MSI2 exerts antitumor effects by regulating T-cell function in kidney renal clear cell carcinoma

Introduction

Renal cell carcinoma (RCC) accounts for 2–3% of all malignant tumors (1), and one type of RCC—kidney renal clear cell carcinoma (KIRC)—accounts for approximately 70–80% of all malignant renal tumors (2). The primary method for treating high-risk metastatic KIRC involves combining novel immune checkpoint inhibitors (ICIs) with targeted therapy (3). However, tumor heterogeneity and evolution often lead to resistance to antitumor drugs. Therefore, there is an urgent need to determine the pathogenesis of RCC and to develop new drugs and therapies that are effective against this disease.

Musashi 2 (MSI2) is one of two members of the Musashi family of RNA-binding proteins. It is involved in post-transcriptional regulation (4), and its overexpression promotes myoblast differentiation (5). In terms of cancer, MSI2 plays an important role in the differentiation and proliferation of multiple tumors, including lymphoma and liver, pancreatic, lung, bladder, and breast cancer (6-9). It also has a close relationship with the tumor microenvironment (TME). In non-small cell lung cancer, MSI2 promotes the invasion and metastasis of cancer cells by stimulating cancer-associated fibroblasts and paracrine interleukin (IL)-6 signaling (10). In colorectal cancer, MSI2 potentiates immune infiltration by regulating the post-translational modification of high mobility group box 1 to promote dendritic cells maturation and migration, leading to better prognoses (11). Notably, given that the MSI2/small nucleolar RNA, C/D box 12B/factor interacting with PAPOLA and CPSF1 (FIP1L1)/zinc finger and BTB domain containing 4 (ZBTB4) positive feedback loop plays a crucial role in regulating glycolipid metabolism in Glioblastoma cells, there is interest in MSI2 as a potential drug target for glioma therapy (12).

Li et al. found that MSI2 expression was significantly reduced in clear cell RCC (ccRCC), and its expression level was positively correlated with overall patient survival, as well as closely related to tumor immune cell infiltration and immune checkpoint expression. It may affect prognosis by regulating metabolic reprogramming and tumor immune microenvironment, and has the potential to serve as a prognostic biomarker for ccRCC (13). However, this study did not further investigate the role of MSI2 in malignant phenotypes such as KIRC cell proliferation, nor did it delve into the cellular heterogeneity of KIRC and the expression characteristics of MSI2 in different cell types at the cellular level. Therefore, based on this, this study further validated the prognostic role of MSI2 in KIRC through bioinformatics methods, and explored the direct effect of MSI2 on the proliferation, invasion, and migration ability of KIRC cells through a cell knockdown assay system. At the same time, this study also utilized single-cell RNA sequencing (scRNA-seq) technology to deeply analyze the cellular heterogeneity of KIRC, revealing the dynamic expression patterns of MSI2 in key cell types, providing new scientific evidence for a comprehensive understanding of the mechanism of MSI2 in the occurrence and development of KIRC. We present this article in accordance with the MDAR reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-437/rc).

Methods

Acquisition of clinical samples

We collected five pairs of KIRC and adjacent tissues from patients at Shenzhen Hospital of Southern Medical University between June and September 2024, and we used a quantitative real-time polymerase chain reaction (RT-qPCR) to detect MSI2 expression. Detailed clinical characteristics are presented in Table 1. We also collected 34 pairs of KIRC and adjacent tissues from patients at the same hospital between January 2022 and September 2024 and used immunohistochemistry (IHC) to detect MSI2 expression. Detailed clinical characteristics are presented in Table 2. All patients were pathologically confirmed to have KIRC and had both tumor and adjacent non-tumorous kidney tissue samples available. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethical Committee of Southern Medical University (No. NYSZYYEC2025K078R001), and all participants provided written informed consent.

Data sources

Expression matrix and clinical information were retrieved from The Cancer Genome Atlas (TCGA) Pan-Cancer (PANCAN) and KIRC datasets via the University of California Santa Cruz Xena data hub (https://xenabrowser.net/datapages/). The TCGA-PANCAN dataset included 33 cancers and 10,535 samples. Cancers without paired normal samples were removed, leaving 23 cancers and 8,725 samples. The TCGA-KIRC dataset included 607 samples, with 598 samples with 01A and 11A retained (KIRC: control =526:72,522 with survival information). The median follow-up duration was 1,416 days (range, 0–4,573 days), with 377 patients included in the follow-up analysis. Detailed clinical characteristics are presented in https://cdn.amegroups.cn/static/public/tau-2025-437-1.xls. Transcriptome data were obtained from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/gds). The GSE76351 (GPL570) dataset included 12 KIRC and 12 normal tissue samples. The GSE242299 scRNA-seq dataset (GPL21697 platform) was also obtained from the GEO database. After duplicate samples were removed, this dataset included eight KIRC tumor samples and nine healthy normal kidney tissue samples.

Differential expression analysis of MSI2 messenger ribonucleic acid (mRNA)

Based on cancer data from TCGA-PANCAN, the Wilcoxon rank-sum test was used to examine the disparity in MSI2 mRNA expression between 23 pairs of cancer and control samples (P<0.05). Subsequently, we analyzed MSI2 mRNA expression across various cancers using the TIMER2.0 database (http://timer.comp-genomics.org/timer/). Finally, we intersected the cancers that showed differences in expression in the two analyses.

Differential expression analysis of MSI2

We used the University of Alabama at Birmingham Cancer Data Analysis Portal (UALCAN; https://ualcan.path.uab.edu/index.html) database to assess the differential expression of MSI2 in cancers that showed significant differences in MSI2 mRNA expression (P<0.05).

Diagnostic and prognostic analysis of MSI2

To determine the diagnostic capacity of MSI2, receiver operating characteristic (ROC) curves were plotted for the KIRC (TCGA-KIRC and GSE76351) datasets using pROC (v 1.18.0) (14). Diagnostic performance was assessed based on the area under the curve (AUC) value, with AUC >0.7 indicating diagnostic capability. Based on the expression of MSI2, the cancer samples were classified into high- and low-expression groups (minprop =0.4) using the surv_cutpoint function. A Kaplan-Meier (KM) survival analysis was performed using survminer (v 0.4.9) (15). The KM survival curves for the high- and low-expression groups were plotted, and the differences between the groups were compared using the log-rank test (P<0.05).

