosis. Current diagnostic modalities—including renal arteriography, Doppler ultrasonography (DUS), computed tomography angiography (CTA), and magnetic resonance angiography (MRA)—are limited by invasiveness or technical constraints. The development of noninvasive biomarkers for ARAS detection is therefore clinically imperative. We hypothesize that ischemic renal injury in ARAS induces pathological alterations that generate unique urinary protein signatures.

Methods

A total of ​138 subjects​ were enrolled and ​randomly divided into the discovery cohort and the validation cohort. Urinary proteins from the discovery cohort were profiled using ​data-independent acquisition mass spectrometry (DIA-MS)​​ to identify differentially expressed proteins. Candidate biomarkers were subsequently validated in the validation cohort via ​quantitative ELISA. Diagnostic performance was assessed by ​receiver operating characteristic (ROC) curve analysis, and associations between target protein levels and clinical parameters were evaluated using ​Spearman correlation​.

Results

As a result of DIA-MS indicated in the discovery cohort, 485 up-regulated urinary proteins and 177 down-regulated urinary proteins were identified in ARAS patients compared to disease controls. The top three proteins CA3, ORM2, and ART3 in AUC were selected as the potential urinary markers for further validation in the validation cohort. The uCA3/Cr level was significantly higher in both ARAS and disease controls than in healthy controls, and the uCA3/Cr level in patients with ARAS was significantly higher than in disease controls. There was no statistical difference in the uCA3/Cr level in subgroups with different severity of ARAS. Taking healthy controls as control values, the ROC area of uCA3/Cr was 0.9649(95%CI: 0.9200-1.000). We evaluated the area under the uCA3/Cr curve with the values of the disease control group as the control, the areas of ROC were 0.7391(95%CI: 0.6154–0.8628). Correlation analysis demonstrated that there was a strong positive correlation between uCA3/Cr concentration and urinary a1-microglobulin(a1MG) in ARAS and disease controls.

Conclusion

CA3 in urine may be related to renal interstitial injury, it could be used as a non-invasive biomarker for ARAS. However, large cohort studies and mechanistic investigations are required for further validation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12882-025-04457-w.

Introduction

Renal artery stenosis (RAS) refers to the stenosis of one or more main renal arteries or their branches, which damages the blood flow of the kidney [1]. The leading etiologic factor for RAS is atherosclerosis, which accounts for more than 90% of the patients age > 40 years [2], followed by fibromuscular dysplasia (FMD) and other less common causes [3]. As patients with RAS are often asymptomatic or without severe hypertension (HTN), the true prevalence is unknown. Within the United States Medicare population, DUS identified atherosclerotic renal artery stenosis (ARAS) in 1%-7% of individuals [4]. The study has shown that ARAS exists in 7% of the general population over 65 years old and 20%ཞ45% of patients with coronary heart disease and aortoiliac artery disease [5].

RAS is related to three main clinical syndromes: secondary hypertension, ischemic nephropathy, and unstable cardiac syndromes [6]. Pathological changes in the kidney caused by abnormal renal blood flow are the key factors in the development of RAS complications [7]. A large of interrelated mechanisms have been reported to explain how hemodynamically significant lesions eventually lead to interstitial fibrosis. For one thing, repeated local ischemia results in tubulointerstitial injury and microvascular damage. For another, global renal hypoperfusion causes changes in endothelial and epithelial factors and activation of the renin-angiotensin aldosterone system as well as subsequent vasoconstriction. These two aspects are believed to lead to oxidative damage, increased production of fibrotic cytokines, and inflammation, eventually contributing to atrophy and fibrosis [8]. The study has found that one third of patients who plan to undergo nephrectomy surgery have kidney damage caused by ARAS [9]. In the study by Wright et al., they followed patients with biopsy-proven atherosclerotic nephropathy for 2 years. In the group with worsening renal function, histological evidence suggested more extensive interstitial atrophy and glomerulosclerosis [10]. These pathological changes in the kidneys caused by abnormal renal blood flow may produce specific protein biomarkers. Therefore, searching for specific protein biomarkers will be helpful for the diagnosis and prognosis monitoring of ARAS.

At present, the golden diagnosing standard for ARAS is renal arteriography, but it is invasive. DUS, CTA, and MRA are all non-invasive imaging modalities [11]. These imaging examinations have drawbacks such as long detection time, high detection costs, invasiveness, additional harm, and are not suitable for screening large-scale population. In laboratory tests, such as fibrinogen, homocysteine, lipoproteins, C-reactive protein and serum creatinine levels, they have a certain predictive effect on ARAS, but their clinical values remain to be further determined [12]. Therefore, there is a lack of sensitive and specific diagnostic biomarkers for ARAS.

