NCBI GEO (http://www.ncbi.nlm.nih.gov/geo) (9) is a public functional genomics data repository with high-throughput gene expression data, chips, and microarrays. We used “lupus nephritis” (keywords) to search, and there were 335 results in the GEO Database. Then, we selected Homo sapiens, and 301 results were left. We browsed 301 links in detail. In this study, we want to explore the DEGs between LN glomerulus and normal glomerulus. In the first step, we selected LN kidney samples. And only 5 datasets were left. They were GSE112943, GSE127797, GSE113342, GSE32591, and GSE69438. Then, only glomeruli and tubulointerstitium separated samples were what we needed, and GSE113342 and GSE32591 were selected. The two gene expression datasets GSE113342 (10) and GSE32591 (11) were downloaded. The GSE113342 dataset contained 14 LN glomerulus biopsy samples and six normal glomerulus biopsy samples. GSE32591 contained 32 LN glomerulus biopsy samples and 14 normal glomerulus biopsy samples.

The GEO2R online tool was used to identify DEGs between LN and normal glomerular biopsy samples. Log fold change (FC) >1 and P-value <0.01 were considered statistically significant. The raw data were checked online using Venn software to detect common DEGs. DEGs with log FC <0 were considered downregulated, whereas DEGs with log FC >0 were considered upregulated.

The Database for Annotation, Visualization, and Integrated Discovery (DAVID; http://david.ncifcrf.gov) (12) is an online biological information database. The Kyoto Encyclopedia of Genes and Genomes (KEGG) is a database resource for understanding high-level functions and biological systems from large-scale molecular datasets generated using high-throughput experimental technologies (13). Gene Ontology (GO) is a major bioinformatics tool for annotating genes and analyzing their biological processes (14). DAVID 6.8 Bioinformatics Resources was used for pathway annotations. Statistical significance was set at P < 0.05. A PPI network was established using the STRING (version 10.0) (15) online search tool, and an interaction with a combined score >0.4 was considered statistically significant. The PPI network was visualized using Cytoscape (version 3.7.2).

To assess the expression changes in immune cells and obtain the proportion of various types of immune cells from the glomerulus, we used the online CIBERSORT algorithm (https://cibersort.stanford.edu/). The GSE113342 and GSE32591 series matrix txt format files were downloaded from the NCBI GEO website, and the glomerular expression data were selected (GSM3103966, GSM3103968, GSM3103970, GSM3103972, GSM3103974, GSM3103976, GSM3103978, GSM3103980, GSM3103982, GSM3103984, GSM3103986, GSM3103988, GSM3103990, GSM3103992, GSM3103994-99, GSM807889-7934). Differences in 22 immune cells between normal and lupus glomeruli were analyzed.

All protocols were approved by the ethics committee of the Fujian Provincial Hospital. Overall, 30 LN patients with a mean age of 32.70 ± 12.17 years were included. Six types of pathological classifications (Classes II, III, IV, V, III+V, and IV+V) were used, with five patients in each pathological classification. Normal renal tissues from 5 patients with a mean age of 55.47 ± 8.82 years who underwent nephrectomy due to renal tumors were used as normal controls.

Renal biopsy specimens were embedded in an OCT mixture (Sakura, Hayward, CA, USA) and sliced into 5-μm frozen sections (16). The mouse anti-S100A8 antibody (Proteintech Group, Inc., Chicago, IL, USA) was used. Rabbit anti-synaptopodin antibody (Proteintech Group, Inc., Chicago, IL, USA) was used as a podocyte marker for double immunofluorescence staining. Goat anti-rabbit IgG/Alexa Fluor 594 antibody and goat anti-mouse IgG/Alexa Fluor 488 antibody (BIOGOT, Nanjing, China) were used to visualize the different proteins. 4',6-diamidino-2-phenylindole (DAPI) was used to stain the nucleus. Images were captured using a fluorescence microscope (Nikon Eclipse C1, Japan). Integral optical density (IOD) and the area ratio (AR) of the positively stained area to the glomerular area were used as semiquantitative values of the expression of S100A8. We used Image-Pro Plus software (17).

