Bioinformatics-based exploration of the role of copper metabolism in chronic glomerulonephritis

Abstract

Background:

Chronic glomerulonephritis is a prevalent renal disorder in clinical practice. A metabolic imbalance in copper, a vital trace element in the human body, is potentially linked to a myriad of diseases. Consequently, this study aimed to elucidate the role of copper metabolism in the onset and progression of chronic glomerulonephritis.

Methods:

Public microarray datasets gene expression omnibus series(GSE) 66494, GSE32591, and GSE116626 were procured from the gene expression omnibus database, encompassing clinical information on chronic glomerulonephritis and normal kidney tissues. Differential gene expression analysis was conducted, followed by the intersection of the differentially expressed genes from the 3 datasets with those related to copper metabolism. Furthermore, STRING and Cytoscape were used to construct a protein–protein interaction (PPI) network; PPI hub genes were identified via the cytoHubba plugin, and gene set enrichment analysis was performed. Tissues were subsequently collected from chronic glomerulonephritis patients at our hospital; hematoxylin and eosin, periodic acid-silver methenamine, and Masson staining were conducted; and transmission electron microscopy was performed. Pivotal genes were selected for real time quantitative polymerase chain reaction (PCR) (RT-qPCR) validation, and serum copper ion levels were measured. Additionally, CIBERSORT analysis was employed to assess immune cell infiltration and its correlation with pivotal genes.

Results:

Differential gene expression analysis revealed 25 differentially expressed copper metabolism-related genes (CMRGs), comprising 11 upregulated genes and 14 downregulated genes. PPI network construction and gene correlation analysis ultimately identified aconitase (ACO1) and superoxide dismutase (SOD2) as pivotal genes for RT-qPCR validation. The results demonstrated that ACO1 was expressed at low levels in lupus nephritis tissues, whereas SOD2 was highly expressed in immunoglobulin A nephropathy and diabetic nephropathy tissues. Moreover, immune infiltration analysis revealed that diverse immune cell types, including T cells, B cells, and macrophages, are intricately associated with the onset and progression of chronic glomerulonephritis in patient tissues. Differential and correlation analyses of CMRGs revealed that T cells, plasma cells, and monocyte-macrophages are involved in the biological processes of copper metabolism.

Conclusion:

This study suggests that the CMRGs ACO1 and SOD2 may modulate the inflammatory response and immune microenvironment in chronic glomerulonephritis by regulating immune cell activation and function, thereby facilitating disease progression.

1. Introduction

Chronic kidney disease (CKD) is a progressive condition primarily characterized by a decline in the glomerular filtration rate.^[1] With a prevalence exceeding 10%, CKD represents a major global public health challenge.^[2] Statistics indicate that over 75% of CKDs result from glomerulonephritis, which includes diabetic nephropathy (DN), immunoglobulin A nephropathy (IgAN), and membranous nephropathy (MN).^[3] Copper is an essential mineral nutrient involved in various cellular processes, including proliferation and apoptosis.^[4] As an indispensable cofactor, copper influences a wide range of enzymes crucial to fundamental cellular functions such as antioxidant defense, mitochondrial respiration, and biosynthesis. Furthermore, copper plays a critical role in regulating oxidative stress, with Cu(II) and Cu(I) being its 2 oxidation states; Cu(I) being the predominant reduced form in the cytoplasm.^[5] Excessive copper directly catalyzes the production of reactive oxygen species (ROS), generating highly reactive hydroxyl radicals. Concurrently, it significantly reduces glutathione levels, enhancing ROS cytotoxicity. This increased catalytic activity of copper perpetuates a vicious cycle of elevated ROS production.^[6] Consequently, copper metabolic imbalance can precipitate a range of pathological conditions, including neurodegenerative diseases and cancer. The increase in ROS coupled with the decrease in antioxidant capacity also intensifies lipid peroxidation, thereby exacerbating kidney disease.^[7] To investigate the relationship between copper metabolism and chronic glomerular disease, we performed bioinformatics analyses using chronic glomerulonephritis data from the NCBI gene expression omnibus (GEO) database, identified differentially expressed copper metabolism-related genes (CMRGs) and further screened hub genes for real time quantitative PCR (RT-qPCR) validation. Additionally, we analyzed the correlation between hub genes and immune cells within the autoimmune microenvironment to elucidate the role and potential mechanisms of CMRGs in the progression of chronic glomerular disease.