Urine, a non-invasive sample, has attracted much attention for large-scale clinical screening programs [13]. Meanwhile, urine can accumulate early changes in biomarkers that refer to biological processes regulated by homeostasis mechanisms [14]. Studies have demonstrated that urinary proteomics has the potential to monitor changes in our body sensitively and provide the possibility of identifying early biomarkers of atherosclerosis [15], coronary artery disease (CAD) [16], kidney disease [17] and hypertension [18].

As mentioned above, we speculate that patients with ARAS will produce specific protein biomarkers in urine due to an array of pathological changes caused by ischemic kidney injury. First, we performed DIA-MS to screen differentially expressed proteins (DEPs) in urine from ARAS patients and disease controls. Second, ELISA was applied to validate DEPs in ARAS patients, disease and healthy controls. Finally, we evaluated the protein’s diagnostic efficacy for ARAS using the receiver operating characteristics curves (ROC) analysis, therefore providing a suitable, sensitive and specific monitoring biomarker for ARAS.

Materials and methods

Study participants

Patients within the age range 40∼80 years, were screened for enrollment between July 2020 and October 2022 at the Aerospace Center Hospital (Beijing, China). One hundred and thirty-eight objects were enrolled and divided into the discovery cohort and the validation cohort. The discovery cohort consisted of 28 ARAS patients with sex and age-matched and 14 disease controls. Another 46 patients with ARAS, 30 disease controls and 20 healthy controls were included in the validation cohort. Diagnostic criteria for ARAS based on the 2017 Chinese Expert Consensus on the Diagnosis and Management of Renal Artery Stenosis: (1) angiography or renal artery ultrasound indicated ARAS: renal angiography can provide images of the distribution, degree of stenosis, and anatomical features of the lesion. Ultrasound morphology was used to determine whether there was lumen stenosis and calculate the stenosis rate [stenosis rate = (1-residual diameter at stenosis/normal lumen diameter)×100%]. The stenosis rate < 50% :mild stenosis, 50%≤stenosis rate < 70%:moderate stenosis, stenosis rate ≥ 70% :severe stenosis [19, 20]; (2) at least one risk factor for atherosclerosis (diabetes, obesity, hyperlipidemia, > 40 years old, long-term smoking); (3) at least 2 imaging manifestations of atherosclerosis (conical stenosis or occlusion of renal artery, eccentric stenosis, irregular plaque, calcification, mainly involving the proximal segment and opening of renal artery; manifestations of other abdominal vascular atherosclerosis). Exclusion criteria were: (1) renal artery stenosis caused by FMD and Takayasu arteritis; (2) malignant tumor. (3) Current use of angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), or sodium-glucose cotransporter-2 (SGLT2) inhibitors prior to enrollment. (4) Subjects exhibiting macroalbuminuria (UACR > 300 mg/g creatinine) were prospectively excluded from subsequent analyses in accordance with KDIGO clinical practice guidelines (2021). Disease controls (DC) were patients hospitalized in the same period as ARAS patients, they met the following inclusion criteria:1) > 40years old; 2) angiography or imaging examination did not show plaque or renal artery stenosis. Healthy controls (HC) came from the same period of the health examination population in our hospital, who met the following inclusion criteria:1) > 40years old; 2) apparently healthy individuals; 3) no medical history of hypertension, diabetes, coronary heart disease, hyperlipidemia, and malignant tumor. The experiment received approval from the Ethics Committee of the Aerospace Center Hospital. All participants signed informed consent before inclusion in this study and specimen collection. All procedures adhered to the ethical standards outlined in the Declaration of Helsinki.

Data collection

The measurement and collection of clinical and laboratory data were applied in the clinical laboratory of the Aerospace Center Hospital. Clinical data included sex, age, smoking, body mass index (BMI), hypertension (HT), diabetes mellitus (DM), coronary heart disease (CHD), cerebrovascular disease (CVD). Laboratory data included creatinine (CREA); estimated glomerular filtration rate (eGFR); alanine aminotransferase (ALT); triglyceride (TG); total cholesterol (TC); high-density lipoprotein (HDL); low-density lipoprotein (LDL); glycosylated hemoglobin (HbA1c); urine albumin (uALB); urine a1 microglobulin (ua1MG); urine immunoglobulin G (uIgG); urine n-acetyl-beta-d-glucosaminidase (uNAG). The eGFR was calculated according to Chronic Kidney Disease Epidemiology Collaboration equation: eGFR = 175×(Scr)-1.154×(Age)-0.203×[0.742(female)].