IHC

Paraffin sections (thickness: 3–5 µm) were heated in a 72 ℃ oven for 2 h and then dewaxed and rinsed thoroughly with tap water for 3 min, the sections were placed in antigen repair solution and heated together with the container in a pressure cooker until boiling. Cover the pressure valve until steam is sprayed and continue for 1–4 min, and the sections were then soaked in tap water for 2 min. The samples were incubated with 3% hydrogen peroxide at room temperature for 4 min to block endogenous peroxidase activity. The samples were washed with tap water for 2 min and phosphate-buffered saline (PBS) for 2 min. After a 10% bovine serum albumin blocking solution was added, the samples were incubated at 37 ℃ for 40 min and were washed with PBS for 2 min. Then, 60 µL of primary antibody (rabbit monoclonal anti-MSI2 antibody, EP1305Y, Abcam, Cambridge, UK) was added, and the samples were incubated overnight at 4 ℃. After washing with PBS, 70 µL of secondary antibody (Goat Anti-Rabbit IgG, Abcam) was added, and the samples were incubated at 37 ℃ for 40 min. After washing with PBS, diaminobenzidine was added. Samples were counterstained with hematoxylin. The H-score was used to evaluate the degree of expression and t-test was used to compare the expression differences between RCC tissue and adjacent tissues. Differences were considered statistically significant when P<0.05. Replaced primary antibody with PBS as negative control and used normal renal tubular tissue expression as positive control.

Cell culture and transfection

We sourced 786-O KIRC cells and ACHN cells from Procell Biotechnology Co. (Wuhan, China). The cells were seeded in 10 cm dishes and cultured in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum (FBS) (Biological Industries, Kibbutz Beit Haemek, Israel) at 37 ℃ in 5% CO₂.

For the transfection step, 786-O cells and ACHN cells were cultured in 24-well plates. When the cell density reached about 200,000 cells/well (2 mL/well), 60 µL of lentivirus (1×10⁸ TU/mL) was added. The MSI2-knockdown (shMSI2) and negative-control (shLacZ) lentiviruses were synthesized by Haitro Biotechnology Co. (Shanghai, China).

RT-qPCR and Western blot

For the RT-qPCR, 100 mg of KIRC or adjacent tissue was ground in a grinder with 1 mL of TRIzol reagent (ThermoFisher, Shanghai, China). The total RNA was reverse transcribed into cDNA using the HiScript IV 1st Strand cDNA Synthesis Kit (Vazyme Biotech, Nanjing, China). The RT-qPCR was performed using the GoTaq qPCR Master Mix (A6001; Promega, Madison, USA) according to the manufacturer’s instructions. The MSI2 primers were TTCGCAGACCCAGCAAGTG (forward) and TCGCAGATAACCCGCCTAC (reverse). Human β-actin was included as an internal reference, and the β-actin primers were TGGCACCCAGCACAATGAA (forward), CTAAGTCATAGTCCGCCTAGAAGCA (reverse). T-test was used to compare the expression differences (P<0.05).

For the Western blot analysis, poured out the culture medium of the sample (786-O cells and ACHN cells), washed three times with PBS, added 400 µL of lysis buffer containing phenylmethanesulfonylfluoride (add 10 µL phenylmethanesulfonylfluoride to 1 mL lysis buffer), lysed on ice for 30 min, then transferred the lysis buffer to a 1.5 mL centrifuge tube and centrifuged at 12,000 rpm for 5 min at 4 ℃, and the supernatants were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; electrophoresis time 2–3 h, voltage 80 V). The separated proteins were transferred to polyvinylidene fluoride membranes, which were incubated in a blocking solution [Tris-Buffered Saline with Tween 20 (TBST) containing 5% skim milk] at room temperature for 2 h. Primary antibody (rabbit monoclonal anti-MSI2 antibody, EP1305Y, Abcam) diluted in blocking solution was then added, and the membranes were shaken gently overnight at 4 ℃. The primary antibody was washed away, and then the membranes were washed three times with TBST (15 min each time). Prepared the secondary antibody (Goat Anti-Rabbit IgG, Abcam) dilution solution using the same method and contacted it with the membrane. Incubated at room temperature for 1–2 h, then washed twice on a decolorization shaker at room temperature using TBST for 10 min each time; Washed again with TBS for 10 min and performed chemiluminescence reaction.

Cell proliferation, invasion, and migration assays

To determine the effect of MSI2 on KIRC cell proliferation, 786-O cells and ACHN cells infected with the MSI2-knockdown lentivirus (shMSI2 cells) or the negative-control lentivirus (shLacZ cells) were separately seeded into 96-well plates at approximately 4,000 cells/well and cultured at 37 ℃ in 5% CO₂. After 24, 48, 72, 96, and 120 h, 10 µL of Cell Counting Kit-8 (CCK-8) reagent (NCM Biotech, Suzhou, China) was added to each well, and the cells were incubated for 3 h. The absorbance at 450 nm was then measured using a microplate reader (GloMax, Promega). Analysis of Variance was used to compare the differences (P<0.05).

786-O cells (shMSI2 and shLacZ) and ACHN cells (shMSI2 and shLacZ) were starved for 24 h in a serum-free medium. The cell density was adjusted to 200,000 cells/mL, 200 µL was added to the upper chamber of 24-well transwell plates, and 500 µL of culture medium containing 10% FBS was added to the lower chamber. The cells were cultured at 37 ℃ in 5% CO₂ for 24 h. The cells in the lower chamber were stained with crystal violet and quantified. Imaging was performed using an inverted microscope (magnification: 100×, Nikon, Tokyo, Japan).

Relationship between MSI2 and clinicopathologic features

To investigate the relationship between MSI2 expression and clinical characteristics in KIRC, we employed Wilcoxon tests to analyse the correlation between MSI2 expression and age, sex, grade, M stage, and T stage based on clinical information from the TCGA-KIRC dataset (P<0.05). Given that 271 cases had “NX” as the N stage, the N stage was not included in the analysis. We also generated a Sankey diagram using ggalluvial (v 0.12.5) (16). Subsequently, to assess whether MSI2 expression serves as an independent prognostic factor, we employed the survival package (v 3.5-3, https://CRAN.R-project.org/package=survival) to conduct univariate and multivariate Cox proportional hazards (PHs) regression analyses, incorporating other clinical and pathological characteristics for adjustment (P<0.05). Concurrently, to ensure the validity of model assumptions and the interpretability of results, PHs assumption tests were performed using the survminer package (v 0.5.0, https://CRAN.R-project.org/package=survminer) (P>0.05).

Enrichment analysis of MSI2

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed using the gene set variation analysis (GSVA) package (v 1.44.5) (17) to explore the biological functions of the MSI2 from the MSigDB database (https://www.gsea-msigdb.org/gsea/msigdb). The enriched pathways were analyzed for differences using limma (v 3.52.4) (18) (|t|>2 and P<0.05), rigorously identifying pathways exhibiting biologically significant differences. To investigate the relationship between MSI2 and single cells in the TME, we analyzed MSI2 expression in KIRC single-cell datasets based on the TME single-cell database (http://tisch1.comp-genomics.org/home/).

