ALOX5, a novel diagnostic and prognostic marker associated with immune cells and ferroptosis in diabetic kidney disease

doi:10.21203/rs.3.rs-3744360/v1

Download PDF

Article

ALOX5, a novel diagnostic and prognostic marker associated with immune cells and ferroptosis in diabetic kidney disease

https://doi.org/10.21203/rs.3.rs-3744360/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Immune infiltration and ferroptosis play pivotal roles in the progression of diabetic kidney disease (DKD). However, the investigation of immune cell-associated ferroptosis genes (ICRFGs) in the context of DKD remains insufficient. This study aims to identify ICRFGs relevant to DKD. In this study, two DKD datasets were utilized, and alongside ferroptosis-related genes were obtained from public database. We employed the CIBERSORT and WGCNA methods to identify immune-related differential genes. The common ICRFGs were derived through the Venn plot and function analysis was performed. We employed the random forest and LASSO techniques to further select hub ICRFGs. The diagnostic efficacy of hub ICRFGs was assessed using K-M survival analysis, ROC curve analysis, and clinical factors of DKD. Hub ICRFGs-related pathways and functions in DKD patients were elucidated via the GSEA algorithm and the Metascape database. Additionally, we validated the expression of hub ICRFGs in high glucose-induced HK-2 cells and other datasets. Through bioinformatics analysis, we identified ALOX5 as a pivotal biomarker linked to immune-related ferroptosis. ALOX5 exhibited significant high expression in HK-2 cells exposed to high glucose and in other independent datasets. Clinical correlation and K-M survival analysis affirmed ALOX5 as the most crucial ICRFGs in DKD. Furthermore, ALOX5 demonstrated significant enrichment in ECM-receptor interaction, regulation of chemokine production, and regulation of inflammatory response. A positive correlation was observed between ALOX5 and M1 macrophage cells, γδ T cells, and monocytes. In conclusion, we identified ALOX5 as a dependable and prospective diagnostic marker associated with immunity and ferroptosis in DKD.

Biological sciences/Immunology

Health sciences/Nephrology

diabetic kidney disease

ferroptosis

immune cells

machine learning

ALOX5

Globally, diabetic kidney disease (DKD) stands as the predominant contributor to end-stage kidney disease (ESKD). The renal tubulointerstitial syndrome has increasingly been linked to the etiology of DKD, correlating with the progressive deterioration of renal function [1]. The etiology of DKD is multifaceted, wherein hyperglycemia is a necessary but not the sole causative factor. Alongside elevated blood glucose levels, other risk factors likely contribute to DKD. The mechanisms driving kidney pathology, triggered by heightened glucose levels, can be potentiated by various factors [2]. Predictive clinical risk factors for the decline in glomerular filtration rate (GFR) among DKD patients encompass diverse parameters such as age, diabetes duration, HbA1c levels, systolic blood pressure (SBP), albuminuria, prior estimated GFR (eGFR), and retinopathy status [3]. However, efforts to formulate and validate prognostic models using clinical information, serving as the primary means of forecasting through biomarkers, have been relatively limited. Consequently, about 40% of patients necessitate kidney replacement therapy due to kidney failure [4]. The current standard of clinically managing DKD involves stringent blood pressure and blood glucose control by blocking the renin-angiotensin system (RAS) and controlling hyperglycemia and hypertension [5, 6], as well as inhibit angiotensin-converting enzyme, angiotensin receptor, and sodium-glucose cotransporter-2, and so on as clinical treatment of DKD [6, 7]. Nevertheless, many patients experience uncontrollable progression of DKD to ESKD. While the mentioned biomarkers offer some predictive evidence for renal function decline and other DKD-related phenotypes, further research into more effective early-stage DKD diagnostic biomarkers is imperative.

Meanwhile, prior research has indicated the potential significance of ferroptosis and infiltrating immune cells in the pathogenesis of DKD [8, 9]. Evidence supports the relevance of inflammation and immune cell populations in DKD [10, 11]. Renal inflammation and fibrosis involve complex pathways resulting in a chronic inflammatory infiltrate composed of macrophages and other immune cells. This infiltrate releases cytokines and pro-fibrotic factors, creating a pro-fibrotic microenvironment through interaction with intrinsic renal cells [11]. Certain studies suggest that immune cells can detect ferroptotic cells, leading to various inflammatory or specific responses [12]. However, the interplay between ferroptosis genes and immune cells in influencing DKD pathogenesis remains unknown. Thus, comprehending the immunological processes and immune cell-associated ferroptosis genes regulating early DKD pathogenesis could offer insights into DKD diagnosis and treatment.

In this study, we employed both bioinformatics and experimental methodologies to explore potential indications of immune cell-associated ferroptosis in DKD. Publicly available datasets from the GEO database were used for both DKD and normal tubule samples. Our findings unveiled a notable increase in arachidonate 5-lipoxygenase (ALOX5) levels in DKD samples, suggesting its potential diagnostic utility in distinguishing DKD from healthy samples. Furthermore, following high glucose stimulation, elevated levels of ALOX5 protein were observed in HK-2 cells. In summary, ALOX5 could play a pivotal role in DKD pathophysiology. The overall research design is illustrated in Fig. 1.

2.1. Dataset acquisition and preprocessing

The study primarily relied on the Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/) database. Data were collected for the training cohort, denoted as GSE30122, and the validation cohort, inclusive of GSE30529, as described in reference [13]. Refer to Table S1 for a summary of the dataset particulars. The Xiantao (https://www.xiantaozi.com/) platform was employed to analyze the GSE30122 and GSE30529 datasets.

