Integrated bioinformatics reveals tubulointerstitial immune microenvironment signatures and machine learning-driven prognostication of clinical treatment response in lupus nephritis

Lupus nephritis (LN) is a common and severe glomerulonephritis that often occurs as an organ manifestation of systemic lupus erythematosus (SLE). However, the complex pathological mechanisms associated with LN have hindered the progress of targeted therapies. We analyzed glomerular tissues from 133 patients with LN and 51 normal controls using data obtained from the GEO database. Differentially expressed genes (DEGs) were identified and subjected to enrichment analysis. Weighted gene co-expression network analysis (WGCNA) was utilized to identify key gene modules. The least absolute shrinkage and selection operator (LASSO) and random forest were used to identify hub genes. We also analyzed immune cell infiltration using CIBERSORT. Additionally, we investigated the relationships between hub genes and clinicopathological features, as well as examined the distribution and expression of hub genes in the kidney. A total of 270 DEGs were identified in LN. Using weighted gene co-expression network analysis (WGCNA), we clustered these DEGs into 14 modules. Among them, the turquoise module displayed a significant correlation with LN (cor=0.88, p<0.0001). Machine learning techniques identified four hub genes, namely CD53 (AUC=0.995), TGFBI (AUC=0.997), MS4A6A (AUC=0.994), and HERC6 (AUC=0.999), which are involved in inflammation response and immune activation. CIBERSORT analysis suggested that these hub genes may contribute to immune cell infiltration. Furthermore, these hub genes exhibited strong correlations with the classification, renal function, and proteinuria of LN. Interestingly, the highest hub gene expression score was observed in macrophages. CD53, TGFBI, MS4A6A, and HERC6 have emerged as promising candidate driver genes for LN. These hub genes hold the potential to offer valuable insights into the molecular diagnosis and treatment of LN.

Background: Lupus nephritis (LN) is one of the most common and serious manifestations of systemic lupus erythematosus (SLE) that causes significant morbidity and mortality. Certain biomarkers for LN are sometimes able to assess treatment response in lupus nephritis. This study aimed to compare serum complement levels (C3 and C4) as markers of treatment response of LN and their relation to the LN class in renal biopsy.&#x0D; Methods: This prospective observational study was conducted in the Department of Nephrology, Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka, Bangladesh from July 2018 to August 2019. Twenty seven patients who were diagnosed with LN after kidney biopsy were included in this study. Serum complement levels (C3 and C4), 24 hours urinary total protein (24-hr UTP) and anti-double-stranded DNA (anti-ds DNA) were measured in all patients at baseline, 3 months and 6 months after treatment initiation. These biomarker values before and after treatment were compared between the proliferative and non-proliferative LN patients.&#x0D; Results: Serum C3 levels were significantly different between patients with proliferative LN (Class III and Class IV) and non-proliferative LN (Class V) at baseline (0.47 ± 0.32 g/l versus 0.89 ± 0.43 g/l, p=0.009) and levels changed significantly 6 months after treatment initiation (p&lt;0.001) and likewise for serum C4 levels (0.10 ± 0.06 g/l versus 0.24 ± 0.26 g/l, p=0.040). The values of 24-hr UTP and anti-ds-DNA were significantly different 6 months after treatment with p value &lt;0.05 in both groups but C3 (p&lt;0.001) and renal Systemic Lupus Erythematosus Disease Activity Index (rSLEDAI) (p&lt;0.001) were only significant in the proliferative group. On the other hand, after 6 months treatment, C4 levels became relatively higher but that was not significant in both groups (p&gt;0.05).&#x0D; Conclusion: After 6 months of treatment, serum C3 and C4 levels increased towards normal in both LN groups. Serum C3 and C4 levels in patients with LN correlate with disease activity. Therefore, serum complement (C3 and C4) levels may be utilized as serological biomarkers for treatment response of LN.&#x0D; Birdem Med J 2021; 11(2): 97-102