The clinical data are from the cohort used for fluorescence staining of S100A8 in our study. Thirty LN patients were used for this correlation analysis. Six types of pathological classifications (Classes II, III, IV, V, III+V, and IV+V) were used, with five patients in each pathological classification.

SPSS (version 21.0; SPSS Inc., Chicago, IL, USA) was used to compare the differences between the two groups using the t-test or Mann–Whitney U test. The ratio was calculated using the χ² test. Spearman correlation analysis was performed between clinical data and glomerular S100A8 expression levels. Statistical significance was set at P < 0.05.

From the GSE113342 dataset, 93 DEGs were successfully identified, including 53 upregulated and 40 downregulated genes. From the GSE32591 dataset, 345 DEGs, involving 97 upregulated and 248 downregulated genes, were observed. Out of all the DEGs, 13 were common between the two datasets, as shown in the Venn diagram (Figure 1A). These 13 DEGs discovered between the LN glomerulus and the normal glomerulus based on the two microarray datasets consisted of 12 downregulated genes and one upregulated gene (S100A8). The 13 DEGs are plotted in Figure 1B, where the red and green dots represent upregulated and downregulated genes, respectively.

To gain deeper insight into the biological roles of these 13 DEGs, functional and pathway enrichment analyses were performed using DAVID. The enriched GO terms and KEGG pathways are shown in Table 1 and Figure 2. KEGG pathway analysis revealed that the DEGs were mainly associated with infections and the Toll-like receptor (TLR) signaling pathway. GO biological process analysis indicated that the 13 DEGs were significantly associated with inflammatory response, innate immune response, neutrophil chemotaxis, leukocyte migration, cell adhesion, and cell–cell signaling. The top 3 significantly enriched terms regarding changes in cell component of DEGs were plasma membrane, extracellular exosome, and integral component of the plasma membrane. Changes in molecular function are primarily associated with protein binding.

A total of 13 DEGs were imported into the PPI network complex, comprising 13 nodes and 57 edges, including 12 downregulated genes and the upregulated gene S100A8 (Figure 3). We then applied Cytotype MCODE for further analysis, and the results are shown in Figure 3B.

We can conclude that monocytes (GSE32591) and eosinophils (GSE113342) accounted for the majority of all infiltrating cells in the glomerulus, especially in patients with LN. The differential expression proportion of immune-infiltrating cells in the LN and normal groups is shown in Figures 4, 5. Based on analyses of the GSE32591 database, we found that memory B cells, follicular helper T cells, regulatory T cells (Tregs), and resting natural killer (NK) cells were downregulated in LN glomeruli. In addition, monocytes and activated NK cells were upregulated. Analyses of the GSE113342 database revealed that naive B cells, plasma cells, and resting mast cells were downregulated, and monocytes and activated NK cells were upregulated, similar to the observations from GSE32591. In addition, activated mast cells and eosinophils were also upregulated.

No significant age- or sex-dependent differences were found among the renal biopsies of different LN groups. There were some differences in Systemic Lupus Erythematosus Disease Activity Index (SLEDAI), Creatinine (Cr), Blood Urea Nitrogen (BUN), C3, C4, and Albumin (Alb) among various ISN/RPS class LN patients. In patients with LN, no significant difference in 24-h urinary protein measurements was found. The details are presented in Table 2.

Using immunofluorescence microscopy, the glomerular staining of S100A8 in the controls was found to be weak, and the staining in patients with classes II and V was similar to that of the controls. Glomerular staining was markedly enhanced in Class IV. We found the S100A8 proteins to be distributed throughout the glomerulus, and S100A8 did not colocalize with the podocyte marker synaptopodin (Figure 6).

We conducted a semiquantitative analysis and found a significant increase in IOD and AR in LN compared with that of the controls. However, no significant differences were found in the glomerular expression of S100A8 between the control and class II groups or control and class V groups (Tables 2, 3).

The IOD of S100A8 positively correlated with the AR of S100A8. However, the IOD and AR of S100A8 did not correlate with clinical and laboratory data. Correlations were observed among anti-ds DNA, SLEDAI, Cr, BUN, C3, C4, Alb, and 24-h urinary total protein (UTP) in various ISN/RPS class LN patients. The details are presented in Figure 7.