2.2. Quantification of immune cell infiltration and immune status

The microenvironment score, an amalgamation of immune and stromal scores, was generated utilizing the ESTIMATE algorithm. The CIBERSORT method, employing 22 distinct infiltrating immune cells, facilitated the quantification of immune infiltration. Immune scores signified the extent of immune cell infiltration, while stromal scores indicated stromal cell infiltration. The microenvironment score's visual depiction through a violin plot comparing DKD and normal samples was produced using GraphPad Prism 9.0.1 software. The CIBERSORT technique, incorporating 1,000 permutations, transformed the genetic expression matrix to calculate infiltrating immune cell abundances. Data were filtered based on CIBERSORT output (p < 0.05) to ensure accurate prediction of immune infiltrating cell composition. Further, box plots and heatmaps illustrating 22 distinct invading immune cell types were constructed using the online analysis and visualization platform accessible at https://www.bioinformatics.com.cn, last accessed on 10 July 2023. The green hue denoted immune cell abundance in normal samples, while the red indicated immune cell proportion in DKD samples.

2.3. Selection of relevant modules and genes with immune infiltration properties

To ascertain the correlation between gene expression data and the quantity of immune cells, the OECloud (https://cloud.oebiotech.com) tool, was utilized. This tool ranked the uppermost 39% (4,868 genes) of variable genes to establish the weighted co-expression network (WGCNA). The objective was to pinpoint significant gene modules linked with infiltrating immune cells. For this purpose, a soft threshold approach was employed, involving the application of a weight function to progressively transform the elements of the adjacency matrix. The selection of the soft threshold (β) could potentially influence both the outcomes of module classification and the corresponding network of each node's random average. To ensure stability, a preference was given to a scale-free network where only a limited number of nodes exhibited considerably higher degrees compared to the general point. Consequently, a soft threshold of R² = 0.82 (β = 22) was chosen for subsequent analysis, resulting in a network exhibiting a nearly scale-free topology.

To establish the connection between the quantity of infiltrating immune cells and the modules, as well as to identify pertinent modules, a Pearson's correlation test was conducted. Additionally, a link between co-expressed genes and the abundance of infiltrating immune cells was established using module membership (MM) and gene significance (GS). Genes with higher MM and GS values were found to have more substantial associations with modules and phenotypes, respectively. For further exploration, genes within modules exhibiting a strong correlation with infiltrating immune cells (correlation coefficient ≥ 0.5 and p ≤ 0.05) were selected, along with genes demonstrating the most robust association with immune cells (GS ≥ 0.5, MM ≥ 0.5).

2.4. Function evaluation and selection of hub ICRFGs

We obtained 487 ferroptosis-related genes from the FerrDb (www.zhounan.org/ferrdb) database. Subsequently, the Venny (https://bioinfogp.cnb.csic.es/tools/venny/index2.0.2.html) tool was utilized to obtain common ICRFGs from ferroptosis-related genes and immune cell-related candidate genes obtained from WGCNA analyses [14]. The resulting overlapping genes underwent functional enrichment analysis employing the Database for Annotation, Visualization, and Integrated Discovery (DAVID, https://david.ncifcrf.gov/) [15], Metascape (https://metascape.org/gp/index.html#/main/step1) [16], and Enrichr (https://maayanlab.cloud/Enrichr/) databases [17], with a significant p < 0.05. Enrichment annotations encompassed biological process (BP), cellular components (CC), and molecular functions (MF) sub-ontologies. Subsequently, Random forest (RF) analysis was executed using the "randomForest" R package to identify the three most significant genes. Following the RF analysis, the least absolute shrinkage and selection operator (LASSO) regression procedure was implemented using the "glmnet" R package to enhance prediction accuracy and identifying hub diagnostic indicators. LASSO analysis incorporated a 10-fold cross-validation with a penalty parameter, and gene expression levels were validated in the GSE30122 training cohort.

2.5. Validation of ALOX5 prognostic value in DKD

To validate the identified diagnostic indicators, the GSE30529 testing cohort was employed, along with DKD expression data and clinical information data, particularly serum creatinine levels and GFR, which were obtained from the Nephroseq v5.0 (http://v5.nephroseq.org/) database. The diagnostic effectiveness of these markers was evaluated using the receiver operating characteristic (ROC) methodology. Subsequently, the predictive value of ALOX5 in DKD was confirmed through the database for prognostic markers of kidney disorders (dbPKD, http://115.28.66.83/dn/index.php) [18]. For the analysis of ALOX5, both univariate and multivariate approaches were selected based on their expression level and risk score.

2.6. Enrichment and interaction analysis of prognostic model

The Gene Set Enrichment Analysis (GSEA) software (v 4.3.2) based on Java software was employed for identifying the statistically important molecular pathways from the molecular signatures database (MsigDB, https://www.gsea-msigdb.org/gsea/msigdb). The gene sets “c2.cp.kegg.v2023.1.Hs.symbols.gmt” was selected as the reference gene set. A pathway with NES ＞ 1 and FDR q < 0.25 was defined as statistically significant. Then, using a plethora of genomics and proteomics data, the GeneMANIA (http://www.genemania.org) database was utilized to locate functionally related genes based on the supplied gene list [19]. We identified genes with expression patterns resembling that of ALOX5 by examining analogous gene functions in GeneMANIA. The Metascape database was used for enrichment analysis.

2.7. Analysis of the relationship between infiltrating immune cells and diagnostic markers

Spearman correlations between infiltrating immune cells and diagnostic markers from the GSE30122 dataset were visualized using the Xiantao platform. Graphical presentation employed the "ggplot2" package (v 3.3.6) in the R package (v 4.2.1).