2. Materials and methods

2.1. Download and preprocessing

The public microarray datasets GSE66494, GSE32591, GSE116626, and GSE104948 (https://www.ncbi.nlm.nih.gov/geo/), which encompass clinical data on chronic glomerulonephritis and normal kidney tissues, were acquired from the GEO database. The GSE66494 dataset comprises 53 chronic glomerulonephritis kidney samples and 8 normal samples, utilizing the Agilent-014850 whole human genome microarray 4 × 44K G4112F from the GPL6480 platform. The GSE116626 dataset contains 74 chronic glomerulonephritis kidney samples and 7 normal samples and is based on the GPL14951 platform’s Illumina Human HT-12 WG-DASL V4.0 R2 expression bead chip. The GSE32591 dataset consists of 64 chronic glomerulonephritis kidney samples and 29 normal samples and is based on the Affymetrix Gene Chip Human Genome HG-U133A Custom CDF of the GPL14663 platform. The GSE104948 dataset comprises 175 chronic glomerulonephritis samples and 21 control samples, utilizing the GPL24120 platform’s Affymetrix Human Genome U133A Array and the GPL22945 platform’s Affymetrix Human Genome U133 Plus 2.0 Array. GSE104948 of them are used as validation sets. CMRGs were identified from previous studies^[4,8] and downloaded from the molecular signatures database v7.5.1. To mitigate potential batch effects arising from different platforms (GSE66494, GSE116626, and GSE32591), we did not directly merge these datasets. Instead, we conducted analyses based on the intersection of CMRGs across all 3 datasets. In each dataset, raw expression matrices were preprocessed using the R package “limma.” Specifically, we performed background correction and log₂ transformation, selected the probe with the highest expression value when multiple probes mapped to the same gene, and applied the normalizeBetweenArrays function to achieve inter-sample normalization.

2.2. Identification and enrichment analysis of differentially expressed copper metabolism-related genes (CMRGs)

First, differential expression analysis between the chronic glomerulonephritis group and the control group was performed using the R package “limma,” with the statistical criteria for RNA expression set at |log FC| > 0.5 and Benjamini-Hochberg adjusted P value <.05 (Fig. 1A, B).The Venn diagram tool was subsequently used to identify overlap among the differentially expressed genes (DEGs) from the 3 datasets and the CMRGs. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed on the overlapping genes using the R packages “org.Hs.e.g.db” and “ClusterProfiler” to elucidate their biological functions and pathways (Fig. 2A–C).^[9]

Figure 1.

Construction of CMRGs in chronic glomerulonephritis. (A) Volcano plots of common DEGs chronic glomerulonephritis versus control samples in 3 GSE datasets. (B) Heatmaps of common DEGs from the 3 datasets. (C) Venn diagram showing the intersection of copper metabolism-related genes (CMRGs) with DEGs in the datasets. (D) PPI network construction for CMRGs (≥3 dataset overlap). (E) CMRGs network built using the DEGREE algorithm in Cytoscape. (F) Correlation analysis of 10 CMRGs identified through the PPI network. CMRGs = copper metabolism-related genes, DEGs = differentially expressed genes, GSE = gene set enrichment, PPI = protein–protein interaction.

Figure 2.

Functional enrichment analysis of CMRGs. (A) KEGG enrichment analysis. (B, C) GO enrichment analysis. CMRGs = copper metabolism-related genes, GO = gene ontology, KEGG = Kyoto Encyclopedia of Genes and Genomes.

2.3. Protein–protein interaction (PPI) network and hub gene analysis

The STRING database (http://string-db.org) and Cytoscape software (version 3.9.1) were used to evaluate the interaction relationships of overlapping genes by constructing and visualizing their biological network, followed by optimization. Next, Cytoscape software was used to represent the results from STRING as a protein–protein interaction (PPI) network, removing nodes that were not connected to any other nodes (Fig. 1D). Finally, the DEGREE algorithm in the Cytohubba plugin was applied to analyze and extract the top 10 ranked genes, identifying them as hub genes within the PPI network for further analysis (Fig. 1E).

2.4. Gene set enrichment analysis (GSEA) of 10 hub genes

To investigate the biological functional characteristics of the 10 hub genes, we employed the R package “clusterProfiler” for gene set enrichment analysis (GSEA), categorizing genes into low expression (<50%) and high expression (≥50%) groups based on their expression levels. The molecular signatures database serves as the source for reference gene set data. The consistent P value for each gene set was calculated using the reference gene set data. The P value was derived by comparing the enrichment score with the results generated from 1000 random permutations of the gene set. If the P value was <.05, the gene set was deemed significantly enriched, after which the gene sets are ranked (Fig. 3A–J).^[10]

Figure 3.

Gene set enrichment analysis of HUB genes. (A) ACO1, (B) ALB, (C) APP, (D) FXN, (E) IDH2, (F) JUN, (G) SDHA, (H) SDHB, (I) SDHC, and (J) SOD2. ACO = aconitase, ALB = albumin, FXN = frataxin, JUN = jun proto-oncogene, SDHA = Succinate Dehydrogenase Complex Flavoprotein Subunit A, SDHB = Succinate Dehydrogenase Complex Iron-Sulfur Subunit B, SDHC = Succinate Dehydrogenase Complex Cytochrome b558 Subunit C, SOD2 = superoxide dismutase.

