Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2025 Jan 2;15:469. doi: 10.1038/s41598-024-84834-x

Comprehensive analysis of heat shock protein 110, 90, 70, 60 families and tumor immune microenvironment characterization in clear cell renal cell carcinoma

Wenjing Liao 2,#, Mao Huang 2,#, Xiaoyi Du 2, Liangdan Tang 2, Junwu Li 2, Qin Tang 1,✉
PMCID: PMC11697189  PMID: 39747468

Abstract

Heat shock proteins (HSPs) are a kind of molecular chaperone that helps protein folding, which is closely related to cancer. However, the association between HSPs and clear cell renal clear cell carcinoma (ccRCC) is uncertain. We explored the prognostic value of HSP110, HSP90, HSP70 and HSP60 families in ccRCC and their role in tumor immune microenvironment. The data obtained from the Cancer Genome Atlas (TCGA) were applied to determine the differential expression of HSPs in normal tissues and ccRCC. We comprehensively analyzed the prognostic value of HSPs in ccRCC and constructed a prognostic signature. We further explored the differences of tumor immune microenvironment and targeted therapy based on the signature. Cell proliferation, invasion and metastasis were detected by CCK8 assay, wound healing and transwell. Three clusters were identified with differences in overall survival and tumor stage. 6-gene signature (HSPA8, HSP90B1, HSPA7, HSPA12B, HSPA4L, HSPA1L) was identified to predict ccRCC patients’ prognosis. The signature was confirmed in the internal cohort. Survival analysis, receiver operating characteristic (ROC) curve, univariate and multivariate COX regression analysis demonstrated the accuracy and independence of signature. The expression of HSPA7, HSPA8 and HSP90B1 were validated with quantitative real-time PCR. Our signature played a pivotal role in predicting tumor immune microenvironment, immune checkpoint gene expression, drug sensitivity, and tumor mutational burden (TMB) in patients with ccRCC. Our cellular experiments confirmed HSPA7 promotes the proliferation, invasion and metastasis of ccCRC cells. The HSPs signature identified in this study could serve as potential biomarkers for predicting prognosis and treatment response in ccRCC patients. It may provide new ideas for the current research on targeted therapy and immunotherapy strategies for ccRCC patients.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-84834-x.

Keywords: Heat shock protein, Prognostic signature, Clear cell renal cell carcinoma, Tumor immune microenvironment, Immune checkpoints, Drug sensitivity

Subject terms: Cancer, Cancer therapy, Urological cancer

Introduction

Renal cell carcinoma (RCC) is one of the most common and aggressive urologic cancers1. The morbidity and mortality rates of RCC have been increasing in recent years. RCC has a variety of pathological types, of which approximately 75% are clear cell renal cell carcinoma2. Surgery is still an effective treatment for localized ccRCC. However, due to the insensitivity of ccRCC to radiotherapy and chemotherapy, the mortality and recurrence rate of ccRCC are still very high3,4. Patients with similar clinicopathological characteristics have different response to treatment and prognosis because of individual gene differences5. Therefore, in addition to the clinicopathological characteristics, it is imperative to identify prognostic indicators at the gene level, integrate individual differences, and develop new treatment models to improve the overall clinical outcomes of patients in ccRCC.

Heat shock proteins (HSPs) generally function as molecular chaperones to assist the folding of nascent polypeptides and maintain the normal structure and function of proteins6–8. Traditionally, according to different molecular weights, HSPs are divided into 6 families, including small heat shock proteins (sHSPs), HSP40 (DNAJ), HSP60 (HSPD), HSP70 (HSPA), HSP90 (HSPC) and large HSPs (HSPH)9–11. HSPs are aberrantly expressed in a variety of cancers, including breast, prostate and ovarian cancers12–14. HSPs have been found to regulate tumor cell proliferation, invasion and migration, as well as tumor sensitivity to radiotherapy and chemotherapy15–17. HSPs can also affect the secretion of IL-10, IL-6, CSF2, IL-12 and other cytokines, thereby affecting tumor immunity18. Changes in HSPs may have implications for tumor immune microenvironment and tumor patient prognosis, and HSPs may be a potential biomarker for cancer diagnosis and prognosis16,19,20. In conclusion, HSPs exerted a crucial role in tumorigenesis and antitumor process. However, the complete understanding of the role of HSPs in ccRCC, including the interaction between HSPs and the tumor microenvironment is still limited.