2.8. Cell culture and treatment

HK-2 cells were procured from Kunming Cell Bank (Chinese Academy of Sciences, China) and separated into control and experimental groups. Both groups of HK-2 cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37°C with 5% CO₂. Upon reaching approximately 50% confluence, the experimental group was treated with 40 mmol/L D (+)-glucose (CAS#50-99-7, Yuanye Bio-Technology Co., Ltd) (HG) was introduced to the experiment group as high glucose (HG) groups, and the cells were treated for 48 h. Cell morphology changes were observed using scanning electron microscopy. Subsequently, both groups underwent western blotting.

2.9. Protein extraction and western blot assay

For protein quantification assessment, the HK-2 cells underwent lysis. The BCA protein assay kit (Cat#PC0020, Solarbio) was utilized for this purpose. Subsequently, the proteins were separated using a 10% SDS-PAGE gel (Cat#P1200, Solarbio). These proteins were then transferred onto PVDF transfer membranes (Cat#IPVH00010, Millipore). Following this, the membranes were blocked at room temperature (RT) for a duration of 2 h using 5% skimmed milk powder (Cat#D8340, Solarbio) in 1 × TBST buffer (Cat#T1081, Solarbio). The blocked membranes were subsequently incubated overnight at 4 ℃ with primary antibodies, specifically ALOX5 (1:500, Cat#R382216, ZEN BIO), and β-actin (1: 1,000, Cat#PTM-5436, PTM biolabs). Following thorough washing with 1 × TBST, the membranes were incubated for 2 h at RT with the secondary antibody (1:8,000, Cat#BS13278, Bio-world). Visualization of the protein bands was achieved using the chemiluminescent HRP substrate kit (Cat#P90719, Millipore). Image acquisition was performed using Bio-Rad Image Lab^™ software (v 5.2.1), and ImageJ software was employed for the determination of protein densities. Notably, β-actin was utilized as an internal control, and the quantity of ALOX5 protein in each sample was normalized to that of β-actin.

2.10. Statistical analysis

Data were expressed as the mean ± SEM. p < 0.05 was considered statistically significant. The unpaired Student’s t-test of variance with Bonferroni test was performed for comparisons between two groups. Statistical analysis was calculated with the GraphPad Prism v 9.0.1 software. The ESTIMATE and CIBERSORT analyses were conducted using the Sangerbox online tool (v 3.0).

3.1. Immune cell infiltration and immune status analysis

Initially, the microarray data from the GSE30122 dataset were submitted to CIBERSORT to estimate the percentages of 22 distinct immune cell types and calculate the differential immune landscape in DKD and normal tissues. Compared to normal tissues, 7 out of 22 infiltrating immune cell types, including M1 macrophage cells, resting mast cells, monocytes, and γδ T cells, were more prevalent in DKD tissues (Fig. 2A). Additionally, the heatmap depicted the abundance of 22 immune cell types in the normal and DKD groups (Fig. 2B). Correlation analysis of the 22 immune cell types revealed positive relationships between M1 macrophage cells and monocyte cells (r = 0.54) and γδ T cells (r = 0.84) and a negative correlation between resting mast cells and activated mast cells (r = -0.61). Furthermore, neutrophils exhibited a positive correlation with resting mast cells (r = 0.64) (Fig. 2C). The ESTIMATE method was employed to determine immune and stromal scores in the GSE30122 dataset. The results indicated higher ESTIMATE scores, immune scores, and stromal scores in DKD compared to normal samples (Figs. 2D-F).

3.2. Identification of immune-related gene modules

To further explore the association between gene expression and the abundance of 22 immune cell populations, the OECloud tools were utilized to establish WGCNA with GSE30122. The top 39% of variable genes (4,868 genes) were included in the WGCNA. A soft threshold (β = 22, scale-free network R² = 0.82) was employed to create a scale-free network (Figs. 3A-D). Except for the gray module, containing unidentified genes, 23 color-coded gene modules were segregated for further analysis (Fig. 3E). For GSE30122, hierarchical clustering indicated that 23 co-expression models were analyzed involving immune cells. The black module was positively correlated with M1 macrophage cells (Cor = 0.75, p < 0.0001), monocytes (Cor = 0.5, p = 0.02), and γδ T cells (Cor = 0.74, p < 0.0001). The red module also positively correlated with resting mast cells (Cor = 0.64, p = 0.001) (Fig. 3F). Pearson’s correlation analysis details are presented in Fig. 3G. Scatter plots demonstrated the relationship between MM and GS in the black and red modules. Genes with MM ≥ 0.5 and GS ≥ 0.5 were chosen for further analysis.

3.3. Function enrichment analysis and selection of ICRFGs

Finally, 11 ICRFGs were identified among ferroptosis-related and immune genes selected from WGCNA (Fig. 4A). Additionally, pathway and functional enrichment analyses were performed within the DAVID, Metascape, and Enrichr databases to assess signaling pathways and biological roles of hub ICRFGs in DKD and normal samples. According to GO enrichment analysis, these hub ICRFGs were primarily enriched in negative regulation of protein kinase activity, intracellular sequestering of iron ion, ferric iron binding, and regulation of angiogenesis (Figs. 4B, C). KEGG pathway analysis demonstrated significant enrichment in ferroptosis, VEGF, PPAR, and Fc epsilon RI signaling pathways (Fig. 4D). A diagnostic model for DKD was constructed using two algorithms to separate DKD patients from healthy controls. The top three genes (ALOX5, NQO1, and FABP4) were chosen as diagnostic genes using random forests (Figs. 4E, F), and three out of eleven diagnostic ICRFGs (ALOX5, NQO1, and FABP4) with non-zero coefficients were selected using LASSO regression (Figs. 4G, H). The expression levels of these genes were validated in the GSE30122 training cohort (Fig. 4I).