Urine sample collection and MS analysis

Clean morning urine samples were collected from the subjects, and the urine was centrifuged (400×g, 5 min) to remove cellular debris. The urine supernatants were then stored at -80℃ for further analysis. One milliliter of each urine sample was diluted with lysis buffer (50 Mm NH4HCO3, pH7.4, 10 Mm MgCl2, 7 M urea, 2 M thiourea), followed by a 10KD ultraffltration (Sartorious, Göttingen, Germany) and centrifugation (12000×g, 15 min, 4℃). The supernatant protein was collected in new tubes and reduced with 20mM Dithiothreitol (DTT) for 30 min at 55℃ and alkylated with 40 Mm iodoacetamide at 25℃ for 15 min. The protein was then digested with trypsin (Promega, Madison, USA) at 1:50 enzyme-to-protein ratio for 16 h at 37℃. The peptides were desalted using C18 tips and then dried by vacuum evaporation. Protein concentration was measured with Bradford assay. Pooled peptide sample was obtained by mixing equal number of digested peptides of each group sample for generating a spectral library.

LC-MS/MS analysis

For analysis by LC-MS/MS (Q Exactive HF-X mass spectrometer, Thermo Fisher Scientific), the processed urine peptides were redissolved in 0.1% FA-H₂O and placed onto a trap column (100 μm×2 cm; particle size, 3 μm; pore size, 120 Å; SunChrom, USA), and then separated by a silica microcolumn (150 μm×10 cm, particle size, 1.9 μm; pore size, 120 Å; SunChrom, USA) with a gradient of 5%-35% mobile phase B (20%H2O/80%ACN, 0.08%FA) at a flow rate of 800nl/min for 30 min. The data dependent acquisition (DDA) mode of the Q Exactive HF-X mass spectrometer was selected to obtain multiple scan. For full scan, the samples were analyzed by the following condition: 300–1400 m/z Orbitrap, 60,000 resolution, 3e⁶ AGC target and 20ms maximum injection time. For the MS/MS scan, the analysis was carried out in a mode with a nominal mass resolution power of 7,500, AGC target of 5e⁴, maximum injection time of 12ms, and dynamic exclusion time of 15 s.

Protein identification and label-free quantification

The MS data were processed on the Firmiana platform. Protein identification was performed using the MASCOT search engine (Matrix Science, version 2.3.01) in the NCBI human RefSeq protein database. The false discovery rate (FDR) of protein identifications was set to 1%. Those with ≥ 1 unique and strict peptides and ≥ 2 strict peptides (ion score>20) or ≥ 3 strict peptides comparable to 1% FDR at the protein level were identified for subsequent analysis. Proteins were quantified using the intensity based absolute quantification (iBAQ) algorithm. To standardize the difference in sample size, the iBAQ value of each protein was converted to iFOT(fraction of total) calculated by dividing the iBAQ value by the total iBAQ of the sample, and then multiplied by 105 to facilitate visualization. All missing values were superseded by zero.

Bioinformatics analysis

Fold Change (FC) > 1.5 or < 0.667 and also p-value < 0.05 were used as cutoffs for screening significantly expressed proteins. Clustering of multivariate data was visualized using Heatmap (http://biit.cs.ut.ee/clustvis) [21]. The Gene ontology (GO) function and the Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations were performed with the ClusterProfiler version 4.4.1 package [22, 23]. Protein-protein interaction (PPI) networks were performed with the STRING online database (https://cn.string-db.org/).

Validation by ELISA analysis

According to the reagent instructions, commercial ELISA kits for carbonic anhydrase 3 (CA3), Ecto-ADP-ribosyltransferase 3 (ART3) and Alpha-1-acid glycoprotein 2 (ORM2) (Zeye Biotechnology Co., Ltd. Shanghai, China) were used for quantitative analysis. Primarily, urine samples were loaded into individual wells and incubated in an incubator for 50 min. Subsequently, the wells were incubated with anti-human CA3, ART3, or ORM2 antibodies for 50 min, and then incubated with anti-human IgG antibodies for 30 min. TMB/peroxide substrate incubation was used to perform colorimetric detection and the OD (450 nm) at 15 min was read on an enzyme calibration (sunrise, TECAN, Austria). Finally, sample concentrations were calculated according to the simultaneous standard curves generated by the standard materials. uCA3, uART3, and uORM2 levels and uALB, ua1MG, uIgG and uNAG levels were normalized with urinary creatinine (uCr), which was measured on AU4800 (Beckman, USA).