2.5. Construction of receiver operating characteristic curves to evaluate diagnostic performance

To validate the predictive value of the identified hub genes, the R package “pROC” was used to plot the receiver operating characteristic (ROC) curve and visualize the results using the R package “GGPLOT2.” The diagnostic value of the hub genes was assessed by calculating the area under the ROC curve (area under the curve [AUC]), which ranges from 0.5 to 1; the closer the AUC is to 1, the more effective the diagnostic performance. The expression levels and diagnostic performance of the candidate biomarkers were further validated via the external dataset GSE104948, which comprises 175 chronic glomerulonephritis samples and 21 control samples (Fig. 8C–E).

Figure 8.

Construction of the CMRGS in Chronic Glomerulonephritis (A) Cu²⁺ levels in serum in patients with chronic glomerulonephritis and normal humans. (B) Expression of ACO1 and SOD2 mRNA in chronic glomerulonephritis kidney tissue and normal kidney tissue. The P value is: * P < .05; **P < .01. (C) ROC curves for the ACO1 and SOD2 genes in the combined (internal) GSE66494 dataset. (D) ROC curves for the ACO1 and SOD2 genes in the combined (internal) GSE32591 dataset. (E) ROC curves for the ACO1 and SOD2 genes in the combined (external) GSE104948 dataset. (F, G) Associations between ACO1 and SOD2 and immune cells. ACO = aconitase, CMRGs = copper metabolism-related genes, GSE = gene set enrichment, ROC = receiver operating characteristic, SOD2 = superoxide dismutase.

2.6. Immune infiltration analysis

To elucidate the relationships between the hub genes and the immune microenvironment and to achieve a reliable assessment of immune infiltration, the immune infiltration of CMRGs in the chronic glomerulonephritis group was analyzed via the CIBERSORT algorithm. CIBERSORT performs deconvolution of 22 immune cell types based on the principle of linear support vector regression, producing a normalized transcriptome expression matrix. CIBERSORT was used to calculate the composition of 22 immune cell types in each sample, and samples with P < .05 were filtered out for subsequent analyses (Fig. 4A). Spearman correlation analysis was used to assess the relationships between the 22 types of infiltrating immune cells (Fig. 4B). Finally, the R package “ggplot2” was used to analyze differences in the expression of the 22 types of infiltrating immune cells between the high- and low-expression groups of hub genes, and the results were visualized with box plots. A P value <.05 indicated a statistically significant difference (Fig. 4C–L).

Figure 4.

Immune cell infiltration and correlational analysis. (A) The ratio of 22 immune cells of each sample of chronic glomerulonephritis. (B) The correlation between each of immune cells. (C–L) Immune cell infiltration comparison between low and high expression groups of ACO1, ALB, APP, FXN, IDH2, JUN, SDHA, SDHB, SDHC, and SOD2 in chronic glomerulonephritis. ACO = aconitase, ALB = albumin, FXN = frataxin, JUN = jun proto-oncogene, SDHA = succinate dehydrogenase complex flavoprotein subunit A, SDHB, Succinate Dehydrogenase Complex Iron-Sulfur Subunit B, SDHC = succinate dehydrogenase complex cytochrome b558 Subunit C, SOD2 = superoxide dismutase.

2.7. Clinical specimen collection

With the approval of the Ethics Committee of the Second Hospital of Dalian Medical University The above serum and tissue samples were collected to obtain informed consent from study participants, and informed consent forms have been presented in writing (KY2024-111-01), tissues were collected from 9 IgAN patients, 6 lupus nephritis (LN) patients, 8 MN patients, 8 DN patients, and 6 control patients at the Second Hospital of Dalian Medical University. The control patients exhibited only mild hematuria (RBC < 5/HP) and were pathologically normal. Additionally, serum samples were collected from 12 patients each with IgAN, LN, MN, and DN and from 12 healthy controls (Table 1).

Table 1.

Clinical characteristics of chronic glomerulonephritis patients included in this study.