Analysis of the immune microenvironment

To investigate the differences in the immune, stromal, and ESTIMATE scores between the KIRC samples with high and low MSI2 expression, the scores were calculated using ESTIMATE (v 1.0.13, https://R-Forge.R-project.org/projects/estimate/) and subjected to the Wilcoxon rank-sum test. To investigate the infiltration of immune cells in the KIRC samples with high and low MSI2 expression, the CIBERSORT algorithm was used to calculate the proportions of 22 immune cells in the KIRC samples from the LM22 gene set (19) (samples with P>0.05 were excluded, which left 400 samples). This algorithm, as a deconvolution technique, enables the precise estimation of cellular composition in complex tissue samples based on gene expression profiles. Subsequently, Spearman’s rank correlation was performed using the psych package (v 2.2.9) (20) on the different immune cells found to express MSI2. Finally, 12 immune checkpoints (21) were subjected to Spearman correlation analysis using MSI2 in the KIRC samples.

Analysis of the correlations between MSI2 expression and tumor and cellular quantitative measures

Tumor immune dysfunction and exclusion (TIDE) scores were calculated for the KIRC samples using the TIDE website (http://tide.dfci.harvard.edu/). The tumor gene mutation profiles of the patients included in the TCGA-KIRC dataset were analyzed using the maftools package (v 2.12.0) (22), and the tumor mutational burden (TMB) and mutant-allele tumor heterogeneity (MATH) values were calculated for the KIRC patients in the TCGA-KIRC dataset. The microsatellite instability (MSI) data for KIRC were obtained from the cBioPortal database (https://www.cbioportal.org/). Correlations between MSI2 expression and the TIDE, TMB, MATH, and MSI values were analyzed using Spearman’s rank correlation.

Construction of regulatory networks of MSI2

To further explore the gene interactions and functions of MSI2, we analyzed MSI2 and constructed a network using the GeneMANIA database (http://genemania.org/). The microRNA (miRNA) targets of MSI2 were predicted using the miRDB and miRTarBase databases, and intersections were used to identify shared miRNAs. miRNAs corresponding to long non-coding RNAs (lncRNAs)were predicted using the miRNet and starBase (ClipExpNum >10) databases, and intersections were used to identify shared lncRNAs. Finally, a competing endogenous RNA (ceRNA) network was constructed. The transcription factors (TFs) of MSI2 were predicted using the ChEA3 database (https://maayanlab.cloud/chea3/), and a TF-mRNA network was constructed.

scRNA-seq analysis

The GSE242299 scRNA-seq dataset was analyzed using the Seurat package (v 5.0.1) (16). Initial quality control (QC) was performed to filter out low-quality cells and genes, thereby avoiding potential noise introduced into the analysis. Specifically, cells with <200 detected genes and genes expressed in fewer than three cells were excluded. The remaining cells were retained if they met the following criteria: 200< nFeature_RNA (genes/cell) <6,000; 500< nCount_RNA (total RNA count/cell) <10,000; and percent.mt (proportion of mitochondrial gene expression) <25%. This procedure aims to effectively exclude low-quality or abnormal cells, thereby enhancing the accuracy of data analysis and ensuring the reliability of subsequent analytical results. The data were then normalized using the NormalizeData function. Highly variable genes (HVGs) were identified using the FindVariableFeatures function, and a principal component analysis (PCA) was performed on the HVGs using the RunPCA function. The number of principal components (PCs) to retain was determined using the JackStrawPlot function (P<0.05). Unsupervised clustering was performed using the FindNeighbors and FindClusters functions (resolution =0.4). Uniform manifold approximation and projection (UMAP) was then applied to visualize the clusters. The FindAllMarkers function was used to identify overexpressed genes in each cell type. By filtering for genes with |log₂fold change (FC)| >1 and adjusted P<0.01, ensuring the selection of biologically significant and significantly differentially expressed genes, while controlling the false positive rate to ensure statistical significance of the results. Cell types were annotated using the singleR package (v 1.0.6) (23), relevant literature (24), and the CellMarker 2.0 website (http://bio-bigdata.hrbmu.edu.cn/CellMarker/).

The proportions of annotated cells in the KIRC samples, normal samples, and all samples were visualized using ggplot2 (v 3.3.6) (25). The Wilcoxon test was applied to assess the differences in the distribution of the annotated cells between the KIRC and normal samples (P<0.05). Annotated cells with significantly different distributions in the KIRC and normal samples were considered differential cells (P<0.05). A Manhattan plot was generated to visualize the top 10 differentially expressed genes, showing the most significant upregulation and downregulation in the differential cells. Additionally, the expression of MSI2 in the annotated cells was analyzed, and the Wilcoxon test was applied to explore the differences between the KIRC and normal samples for each annotated cell type (P<0.05). The results of the annotated cell distribution differences, MSI2 expression, and MSI2 expression differences analyses were combined to identify key cells in KIRC.

Function enrichment, cell-to-cell communication, and pseudotemporal analyses

To investigate the biological functions of the annotated cells in the KIRC and normal samples, a functional enrichment analysis was performed using the ReactomeGSA package (v 1.12.0) (26). The top 20 enriched pathways were identified and visualized using the pathways function. For a deeper understanding of the annotated cell clusters, a cell-to-cell communication analysis was conducted using the CellChat package (v 1.6.1) (27). We constructed cell-to-cell communication networks to examine the number and strength of interactions between the annotated cell types in both the KIRC and normal samples.

A secondary dimensionality reduction was performed to further examine the key cells. The methodology was the same as that outlined in the “scRNA-seq analysis” section. The key cells were then subdivided into differential subclusters. To explore the differentiation states and trajectories of the key cells, a pseudotemporal analysis was performed using the Monocle package (v 2.26.0) (28). The trajectories of the key cells were visualized using the DDRTree package (v 0.1.5) (https://CRAN.R-project.org/package=DDRTree).

Statistical analysis

Bioinformatics analyses were conducted using the R program (R version 4.3.1). Data from different groups were compared using Wilcoxon’s test. Differences were considered statistically significant when P<0.05.

Results

MSI2 was downregulated in the KIRC samples

When we compared MSI2 expression in the cancer and normal samples from the TCGA-PANCAN dataset, we observed significant variation in the expression across 18 cancer types (Figure S1). In addition, MSI2 mRNA expression differed between the cancer and normal groups in 19 cancer types from the TIMER2.0 database (Figure S2). Among the cancer types found to have differential expression of MSI2 or its mRNA, the following 18 were identified as intersecting cancer types: BRCA, CESC, CHOL, COAD, ESCA, HNSC, KICH, KIRC, KIRP, LIHC, LUAD, LUSC, PCPG, PRAD, READ, STAD, THCA, and UCEC. Seven were identified by comparing the MSI2 expression differences between the cancer and normal groups: KIRC, BRCA, COAD, LIHC, LUAD, LUSC, and UCEC. The KIRC samples exhibited lower MSI2 expression than the associated normal samples (Figure S3A); however, the other six cancer types exhibited higher MSI2 expression than the associated normal samples, which was consistent with the MSI2 gene expression trend (Figure S3B-S3G).