Immunoglobulin A nephropathy (IgAN) and lupus nephritis (LN) are the most prevalent primary and secondary glomerular diseases, respectively, with several similarities in clinical presentations. Common pathogenic mechanisms in IgAN and LN have been well investigated by previous studies. However, the manifestation mechanism of these two independent diseases carrying distinct immunofluorescent pathological features is still unknown considering the similarities between them. Therefore, differences in pathogenic mechanisms between IgAN and LN were compared in this study. R packages were used for processing the glomerular gene expression datasets acquired from the Gene Expression Omnibus (GEO) database. Least Absolute Selection and Shrinkage Operator (LASSO) and multivariate logistic regression analysis were used to construct models predicting IgAN and LN. Cibersort was used to process the immune cell infiltration analysis. Immunochemistry was used to validate the findings by bioinformatics analysis. In the predicting models based on differentially expressed genes (DEG) and weighted correlation network analysis (WGCNA), retinoic acid receptor γ (RARG) and prolactin releasing hormone (PRLH) were independent risk factors for IgAN, and HECT domain and RCC1-like domain-containing protein 5 (HERC5) and interferon stimulated exonuclease gene 20 (ISG20) were independent risk factors for LN. Gene Ontology (GO) analysis revealed that DEGs mostly correlated to IgAN were enriched in ligand-receptor activity-induced cellular growth and development, while DEGs mostly correlated to LN were enriched in nucleic acid/nucleotide binding-induced type I interferon-related activity and response to virus infection. Immune infiltration analysis showed CD4+ T-cells and M2 macrophage abundance in the glomerular compartment in IgAN and LN, respectively. Immunochemistry validated the predicting models for IgAN and LN and revealed different expression patterns of RARG, PRLH, HERC5, and ISG20. We investigated key differences in the pathogenesis between IgAN and LN and provided validated predicting models to distinguish IgAN and LN. RARG and PRLH, HERC5 and ISG20 might play an essential role in the formation of IgAN and LN, respectively.

Nonstandard and adjunctive medical therapies for systemic lupus erythematosus

BackgroundIdentifying candidates responsive to treatment is important in lupus nephritis (LN) at the renal flare (RF) because an effective treatment can lower the risk of progression to end-stage kidney disease. However, machine learning (ML)-based models that address this issue are lacking.MethodsTranscriptomic profiles based on DNA microarray data were extracted from the GSE32591 and GSE112943 datasets. Comprehensive bioinformatics analyses were performed to identify disease-defining genes (DDGs). Peripheral blood samples (GSE81622, GSE99967, and GSE72326) were used to evaluate the effect of DDGs. Single-sample gene set enrichment analysis (ssGSEA) scores of the DDGs were calculated and correlated with specific immunology genes listed in the nCounter panel. GSE60681 and GSE69438 were used to examine the ability of the DDGs to discriminate LN from other renal diseases. K-means clustering was used to obtain the separate gene sets. The clustering results were extended to data derived using the nCounter technique. The least absolute shrinkage and selection operator (LASSO) algorithm was used to identify genes with high predictive value for treatment response after the first RF in each cluster. LASSO models with tenfold validation were built in GSE200306 and assessed by receiver operating characteristic (ROC) analysis with area under curve (AUC). The models were validated by using an independent dataset (GSE113342).ResultsForty-five hub genes specific to LN were identified. Eight optimal disease-defining clusters (DDCs) were identified in this study. Th1 and Th2 cell differentiation pathway was significantly enriched in DDC-6. LCK in DDC-6, whose expression positively correlated with various subsets of T cell infiltrations, was found to be differentially expressed between responders and non-responders and was ranked high in regulatory network analysis. Based on DDC-6, the prediction model had the best performance (AUC: 0.75; 95% confidence interval: 0.44–1 in the testing set) and high precision (0.83), recall (0.71), and F1 score (0.77) in the validation dataset.ConclusionsOur study demonstrates that incorporating knowledge of biological phenotypes into the ML model is feasible for evaluating treatment response after the first RF in LN. This knowledge-based incorporation improves the model's transparency and performance. In addition, LCK may serve as a biomarker for T-cell infiltration and a therapeutic target in LN.