Statistical analysis

All statistical analyzes were performed with SPSS (version 27.0; IBM Corporation, USA) and GraphPad Prism 10.0 software (La Jolla, CA, USA). In continuous data, normal distribution data were compared using the Independent-Sample T-Test between two groups or the one-way ANOVA test between multiple groups, and skewed distribution data were compared using the Mann-Whitney U test between two groups or the Kruskal-Wallis H test between multiple groups. Pearson’s chi-squared test was used for the comparison of categorical data between groups. The receiver operating characteristics (ROC) curves was used to analyze the diagnostic efficacy of selected biomarkers. Spearman correlation analysis was used to analyze the correlation between target biomarkers and clinical indicators. P < 0.05 was considered statistically significant.

Results

Patients

A summary of the clinical characteristics of the subjects and their laboratory values is shown in Table 1. There were no significant differences in all clinical and laboratory values between ARAS patients and the disease controls in the discovery cohort (P > 0.05). For the validation cohort, there were significant differences in CHD, BMI, and ua1MG between the three groups.

Table 1.

Clinical characteristics and laboratory values in discovery and validation cohort

Variable	discovery cohort			validation cohort
	ARAS (n = 28)	Disease control (n = 14)	P value	ARAS group (n = 46)	Disease control (n = 30)	Healthy control (n = 20)	P value ARAS vs. DC	P value ARAS vs. HC	P value DC vs. HC
Male (%)	14(50)	7(50)	0.378	22(48)	18(60)	10(50)	0.225	0.774	0.209
Age(years)	73 ± 9	67 ± 13	0.086	70 ± 7	67 ± 7	46 ± 5	0.053	< 0.001	< 0.001
HT (%)	27(96)	14(100)	0.487	44(96)	30(100)	-	0.465	-	-
DM (%)	13(46)	6(43)	0.824	16(35)	13(43)	-	0.505	-	-
CHD (%)	17(61)	9(64)	0.824	32(70)	11(37)	-	0.005	-	-
CVD (%)	14(50)	4(29)	0.171	14(30)	7(23)	-	0.469	-	-
smoking (%)	13(46)	5(36)	0.498	15(33)	15(50)	-	0.153	-	-
BMI (kg/m²)	24.9 ± 4.4	26.3 ± 4.5	0.421	25.5 ± 3.4	27.6 ± 3.5	23.8 ± 3.6	0.016	0.082	< 0.001
CREA(µmol/L)	87.3 ± 29.4	81.9 ± 24.7	0.575	79.8 ± 33.3	75.6 ± 19.0	70.7 ± 12.9	0.504	0.199	0.513
eGER	84.47 ± 34.44	90.33 ± 9.67	0.667	86.68 ± 33.10	93.05 ± 19.93	96.29 ± 12.04	0.347	0.213	0.518
ALTa(IU/L)	15.1(11.7,24.0)	17.8(15.2,22.0)	0.233	17.2(12.1,29.0)	21.8(17.1,32.4)	19.4(13.5,23.0)	0.041	0.989	0.048
TGa(mmol/L)	1.4(1.1,1.8)	1.4(1.1,2.0)	0.864	1.4(1.1,1.8)	1.8(1.2,2.7)	1.0(0.8,1.1)	0.071	< 0.001	< 0.001
TC(mmol/L)	4.9 ± 1.8	4.9 ± 1.0	0.869	4.7 ± 1.6	4.5 ± 1.2	4.4 ± 0.9	0.479	0.372	0.805
HDL(mmol/L)	1.1 ± 0.2	1.2 ± 0.2	0.354	1.0 ± 0.3	1.0 ± 0.2	1.2 ± 0.3	0.194	0.038	0.003
LDL(mmol/L)	2.7 ± 1.3	2.6 ± 0.9	0.822	2.6 ± 1.2	2.5 ± 0.8	2.6 ± 0.6	0.465	0.925	0.600
HbA1c (%)	7.0 ± 1.4	6.5 ± 2.1	0.485	6.6 ± 1.2	7.0 ± 2.2	5.6 ± 0.4	0.285	0.031	0.004
uALB(mg/L)	36.0(13.4,216.8)	11.0(3.5,15.0)	0.014	17.8(8.5,35.9)	15.7(9.7,36.9)	9.1(5.5,11.7)	0.683	0.114	0.250
ua1MG (mg/L)	18.8(12.1,45.1)	2.8(2.4,6.5)	< 0.001	12.6(4.4,27.9)	6.8(3.5,14.4)	5.4(2.4,9.0)	0.007	0.008	0.766
uIgG(mg/L)	7.4(3.6,36.3)	3.9(2.6,4.1)	0.044	5.2(3.2,10.0)	4.2(3.1,10.2)	3.2(2.6,4.6)	0.644	0.180	0.378
uNAG(IU/L)	4.7(2.1,9.9)	1.3(0.6,2.3)	0.048	2.7(1.6,8.1)	3.0(1.8,4.7)	2.4(0.9,3.5)	0.112	0.073	0.707