3.4. Identification and validation of ALOX5 in datasets and in vitro study

Further evaluation of differential expression of hub ICRFGs was conducted using the Nephroseq v5.0 database and the GSE30529 dataset to validate the function of the 3 hub ICRFGs (ALOX5, NQO1, and FABP4). The results demonstrated significantly enhanced expression of ALOX5, NQO1, and FABP4 in DKD samples within the GSE30529 testing cohort. ROC analysis indicated that ALOX5, NQO1, and FABP4 possessed diagnostic values for DKD patients compared to normal patients in the GSE30529 dataset (Figs. 5A-D). ROC curve analysis and gene expression data revealed significantly elevated ALOX5 and NQO1 expression in DKD samples within the Nephroseq v5.0 database, except for FABP4 (Figs. 5E, F). These findings suggest that ALOX5 and NQO1 may serve as biomarkers to measure DKD activity and assess therapeutic efficacy. Following the identification of significantly elevated expression and strong diagnostic value of ALOX5 in DKD, ALOX5 protein levels were assessed in control and HK-2 cells after 48 hours of high glucose stimulation. Results indicated a clear increase in ALOX5 protein levels after 24 hours of high glucose exposure, as determined by western blot analysis. Moreover, the relative protein expression of ALOX5 compared to β-actin was notably higher in the high glucose group than in the control group (p = 0.005) (Fig. 5G). The in vitro validation results enhance confidence in the use of ALOX5 as an early diagnostic indicator and potential therapeutic target for DKD.

3.5. Correlation of ALOX5 expression with serum creatinine level/GFR and the prognostic significance of ALOX5 in DKD

From Nephroseq v5.0 database analysis, ALOX5 expression was positively correlated with increasing serum creatinine levels (r = 0.679, p = 0.003) and negatively correlated with decreasing GFR (r = 0.689, p = 0.028) in DKD patients (Figs. 6A, B). This suggests a close correlation between ALOX5 expression and renal function in DKD patients. Furthermore, the prognostic value of ALOX5 was analyzed using Kaplan-Meier survival analysis (log-rank test) and ROC curve. Based on ALOX5 expression levels, DKD patients from the dbPKD database were categorized into high ALOX5 expression (n = 22) and low ALOX5 expression (n = 21) groups. Notably, high ALOX5 expression was associated with poor prognosis for DKD patients according to univariate analysis (p = 0.00194). Moreover, DKD patients from the dbPKD database were divided into high-risk (n = 21) and low-risk (n = 22) groups based on ALOX5 risk scores, revealing that high-risk groups were linked to poor survival probability via multivariate analysis (p = 2.3e-05). The AUC values of ALOX5 via ROC analysis suggested its potential to distinguish DKD (Figs. 6C, D).

3.6. Enrichment and interaction analysis of ALOX5

To further validate the role of ALOX5 in the development and pathogenesis of DKD, we conducted a GSEA to explore pathways associated with ALOX5 in the GSE30122 dataset. Our analysis revealed three significant DKD-related pathways: focal adhesion, ECM-receptor interaction, and NOD-like receptor signaling pathway (Fig. 7A). We also constructed a gene-gene interaction network using the GeneMANIA database, which demonstrated strong correlations between ALOX5 and the top 20 most frequently modified genes. Notably, the Coactosin Like F-Actin Binding Protein 1 (COTL1) exhibited the strongest correlation with ALOX5 (Fig. 7B). Additionally, the functional analysis highlighted the connection between ALOX5 and its related genes in the regulation of inflammatory response, chemokine production, and their regulation (Figs. 7C-D).

3.7. Correlation between ALOX5 expression and infiltrated immune cells

We further explored the correlation between ALOX5 expression and four specific infiltrated immune cells. To investigate this, we categorized ALOX5 expression levels as high or low. The co-expression heatmap displayed a higher infiltration of the four selected immune cells in the high-expression group of ALOX5 (Fig. 8A). To comprehensively understand the relationship between ALOX5 expression and these immune cells, we employed Spearman correlation analysis. Our findings revealed a significant positive association between ALOX5 expression and the four immune cells (Fig. 8B). Specifically, ALOX5 expression was positively correlated with M1 macrophage cells (r = 0.686, p < 0.001), monocytes (r = 0.556, p = 0.007), and γδ T cells (r = 0.582, p = 0.004), but showed no significant correlation with resting mast cells (r = 0.325, p = 0.140) (Fig. 8C).

DKD is one of the leading causes of ESKD, and most patients die without progressing to ESKD. Multiple factors revealed that inflammation and oxidative stress contribute to the development and progression of DKD [9, 20]. Tubular cell hypertrophy, followed by an increase in tubular basement membrane thickness, represents an early pathogenic change in DKD [9]. The tubulointerstitial aspect of the kidney has long been recognized as crucial in DKD. Renal pathological alterations appear in diabetes patients who have experienced extended periods without developing microalbuminuria [21]. Recent studies have indicated the involvement of ferroptosis in DKD by enhancing renal tubular cell death, promoting promoted fibrosis, and inducing inflammation [9, 22, 23]. Ferroptosis, a recently identified form of programmed cell death, is intertwined with metabolism, redox biology, and various diseases. This phenomenon was initially observed and named by Brent R. Stockwell in 2012. Ferroptosis's components encompass nicotinamide adenine dinucleotide phosphate (NADPH) and glutathione (GSH). It exhibits a distinct morphological, biochemical, and genetic differing from other controlled cell death types [24–26]. The complicated processes of renal inflammation and fibrosis formation involve several interacting pathways leading to a persistent inflammatory infiltrate with macrophages and other immune cells that produce cytokines and pro-fibrotic factors [11]. Several researchers have proposed a potential association between specific immune cells and ferroptosis. Consequently, comprehending the pathogenic and molecular pathways leading to DKD is pivotal for early clinical assessment and treatment. Furthermore, it could aid in the identification of molecular markers for early DKD detection.