The RT-qPCR and IHC results obtained with the KIRC and paired tissue samples reflected the above findings, showing that the expression of MSI2 in the adjacent tissues was significantly higher than that in the KIRC tissues (Figure 1A,1B). Interestingly, a small number of renal tubule cells exhibited low or no MSI2 expression, similar to the tumor cells (Figure 1C). Hence, our expression analysis of data from the TCGA-KIRC and GSE76351 datasets showed a consistent trend in terms of MSI2 expression in KIRC (Figure 1D).

Figure 1 MSI2 expression was downregulated in KIRC. (A) Expression of MSI2 mRNA in KIRC and adjacent normal tissues determined via RT-qPCR (P<0.001). (B) Expression of MSI2 in KIRC and adjacent normal tissues determined via immunohistochemistry. (C) Cells in the renal tubules showed significantly lower MSI2 expression compared to surrounding cells in normal tissue (indicated by red arrows). (D) Expression of MSI2 in KIRC and normal tissues in the TCGA-KIRC and GSE76351 datasets (Wilcoxon rank-sum test) (*, P<0.05; ***, P<0.001; ****, P<0.0001). GSE, Gene Expression Omnibus Series; KIRC, kidney renal clear cell carcinoma; MSI2, Musashi 2; RT-qPCR, quantitative real-time polymerase chain reaction; TCGA, The Cancer Genome Atlas.

MSI2 expression indicates a good prognosis in patients with KIRC

We assessed the diagnostic efficacy of MSI2 in KIRC by generating ROC curves using the TCGA-KIRC and GSE76351 datasets. The AUC value for MSI2 was >0.7 in both cases (Figure S4A,S4B). The KM survival analysis based on TCGA-KIRC data revealed that higher expression of MSI2 correlated with improved survival among patients with KIRC (Figure S4C). Next, we analyzed the relationship between MSI2 expression and the clinical features of KIRC (Figure S4D), and the results showed that MSI2 mRNA expression correlated with sex and partially correlated with T stage, pathologic stage, and histologic grade (Figure S5A-S5F). Furthermore, both MSI2 expression and age were confirmed as independent prognostic factors for KIRC through univariate and multivariate Cox regression analyses, with both satisfying the PH assumption test (Figure S5G-S5I).

Knocking down MSI2 promoted the proliferation, invasion, and migration of KIRC cells in vitro

To determine the role that MSI2 plays in the inhibition of KIRC, we analyzed the proliferation, migration, and invasion of 786-O cells and ACHN cells with MSI2 knocked down. First, we confirmed that the shMSI2 cells had reduced MSI2 expression via Western blot, and the results are shown in Figure 2A,2B. Then, we used a CCK-8 assay and a transwell-based assay to assess the capacity of the shMSI2 cells to proliferate and migrate. The results showed that the shMSI2 cells had significantly enhanced abilities to proliferate, invade, and migrate compared to the control cells (Figure 2C-2H).

Figure 2 Knocking down MSI2 promoted the proliferation, migration, and invasion of RCC cells in vitro. 786-O cells and ACHN cells were separately infected with either a control (shLacZ) or MSI2-knockdown (shMSI2) lentivirus. RNA and proteins were isolated to determine MSI2 expression. (A,B) Western blotting with an anti-MSI2 antibody was conducted to analyze the level of MSI2 in 786-O cells and ACHN cells, and GAPDH was used as the loading control. (C-F) Control (shLacZ) or MSI2-depleted (shMSI2) 786-O cells and (shMSI2) ACHN cells were separately incubated for 72 h to determine the proliferation rate via a CCK-8 assay. The data shown are from two to three experiments. (G,H) Control (shLacZ) or MSI2-depleted (shMSI2) 786-O cells and ACHN cells were separately seeded in Matrigel-uncoated and coated Transwells and incubated for 18 and 24 h to assess migration and invasion, respectively. Cells that migrated to the lower chamber were stained with crystal violet staining solution and quantified. Imaging was performed using an inverted microscope (magnification: 100×), and representative images are shown. The data shown are from three experiments (*, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001). ①, specimen 1; ②, specimen 2. CCK-8, Cell Counting Kit-8; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; MSI2, Musashi 2; OD, optical density; RCC, renal cell carcinoma.

Determination of the function and regulatory relationships of MSI2 in KIRC

A GO analysis showed that MSI2 is mainly involved in the following biological processes: intracellular pH reduction, synaptic vesicle lumen acidification, proton transmembrane transport, and Golgi lumen acidification (Figure 3A). The genes involved in these biological processes include ATP6V1A, ATP6AP2, ATP6V1H, ATP6V0A4, and ATP6V0D2, which suggests that MSI2 is involved in cellular acidification and hydrogen ion secretion. In KIRC, 135 pathways were significantly enriched, particularly the non-homologous end-joining, lysine degradation, and ribosome pathways (Figure 3B).

Figure 3 Analysis of the function and regulatory relationships of MSI2 in KIRC. (A) GO analysis of MSI2 using data from the MSigDB database. (B) KEGG analysis of MSI2 using data from the MSigDB database (|t|>2 and P<0.05). (C) The GGI network of MSI2. (D) Prediction of miRNA targets of MSI2 using the miRDB and miRTarBase databases. (E) Prediction of the lncRNAs corresponding to the predicted miRNAs targets of MSI2 using the miRNet and Starbase (ClipExpNum >10) databases. (F) Relationships between MSI2 and competitive endogenous RNAs in KIRC, determined using data from the TCGA database. (G) The transcription factors of MSI2 were predicted using data from the ChEA3 database. GGI, gene-gene interaction; GO, Gene Ontology; GSVA, gene set variation analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes; KIRC, kidney renal clear cell carcinoma; MSI2, Musashi 2; TCGA, The Cancer Genome Atlas.

We then performed a gene-gene interaction (GGI) network analysis to enhance our understanding of the functions and interactions of MSI2. The analysis revealed that MSI2 is predominantly linked to the cellular response to heat, the nuclear replisome, trans-damage synthesis, and other biological processes (Figure 3C). A total of 29 shared miRNAs and 64 lncRNAs were predicted (Figure 3D,3E). Thus, we constructed a ceRNA network containing one mRNA, eight miRNAs, and eight lncRNAs, and an MSI2-hsa-let-7f-5p-SNHG4 link was identified (Figure 3F). Subsequently, a TF-mRNA network was established. MSI2 was found to be regulated by 12 proteins, among which USF1, USF2, CEBPB, MAZ, ZNF263, MYOG, MAFK, ZKSCAN1, and NR3C1 are TFs, regulatory TFs, or have transcription-related roles, and BHLHE40, MYC, TAL1, and GATA1, which play important roles in cell differentiation (Figure 3G).