Note: ^a median (interquartile range); HT: Hypertension; DM: Diabetes Mellitus; CHD: Coronary Heart Disease; CVD: cerebrovascular disease; BMI: Body mass index; CREA: creatinine; eGFR: estimated glomerular filtration rate; ALT: alanine aminotransferase; TG: triglyceride; TC: total cholesterol; HDL: high density lipoprotein; LDL: low density lipoprotein; HbA1c: glycosylated hemoglobin; uALB: urine albumin; ua1MG: urine a1 microglobulin; uIgG: urine Immunoglobulin G; uNAG: urine n-acetyl-beta-d-glucosaminidase

Proteomics results

A total of 485 up-regulated urinary proteins and 177 down-regulated urinary proteins were identified in ARAS patients compared to disease controls in the discovery cohort (Fig. 1A). We chose the up-regulated proteins for further analysis. Fold change > 1.5 and P value < 0.05 were considered as significant differences in the up-regulated proteins, and then the diagnostic performances of the up-regulated proteins were evaluated using ROC analysis. The top 10 areas under the ROC curve of the up-regulated proteins are shown in Table 2. The AUC ranged from 0.99408979 to 0.895408163, followed by CA3, ART3, ORM2, SNX3, OGN, TK2, ORM1, SCGB1A1, RNF213 and CA1. The hierarchical clustering heat map showed the differential expression of these top 10 urinary proteins between the two groups in the discovery cohort (Fig. 1B).

Fig. 1.

Analysis of differentially expressed proteins. (A) Volcano plots analysis of all DEPs compared to disease controls. (B) Hierarchical clustering of top 10 DEPs in different groups. (C) Gene Ontology (GO) enrichment analysis of up-expressed top 10 DEPs: Biological Process (BP), Cellular Component (CC) and Molecular Function (MF). (D) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of up-expressed top 10 DEPs. (E) Protein interaction networks of top 10 up-expressed proteins

Table 2.

The top 10 area under the ROC curve of the up-regulated proteins

No.	Gene Name	Protein Name	FoldChange	P value	AUC
1	CA3	Carbonic anhydrase 3	1028.517	< 0.001	0.994089796
2	ART3	Ecto-ADP-ribosyltransferase 3	48.57	0.006	0.954139592
3	ORM2	Alpha-1-acid glycoprotein 2	9.191	< 0.001	0.938826531
4	SNX3	Sorting nexin-3	2.199	< 0.001	0.931122449
5	OGN	Mimecan	23.635	0.043	0.928571429
6	TK2	Thymidine kinase 2, mitochondrial	10.788	0.017	0.920918367
7	ORM1	Alpha-1-acid glycoprotein 1	9.129	< 0.001	0.918367347
8	SCGB1A1	Uteroglobin	12.959	0.003	0.908163265
9	RNF213	E3 ubiquitin-protein ligase RNF213	3.345	0.001	0.899234694
10	CA1	Carbonic anhydrase 1	11.806	0.018	0.895408163

GO and KEGG pathway analysis and PPI analysis

We further performed a GO and KEGG pathway analysis with the 10 urinary proteins, and the results are shown in Fig. 1C-D. Biological process analysis showed that these proteins were closely related to one-carbon metabolic process and bicarbonate transport. Most cellular components were located in specific granule lumen, and their molecular function was mainly related to carbonate dehydratase activity. By KEGG pathway analysis, nitrogen metabolism was the most enriched function. In the meantime, the STRING database was used to predict the interactions of these 10 proteins in this study. As shown in Fig. 1E, there were strong interactions between CA3 and CA1, ORM2 and ORM1, and there was no interaction between the other six proteins. Meanwhile, the AUC of ART3 in the diagnosis of ARAS was the second. Therefore, the top three urinary proteins CA3, ORM2, and ART3 in AUC were selected as potential urinary biomarkers for further validation. The expression quantities and AUC of CA3, ORM2, and ART3 in proteomics are shown in Fig. 2A-C and Fig. 2D-F.

Fig. 2.

Expressions and AUCs of CA3, ORM2 and ART3 in proteomics. (A) Expressions of CA3 in ARAS and DC. (B) Expressions of ART3 in ARAS and DC. (C) Expressions of ORM2 in ARAS and DC. (D) ROC curves of CA3 for the discrimination of ARAS vs. DC groups. (E) ROC curves of ART3 for the discrimination of ARAS vs. DC groups. (F) ROC curves of ORM2 for the discrimination of ARAS vs. DC groups

Increased uCA3, uORM2, and uART3 in ARAS displayed by ELISA in validation cohort

We performed ELISA tests for uCA3, uORM2, and uART3 in an independent validation cohort, which included 46 patients with ARAS (13 moderate to severe ARAS and 33 mild ARAS), 30 disease controls and 20 healthy controls. The uCA3/Cr levels were significantly higher in the ARAS and DC groups than in healthy controls, and there were also significant differences between the ARAS group and the DC group. The uART3/Cr levels were higher in the ARAS group than in disease and healthy controls, but there were no significant differences between the ARAS and DC groups. There was no statistical difference in uORM2/Cr levels between the three groups (Fig. 3A-C). We also compared the three proteins in subgroups with different severity of ARAS and there was no statistical difference. (Fig. 3D-F).