The present study employed the CIBERSORT method to assess the immune status and immune cell infiltration in the GSE30122 dataset from the GEO database. Notably, proportions of M1 macrophage cells, mast cells, monocytes, γδ T cells, etc., were notably higher in tubule samples from DKD patients than in normal samples. Subsequently, the ESTIMATE method generated ESTIMATE, immune, and stromal scores in the GSE30122 dataset, all of which were higher in DKD samples than in normal ones. Microarray and bioinformatics analysis has enhanced our comprehension of the molecular mechanisms underlying disease progression and pathogenesis. As such, it is imperative to investigate genetic changes and pinpoint potential clinical biomarkers. The WGCNA was developed using CIBERSORT algorithm outcomes to elucidate the relationship between gene expression and infiltrating immune cell fractions. Correlation coefficients were used to identify crucial immune cell-related gene modules, and genes within hub modules displaying MM ≥ 0.5 and GS ≥ 0.5 were designated as hub biomarkers. Ultimately, twenty-three modules were obtained, and gene modules linked to high correlation patterns with four immune cells, namely M1 macrophage cell, resting mast cells, monocytes, and γδ T cells, were identified. In this study, 11 ICRFGs were initially screened based on candidate genes from WGCNA and ferroptosis-related genes from the FerrDb database.

Subsequently, GO and KEGG enrichment analysis demonstrated significant associations of the 11 ICRFGs with ferroptosis and immune processes, encompassing protein kinase activity, intracellular iron ion sequestering, ferroptosis, VEGF signaling pathway, and PPAR signaling pathway. This suggests the potential significance of these 11 ICRFGs in DKD development. Employing two effective algorithms (LASSO and RF), three genes (ALOX5, NQO1, and FABP4) were identified as potential candidate genes for DKD diagnostic biomarkers. These three ICRFGs exhibited upregulation in DKD patients compared to controls in both the training cohort GSE30122 and the testing cohort GSE30529. ALOX5 and NQO1 displayed high expression in DKD patients within the Nephroseq v5.0 database, with ALOX5 and NQO1 emerging as prognostic biomarkers for DKD based on AUC values from ROC analysis. Furthermore, both K-M survival and ROC analysis affirmed the association between high ALOX5 expression and poor survival probability in DKD via univariate and multivariate analyses in the dbPKD database. This finding was subsequently confirmed in the HK-2 cell line under high glucose stimulation, with ALOX5 protein expression levels assessed through a western blot experiment. The research unveiled that arachidonate 5-lipoxygenase (ALOX5) plays a crucial role, being a non-heme iron-containing enzyme encoded by the ALOX5 gene in humans and a member of the lipoxygenase enzyme family [27]. ALOX5's importance in inflammation is well-documented, encoding 5-lipoxygenase (5-LO) present in leukocytes and catalyzing leukotriene (LT) synthesis [28]. Badr et al. demonstrated that a 5-LO-activating protein antagonist could reduce proteinuria in glomerulonephritis patients, affirming leukotrienes' role in proteinuria [29]. These findings substantiate ALOX5's potential as both a diagnostic marker and therapeutic target for DKD.

Further investigations demonstrated a correlation between ALOX5 expression, declining glomerular filtration rate (GFR), and increased serum creatinine levels in the Nephroseq v5.0 database. Additionally, GSEA, GeneMANIA, and Metascape analyses indicated significant enrichment of ALOX5 in ECM-receptor interaction, NOD-like receptor signaling pathway, chemokine production, regulation of chemokine production, and regulation of inflammatory response. The mesangial extracellular matrix (ECM) is constituted of functional macromolecules and an intricate blend of structural elements. ECM signaling pathways are renowned not just for tissue and organ development, but also for maintaining cellular and tissue structures and functions [30, 31]. ECM-receptor interaction is a microenvironmental conduit that sustains cellular and tissue form and function. Essential structural and functional macromolecules found in the ECM include collagen, fibronectin (FN), and laminin. Notably, ECM and ECM receptors, particularly FN, play vital roles in kidney function [32]. DKD, characterized by proteinuria and declining GFR, is linked to chronic inflammation. The progression of DKD prominently involves chemokines and their corresponding receptors, significantly contributing to the inflammatory process [33]. In DKD animal models, various chemokines, particularly CCL2, CCL20, CXCL5, CXCL7, and CXCL12, display elevated levels within proximal tubules [34]. Among these, CCL2, also known as macrophage chemokine-1 (MCP-1), stands out as the most prominent CC chemotactic chemokine family member, closely tied to tubular damage and renal inflammation in DKD [34, 35]. Finally, co-expression and correlation analysis underscored heightened expression of M1 macrophages, monocytes, and γδ T cells in ALOX5 high-expression groups, positively associated with ALOX5. Macrophage accumulation and activation align with hyperglycemia and tubular damage in DKD [36], leading to deteriorating renal function [37, 38]. Macrophages are classified into M1, generally pro-inflammatory and active, and M2, alternatively activated and anti-inflammatory [39, 40]. M1 macrophage cells accumulate at the site of diabetic kidney injury, and mice lacking macrophage cyclooxygenase-2 (COX-2) revealed a link between M1 and DKD progression, with increased M1 polarization leading to more extensive renal damage [41]. Other immune cells also contribute to DKD progression, including γδ T cells, an innate immune subset with vital mucosal barrier functions. γδ T cells release interleukin (IL)-17A, a Th17 effector cytokine [42]. IL-17A plays a confirmed role in developing immunological and chronic inflammatory disorders, including cardiovascular and renal diseases [43]. Administering an IL-17A neutralizing antibody in BTBR ob/ob mice with renal damage showcased a reversal of structural abnormalities associated with DKD, highlighting IL-17A's involvement in diabetes-mediated renal impairment and its potential as a DKD treatment target [44].