In this study, we performed a comprehensive retrospective analysis of the expression, prognosis, mutation and copy number variation (CNV) of HSP110, HSP90, HSP70, and HSP60 families in ccRCC based on public databases. The least absolute shrinkage and selection operator (LASSO) regression analysis and consensus clustering were applied to distinguish subgroups. Then the relationship between HSPs and the prognosis, drug sensitivity and tumor immune microenvironment of patients with ccRCC was explored. We performed this study to provide a new perspective for finding potential prognostic markers of ccRCC, to provide new ideas for exploring the role of HSPs on the tumor immune microenvironment of ccRCC, and to provide some clinical guidance for individualized treatment of ccRCC.

Materials and methods

Datasets

This study’s workflow was presented in Supplementary Fig. S1 online. The information of 539 ccRCC and 72 normal kidney tissues were obtained from the TCGA database (http://cancergenome.nih.gov/). ccRCC patients without clinical information were excluded from analysis. We used the limmar R package to identify the genes with differential expression between tumor and normal tissue. The somatic mutation data were obtained from the TCGA database. The maftools Bioconductor package was used to read the MAF files to count the variants in each sample and visualize it. Copy number data was acquired from UCSC Xena database (https://xena.ucsc.edu/public) for CNV analysis. The R package of RCircos was used to plot the copy number variation landscape of HSP genes in 23 pairs of chromosomes. In this study, there were 4 HSP families, including HSP110 family (HSPH1, HYOU1), HSP90 family (HSP90AA1, HSP90AB1, HSP90B1, TRAP1), HSP70 family (HSPA2, HSPA4, HSPA5, HSPA6, HSPA1A, HSPA1B, HSPA1L, HSPA4L, HSPA7, HSPA8, HSPA9, HSPA13, HSPA14, HSPA12A, HSPA12B), and HSP 60 family (HSPD1), with a total of 22 HSPs.

Consensus clustering

Univariate COX regression analysis was performed to preliminarily screen out the prognosis-related HSPs in ccRCC. Consensus clustering was used to distinguish subgroups based on the expression levels of each prognostic-related HSPs in ccRCC patients. Survival analysis was performed on patients in each subgroup, and heatmaps visualized clinical information and prognostic-related HSPs expression in each subgroup.

Establishment of HSP genes prognostic signature and independent prognostic analysis

Patients were randomized into training cohorts and testing cohorts for risk signature construction and verification. Based on the prognostic-related genes that were initially screened, by using LASSO regression analysis with the training cohort, a prognostic signature was constructed. We could get the optimal lambda value and a list of prognostic genes with coefficients generated by the LASSO model. Calculated the risk score for each sample in the cohort by the formula:Inline graphic, here xi represents the expression of each prognostic gene, coef represents the coefficient value, and k represents the amount of the prognostic gene. The median risk score was used as a cutoff value to classify ccRCC patients into high-risk and low-risk groups, and survival analysis was performed using the Kaplan-Meier (KM) method. In order to analyze the prediction effect of signature, ROC curve was constructed and the AUC was computed. The same algorithms were used for the test cohort and the entire cohort. Heatmaps were performed to show the distribution of clinicopathological features in low-risk and high-risk groups. To verify the independence of prognostic signature, univariate and multivariate COX regression analyses were used. Then, the RMS R package was used to draw nomogram for clinical practice. The IC50 of ccRCC patients was calculated for drug sensitivity assessment using the pRRophetic R software package based on the Genomics of Cancer Drug Sensitivity (GDSC) database.

Evaluation of the TMB and immune microenvironment

Enrichment levels of immune cells in samples analyzed by ssGSEA using the gsva R package. Moreover, we assessed the level of tumor immune infiltration in ccRCC patients by CIBERSORT. Immune checkpoint genes expression differences were analyzed based on prognostic signatures. The correlation between risk score and TMB was explored.

Cell culture

The ccRCC cell line (769-P and ORSC2) and normal human tubular epithelial cell line (HK-2) were obtained from the American Type Culture Collection (Rockville, MD, USA). The ccRCC cell lines (A498 and 786-O) were obtained from Procell Life Science & Technology Co. Ltd. (Wuhan, China). A498,769-P, ORSC2, HK-2 and 786-O were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium (Gibco, CA, USA) supplemented with 10% FBS (FBS Premium, Pan Biotech, Germany) and 1% penicillin/streptomycin (Invitrogen, Carlsbad, CA, USA). The cells were incubated at 37◦C in 5% CO2.