Identification of T cells as key cells for KIRC

In the next phase of this study, we employed the single-cell dataset GSE242299 to identify the key cells that contribute to the development and progression of KIRC. After filtering out ineligible cells and genes, 59,236 cells and 20,134 genes were retained for further analysis (Figure S6A,S6B). From these, 2,000 HVGs and the top 10 HVGs were identified (Figure S6C). A PCA was performed, and the top 50 PCs were selected for downstream analysis (Figure S6D). UMAP was used to classify the cells into 17 clusters (resolution =0.4) (Figure 4A). These clusters were then annotated as cell types, including macrophages, endothelial cells, epithelial cells, T cells, natural killer (NK) cells, fibroblasts, mast cells, monocytes, B cells, neural progenitor cells, and cancer stem cells (Figure 4B,4C and Figure S6E).

Figure 4 Identification of T cells as key cells in KIRC. (A) Cells were classified into 17 clusters (resolution =0.4). (B) The cell clusters were annotated according to the marker genes. (C) The annotated cell clusters in the normal and KIRC tissue samples. (D) Cumulative bar chart showing the abundance of each cell type in the KIRC and normal tissue samples. (E) The proportion of each cell type in each cell cluster in the scRNA-seq dataset. (F) The differences in the proportions of each cell type between the normal and KIRC tissue samples (Wilcoxon test; ns, P>0.05; *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001). (G) MSI2 expression in the different cell types. (H) Difference in MSI2 expression in each cell type between the KIRC and normal tissue samples (Wilcoxon test). CON, control; KIRC, kidney renal clear cell carcinoma; MSI2, Musashi 2; NK, natural killer; scRNA-seq, single-cell RNA sequencing; UMAP, uniform manifold approximation and projection.

When we analyzed the proportions of these cell types in the KIRC and normal tissue samples, we observed that epithelial cells were more abundant in the normal tissue samples, while the tumor samples exhibited higher proportions of T cells and macrophages (Figure 4D,4E). Similar results were observed when we examined the individual cell subpopulations of the KIRC_GSE111360, KIRC_GSE121636, KIRC_GSE139555, and KIRC_GSE145281 datasets (Figure S6F). Additionally, significant differences in the proportions of eight cell types (macrophages, T cells, epithelial cells, NK cells, cancer stem cells, B cells, neural progenitor cells, and mast cells) were found between the KIRC and normal tissue samples (P<0.05) (Figure 4F). We further investigated these differential cell types, including the top 10 differentially upregulated and downregulated genes (Figure S6G). MSI2 expression was primarily observed in T cells and neural progenitor cells (Figure 4G), and it significantly differed between the KIRC and normal tissue samples, particularly in T cells and epithelial cells (P<0.001) (Figure 4H). Based on these results, we identified T cells as key cells in the context of KIRC.

Functional enrichment analysis and cell-to-cell communication analysis of key cells

The functional enrichment analysis revealed that the annotated cells in both the KIRC and normal tissue samples were significantly enriched in several pathways, such as the “TWIK-related acid-sensitive K⁺ channel”, “multifunctional anion exchangers”, and “biogenic amines are oxidatively deaminated to aldehydes by MAOA and MAOB” pathways (Figure 5A). The cell-to-cell communication analysis revealed frequent interactions among the annotated cell types. Notably, epithelial cells exhibited more interactions with other cells (e.g., T cells, NK cells, and macrophages) in the tumor samples than in the normal tissue samples (Figure 5B-5E and Figure S7A,S7B). In addition, T cells primarily received signals in the normal tissue samples, whereas in the KIRC samples, they not only received signals but also sent signals to neural progenitor cells.

Figure 5 Functional enrichment analysis and cell-to-cell communication analysis of the key cells. (A) Functional enrichment analysis results for each cell in the single-cell dataset. (B) The number of cell-to-cell communications in the normal tissue samples. (C) The number of cell-to-cell communications in the KIRC tissue samples. (D) The strength of the cell-to-cell communications in the normal tissue samples. (E) The strength of the cell-to-cell communications in the KIRC tissue samples. The size of each colored circle indicates the number of cells, with larger circles indicating more cells. The cells that emit arrows express ligands, while the cells targeted by the arrows express receptors. The more ligand-receptor pairs, the thicker the line. KIRC, kidney renal clear cell carcinoma; NK, natural killer.

The differentiation trajectories of T cells

A secondary dimension-reducing clustering analysis resulted in the T cells being grouped into eight subclusters (T cells 1–8) (Figure S8A-S8C). We then performed a pseudotemporal analysis to determine the T-cell differentiation trajectories, and the results revealed a progressive differentiation process, with cells transitioning from early to later stages (Figure 6A). The T cells were classified into nine distinct developmental states, with state 1 representing the earliest stage of differentiation (Figure 6B-6E). During T-cell differentiation, the expression of MSI2 exhibited dynamic changes: no change was observed in the early stages, and an upward trend was then observed, followed by stabilization during the later stages (Figure 6F).

Figure 6 The differentiation trajectories of T cells. (A) Representation of the T-cell differentiation process, with cells transitioning from early stages (dark blue) to later stages (light blue). (B) The nine developmental states of the T cells in the KIRC tissue samples. (C) The differentiation stages of the T cells in the normal and KIRC tissue samples. (D) The T-cell distribution in the normal and KIRC tissue samples. (E) The eight T-cell subclusters in the KIRC tissue samples. (F) The MSI2 expression exhibited dynamic changes during T-cell differentiation. KIRC, kidney renal clear cell carcinoma; MSI2, Musashi 2.

MSI2 expression in CD4+ T cells significantly differed between the KIRC and normal tissue samples

To further investigate the specific types of T cells in each T cell cluster, we obtained marker genes (Table 3) from the CellMarker 2.0 database (http://bio-bigdata.hrbmu.edu.cn/CellMarker/index.html) and used them to annotate CD4⁺ and CD8⁺ T cells.

Different colors were used to represent CD4⁺ T cells and CD8⁺ T cells, and the results are shown in Figure 7A and Figure S8B. The proportions of the two types of cells in the KIRC and normal tissue samples were determined, and the KIRC tissue samples had higher proportions of CD8⁺ T cells, while the normal tissue samples had higher proportions of CD4⁺ T cells (Figure 7B,7C). When we screened for significantly different proportions of cells between the KIRC and normal tissue samples, the results showed that CD8⁺ T cells were in significantly higher proportions in the KIRC tissue samples and that CD4⁺ T cells were in significantly higher proportions in the normal tissue samples (Figure 7D). To clarify the expression of key genes in the different cell types, we generated a dot matrix plot of the key genes detected in the scRNA-seq analysis. MSI2 was expressed in both CD4⁺ and CD8⁺ T cells (Figure 7E). When we comparatively analyzed the expression of the key genes in each cell type between the KIRC and normal tissue samples, MSI2 expression was found to significantly differ in CD4⁺ T cells (Figure 7F).