In conclusion, this study furnishes valuable insights into the interplay between gene expression data and infiltrating immune cell quantities through bioinformatics analysis and validation experiments. Furthermore, it identifies ALOX5 as a ferroptosis-related biomarker intertwined with the immune system in DKD progression. This research contributes fresh evidence and avenues for future investigations that delve into the immunological mechanisms underpinning DKD.

Author Contributions: Conceptualization and Methodology, N.N.Z. and Y.Z; Writing-original draft, Y.Z; Visualization and Software, Y.Z and N.N.Z; In vitro experiments, F.H., M.Y.A., and X.Y.; Writing-review, Editing, and Funding acquisition, N.N.Z., X.J.L., and J.L. All authors have read and agreed to the published version of the manuscript.

Funding: This work was supported by the Guizhou Provincial Basic Research Program (Natural Science) (No.: Qiankehe Foundation-ZK [2023] General 428) and Academic New Seedling Cultivation Project of Guizhou University of Traditional Chinese Medicine ((2023)-28) to N.N.Z. The Science and Technology Foundation of Guizhou Provincial Health Commission (GZWKJ 2023-266) to J.L. The National Natural Science Funding of China (81874318; 82073878; 81673453) to X.J.L.

Data Availability Statement: The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Acknowledgments: We appreciate the Sangerbox, OECloud, Shanghai NewCore Biotechnology Co., Ltd. (https://www.bioinformatics.com.cn, last accessed on 10 July 2023), and Xian Tao platform for providing support in bioinformatics analysis and plotting, GEO database for providing available data.