Fig. 3.

Urinary CA3/Cr, ART3/Cr and ORM2/Cr levels in different groups. (A) uCA3/Cr levels in ARAS, DC and HC groups. (B) uART3/Cr levels in ARAS, DC and HC groups. (C) uORM2/Cr levels in ARAS, DC and HC groups. (D) uCA3/Cr levels in mild and severe ARAS, DC and HC groups. (E) uART3/Cr levels in mild and severe ARAS, DC and HC groups. (F) uORM2/Cr levels in mild and severe ARAS, DC and HC groups

Receiver operator characteristic analysis for ARAS

ROC analysis was performed to access the diagnostic efficiency of uCA3/Cr and uART3/Cr in the validation cohort. Taking the healthy controls as the control values, the areas of the ROC curve were 0.9649(95%CI: 0.9200-1.000),0.9487(95%CI:0.8952-1.000), respectively (Fig. 4A-B). Since there was no difference in uART3/Cr levels between the ARAS group and the DC group, we only evaluated the area under the uCA3/Cr curve with the values of the disease control group as the control. The areas of ROC were 0.7391(95%CI: 0.6154–0.8628) (Fig. 4C).

Fig. 4.

ROC curves of the DEPs. (A) ROC curves of uCA3/Cr for the discrimination of ARAS vs. HC groups. (B) ROC curves of uART3/Cr for the discrimination of ARAS vs. HC groups. (C) ROC curves of uCA3/Cr for the discrimination of ARAS vs. DC groups

Correlation analysis between uCA3/Cr and clinical indicators

Based on the ELISA data of the validation cohort, we explored the correlation between uCA3/Cr concentration and clinical indicators, including age, uALB/Cr, ua1MG/Cr, uIgG/Cr, uNAG/Cr, etc. The results are shown in Table 3. There was a statistically positive correlation between uCA3/Cr concentration and uALB/Cr, ua1MG/Cr and uNAG/Cr in ARAS and disease controls, no statistically significant correlation was observed between uCA3/Cr and uALB/Cr, ua1MG/Cr and uNAG/Cr in healthy controls. Pooled analysis of the entire cohort revealed a statistically positive correlation between uCA3/Cr concentration and uALB/Cr, ua1MG/Cr and uNAG/Cr. However, no statistically significant correlation was observed between uCA3/Cr and uIgG/Cr, lipid profile. Detailed results are presented in Table 4.

Table 3.

Correlation between uCA3/Cr concentration and clinical indicators

	uCA3/Cr (ARAS + DC)		uCA3/Cr (HC)
	r	P	r	P
Age	0.049	0.687	-0.401	0.080
CREA	0.027	0.826	-0.075	0.752
eGFR	-0.066	0.589	0.197	0.405
uALB/Cr	0.344	< 0.001	0.118	0.621
ua1MG/Cr	0.277	< 0.001	0.336	0.148
uIgG/Cr	0.101	0.213	0.235	0.318
uNAG/Cr	0.304	< 0.001	0.080	0.738
ALT	-0.211	0.082	-0.135	0.572
TG	-0.127	0.317	0.140	0.556
TC	0.225	0.074	0.378	0.101
HDL	0.026	0.842	0.134	0.573
LDL	0.168	0.184	0.416	0.068
HbA1c	-0.062	0.641	-0.241	0.369

Bold italics represent correlations with P values less than 0.05

Table 4.