Figure 7 Analysis of CD4⁺ and CD8⁺ T cells in the KIRC tissue samples. (A) UMAP diagram of the annotated CD4⁺ and CD8⁺ T cells. (B,C) The proportions of the two cell types in the normal and KIRC tissue samples. (D) Summary of the proportions of the two cell types in the KIRC and normal tissue samples. (E) MSI2 expression in the CD4⁺ and CD8⁺ T cells. (F) Differential expression of MSI2 in the two cell types between the KIRC and normal tissue samples (ns, P>0.05; *, P<0.05; ***, P<0.001). CON, control; KIRC, kidney renal clear cell carcinoma; MSI2, Musashi 2; UMAP, uniform manifold approximation and projection.

Correlation analysis revealed immune infiltration patterns in KIRC

Finally, we conducted a comparative analysis of the immune, stromal, and ESTIMATE scores of the TCGA-KIRC tissue samples showing high and low MSI2 expression. All three scores were significantly lower in the high-expression samples (Figure 8A). The abundance of regulatory T cells (Tregs), monocytes, M0 macrophages, and M1 macrophages also differed between the low- and high-expression samples (Figure 8B,8C). Additionally, MSI2 showed a positive correlation with M1 macrophages (cor =0.11, P<0.05) and a negative correlation with M0 macrophages (cor =−0.21, P<0.01). MSI2 negatively correlated with Tregs and positively correlated with monocytes (Figure 8D). A significant negative correlation between MSI2 gene expression and TIDE and TMB was also observed in the KIRC samples. A positive correlation was observed between MATH and MSI2 expression. However, there was no statistically significant correlation between MSI and MSI2 expression (Figure 8E).

Figure 8 Correlation analysis showing immune infiltration patterns in KIRC. (A) Immune, stromal, and ESTIMATE scores of the TCGA-KIRC samples with high and low MSI2 expression. (B) Abundance (%) of 22 immune cells in the MSI2 high- and low-expression samples. (C) Summary of the abundance of 22 immune cells in the MSI2 high- and low-expression samples from the TCGA-KIRC dataset. (D) Correlations between MSI2 and regulatory T cells, monocytes, M0 macrophages, and M1 macrophages in TCGA-KIRC dataset samples. (E) Correlations between MSI2 and TIDE, TMB, MATH, and MSI in KIRC. (F) Relationships between MSI2 and immune checkpoint molecules in TCGA-KIRC dataset samples (ns, P>0.05; *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001). KIRC, kidney renal clear cell carcinoma; MATH, mutant-allele tumor heterogeneity; MSI, microsatellite instability; MSI2, Musashi 2; NK, natural killer; TCGA, The Cancer Genome Atlas; TIDE, tumor immune dysfunction and exclusion; TMB, tumor mutational burden.

An immune checkpoint correlation analysis showed significant associations between some immune checkpoint molecules and MSI2 in KIRC. Specifically, galectin 9 (GAL9) had the strongest negative correlation with MSI2 (cor =−0.34, P<0.01), and CD274 (PD-L1) had the strongest positive correlation with MSI2 (cor =0.36, P<0.01) (Figure 8F).

Discussion

ICIs are crucial therapeutic agents used to treat advanced RCC; however, most RCCs develop resistance to ICIs over time (29). The development of resistance is often related to tumor heterogeneity and the composition of the TME (30). Hence, there is a need for new therapeutic agents for the treatment of RCC. MSI2 is known to play a role in various tumors; however, its mechanism of action is not fully understood, and there are limited reports on its function in KIRC. In this study, we used bioinformatics and cell-based experiments to analyze the mechanism of action of MSI2 in KIRC, and we identified MSI2 as a potential prognostic biomarker and regulatory factor for immune infiltration in KIRC.

Our KM survival analysis of the TCGA-KIRC dataset samples suggests that MSI2 has good predictive power for the prognosis of KIRC. Moreover, the expression of MSI2 in normal kidney tissues was significantly higher than that in KIRC tissue samples. These results indicate that MSI2 exhibits an anticancer effect against KIRC and align with the findings reported by Li et al. (31) in their study on triple-negative breast cancer (TNBC). Their research showed that the downregulation of MSI2 subtypes is closely related to disease progression and poor prognosis in patients with TNBC. In contrast, other studies have shown that MSI2 exhibits a pro-cancer effect in the vast majority of solid tumors (6,7).

We then turned our attention to identifying the molecular functions and regulation of MSI2. MSI2 is likely involved in ubiquitin-mediated proteolysis and lysine degradation, which contribute to the regulation of cell cycle progression, proliferation, and differentiation (32). Through a GGI analysis, we found that MSI2 is regulated by 12 TFs that play important roles in promoting cell differentiation. These results suggest that MSI2 promotes cell synthesis and differentiation.

In terms of the expression of MSI2 in renal tissues, our IHC experiments revealed that it was mainly expressed in the renal tubules and collecting tubules. Relatively less expression was observed in KIRC tissues. Given that our results indicated that MSI2 has an antitumor effect, promotes cell synthesis and differentiation, and is expressed at higher levels in normal renal tissue than in tumor tissue, we theorized that MSI2 may not act directly on tumor cells.

To explore the mechanism of action of MSI2 in KIRC, we investigated the cellular composition of the TME in KIRC tissue samples. The TME plays a crucial role in immune suppression or activation and thus influences tumor progression and the efficacy of immunotherapy (33). Research has shown that the presence of a tumorigenic TME is closely related to tumor occurrence, development, and metastasis (34). We found that the immune, stromal, and estimate scores in the KIRC tissue samples with high MSI2 expression were lower than those in the samples with low MSI2 expression, which indicates a close relationship between MSI2 and TME. Furthermore, our single-cell analysis showed that MSI2 is mainly expressed in T cells and plays an important role in the activation and maintenance of T-cell function. In KIRC, tumor recurrence is associated with lower levels of T cell infiltration, while a higher ratio of effector T cells (Teff) to Treg can reduce the risk of recurrence (35). Further analysis shows that MSI2 is mainly expressed on CD4⁺ and CD8⁺ T cells, and CD4⁺ T cells show differential expression in KIRC. Yao et al. found that the LFA-1/ICAM-1 interaction between CD8⁺ and CD4⁺ T cells can promote CD4⁺ Th1 dominant differentiation and enhance the cytotoxicity of CD8⁺ T cells, thereby strengthening the anti-tumor immune response (36). In addition, this study also found that MSI2 is associated with the TF MYC. Previous studies have suggested that MSI2 can enhance its stability by binding to MYC mRNA, thereby upregulating the expression level of MYC (37,38). As a key regulatory factor, MYC is widely involved in regulating the effector functions of CD8⁺ T cells, including the expression of cytotoxic molecules and secretion of cytokines (39,40). Based on the above evidence, we speculate that MSI2 may potentially affect the activation status and effector function of CD4⁺ and CD8⁺ T cells by regulating the expression of MYC (41,42), thus playing an important role in shaping the immune microenvironment and disease progression of KIRC. However, this speculation still needs to be verified through in vitro co culture experiments and other experiments.