Conflicts of Interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Hills, C.E.; Squires, P.E. The Role of TGF-β and Epithelial-to Mesenchymal Transition in Diabetic Nephropathy. Cytokine Growth Factor Rev. 2011, 22, 131–139, doi:10.1016/j.cytogfr.2011.06.002.
Thomas, M.C.; Brownlee, M.; Susztak, K.; Sharma, K.; Jandeleit-Dahm, K.A.M.; Zoungas, S.; Rossing, P.; Groop, P.-H.; Cooper, M.E. Diabetic Kidney Disease. Nat. Rev. Dis. Primer 2015, 1, 1–20, doi:10.1038/nrdp.2015.18.
Colhoun, H.M.; Marcovecchio, M.L. Biomarkers of Diabetic Kidney Disease. Diabetologia 2018, 61, 996–1011, doi:10.1007/s00125-018-4567-5.
KDOQI KDOQI Clinical Practice Guidelines and Clinical Practice Recommendations for Diabetes and Chronic Kidney Disease. Am. J. Kidney Dis. Off. J. Natl. Kidney Found. 2007, 49, S12-154, doi:10.1053/j.ajkd.2006.12.005.
Chan, G.C.W.; Tang, S.C.W. Diabetic Nephropathy: Landmark Clinical Trials and Tribulations. Nephrol. Dial. Transplant. Off. Publ. Eur. Dial. Transpl. Assoc. - Eur. Ren. Assoc. 2016, 31, 359–368, doi:10.1093/ndt/gfu411.
Johnson, S.A.; Spurney, R.F. Twenty Years after ACEIs and ARBs: Emerging Treatment Strategies for Diabetic Nephropathy. Am. J. Physiol. Renal Physiol. 2015, 309, F807-820, doi:10.1152/ajprenal.00266.2015.
Garcia-Ropero, A.; Badimon, J.J.; Santos-Gallego, C.G. The Pharmacokinetics and Pharmacodynamics of SGLT2 Inhibitors for Type 2 Diabetes Mellitus: The Latest Developments. Expert Opin. Drug Metab. Toxicol. 2018, 14, 1287–1302, doi:10.1080/17425255.2018.1551877.
Kim, S.; Kang, S.-W.; Joo, J.; Han, S.H.; Shin, H.; Nam, B.Y.; Park, J.; Yoo, T.-H.; Kim, G.; Lee, P.; et al. Characterization of Ferroptosis in Kidney Tubular Cell Death under Diabetic Conditions. Cell Death Dis. 2021, 12, 160, doi:10.1038/s41419-021-03452-x.
Zheng, Z.; Zheng, F. Immune Cells and Inflammation in Diabetic Nephropathy. J. Diabetes Res. 2016, 2016, 1841690, doi:10.1155/2016/1841690.
Wada, J.; Makino, H. Inflammation and the Pathogenesis of Diabetic Nephropathy. Clin. Sci. Lond. Engl. 1979 2013, 124, 139–152, doi:10.1042/CS20120198.
Tang, S.C.W.; Yiu, W.H. Innate Immunity in Diabetic Kidney Disease. Nat. Rev. Nephrol. 2020, 16, 206–222, doi:10.1038/s41581-019-0234-4.
Chen, X.; Kang, R.; Kroemer, G.; Tang, D. Ferroptosis in Infection, Inflammation, and Immunity. J. Exp. Med. 2021, 218, e20210518, doi:10.1084/jem.20210518.
Zhao, P.; Hu, W.; Wang, H.; Yu, S.; Li, C.; Bai, J.; Gui, S.; Zhang, Y. Identification of Differentially Expressed Genes in Pituitary Adenomas by Integrating Analysis of Microarray Data. Int. J. Endocrinol. 2015, 2015, 164087, doi:10.1155/2015/164087.
Zhou, N.; Bao, J. FerrDb: A Manually Curated Resource for Regulators and Markers of Ferroptosis and Ferroptosis-Disease Associations. Database J. Biol. Databases Curation 2020, 2020, baaa021, doi:10.1093/database/baaa021.
Huang, D.W.; Sherman, B.T.; Tan, Q.; Kir, J.; Liu, D.; Bryant, D.; Guo, Y.; Stephens, R.; Baseler, M.W.; Lane, H.C.; et al. DAVID Bioinformatics Resources: Expanded Annotation Database and Novel Algorithms to Better Extract Biology from Large Gene Lists. Nucleic Acids Res. 2007, 35, W169-175, doi:10.1093/nar/gkm415.
Zhou, Y.; Zhou, B.; Pache, L.; Chang, M.; Khodabakhshi, A.H.; Tanaseichuk, O.; Benner, C.; Chanda, S.K. Metascape Provides a Biologist-Oriented Resource for the Analysis of Systems-Level Datasets. Nat. Commun. 2019, 10, 1523, doi:10.1038/s41467-019-09234-6.
Xie, Z.; Bailey, A.; Kuleshov, M.V.; Clarke, D.J.B.; Evangelista, J.E.; Jenkins, S.L.; Lachmann, A.; Wojciechowicz, M.L.; Kropiwnicki, E.; Jagodnik, K.M.; et al. Gene Set Knowledge Discovery with Enrichr. Curr. Protoc. 2021, 1, e90, doi:10.1002/cpz1.90.
Wang, G.; Ouyang, J.; Li, S.; Wang, H.; Lian, B.; Liu, Z.; Xie, L. The Analysis of Risk Factors for Diabetic Nephropathy Progression and the Construction of a Prognostic Database for Chronic Kidney Diseases. J. Transl. Med. 2019, 17, 264, doi:10.1186/s12967-019-2016-y.
Franz, M.; Rodriguez, H.; Lopes, C.; Zuberi, K.; Montojo, J.; Bader, G.D.; Morris, Q. GeneMANIA Update 2018. Nucleic Acids Res. 2018, 46, W60–W64, doi:10.1093/nar/gky311.
Jha, J.C.; Banal, C.; Chow, B.S.M.; Cooper, M.E.; Jandeleit-Dahm, K. Diabetes and Kidney Disease: Role of Oxidative Stress. Antioxid. Redox Signal. 2016, 25, 657–684, doi:10.1089/ars.2016.6664.
Adler, S.G.; Kang, S.-W.; Feld, S.; Cha, D.R.; Barba, L.; Striker, L.; Striker, G.; Riser, B.L.; LaPage, J.; Nast, C.C. Can Glomerular MRNAs in Human Type 1 Diabetes Be Used to Predict Transition from Normoalbuminuria to Microalbuminuria? Am. J. Kidney Dis. Off. J. Natl. Kidney Found. 2002, 40, 184–188, doi:10.1053/ajkd.2002.33928.
Wang, Y.; Bi, R.; Quan, F.; Cao, Q.; Lin, Y.; Yue, C.; Cui, X.; Yang, H.; Gao, X.; Zhang, D. Ferroptosis Involves in Renal Tubular Cell Death in Diabetic Nephropathy. Eur. J. Pharmacol. 2020, 888, 173574, doi:10.1016/j.ejphar.2020.173574.
Feng, X.; Wang, S.; Sun, Z.; Dong, H.; Yu, H.; Huang, M.; Gao, X. Ferroptosis Enhanced Diabetic Renal Tubular Injury via HIF-1α/HO-1 Pathway in Db/Db Mice. Front. Endocrinol. 2021, 12, 626390, doi:10.3389/fendo.2021.626390.
Qiu, Y.; Cao, Y.; Cao, W.; Jia, Y.; Lu, N. The Application of Ferroptosis in Diseases. Pharmacol. Res. 2020, 159, 104919, doi:10.1016/j.phrs.2020.104919.
Stockwell, B.R.; Friedmann Angeli, J.P.; Bayir, H.; Bush, A.I.; Conrad, M.; Dixon, S.J.; Fulda, S.; Gascón, S.; Hatzios, S.K.; Kagan, V.E.; et al. Ferroptosis: A Regulated Cell Death Nexus Linking Metabolism, Redox Biology, and Disease. Cell 2017, 171, 273–285, doi:10.1016/j.cell.2017.09.021.
Dixon, S.J.; Lemberg, K.M.; Lamprecht, M.R.; Skouta, R.; Zaitsev, E.M.; Gleason, C.E.; Patel, D.N.; Bauer, A.J.; Cantley, A.M.; Yang, W.S.; et al. Ferroptosis: An Iron-Dependent Form of Nonapoptotic Cell Death. Cell 2012, 149, 1060–1072, doi:10.1016/j.cell.2012.03.042.
Funk, C.D.; Hoshiko, S.; Matsumoto, T.; Rdmark, O.; Samuelsson, B. Characterization of the Human 5-Lipoxygenase Gene. Proc. Natl. Acad. Sci. U. S. A. 1989, 86, 2587–2591, doi:10.1073/pnas.86.8.2587.
Poirier, S.J.; Boudreau, L.H.; Flamand, N.; Surette, M.E. LPS Induces ALOX5 Promoter Activation and 5-Lipoxygenase Expression in Human Monocytic Cells. Prostaglandins Leukot. Essent. Fatty Acids 2020, 154, 102078, doi:10.1016/j.plefa.2020.102078.
Badr, K.F. Lipoxygenases as Therapeutic Targets in the Acute and Subacute Phases of Glomerular Immune Injury. Contrib. Nephrol. 1996, 118, 113–125, doi:10.1159/000425084.
Rosso, F.; Giordano, A.; Barbarisi, M.; Barbarisi, A. From Cell-ECM Interactions to Tissue Engineering. J. Cell. Physiol. 2004, 199, 174–180, doi:10.1002/jcp.10471.
Daley, W.P.; Peters, S.B.; Larsen, M. Extracellular Matrix Dynamics in Development and Regenerative Medicine. J. Cell Sci. 2008, 121, 255–264, doi:10.1242/jcs.006064.
Müller, U.; Brändli, A.W. Cell Adhesion Molecules and Extracellular-Matrix Constituents in Kidney Development and Disease. J. Cell Sci. 1999, 112 ( Pt 22), 3855–3867, doi:10.1242/jcs.112.22.3855.
Chang, T.-T.; Chen, J.-W. The Role of Chemokines and Chemokine Receptors in Diabetic Nephropathy. Int. J. Mol. Sci. 2020, 21, 3172, doi:10.3390/ijms21093172.
Satirapoj, B. Tubulointerstitial Biomarkers for Diabetic Nephropathy. J. Diabetes Res. 2018, 2018, 2852398, doi:10.1155/2018/2852398.
Fadini, G.P.; Bonora, B.M.; Cappellari, R.; Menegazzo, L.; Vedovato, M.; Iori, E.; Marescotti, M.C.; Albiero, M.; Avogaro, A. Acute Effects of Linagliptin on Progenitor Cells, Monocyte Phenotypes, and Soluble Mediators in Type 2 Diabetes. J. Clin. Endocrinol. Metab. 2016, 101, 748–756, doi:2016020312184700850.
Chow, F.; Ozols, E.; Nikolic-Paterson, D.J.; Atkins, R.C.; Tesch, G.H. Macrophages in Mouse Type 2 Diabetic Nephropathy: Correlation with Diabetic State and Progressive Renal Injury. Kidney Int. 2004, 65, 116–128, doi:10.1111/j.1523-1755.2004.00367.x.
Nguyen, D.; Ping, F.; Mu, W.; Hill, P.; Atkins, R.C.; Chadban, S.J. Macrophage Accumulation in Human Progressive Diabetic Nephropathy. Nephrol. Carlton Vic 2006, 11, 226–231, doi:10.1111/j.1440-1797.2006.00576.x.
Calle, P.; Hotter, G. Macrophage Phenotype and Fibrosis in Diabetic Nephropathy. Int. J. Mol. Sci. 2020, 21, 2806, doi:10.3390/ijms21082806.
Gordon, S.; Martinez, F.O. Alternative Activation of Macrophages: Mechanism and Functions. Immunity 2010, 32, 593–604, doi:10.1016/j.immuni.2010.05.007.
Kennedy, A.; Fearon, U.; Veale, D.; Godson, C. Macrophages in Synovial Inflammation. Front. Immunol. 2011, 2, doi:10.3389/fimmu.2011.00052.
Wang, X.; Yao, B.; Wang, Y.; Fan, X.; Wang, S.; Niu, A.; Yang, H.; Fogo, A.; Zhang, M.-Z.; Harris, R.C. Macrophage Cyclooxygenase-2 Protects Against Development of Diabetic Nephropathy. Diabetes 2017, 66, 494–504, doi:10.2337/db16-0773.
Cua, D.J.; Tato, C.M. Innate IL-17-Producing Cells: The Sentinels of the Immune System. Nat. Rev. Immunol. 2010, 10, 479–489, doi:10.1038/nri2800.
Lavoz, C.; Rayego-Mateos, S.; Orejudo, M.; Opazo-Ríos, L.; Marchant, V.; Marquez-Exposito, L.; Tejera-Muñoz, A.; Navarro-González, J.F.; Droguett, A.; Ortiz, A.; et al. Could IL-17A Be a Novel Therapeutic Target in Diabetic Nephropathy? J. Clin. Med. 2020, 9, 272, doi:10.3390/jcm9010272.
Lavoz, C.; Matus, Y.S.; Orejudo, M.; Carpio, J.D.; Droguett, A.; Egido, J.; Mezzano, S.; Ruiz-Ortega, M. Interleukin-17A Blockade Reduces Albuminuria and Kidney Injury in an Accelerated Model of Diabetic Nephropathy. Kidney Int. 2019, 95, 1418–1432, doi:10.1016/j.kint.2018.12.031.