Correlation between uCA3/Cr concentration and clinical indicators

	uCA3/Cr (ARAS + DC + HC)
	r	P
Age	0.245	0.018
DM	0.378	< 0.001
HT	-0.034	0.773
CREA	0.109	0.292
eGFR	-0.045	0.665
uALB /Cr	0.361	< 0.001
ua1MG/Cr	0.432	< 0.001
uIgG/Cr	0.101	0.213
uNAG/Cr	0.284	0.006
ALT	0.036	0.731
TG	0.081	0.449
TC	0.052	0.631
HDL	-0.196	0.068
LDL	-0.029	0.784

Bold italics represent correlations with P values less than 0.05

Discussion

ARAS is a progressive disease characterized by increased stenosis and eventual occlusion, which has a greater impact on organ function and patient prognosis. In this study, a new generation of proteomic technology, DIA-MS, was used for the first time to screen urine proteins suitable for the early screening of ARAS. A total of 485 up-regulated and 177 down-regulated urinary proteins were identified in patients with ARAS compared to disease controls in the discovery cohort, and the top 10 up-regulated proteins of AUC were selected for bioinformatics analysis. CA3, ART3, and ORM2 were screened for ELISA verification in the validation cohort. The results showed that the level of uCA3/Cr was significantly higher in both ARAS and disease controls than in healthy controls, and the levels of uCA3/Cr in ARAS patients were significantly higher than those in disease controls, suggesting that uCA3/Cr may be related to the occurrence and development of ARAS. Further ROC analysis showed that the area under the uCA3/Cr curve in the diagnosis of ARAS reached 0.9649. uCA3/Cr may be used as a potential biomarker for monitoring ARAS. The correlation analysis shows that uCA3/Cr was positively correlated with uALB/Cr, ua1MG/Cr and uNAG/Cr in ARAS and disease controls. The results suggest that the change in uCA3/Cr may be related to tubular and glomerular injury.

Carbonic anhydrases (CAs) are zinc-containing metalloenzymes that are highly expressed in vertebrates including humans. Five separate gene families have been found to encode these enzymes, namely α-CAs, β-CAs, γ-CAs, δ-CAs and η-CAs [24]. To date, the α-CAs, with 14 different CA isozymes being described in humans, are the only group found in mammals. They have evolved to catalyze the conversion of CO₂/HCO₃ in different macro- and micro-environments of the human body, thereby participating in a series of physiologic and pathologic processes, for instance, respiration and CO₂ excretion, electrolytes secretion, pH homeostasis, lipogenesis, ureagenesis, gluconeogenesis, bone resorption and calcification, and are also involved in tumorigenicity.

CA3(29.4 kDa) is a low activity cytoplasmic isoenzyme of CAs [25]. It is generously expressed in the cytosol of skeletal muscle cells, hepatocytes and adipocytes [26], and in small amounts in the kidney, uterus, red cells, prostate, lung, colon and testis [27, 28]. In our study, bioinformatics analysis illustrated that CA3 is closely related to carbon and nitrogen metabolism. Study [29] showed that CA3 regulate intracellular pH by delivering CO₂ produced during cellular metabolism. The renal blood flow is reduced in patients with ARAS [8]. The recurrent renal ischemia induces tubulointerstitial injury and microvascular damage, while global hypoperfusion triggers endothelial-epithelial dysfunction and activates the renin-angiotensin-aldosterone system (RAAS), leading to vasoconstriction. These dual mechanisms synergistically promote oxidative stress, upregulation of profibrotic cytokines, and inflammatory responses, ultimately driving renal atrophy and fibrosis. A previous study had shown that CA3 protects cells from oxidative damage in both the liver and skeletal muscle, as it plays a role as an oxyradical scavenger of reactive oxygen species [26]. Based on the clinical investigations, the profound role of increased oxidative stress, nuclear translocation of NF-kB, production of proinflammatory cytokines such as TNF-α, IL-6, increased apoptosis, reduced angiogenesis, and histological aberrations in CA1, CA2, CA3 of the hippocampus have been reported [30].The study by P Gailly et al. demonstrated that the induction of CA3 may be part of the cellular response to oxidative damage. The expression of CA3 was induced to increase when the kidneys were characterized by increased cell proliferation and oxidative stress [31]. The study [32] manifested that elevated levels of angiotensin II caused by renal artery stenosis may lead to significant excessive oxidative stress and may seriously affect endothelial function. Meanwhile, renal angioplasty improves endothelial dependent vasodilation in patients with renal vascular hypertension by reducing oxidative stress. Therefore, we speculate that elevated levels of CA3 in the ARAS group may be related to increased oxidative stress caused by renal artery stenosis.