The expression of the MSI2 gene significantly negatively correlated with the TIDE score, indicating that an increase in MSI2 gene expression leads to a decrease in the possibility of immune escape. This discovery is consistent with the finding that T cells with higher expression of MSI2 may have stronger anti-cancer activity and reduce the chance of immune escape. On further examination, we found significant correlations between MSI2 and immune checkpoints, with MSI2 showing a significant negative correlation with LGALS9 and a significant positive correlation with PD-L1. This indicates that MSI2 expression in T cells may be comprehensively regulated by immune checkpoints.

Normal renal tubule epithelial cells secrete hydrogen ions, and our results indicate that MSI2 is significantly expressed in these cells and is involved in hydrogen ion secretion. In KIRC, the expression of MSI2 is significantly reduced; consequently, hydrogen ion secretion is also significantly reduced or absent.

KIRC originates from renal tubular epithelial cells (43), and MSI2 is significantly expressed in these cells but not in KIRC cells. This tends to support that carcinogenesis occurs in renal tubular epithelial stem cells and affects the expression of MSI2. Interestingly, a small number of cells in the renal tubules expressed MSI2 in a manner similar to KIRC cells, showing lower or no MSI2 expression. These may be examples of renal tubular epithelial stem cells. It is also worth mentioning that KIRC cells with higher MSI2 expression may more closely resemble normal renal tubular epithelial cells than KIRC cells with lower MSI2 expression and exhibit a lower degree of malignancy. To further explore the function of MSI2 in KIRC, we knocked down MSI2 in 786-O and ACHN cells. This resulted in tumor cells with enhanced abilities to proliferate, invade, and migrate. Based on these results, we concluded that MSI2 may inhibit tumor development by affecting the differentiation of renal cells and thus may be an important marker for the differentiation of renal tubular epithelial stem cells into renal tubular epithelial cells and for distinguishing KIRC cells from renal tubular epithelial cells.

This study investigated the differential expression of MSI2 in KIRC and normal kidney tissue, further revealing the anticancer mechanism of MSI2 in KIRC. However, this study also has some limitations. Although preliminary analysis suggests that MSI2 may be involved in regulating T cell differentiation, there is still a lack of direct functional experimental evidence to validate it. Moreover, the mechanism of the significant decrease in MSI2 expression in renal clear cell carcinoma is not yet clear. In future studies, we plan to explore the mechanism of MSI2 expression reduction in renal clear cell carcinoma to investigate its role in tumor occurrence and development. At the same time, we will design and conduct functional experiments to directly detect the effects of MSI2 on T cell activity and differentiation, in order to clarify its immunomodulatory role.

Conclusions

In summary, the results of this study revealed that MSI2 exerts antitumor effects by regulating T-cell function in KIRC and that it may be a prognostic indicator in patients with KIRC. The results also indicate that MSI2 may be a marker of differentiation and maturation in renal tubular epithelial cells and potentially useful to distinguish KIRC cells from renal tubular epithelial cells.

Acknowledgments

We thank Dr. Kristen Sadler from Scribendi (www.scribendi.com) for editing a draft of this manuscript. We would like to express our sincere gratitude to all individuals and organizations who supported and assisted us throughout this research. In conclusion, we extend our thanks to everyone who has supported and assisted us along the way. Without your support, this research would not have been possible.

Footnote

Reporting Checklist: The authors have completed the MDAR reporting checklist. Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-437/rc

Data Sharing Statement: Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-437/dss