IntroductionThere is a need for sensitive and specific biomarkers to predict kidney damage and therapeutic response in lupus nephritis (LN). Monocyte chemoattractant protein-1 (MCP-1) and epidermal growth factor (EGF) are cytokines with divergent roles. EGF or EGF/MCP1 ratio have been shown to correlate with prognosis in primary glomerulonephritis, but there is limited information in lupus nephritis (LN). This study evaluated the roles of MCP-1, EGF or their ratio as biomarkers of histopathology and response to treatment in LN.MethodsThis was a cross-sectional and observational study. Baseline urine MCP-1 and EGF levels in systemic lupus erythematosus (SLE) patients and controls (total n = 101) were compared, and levels were correlated with clinicopathological findings and subsequent response to treatment.ResultsMCP-1 was higher in active LN (n = 69) compared to other SLE groups and controls, whereas EGF was not different. MCP-1 correlated with disease activity (proteinuria, renal SLEDAI, classes III/IV/V, and high activity index.) By contrast, EGF correlated with eGFR, but not with proteinuria, activity index, or class III/IV/V. MCP-1 was higher, and EGF was lower in high chronicity index. EGF/MCP-1 decreased with greater clinicopathological severity. In a subgroup with proliferative LN who completed six months of induction therapy (n = 41), EGF at baseline was lower in non-responders compared to responders, whereas MCP-1 was similar. By multivariable analysis, baseline EGF was independently associated with subsequent treatment response. Area under the curve for EGF to predict response was 0.80 (0.66–0.95). EGF ≥ 65.6 ng/ mgCr demonstrated 85% sensitivity and 71% specificity for response. EGF/MCP-1 did not improve the prediction for response compared to EGF alone.ConclusionMCP-1 increased with disease activity, whereas EGF decreased with low GFR and chronic damage. Urine EGF may be a promising biomarker to predict therapeutic response in LN. EGF/MCP-1 did not improve the prediction of response.

Systemic lupus erythematosus (SLE) is a complex autoimmune disease that has significant impacts on patients' quality of life and poses a substantial economic burden on society. This study aimed to elucidate the molecular mechanisms underlying SLE by analyzing glucocorticoid-related genes (GRGs) expression profiles. We examined the expression profiles of GRGs in SLE and performed consensus clustering analysis to identify stable patient clusters. We also identified differentially expressed genes (DEGs) within the clusters and between SLE patients and healthy controls. We conducted Gene Set Enrichment Analysis (GSEA) and Gene Set Variation Analysis (GSVA) to investigate biological functional differences, and we also conducted CIBERSORTx to estimate the number of immune cells. Furthermore, we utilized least absolute shrinkage and selection operator (LASSO) regression and Random Forest (RF) algorithms to screen for hub genes. We then validated the expression of these hub genes and constructed nomograms for further validation. Moreover, we employed single-sample Gene Set Enrichment Analysis (ssGSEA) to analyze immune infiltration. We also constructed an RNA-binding protein (RBP)-mRNA network and conducted drug sensitivity analysis along with molecular docking studies. Patients with SLE were divided into two subclusters, revealing a total of 2,681 DEGs. Among these, 1,458 genes were upregulated, while 1,223 were downregulated in cluster_1. GSVA showed significant changes in the pathways associated with cluster_1. Immune infiltration analysis revealed high levels of monocyte in all samples, with greater infiltration of various immune cells in cluster_1. A comparison of SLE patients to control subjects identified 269 DEGs, which were enriched in several pathways. Hub genes, including PTX3, DYSF and F2R, were selected through LASSO and RF methods, resulting in a well-performing diagnostic model. Drug sensitivity and docking studies suggested F2R as a potential new therapeutic target. PTX3, DYSF and F2R are potentially linked to SLE and are proposed as new molecular markers for its onset and progression. Additionally, monocyte infiltration plays a crucial role in advancing SLE.

Ulcerative colitis (UC) is a chronic nonspecific inflammatory intestinal disease affecting the mucosa and submucosa, characterized by continuous and diffuse active inflammation. However, its underlying pathogenesis remains unclear. This study aimed to identify potential UC biomarkers by integrating weighted gene co-expression network analysis (WGCNA) with machine learning, followed by validation in an experimental UC mouse model. The Gene Expression Omnibus database was systematically queried, and the GSE87466 dataset, comprising of colonic tissues from 87 patients with UC and 21 healthy controls, was retrieved. Differentially expressed genes (DEGs) were identified and subjected to Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses. WGCNA was used to extract UC-related DEGs. Two machine learning algorithms, the Least Absolute Shrinkage and Selection Operator (LASSO) and Support Vector Machine Recursive Feature Elimination (SVM-RFE), were used to screen potential biomarkers. These biomarkers were then validated using animal experiments. A total of 1,097 DEGs were identified. WGCNA constructed nine co-expression gene modules, with the turquoise module (520 genes) exhibiting the highest relevance to UC. LASSO and SVM-RFE analysis identified poly(ADP-ribose) polymerase family member 8 (PARP8) as a potential biomarker of UC. Immunological analysis revealed significantly higher proportions of naive B cells, activated CD4+ memory T cells, follicular helper T cells, γδT cells, M0 macrophages, M1 macrophages, activated mast cells, and neutrophils in UC samples compared to controls. PARP8 expression positively correlated with neutrophils, M1 macrophages, and activated CD4+ T cells, but negatively correlated with plasma cells. In vivo validation confirmed elevated PARP8 expression in dextran sulfate sodium-induced UC mice compared to controls. PARP8 may contribute to UC pathogenesis via immune-related pathways and holds promise as a diagnostic and predictive biomarker.