Early studies showed that in tissue distribution studies of ¹²⁵I-CA isoenzymes with whole body autoradiography in rats, it has been shown that CA1 and CA3 are localized in the renal cortex. This finding therefore indicated that small amounts of CA1 and CA3 isoenzymes are actually filtered into the glomeruli and thereafter reabsorbed and degraded by the proximal tubules [33]. In our study, the STRING analysis showed that CA3 is related to CA1 in urine proteomics. Our correlation analysis shows that uCA3/Cr was statistically positively correlated with uALB/Cr, ua1MG/Cr and uNAG/Cr in ARAS and disease controls. The uALB (66.5 kDa) is a medium-molecular-weight proteins, which is generally excluded from GFB under physiological conditions. Elevated uALB levels may indicate early glomerular injury. Our study found no statistically significant differences in uALB levels among the three groups, and the correlation coefficient was 0.344(< 0.5), indicating a weak association. This observation leads us to hypothesize that patients with ARAS may present with elevated serum CA3 levels, which, following glomerular filtration, could consequently lead to increased urinary CA3 excretion. a1MG is a globular protein present in human plasma, which is mainly synthesized in the liver and then distributed all over the body. The a1MG is a low-molecular-weight protein (approximately 33 kDa) that freely passes through the glomerular filtration barrier. Under normal conditions, it is almost entirely reabsorbed and metabolized by the proximal renal tubules. It has been confirmed that a1MG is a sensitive marker of proximal tubular dysfunction. The excretion of urinary a1MG in chronic interstitial nephropathy and glomerulonephritis has been shown to increase. The a1MG in urine is elevated at very early stages, even when histological damage cannot be detected [34]. When renal tubular cells are injured, the NAG is released into the urine. These results indicate that patients with ARAS may have experienced damage to the kidneys, resulting in reduced reabsorption of CA3 by the proximal tubules and an increase in CA3 levels in urine.

Taken together, CA3 is a cytosolic protein with a molecular weight of approximately 30 kDa. Therefore, CA3 can freely pass through the glomerular basement membrane and predominantly reabsorbed by renal tubules. Prior studies have demonstrated elevated serum carbonic anhydrase III (CA3) levels in patients with renal failure [35]. Our study revealed statistically significant correlations between uCA3 and uALB (P < 0.05) in ARAS and DC groups, which is established indicator of early glomerular injury. This observation leads us to hypothesize that patients with ARAS may present with elevated serum CA3 levels, which, following glomerular filtration, could consequently lead to increased urinary CA3 excretion. Current studies have demonstrated through immunohistochemical staining that renal tubule tissues exhibit positive expression of CA3. Moreover, upregulated expression of CA3 was specifically detected in proximal tubule cell lines exposed to oxidative stress. Our study demonstrated a statistically significant correlation between uCA3 and uα1MG and uNAG in ARAS and DC groups. When renal tubular cells are injured, the α1MG and NAG are released into the urine. Therefore, our study speculates that the increased urinary CA3 levels may be associated with reduced CA3 reabsorption due to ischemic tubular injury and increased CA3 secretion induced by oxidative stress. However, the precise mechanisms warrant further investigation.

There are still quite a few limitations in our study. Due to the limited size of the cohort, the biomarker found in this study needs to be further evaluated and verified by a large number of samples before being applied clinically. Our current results show that uCA3 cannot be used for a differential diagnosis of the severity of ARAS. Our analysis results require a larger sample size to reduce variation.

Conclusions

In summary, we found that there are a variety of DEPs in the urine of patients with ARAS by DIA-MS, and further research on these proteins will improve the diagnosis and treatment of ARAS. In addition, this study initially confirmed that uCA3 can be used as a non-invasive urine biomarker for the diagnosis of ARAS, and large cohort studies are required for further validation. Finally, correlation analysis showed that the increase in uCA3 was related to renal interstitial injury, but the exact mechanism was still uncertain. Large cohort studies and mechanistic investigations are required for further validation, so as to provide new ideas for the research of the mechanism and clinical intervention of ARAS.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ARAS

Atherosclerotic renal artery stenosis

LC-MS/MS

Liquid Chromatography Tandem-mass Spectrometry

DIA-MS

data independent acquisition-based liquid chromatography mass spectrometry

ELISA

Enzyme linked immunosorbent assay

FDR

False discovery rate

FC

Fold change

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

ROC

Receiver operating characteristics curves

AUC

Area under the curve

BP

Biological process

CC

Cellular component

MF

Molecular function

CA3

Carbonic anhydrase 3

ART3

Ecto-ADP-ribosyltransferase 3

ORM2

Alpha-1-acid glycoprotein 2

eGFR

estimated glomerular filtration rate

DC

Disease control

HC

Healthy control

Author contributions

Man Zhang designed the study and Xiufeng Li wrote the main manuscript text. All authors reviewed the manuscript.

Funding

Source of Support: the project is supported by the Open Research Funding of Beijing. Key Laboratory of Urinary Cellular Molecular Diagnostics (2020-KF23).

Data availability

The experimental data used and analyzed in the current study were available from thesupplementary material.

Declarations

Ethics approval and consent to participate

The study was conducted in compliance with the declaration of Helsinki principles and followed the recommendations of Medical Ethics Committee of Aerospace Center. Hospital (20201113-QTKT-01) and all subjects provided written consent before inclusion.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.