Peer Review File: Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-437/prf

Funding: This study was supported by the Science and Technology Planning Project of Shenzhen, China (Nos. JCYJ20210324130801004 and JCYJ20240813145205008, to Y.Y.), and the Postdoctoral Research Foundation of Shenzhen (No. UN-KC-BHKY202205, to Y.Y.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-437/coif). Y.Y. declares this study was supported by the Science and Technology Planning Project of Shenzhen, China (Nos. JCYJ20210324130801004 and JCYJ20240813145205008), and the Postdoctoral Research Foundation of Shenzhen (No. UN-KC-BHKY202205). The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethical Committee of Southern Medical University (No. NYSZYYEC2025K078R001), and all participants provided written informed consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bahadoram S, Davoodi M, Hassanzadeh S, et al. Renal cell carcinoma: an overview of the epidemiology, diagnosis, and treatment. G Ital Nefrol 2022;39:2022-vol3.
2. Govindarajan A, Salgia NJ, Li H, et al. Characterization of papillary and clear cell renal cell carcinoma through imaging mass cytometry reveals distinct immunologic profiles. Front Immunol 2023;14:1182581. [Crossref] [PubMed]
3. Chen YW, Wang L, Panian J, et al. Treatment Landscape of Renal Cell Carcinoma. Curr Treat Options Oncol 2023;24:1889-916. [Crossref] [PubMed]
4. Yang Y, Zhao M, Ding T, et al. Advances in research of Musashi2 in solid tumors. Nan Fang Yi Ke Da Xue Xue Bao 2022;42:448-56. [Crossref] [PubMed]
5. Wang R, Kato F, Watson RY, et al. The RNA-binding protein Msi2 regulates autophagy during myogenic differentiation. Life Sci Alliance 2024;7:e202302016. [Crossref] [PubMed]
6. Wang X, Wang R, Bai S, et al. Musashi2 contributes to the maintenance of CD44v6+ liver cancer stem cells via notch1 signaling pathway. J Exp Clin Cancer Res 2019;38:505. [Crossref] [PubMed]
7. Kudinov AE, Karanicolas J, Golemis EA, et al. Musashi RNA-Binding Proteins as Cancer Drivers and Novel Therapeutic Targets. Clin Cancer Res 2017;23:2143-53. [Crossref] [PubMed]
8. Huang L, Sun J, Ma Y, et al. MSI2 regulates NLK-mediated EMT and PI3K/AKT/mTOR pathway to promote pancreatic cancer progression. Cancer Cell Int 2024;24:273. [Crossref] [PubMed]
9. Jadhao M, Hsu SK, Deshmukh D, et al. Prolonged DEHP exposure enhances the stemness and metastatic potential of TNBC cells in an MSI2-dependent manner. Int J Biol Sci 2025;21:1705-29. [Crossref] [PubMed]
10. Samart P, Heenatigala Palliyage G, Issaragrisil S, et al. Musashi-2 in cancer-associated fibroblasts promotes non-small cell lung cancer metastasis through paracrine IL-6-driven epithelial-mesenchymal transition. Cell Biosci 2023;13:205. [Crossref] [PubMed]
11. Meng X, Na R, Peng X, et al. Musashi-2 potentiates colorectal cancer immune infiltration by regulating the post-translational modifications of HMGB1 to promote DCs maturation and migration. Cell Commun Signal 2024;22:117. [Crossref] [PubMed]
12. Dong W, Liu X, Yang C, et al. Glioma glycolipid metabolism: MSI2-SNORD12B-FIP1L1-ZBTB4 feedback loop as a potential treatment target. Clin Transl Med 2021;11:e411. [Crossref] [PubMed]
13. Li H, Meng X, You X, et al. Increased expression of the RNA-binding protein Musashi-2 is associated with immune infiltration and predicts better outcomes in ccRCC patients. Front Oncol 2022;12:949705. [Crossref] [PubMed]
14. Robin X, Turck N, Hainard A, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics 2011;12:77. [Crossref] [PubMed]
15. Liu TT, Li R, Huo C, et al. Identification of CDK2-Related Immune Forecast Model and ceRNA in Lung Adenocarcinoma, a Pan-Cancer Analysis. Front Cell Dev Biol 2021;9:682002. [Crossref] [PubMed]
16. Brunson JC. ggalluvial: Layered Grammar for Alluvial Plots. J Open Source Softw 2020;5:2017. [Crossref] [PubMed]
17. Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 2013;14:7. [Crossref] [PubMed]
18. Ritchie ME, Phipson B, Wu D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. [Crossref] [PubMed]
19. Newman AM, Liu CL, Green MR, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods 2015;12:453-7. [Crossref] [PubMed]
20. Robles-Jimenez LE, Aranda-Aguirre E, Castelan-Ortega OA, et al. Worldwide Traceability of Antibiotic Residues from Livestock in Wastewater and Soil: A Systematic Review. Animals (Basel) 2021;12:60. [Crossref] [PubMed]
21. Shibru B, Fey K, Fricke S, et al. Detection of Immune Checkpoint Receptors - A Current Challenge in Clinical Flow Cytometry. Front Immunol 2021;12:694055. [Crossref] [PubMed]
22. Mayakonda A, Lin DC, Assenov Y, et al. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Res 2018;28:1747-56. [Crossref] [PubMed]
23. Zheng M, Hu Y, Liu O, et al. Oxidative Stress Response Biomarkers of Ovarian Cancer Based on Single-Cell and Bulk RNA Sequencing. Oxid Med Cell Longev 2023;2023:1261039. [Crossref] [PubMed]
24. Zvirblyte J, Nainys J, Juzenas S, et al. Single-cell transcriptional profiling of clear cell renal cell carcinoma reveals a tumor-associated endothelial tip cell phenotype. Commun Biol 2024;7:780. [Crossref] [PubMed]
25. Gustavsson EK, Zhang D, Reynolds RH, et al. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinformatics 2022;38:3844-6. [Crossref] [PubMed]
26. Griss J, Viteri G, Sidiropoulos K, et al. ReactomeGSA - Efficient Multi-Omics Comparative Pathway Analysis. Mol Cell Proteomics 2020;19:2115-25. [Crossref] [PubMed]
27. Luo J, Deng M, Zhang X, et al. ESICCC as a systematic computational framework for evaluation, selection, and integration of cell-cell communication inference methods. Genome Res 2023;33:1788-805. [Crossref] [PubMed]
28. Qiu X, Mao Q, Tang Y, et al. Reversed graph embedding resolves complex single-cell trajectories. Nat Methods 2017;14:979-82. [Crossref] [PubMed]
29. Rose TL, Kim WY. Renal Cell Carcinoma: A Review. JAMA 2024;332:1001-10. [Crossref] [PubMed]
30. DA Silva Prade J. An Overview of Renal Cell Carcinoma Hallmarks, Drug Resistance, and Adjuvant Therapies. Cancer Diagn Progn 2023;3:616-34. [Crossref] [PubMed]
31. Li M, Li AQ, Zhou SL, et al. RNA-binding protein MSI2 isoforms expression and regulation in progression of triple-negative breast cancer. J Exp Clin Cancer Res 2020;39:92. [Crossref] [PubMed]
32. Yuan H, Wu X, Wu Q, et al. Lysine catabolism reprograms tumour immunity through histone crotonylation. Nature 2023;617:818-26. [Crossref] [PubMed]
33. Wu K, Lin K, Li X, et al. Redefining Tumor-Associated Macrophage Subpopulations and Functions in the Tumor Microenvironment. Front Immunol 2020;11:1731. [Crossref] [PubMed]
34. Sa P, Sahoo SK, Dilnawaz F. Responsive Role of Nanomedicine in the Tumor Microenvironment and Cancer Drug Resistance. Curr Med Chem 2023;30:3335-55. [Crossref] [PubMed]
35. Ghatalia P, Gordetsky J, Kuo F, et al. Prognostic impact of immune gene expression signature and tumor infiltrating immune cells in localized clear cell renal cell carcinoma. J Immunother Cancer 2019;7:139. Erratum in: J Immunother Cancer 2019;7:273. [Crossref] [PubMed]
36. Yao Y, Zhang Z, Wang S, et al. LFA-1/ICAM-1 Interactions Between CD8(+) and CD4(+) T Cells Promote CD4(+) Th1-Dominant Differentiation and CD8(+) T Cell Cytotoxicity for Strong Antitumor Immunity After Cryo-Thermal Therapy. Cells 2025;14:620. [Crossref] [PubMed]
37. Yeh DW, Zhao X, Siddique HR, et al. MSI2 promotes translation of multiple IRES-containing oncogenes and virus to induce self-renewal of tumor initiating stem-like cells. Cell Death Discov 2023;9:141. [Crossref] [PubMed]
38. Jiang P, Zhang T, Wu B, et al. Musashi-2 (MSI2) promotes neuroblastoma tumorigenesis through targeting MYC-mediated glucose-6-phosphate dehydrogenase (G6PD) transcriptional activation. Med Oncol 2023;40:332. [Crossref] [PubMed]
39. Chou C, Pinto AK, Curtis JD, et al. c-Myc-induced transcription factor AP4 is required for host protection mediated by CD8+ T cells. Nat Immunol 2014;15:884-93. [Crossref] [PubMed]
40. Dar AA, Kim DD, Gordon SM, et al. c-Myc uses Cul4b to preserve genome integrity and promote antiviral CD8(+) T cell immunity. Nat Commun 2023;14:7098. [Crossref] [PubMed]
41. Morales-Sánchez A, Lavaert M, Vacchio MS, et al. Enhancing thymic function improves T-cell reconstitution and immune responses in aged mice. PLoS Biol 2025;23:e3003283. [Crossref] [PubMed]
42. Liang L, Liang H, He M, et al. Integrative multi-omics analysis reveals the interaction mechanisms between gut microbiota metabolites and ferroptosis in rheumatoid arthritis. Front Immunol 2025;16:1608262. [Crossref] [PubMed]
43. Liu S, Li G, Yin X, et al. Comprehensive investigation of malignant epithelial cell-related genes in clear cell renal cell carcinoma: development of a prognostic signature and exploration of tumor microenvironment interactions. J Transl Med 2024;22:607. [Crossref] [PubMed]