POS1430 EPIDEMIOLOGY OF LUPUS NEPHRITIS IN BRAZIL: FINDINGS FROM THE MACUNAÍMA STUDY - A NATIONWIDE MULTICENTRIC STUDY

The interferon-induced protein with tetratricopeptide repeats 1 (IFIT1) gene is strongly associated with disease activity index of childhood systemic lupus erythematosus (SLE). However, whether IFIT1-regulated gene expression is the molecular basis of the pathogenesis of SLE has not been fully investigated. Dataset GSE11909 was used to analyze the expression profiles of IFIT1 gene in 103 SLE cases and 12 healthy individuals. Differentially expressed genes (DEGs)-affected by IFIT1 gene were screened between the case group and control group, followed by gene function analysis. The clinical diagnostic potential of the least absolute shrinkage and selection operator (LASSO) model, established based on the expression profiles of IFIT1 and IFIFT1-affected DEGs, was evaluated. Analysis of association between IFIFT1-affected DEGs and immune infiltration was performed. IFIT1 was highly expressed in childhood SLE patients. IFIT1 and IFIT1-affected DEGs showed the potential to serve as a diagnostic marker for childhood SLE with area under the curve (AUC) value of 0.947. Childhood SLE patients showed 826 upregulated DEGs and 4,111 downregulated DEGs compared to the control group. Among them, 208 upregulated DEGs and 214 downregulated DEGs were identified in the IFIT1-high group compared to the IFIT1-low group. The LASSO model for the diagnosis of childhood SLE involved 7 marker genes that were related to immune checkpoint and tertiary lymphoid structure in SLE. Our results confirmed the clinical diagnostic potential of IFIT1 and IFIT1-affected genes in childhood SLE. Moreover, this study elucidated that IFIT1-induced changes in the transcriptome are involved in immune checkpoint and tertiary lymphoid structure in childhood.

Lupus nephritis (LN) is one of the most frequent and serious organic manifestations of systemic lupus erythematosus (SLE). Autophagy, a new form of programmed cell death, has been implicated in a variety of renal diseases, but the relationship between autophagy and LN remains unelucidated. We analyzed differentially expressed genes (DEGs) in kidney tissues from 14 LN patients and 7 normal controls using the GSE112943 dataset. Key modules and their contained genes were identified utilizing weighted gene co-expression network analysis (WGCNA). Differentially expressed autophagy-related genes (DE-ARGs) among DEGs, key module genes and autophagy-related genes (ARGs) were obtained by venn plot, and subjected to protein-protein interaction network construction. Two machine learning methods were applied to identify signature genes. The area under the receiver operating characteristic (ROC) curves was used to assess the accuracy of the signature genes. We also analyzed immune cell infiltration in LN. Additionally, the association between key genes and kidney diseases was predicted. Finally, key genes expression in kidney was verified by clinical samples and animal experiments. A total of 10304 DEGs were identified in GSE1129943 and 29 modules were identified in WGCNA. Among them, the brown module and coral 2 module exhibited significant correlation with LN (cor = 0.86, -0.84, p<0.001). Machine learning techniques identified 5 signature genes, but only 2 were validated in the external dataset GSE32591, namely MAP1LC3B (AUC = 0.920) and TNFSF10 (AUC = 0.937), which are involved in autophagy and apoptosis. Immune infiltration analysis suggested that these key genes may be associated with immune cell infiltration in LN. In addition, these genes have been linked to a variety of renal diseases, and their expression was verified in kidney tissues in LN patients and lupus mice. MAP1LC3B and TNFSF10 may be key autophagy-related genes in LN. These key genes have the potential to provide new insights into the molecular diagnosis and treatment of LN.