	Gender	Age	Urine protein	URBC/HPF	u-TP (mg/24 h)	ALB	BUN	SCr	eGFR (Cr)	ALT	AST	PLA2R	C3	C4	IgA	IgM
IgAN	Female	59	+++	25–30	1574.83	36.08	5.6	60	>90	22.01	20.88	–	1.31	0.305	5.16	0.84
	Female	25	+	70–80	165.5	38.98	4.21	53.38	>90	8.72	15.95	–	0.931	0.237	5.14	1.6
	Female	38	++	20–25	1489.04	39.22	3.28	64.46	>90	11.62	15.6	–	1.07	0.326	2.96	1.12
	Female	66	±	15–20	636.31	39.37	4.72	78.81	>90	41.37	34.76	–	1.52	0.231	3.82	0.69
	Male	48	±	15–20	553.48	36.47	4.94	95.18	84.96	13.59	14.01	–	0.858	0.39	4.6	0.67
	Male	68	++	30–35	1419.55	38.11	8.99	204.38	29.99	13.77	14.66	–	1.25	0.429	4.54	0.38
	Female	42	++++	15–20	544.5	31.27	5.8	121.51	49.24	21.95	27.96	–	0.987	0.385	3.64	0.69
	Male	24	+++	15–20	4529.36	31.09	11.55	172.6	48.29	19.43	21.73	–	1.11	0.348	3.72	0.89
	Male	24	++	30–35	928.82	35.48	4.21	84.14	>90	8.56	16.58	–	0.903	0.323	2.75	0.81
LN	Male	26	+++	12–15	2749.3	23.63	4.65	75.14	>90	9.91	15.56	–	0.546	0.09	2.94	0.41
	Female	54	+	8.1	977.79	31.41	12.47	211.07	23.03	16.37	11.71	–	0.857	0.405	5.38	0.25
	Female	65	++	fv	2326.93	23.73	9.95	89.22	61.83	21.01	20.68	–	0.385	0.051	5.07	0.88
	Female	14	++	7.5	5771.28	48.03	4.62	33.54	>90	26.59	14.38	–	1.39	0.313	2.16	1.61
	Female	12	+	84.9	717.46	31.23	2.85	44.4	>90	28.13	31.48	–	0.583	0.11	3.93	2.56
	Female	11	±	20.1	403.25	36.95	5.51	37.8	>90	28.8	23.76	–	0.368	0.028	1.79	1.11
MN	Male	33	+	Norm	1657.22	44.13	4.63	56.13	>90	28.71	20.01	–	1.15	0.272	2.39	0.64
	Male	59	+++	0–1	4069.24	25.91	5.76	94.94	79.58	20.19	18.42	1:100 (+)	1.01	0.316	4.98	0.87
	Male	60	+++	13.5	2291.45	16.57	7.16	69.74	>90	17.05	22.41	1:100 (+)	1.27	0.256	1.96	1.05
	Female	82	++++	15–20	3366.3	48.22	9.53	151.35	29.5	15.81	25.53	1:10 (+)	0.933	0.348	2.49	0.36
	Male	38	+++	Norm	2041.85	36.29	3.39	76.87	>90	19.4	17.51	−	1.37	0.332	3.61	1.82
	Male	50	+++	0.9	7425.34	21.77	5.49	65.9	>90	23.79	24.43	1:100 (+)	1.21	0.4	2.41	1.45
	Male	54	+	Norm	769.62	31.8	5.82	76.81	>90	17.16	16.43	1:100 (+)	1.01	0.413	1.69	0.44
	Male	35	++	2.1	1014.81	32.24	3.92	86.07	>90	19.5	22.13	1:10 (+)	1.34	0.285	2.89	0.58
DN	Male	62	+++	6	8517.91	20.09	8.72	140.23	48.92	33.7	24.62	–	1.21	0.273	3.01	1.75
	Female	63	++	1–2	1812.36	25.49	18.59	179.65	27.03	11.79	15.75	–	0.772	0.244	1.16	1.23
	Male	56	++	Norm	6042.63	42.36	14.02	285	21.68	11.31	14.94	–	0.907	0.439	6.57	0.24
	Male	57	++++	25–30	1731.88	27.86	4.57	82.43	>90	28.56	34.94	–	0.891	0.248	2.59	1.4
	Male	25	++	Norm	1255.43	39.82	8.3	60.4	>90	20.94	17.76	–	1.33	0.277	1.64	0.28
	Female	50	+++	0–1	6658.22	15.66	8.56	143.37	38.42	12.38	17.08	–	0.926	0.336	1.9	0.78
	Male	64	+++	0–1	3987.46	33.31	6.92	89.35	82.97	19.97	18.63	–	1.08	0.225	2.07	0.37
	Male	48	++++	3–5	5535.16	22.49	19.44	272.18	24.08	50.97	98.23	–	0.877	0.327	2.13	0.44

ALB = albumin, ALT = alanine aminotransferase, AST = aspartate aminotransferase, DN = diabetic nephropathy, eGFR = estimated glomerular filtration rate, IgAN = immunoglobulin A nephropathy, LN = lupus nephritis, MN = membranous nephropathy, SCr = serum creatinine, URBC/HPF = urinary red blood cell count/high power field, u-TP = urinary total protein.

2.8. Histochemical staining

Kidney tissue sections were examined via hematoxylin and eosin, periodic acid-silver methenamine, and Masson’s trichrome staining techniques under a light microscope. The kidney tissues were initially fixed with 4% paraformaldehyde, embedded in paraffin, and sectioned into 2 μm thin slices. The tissues were stained via standard hematoxylin and eosin, periodic acid-silver methenamine, and Masson protocols, and the renal pathological features were examined under a light microscope (Figs. 5 and 6).