Total RNA extraction and qPCR

Total RNA from four cell lines were extracted using the Steady Pure Quick RNA Extraction Kit (AG21023, Accurate Biotechnology, Hunan, China) and reverse transcribed (Takara, Dalian, China) to acquire cDNAs. qPCR was performed with BlazeTaq™ SYBR Green qPCR Mix (Genecopoeia, Guangzhou, China) on a BIO-RAD PCR System. The internal control for this qPCR is human β-actin. The relative expression levels of these genes were calculated by the 2-ΔΔCt method. Primers for HSPA7: Forward primer 5’-3’ TGAACTCATTGGCATCCCTCC; Reverse primer 5’-3’ CCCTTGTCATTGGTGATCTTGTTAG.

Western blot analysis

Total protein was extracted from ccCRC cells using RIPA buffer (Beyotime, Shanghai, China) with a supplementary protease inhibitor (Proteintech, Wuhan, China). Then, the protein concentration was measured by the BCA protein assay kit (Beyotime, Shanghai, China) and 30 µg of lysates were electrophoresed on 10% SDS-PAGE and transferred to a polyvinylidene fluoride membrane (Millipore, Darmstadt, Germany). After blocking with 5% BSA, the membranes were incubated overnight with the indicated primary antibody at 4 °C, followed by incubated with Goat anti-Rabbit IgG (H + L) Secondary Antibody, HRP conjugate (1:10000, #BA1039, BOSTER) or Goat anti-Mouse IgG (H + L) Secondary Antibody, HRP conjugate (1:10000, # BA1038, BOSTER) for 1 hat room temperature. Protein bands were visualized using the Enhanced Chemiluminescence System (BIOMIKY, Shanghai, China). The primary antibodies used for the Western blot were as follows: Vimentin antibody (1:1000, # 5741 S, Cell Signaling Technology), E-Cadherin Antibody (1:1000, # 3195 S, Cell Signaling Technology), and anti N-Cadherin antibody (1:10000, # ab76011, Abcam).

Cell counting kit-8 (CCK-8)

A cell counting kit-8 (CCK-8) assay (MeilunBio, Dalian, China) was performed to analyze cell proliferation. Briefly, a 5 × 103/well of cells were seeded onto a 96-well plate and cultured overnight. Then, the cells were incubated with 10µL of CCK-8 and 100 µl culture medium at 37 °C for 1 h. The absorbance values at 450 nm were analyzed. Each sample was tested three times.

Scratch wound healing assay

Cells were seeded into 6-well culture plates and allowed to reach approximately 100% confluence before the monolayer was scratched using a sterile p200 pipette tip. Unattached cells were then removed through gentle washing. Images of the cells at various time points were captured under an inverted microscope for analysis. The migration area (%) was calculated using the formula: migration area (%) = (A0 − Ar)/A0 × 100, where A0 represents the initial wound area and Ar represents the remaining wound area at the indicated time point.

Transwell assay

Cells were seeded into the apical chamber with 5% FBS-containing medium, while the basolateral chamber was filled with 15% FBS medium. Meanwhile, cells were placed on the top chamber of membrane pre-coated with Matrigel for invasion assay. Following a 24–48 h incubation period, migrated and invaded cells were fixed with para-formaldehyde fixativel (Biosharp, Hefei, China) and stained with crystal violet (Beyotime, Shanghai, China). Subsequently, an inverted microscope (Olympus, Tokyo, Japan) was utilized for image capture and cell count determination. Bio-triple replicates were performed for the transwell assay.

Flow cytometry

Cells were digested with trypsin, harvested, and washed three times with PBS. Subsequently, the cells were fixed by slowly adding 70% ethanol precooled at 4 °C and placed at -20 °C for more than 4 h. After fixation, the ethanol was removed and the cells were washed with PBS. Following this, each tube received 500 µL of Rnase A and propidium iodide (PI) staining solution (Rnase A∶PI = 1∶9), which was then incubated at room temperature in the dark for 30 min. The samples were analyzed using a CytoFLEX flow analyzer for cell cycle detection. Fluorescence was recorded at 488 nm, and the experiment was repeated three times. Data analysis statistics were performed using FlowJo 10 software to obtain the proportion of cells in the G0/G1, S, and G2/M phases.

Statistical analysis

The Wilcox test and one-way ANOVA test were used to compare the differences in gene expression levels. The differences in clinical information between the different groups were compared by chi-squared test. Pearson or Spearman analysis were performed to evaluate the correlation. Statistical analysis of data was conducted by R software (v4.0.5), PERL programming language (version 5.32.1.1), GraphPad Prism software, and P < 0.05 was considered statistically significant.