FRI0310 Serum Insulin-like Growth Factor-Binding Protein 2 (IGFBP2) as A Biomarker of Clinical and Histopathological Treatment Response in Lupus Nephritis

Lupus nephritis (LN) is one of the most common and severe complications of systemic lupus erythematosus (SLE). Heme metabolism, a critical component of energy metabolism and redox homeostasis, has been strongly implicated in the pathogenesis and progression of various tissue diseases. This study aims to develop accurate and effective diagnostic biomarkers for LN based on heme metabolism-related genes (HMGs) to enhance early diagnosis and precision treatment of LN. This study is based on transcriptomic data and clinical information from LN patients (human kidney tissue samples) in the GEO database, diagnostic genes for LN were identified through differential analysis, weighted gene co-expression network analysis (WGCNA), and least absolute shrinkage and selection operator (LASSO) regression analysis. Further, single-gene gene set enrichment analysis (GSEA) investigated the potential biological functions and signaling pathways associated with diagnostic genes in LN. Immunocellular infiltration levels in LN and Control groups were assessed utilizing CIBERSORT and ssGSEA. To elucidate the regulatory mechanisms of diagnostic genes, corresponding transcription factor and microRNA regulatory networks were constructed. Additionally, based upon the expression profiles of diagnostic genes, LN patients were molecularly subtyped utilizing consensus clustering analysis. This study identified four diagnostic genes for LN (BTG2, CD163, UCP2, and LMO2) and constructed a diagnostic model with robust predictive performance for LN. ssGSEA immune infiltration analysis indicated that most immune related functions and immune cell infiltration levels were significantly elevated in the LN group. KEGG enrichment analysis further revealed that diagnostic genes were enriched in the Toll-like receptor signaling pathway. Through consensus clustering analysis, LN samples were divided into two molecular subtypes with significant differences. The diagnostic model constructed drawing on HMGs can effectively distinguish LN patients and their immune characteristics, thereby providing a new perspective on the relationship between HMGs and LN.

Lupus nephritis (LN) remains one of the most severe manifestations in patients with systemic lupus erythematosus (SLE). Onset and overall LN risk among SLE patients remains considerably difficult to predict. Utilizing a territory-wide longitudinal cohort of over 10 years serial follow-up data, we developed and validated a risk stratification strategy to predict LN risk among Chinese SLE patients - Risk and Factors associated with disease manifestations in systemic Lupus Erythematosus - Lupus Nephritis (RIFLE-LN). Demographic and longitudinal data including autoantibody profiles, clinical manifestations, major organ involvement, LN biopsy results and outcomes were documented. Association analysis was performed to identify factors associated with LN. Regression modelling was used to develop a prediction model for 10-year risk of LN and thereafter validated. A total of 1652 patients were recruited: 1382 patients were assigned for training and validation of the RIFLE-LN model; while 270 were assigned for testing. The median follow-up duration was 21 years. In the training and validation cohort, 845 (61%) of SLE patients developed LN. Cox regression and log rank test showed significant positive association between male sex, age of SLE onset and anti-dsDNA positivity. These factors were thereafter used to develop RIFLE-LN. The algorithm was tested in 270 independent patients and showed good performance (AUC = 0·70). By using male sex, anti-dsDNA positivity, age of SLE onset and SLE duration; RIFLE-LN can predict LN among Chinese SLE patients with good performance. We advocate its potential utility in guiding clinical management and disease monitoring. Further validation studies in independent cohorts are required.

A comprehensive bioinformatics analysis was conducted to investigate potential new diagnostic biomarkers and immune infiltration characteristics associated with tubulointerstitial injury in lupus nephritis (LN), and to examine possible correlations between key genes and infiltrating immune cells. The GSE32591, GSE113342, and GSE200306 datasets were downloaded from the Gene Expression Omnibus database and differentially expressed genes (DEGs) were identified in the pooled dataset. Support vector machine-recursive feature elimination analysis and the least absolute shrinkage and selection operator regression model were used to screen for possible markers, and the compositional patterns of the 22 types of immune cell fractions in LN were determined using CIBERSORT. Finally, Western blotting, quantitative real-time polymerase chain reaction, and multiple immunofluorescence methods were used to confirm the significance of these feature genes in MRL/lpr mice and patients with LN. Seventeen DEGs were identified, of which 11 were considerably upregulated and six were markedly downregulated. Kyoto Encyclopedia of Genes and Genomes pathway analysis revealed significant enrichment in pertussis, complement and coagulation cascades, systemic lupus erythematosus, and other pathways. Based on the machine learning results, we identified IFI16, EGR1 and MX1 were key diagnostic genes for tubulointerstitial injury associated with LN. Immune cell infiltration analysis revealed that IFI16, EGR1 and MX1 were associated with M1 macrophages. Finally, the association between IFI16, EGR1, MX1 and macrophages in MRL/lpr mice and patients with LN were verified. This study suggests that IFI16, EGR1 and MX1 which are highly expressed in renal tubular epithelial cells in LN and are associated with macrophage infiltration, may be a novel diagnostic and therapeutic target.