Figure 5.

Representative histochemical staining micrographs in the kidney showing the histopathology of chronic glomerulonephritis (staining methods, HE and PASM). (A) IgA, (B) SLE, (C) MN, and (D) DN. DN = diabetic nephropathy, HE = hematoxylin and eosin, IgA = immunoglobulin A, PASM = periodic acid-silver methenamine, SLE = systemic lupus erythematosus.

Figure 6.

Representative Masson stained micrographs in the kidneys show histopathology of chronic glomerulonephritis. (A) IgAN, (B) LN, (C) MN, and (D) DN. DN = diabetic nephropathy, IgAN = immunoglobulin A nephropathy, LN = lupus nephritis, MN = membranous nephropathy.

2.9. Transmission electron micrographs (TEM)

All kidney tissue samples for transmission electron microscopy were fixed with 2.5% glutaraldehyde, rinsed 3 times with 0.1 M phosphate buffer (pH 7.2), and subsequently fixed at room temperature with 1% osmium tetroxide in 0.1 M phosphate buffer (pH 7.2). Following dehydration through an ethanol gradient, the samples were infiltrated with varying ratios of acetone and epoxy resin. The infiltrated samples were then placed in embedding molds and embedded in epoxy resin. Finally, the samples were sectioned into ultrathin slices 70 to 100 nm thick. The sections were stained with uranyl acetate and lead citrate, air dried at room temperature overnight, and subsequently examined under a transmission electron microscope (Fig. 7).

Figure 7.

Representative images of ultrastructural changes in patients with chronic glomerulonephritis by transmission electron microscopy (TEM). (A) IgAN, (B) LN, (C) MN, and (D) DN. DN = diabetic nephropathy, IgAN = immunoglobulin A nephropathy, LN = lupus nephritis, MN = membranous nephropathy, TEM = transmission electron microscopy.

2.10. RNA extraction and real time quantitative PCR (RT-qPCR)

The expression of hub genes was quantified using RT-qPCR (Fig. 8B). Total RNA was isolated from kidney disease patient tissue samples and adjacent normal tissue samples using Trizol (TIANGEN, Beijing, China). The RNA was then reverse transcribed and amplified using the Evo M-MLV RT Kit (AG11705, Accurate Biotechnology (Hunan) Co., Ltd, Changsha, China.) and SYBR Green Pro Taq HS premixed qPCR Kit (AG11701, Accurate Biotechnology (Hunan) Co., Ltd.). RT-qPCR was conducted on a LightCycler96, with 18S rRNA as the internal reference. The 2−ΔΔCT method was used to quantify the number of cDNA cycles of the hub genes. A P value <.05 was considered statistically significant. The primer sequences used were as follows:

AOC1: F:5′-GTCCTGCTGCTCGCTACTT-3′.

R:5′-TCTGTGGTGCCTGCTTGTT-3′.

SOD2: F:5′-GGAAGCCATCAAACGTGACTT-3′.

R:5′-CCCGTTCCTTATTGAAACCAAGC-3′.

18S: F: 5′-AGTCCCTGCCCTTTGTACACA-3′.

R: 5′-CGATCCGAGGGCCTCACT-3′.

2.11. Copper ion detection

The Cu²⁺ levels in the serum of 12 IgAN patients, 12 LN patients, 12 MN patients, 12 DN patients, and 12 healthy individuals were quantified using a Cu colorimetric assay kit (E-BC-K300-M; Elabscience, Wuhan, China). A standard curve was plotted, and the Cu²⁺ concentrations were determined based on the absorbance value at 580 nm (Fig. 8A).

2.12. Statistical analysis

All statistical analyses were conducted using GraphPad Prism 8.0 software. Prior to conducting intergroup comparisons, the data were assessed for normality using the Shapiro–Wilk test. For data that followed a normal distribution, comparisons between groups were performed using a Student t test. For data that did not meet the assumption of normality, nonparametric Mann–Whitney U tests were used for validation. The data are presented as the mean ± standard error of the mean (mean ± SEM). A P value of <.05 was considered indicative of a statistically significant difference.