Results

Overview of genetic changes of HSP genes in ccRCC

We estimated the CNV frequency and somatic mutations of 22 HSP genes in ccRCC. The results suggested that the mutation frequency of HSP90AA1 was 2%, the mutation frequency of HSPA4, HSPA12A, HSPA5, HSP90B1, HYOU1, HSPA9, HSP90AB1, HSPA8 was 1%, the other 13 HSP genes were not found mutations in TCGA ccRCC samples (Fig. 1A). As shown in Fig. 1B, HSPA9 and HSPA4 have the highest CNV alterations. The CNV alteration of the HSPs in the chromosome were shown in Fig. 1C. The difference in mRNA expression levels of 22 HSPs between normal tissues and ccRCC were presented in Fig. 1D. 8 genes (HSP90AA1, HSPA1A, HSPA1B, HSPA2, HSPA12A, HSPA8, HSPA9, HSPA12A, HSPA4) were downregulated while 8 genes (HSP90AB1, HSP90B1, TRAP1, HSPA6, HSPA7, HSPA13, HSPA14, HSPA5) were upregulated in tumor tissues. The correlation between the 22 HPSs was explored, and a wide range of correlations existed between the different HSPs (see Supplementary Fig. S2 online).

Fig. 1.

Fig. 1

Open in a new tab

The characteristics, correlations and differences of heat shock proteins (HSPs) in clear cell renal cell carcinoma (ccRCC). (A) The mutation frequency of 22 HSP genes in 336 ccRCC patients from TCGA-STAD cohort. The right number indicated the mutation frequency in these genes. The annotations below the waterfall chart indicate different mutation types represented by different colors (B) The CNV variation frequency of gain and lost for HSP genes in TCGA cohort. The height of the column represented the alteration frequency. (C) The location of CNV alteration of HSP genes on 23 chromosomes using TCGA cohort. (D) The expressions of HSP genes between normal tissues (blue) and ccRCC tissues (red). The asterisks represented the statistical p value (*: P < 0.05; **: P < 0.01; ***: P < 0.001).

Tumor classification based on differential expression of HSP genes

8 prognosis-related genes (HSPA8, HSPA9, HSPA4L, HSP90B1, HSPA12B, TRAP1, HSPA1L, HSPA7) were initially obtained (see Supplementary Fig. S3 online). The comprehensive landscape of HSPs interactions, HSPs connection and their prognostic significance for ccRCC patients was depicted with the HSPs network (Fig. 2A). The results further illustrated that not only the HSPs in the same family presented a remarkably correlation in expression, but also a significant correlation was shown among the HSP110 family, HSP90 family, HSP70 family, and HSP60 family. We performed consensus cluster analysis to investigate the association between the expression of 8 genes and ccRCC subtypes. The clustering variable (k) ranged from 2 to 9, when k = 3, the intragroup correlation was high and the intergroup correlation was low. Therefore, 530 ccRCC patients could be suitable divided into three clusters based on 8 genes (Fig. 2B and C). The OS of the three clusters showed a particularly significant survival advantage in cluster A (P < 0.001, Fig. 2D). Then the clinicopathological characteristics of the three clusters were further compared, N stage, T stage and AJCC stage were significant differences among the three clusters (Fig. 2E).

Fig. 2.

Fig. 2

Open in a new tab

Biological characteristics and consensus clustering analysis of HSPs in ccRCC. (A) Interaction between HSPs in ccrcc. The circle size represents the impact of each HSP on the prognosis. By Cox regression test, the value ranges are P < 0.0001, P < 0.001, P < 0.01, P < 0.05 and P < 0.1 respectively. Right semicircle green, prognostic protective factors; Right semicircular purple, prognostic risk factors. The left semicircle indicates different HSP families with red, black, yellow and gray, respectively. The line connecting HSPs represents the interaction between them, and the thickness represents the correlation strength between HSPs. It is positively correlated with pink and negatively correlated with blue. (B) Consensus clustering matrix for k = 3. (C) CDF curves for k = 2–9. (D) Kaplan-Meier curves of overall survival (OS) for ccRCC in three clusters. (E) Heatmap and the clinicopathologic characters of the three clusters classified by these differentially expressed genes.

Establishment and validation of prognostic signature for HSP genes

LASSO regression analysis was performed on 8 genes in the training cohort, and the prognostic signature of 6 genes (HSPA8, HSPA4L, HSP90B1, HSPA12B, HSPA1L, HSPA7) were constructed and the corresponding coefficients were obtained (Fig. 3A and C). The risk score of ccRCC patients could be calculated by the formula, and the median risk score was used to divide patients into high-risk group and low-risk groups. To explore the prediction ability of the model, we applied it to the entire cohort and the testing cohort. Survival analysis revealed that compared with the low-risk group, patients in the high-risk group tended to have a poorer prognosis (P < 0.001, Fig. 3D and F). The survival status and risk score distributions of each ccRCC patients were integrated and displayed on the dot polt, and it was observed that OS rate and OS time decreased with increasing risk score. The expression profiles of the 6 prognostic genes indicated that the low-risk group exhibited elevated expression levels of HSPA1L, HSPA12B, HSPA4L, and HSPA8, while demonstrating decreased expression levels of HSP90B1 and HSPA7 (Fig. 3G and I). Finally, receiver operating characteristic (ROC) curve and the area under the curve (AUC) showed that our signature had good accuracy (Fig. 3J and L).