No competing interests reported.

DataSheet.xlsx

Download PDF

Version 1

posted

You are reading this latest preprint version

ALOX5, a novel diagnostic and prognostic marker associated with immune cells and ferroptosis in diabetic kidney disease

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and Methods

2.1. Dataset acquisition and preprocessing

2.2. Quantification of immune cell infiltration and immune status

2.3. Selection of relevant modules and genes with immune infiltration properties

2.4. Function evaluation and selection of hub ICRFGs

2.5. Validation of ALOX5 prognostic value in DKD

2.6. Enrichment and interaction analysis of prognostic model

2.7. Analysis of the relationship between infiltrating immune cells and diagnostic markers

2.8. Cell culture and treatment

2.9. Protein extraction and western blot assay

2.10. Statistical analysis

3. Results

3.1. Immune cell infiltration and immune status analysis

3.2. Identification of immune-related gene modules

3.3. Function enrichment analysis and selection of ICRFGs

3.4. Identification and validation of ALOX5 in datasets and in vitro study

3.5. Correlation of ALOX5 expression with serum creatinine level/GFR and the prognostic significance of ALOX5 in DKD

3.6. Enrichment and interaction analysis of ALOX5

3.7. Correlation between ALOX5 expression and infiltrated immune cells

4. Discussion

5. Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1