3. Results

3.1. Identification of CMRGs in CGN

The public microarray datasets GSE66494, GSE32591, and GSE116626, which contain clinical information on chronic glomerulonephritis and normal kidney tissues, were acquired from the GEO database. These datasets include 191 chronic glomerulonephritis tissues and 44 normal tissues. The number of DEGs identified from the GSE66494, GSE32591, and GSE116626 datasets was 2128, 323, and 51, respectively (Fig. 1A, B). By intersecting these with 194 CMRGs, 25 differentially expressed CMRGs were identified, comprising 11 upregulated and 14 downregulated CMRGs (≥3 intersections; |log FC| > 0.5, Benjamini-Hochberg adjusted P value <.05; Fig. 1C). To systematically analyze the relationships among CMRGs, a protein–protein interaction (PPI) network was constructed using the STRING database and visualized with Cytoscape. Ten hub genes were identified: jun proto-oncogene, aconitase (ACO1), APP, albumin, superoxide dismutase (SOD2), Succinate Dehydrogenase Complex Flavoprotein Subunit A , Succinate Dehydrogenase Complex Iron-Sulfur Subunit B, Succinate Dehydrogenase Complex Cytochrome b558 Subunit C, IDH2, and frataxin. The identification of these CMRGs provides a foundation for further analysis of the role of copper metabolism in the progression of chronic glomerulonephritis (Fig. 1D, E).

3.2. Function and pathway enrichment analysis of CMRGs

KEGG and GO enrichment analyses of the intersecting upregulated and downregulated genes were conducted using R software. As illustrated in the figure, GO analysis of upregulated genes revealed significant enrichment in numerous biological processes associated with immune and metal responses, including the regulation, activation, and modulation of immune responses and innate immune reactions as well as copper ion binding and metal ion redox activity. KEGG pathway analysis revealed significant increases in pathways related to viral infections, lipids and atherosclerosis, the advanced glycation end (AGE)-bind to their receptors (RAGE) signaling pathway in diabetic complications, endocytosis, and Toll-like receptor signaling. For downregulated genes, GO analysis demonstrated enrichment in circulatory metabolism and metal ion responses, encompassing tricarboxylic acid and steroid metabolic processes, responses to metal ions, redox activity involving CH-CH groups, and binding of iron–sulfur clusters and copper ions. KEGG pathway analysis revealed significant increases in the mitogen-activated protein kinase signaling pathway, ROS, and the citric acid cycle. These results suggest that immune responses, metal ion binding, and energy metabolism are intricately linked to the occurrence and development of chronic glomerulonephritis (Fig. 2A–C).

3.3. GSEA of hub genes

To gain deeper insights into the potential biological processes of the 10 hub genes identified in this study and predict their signaling pathways, we conducted GSEA. The results of the top 5 GO biological process terms indicate that these hub genes are involved in various pathways related to the development of chronic glomerulonephritis, primarily including the adaptive immune response, the lymphocyte-mediated immune response, regulation of immune cell activation, G-protein coupled receptor activity, amino acid organic acid metabolic processes, ATP synthesis coupled electron transport, the respiratory electron transport chain, mitochondrial translation termination, detection of chemical stimuli, response to interferon, and cotranslational protein targeting to the membrane. Pathways such as adaptive immune response, regulation of immune cell activation, amino acid organic acid metabolic processes, and G-protein coupled receptor activity were significantly enriched in multiple hub genes. These results suggest that immune and energy metabolism play crucial roles in the mechanisms by which copper metabolism-related hub genes contribute to the pathogenesis of chronic glomerulonephritis (Fig. 3A–J).

3.4. Determination of the serum concentration of copper ions

To observe changes in Cu²⁺ levels in chronic glomerulonephritis patients, serum Cu²⁺ concentrations in patients with IgAN, LN, MN, and DN were measured. The results indicated that serum copper ion levels in IgAN patients were greater than those in the control group, whereas the serum copper ion levels in DN patients were lower than those in the control group (Fig. 8A).

3.5. Expression of hub genes in kidney tissue of patients with chronic glomerulonephritis

To validate the expression of hub genes in chronic glomerulonephritis, RT-qPCR validation was performed using kidney tissues. Compared with that in the control group, ACO1 was underexpressed in LN tissues, whereas SOD2 was overexpressed in IgAN and DN tissues. These findings are consistent with the above bioinformatics predictions (Fig. 8B).

3.6. Diagnostic value and validation of hub genes in chronic glomerulonephritis

The diagnostic predictive value of hub gene expression for chronic glomerulonephritis was evaluated using ROC analysis. The results indicated that in the external dataset GSE104948, the AUC for the hub genes ACO1 and SOD2 were 0.793 and 0.827, respectively. Moreover, in the internal dataset GSE32591, the AUC for ACO1 and SOD2 were 0.787 and 0.772, respectively. In the internal dataset GSE66494, the AUC for ACO1 and SOD2 were 0.509 and 0.962, respectively. A comprehensive analysis of the results from both the external and internal datasets revealed that the hub genes ACO1 and SOD2 are involved in the occurrence and development of chronic glomerulonephritis (Fig. 8C).

3.7. Immune infiltration analysis and correlation analysis

These results indicate that the hub genes are highly enriched in immune-related pathways. To explore the potential mechanisms of hub genes in immune pathways, we extracted the gene expression matrix from chronic glomerulonephritis samples in the GSE66494 dataset and employed the CIBERSORT algorithm to calculate the correlation between these genes and the infiltration of 22 types of human immune cells in chronic glomerulonephritis samples. According to the immune infiltration results, various types of immune cells, including T cells, B cells, and macrophages in the tissues of chronic glomerulonephritis patients are closely associated with the occurrence and development of chronic glomerulonephritis. In the differential and correlation analyses of CMRGs, T cells, plasma cells, and monocyte-macrophages were found to be involved in the biological processes of copper metabolism (Fig. 4A–L).

4. Discussion

Chronic glomerulonephritis is a prevalent disease of the urinary system characterized by hematuria, proteinuria, edema, and hypertension as its primary clinical manifestations; an insidious onset; and a prolonged course. It is often accompanied by varying degrees of renal function decline, eventually progressing to chronic renal failure.^[11] Therefore, investigating the potential mechanisms underlying the occurrence and development of chronic glomerulonephritis holds significant clinical importance for improving disease outcomes.

Copper is an essential trace element in the body that plays crucial roles in various biological processes such as the oxidative stress response, apoptosis, and immune regulation, all of which are closely associated with the occurrence and development of chronic glomerulonephritis. Therefore, this study aimed to explore alterations in copper homeostasis and the role of CMRGs in the occurrence and development of chronic glomerulonephritis. In this study, by analyzing 3 public gene expression datasets of patients with chronic glomerulonephritis and healthy controls in conjunction with CMRGs, we identified ten key CMRGs, including jun proto-oncogene, ACO1, APP, albumin, and SOD2. Following the construction of the PPI network and the identification of ten hub genes using the Degree algorithm within the CytoHubba plugin (Fig. 1E), we performed a gene correlation analysis (Fig. 1F). Among these genes, ACO1 exhibited the highest correlation with other CMRGs, suggesting its potential central role within the copper metabolism regulatory network. Additionally, GSEA indicated that ACO1 was highly associated with key pathways related to energy metabolism. The selection of SOD2 was based on its well-established biological role. As a copper-dependent SOD2 plays a critical function in mitigating oxidative stress.^[12] Degree algorithm analysis ranked SOD2 among the top hub genes, underscoring its potential biological importance within the copper metabolism network. Furthermore, SOD2 was significantly differentially expressed in chronic glomerulonephritis and demonstrated strong diagnostic performance across independent datasets. Literature evidence also supports a close association between SOD2, glomerulonephritis, and oxidative damage.^[13–15] Consequently, based on network centrality, known biological function, and external validation, ACO1 and SOD2 were selected as key genes for subsequent RT-qPCR validation. Enrichment analysis further revealed that both genes are significantly involved in pathways related to energy metabolism and oxidative stress. While other genes also exhibited diagnostic potential and may serve as candidate therapeutic targets, ACO1 and SOD2 were prioritized in this study due to their pivotal functional relevance. Future investigations will further explore the roles of additional genes to elucidate their involvement in copper metabolism-related pathways.The results demonstrated that ACO1 was underexpressed in LN tissues, whereas SOD2 was overexpressed in IgAN and DN tissues (Fig. 8B). Additionally, serum copper ion levels were measured in the control group and the 4 disease groups. The results indicated that the serum copper ion levels in IgAN patients were greater than those in the control group, whereas the serum copper ion levels in DN patients were lower than those in the control group (Fig. 8A). To fully understand the dysfunctional inflammatory cells in chronic glomerulonephritis, immune infiltration analysis was performed. The analysis revealed that T cells, B cells, and macrophages were present in greater proportions in chronic glomerulonephritis tissues, whereas eosinophils and mast cells were present in relatively lower proportions (Fig. 4A). Furthermore, various immune cells were correlated with the 2 hub genes (ACO1 and SOD2). ACO1 was strongly correlated with T cells, plasma cells, and monocytes, whereas SOD2 was strongly significantly correlated with T cells and mast cells (Fig. 8F, G). Chronic glomerulonephritis is closely associated with immune responses. In IgAN and LN patients, the activation of Th1 and Th17 cells and increased secretion of related cytokines, along with decreased and defective Tregs, are closely associated with disease severity.^[16] By secreting cytokines and presenting antigens, B cells activate T cells and further exacerbate the inflammatory response.^[17] Additionally, in MN, B cells produce many autoantibodies (such as anti-PLA2R antibodies), which bind to antigens on the glomerular basement membrane to form immune complexes, causing inflammation and damage.^[18] Given the crucial role of B cells in the development of MN, rituximab is commonly used in clinical treatment. Rituximab binds to the CD20 antigen on the surface of B cells, inducing NK cells to release cytotoxic substances that kill the B cells, thereby suppressing the immune response.^[19,20] Recent studies have shown that infiltration of cluster of differentiation 68+ (CD68+) and CD206+ macrophages in the glomeruli are increased in IgAN patients and that the intensity of this infiltration can predict the response of IgAN patients to immunosuppressive therapy. Therefore, macrophage infiltration can serve as a biomarker for IgAN.^[21] Furthermore, in disease states, proinflammatory cytokines and MMPs secreted by monocytes and macrophages lead to glomerular structural damage and functional impairment.^[22,23] DN is a major cause of end-stage renal disease. Hyperglycemia is considered the primary pathogenic factor for DN. Hyperglycemia promotes the nonenzymatic glycation of proteins and lipids, resulting in the formation of AGEs. AGEs RAGEs activating a series of cell signaling pathways and triggering an inflammatory response.^[24] The interaction of AGEs with RAGEs can promote the production of inflammatory cytokines (such as TNF-α and IL-6), activating macrophages and T cells and leading to renal inflammation and damage. Additionally, hyperglycemia and AGEs upregulate the expression of chemokines (such as MCP-1), attracting monocytes to renal tissue where they differentiate into macrophages, further promoting the inflammatory response.^[25] In summary, immune cells (such as B cells, T cells, and macrophages) play crucial roles as potential pathogenic cells in common types of chronic glomerulonephritis.

Moreover, oxidative stress plays a crucial role in the pathogenesis and progression of DN.^[26] Under hyperglycemic conditions, mitochondria and NADPH oxidase produce substantial amounts of ROS, leading to oxidative stress. Hyperglycemia can inhibit the functioning of antioxidant systems (such as SOD and glutathione peroxidase), resulting in an oxidative stress imbalance. Additionally, the interaction of AGEs RAGEs can activate signaling pathways such as the NF-κB pathway, increasing ROS production and further exacerbating renal damage.^[27] An imbalance in copper metabolism may influence disease occurrence and progression through pathways such as oxidative stress and the immune response. Copper-dependent SOD plays a pivotal role in antioxidant defense by converting superoxide radicals into hydrogen peroxide, thereby reducing oxidative stress damage. An imbalance in copper metabolism may lead to decreased SOD activity, increased oxidative stress levels, and renal cell damage and fibrosis.^[28] Under hyperglycemic conditions, the increase in protein glycation reactions may affect copper binding and metabolism, thereby altering the distribution and function of copper in the body.^[29] Additionally, copper is a crucial component of many immune-related enzymes and is involved in the activation, proliferation, and functional regulation of immune cells. In the immune analysis, significant differences in copper-related genes were found in T cells, B cells, monocytes, and macrophages between the high- and low-expression groups, and these differences were strongly correlated with T cells, plasma cells, and monocytes. Therefore, copper metabolism imbalance may alter immune response patterns in chronic glomerulonephritis by affecting the function and activity of immune cells, resulting in increased inflammation and renal damage. Due to the study being based on single-center tissue sample collection, the actual number of valid tissue and serum samples obtained was limited, which, to some extent, affected the statistical power of the results and may have increased the risk of false negatives, thus limiting the detection of certain significant differences. Additionally, the imbalance in sample sizes across different nephropathy subtypes may have introduced bias into the results. To maximize statistical power, we have made full use of the available dataset for analysis and data validation. In future research, we plan to expand the sample size and collaborate with multiple centers to enhance statistical power and further validate the clinical application and prognostic value of ACO1 and SOD2 in clinical subtyping and progression prediction. Currently, research on the specific associations between chronic glomerulonephritis, copper metabolism, and the immune response is relatively limited, but existing evidence suggests a potentially complex relationship between them. Future research should further explore these associations to elucidate the specific mechanisms of copper metabolism in chronic glomerulonephritis and to identify new therapeutic strategies.

5. Conclusions

In summary, copper metabolism plays a crucial role in the pathogenesis of chronic glomerulonephritis. Our study revealed that the CMRGs ACO1 and SOD2 are closely associated with the infiltration of various immune cells, including T cells, B cells, and monocyte-macrophages. These immune cells also play pivotal roles in chronic glomerulonephritis. Therefore, CMRGs may influence the inflammatory response and immune microenvironment of chronic glomerulonephritis by regulating the activation and function of immune cells, thereby promoting disease progression. Future research should further explore the specific mechanisms of CMRGs in chronic glomerulonephritis and investigate their feasibility as potential therapeutic targets, thereby providing new insights and strategies for the diagnosis and treatment of chronic glomerulonephritis.

Acknowledgments

We thank the GEO databases for providing meaningful data sets.

Author contributions

Data curation: Lingjun Sun, Xiaoying Deng

Formal analysis: Qihui Ren.

Funding acquisition: Xiaoyan Liu.

Software: Shicheng Liu.

Validation: Lingjun Sun, Beibei Yang.

Visualization: Chen Wang.

Writing – original draft: Lingjun Sun.

Writing – review & editing: Ruqiang Yuan, Xinyan Feng, Xiaoyan Liu.

Supplemental digital content “Experimental Original data” is available for this article (https://links.lww.com/MD/Q527).