Fig. 3.

Fig. 3

Open in a new tab

Construction of risk signature in the TCGA cohort. (A, B) Least absolute shrinkage and selection operator (LASSO) Cox regression identified a risk prognosis model. (C) Coefficients of 6 HSP genes. (D-F) In the train cohort (D), test cohort (E) and entire cohort (F), Kaplan-Meier curves suggested that the high-risk subgroup had worse overall survival (OS) than the low-risk subgroup. (G-I) The risk scores distribution, ccRCC patients’ survival status and expression heatmap in the train cohort (G), test cohort (H) and entire cohort (I). (J-L) The train cohort (J) and test cohort (K): 1-, 3- and 5-year ROC curves of risk signature.

Association of prognostic signature with clinical information in the entire cohort

To further validate the reliability of this signature, we explored the relationship among clinical features, genes expression and risk scores. A heatmap was used to visualize differences among clinical features, genes expression, and risk scores (Fig. 4A). Additionally, there were significant differences in risk scores among the diverse groups in terms of the gender, T stage, N stage, M stage, AJCC stage, and grade; no obvious differences were observed in age groups (≤ 65 and > 65) (Fig. 4B and Supplementary Fig. S4A-4C). At the same time, we found that the risk score was the lowest in cluster A (see Supplementary Fig. S4D online). These results showed that risk scores were correlated with subtype groupings. Next, to examine the predictive ability of our signature, stratified by gender (male and female), age (≤ 65 and > 65), stage (stages I-II and stages III-IV), grade (grades 1–2 and grades 3–4), T stage (T1-2 and T3-4), N stage (N0 and N1), and M stage (M0 and M1), survival analysis was carried out in different groups. Survival analysis in subgroups found that patients in the high-risk score group had a worse prognosis than those in the low-risk score group, except that there was no obvious difference in the M1 group (Fig. 4C and Supplementary Fig. S4E-4 J). The above results further indicated that this risk scoring system performed well in predicting OS in ccRCC patients, and associated with clinicopathological characteristics in ccRCC patients.

Fig. 4.

Fig. 4

Open in a new tab

Clinical evaluation of the prognosis risk signature. (A) Heatmap of the expression of 6 genes and distribution of clinical characters between the high-risk and low-risk groups. (B) Correlation between risk signature and clinicopathological. (C) Kaplan Meier survival analysis stratified by clinicopathological features based on risk signature.

Independence verification and nomogram construction

The risk score of HSP genes was an independent prognostic factor in ccRCC patients as shown in Fig. 5A and B. Multivariate ROC curves indicated that our risk scores had higher AUC values relative to other clinicopathological characteristics (Fig. 5C). The nomogram can be conveniently and effectively used to predict the prognosis of patients in clinical practice21. Independent prognostic factors including risk score, age, grade, AJCC stage were selected to construct nomogram predicting the probability of survival at 1, 3, and 5 years (Fig. 5D). Calibration plots were utilized to assess the predictive performance of the nomogram, and showed that the predicted 1-, 3-, and 5-year survival rates were highly consistent with the actual survival rates (Fig. 5E). The pRRophetic R package was applied to explore the drug response in different risk groups. The half maximal inhibitory concentration (IC50) of sunitinib, tesimox, and rapamycin were statistically different between the two groups, while Axitinib, Gemcitabine, Pazopanib, and Sorafenib showed no obvious difference between the two groups (Fig. 5F and H and Supplementary Fig. S5A-5D). These results may help clinicians to provide certain reference value for individualized clinical decision-making of ccRCC patients.

Fig. 5.

Fig. 5

Open in a new tab

The risk model is an independent prognostic factor. (A, B) Univariate (A) and multivariate (B) Cox regression analyses were used to verify the independent prognostic value of the signature. (C) ROC curve analysis showed the prognostic accuracy of risk signature and other clinical features. (D) Nomogram was plotted for the prediction of overall survival time. (E) Calibration curves were drawn to determine the accuracy of nomogram for OS at 1-, 3-, and 5-years. (F-H) Comparison of drug sensitivity between high-risk and low risk groups. Sunitinib (F); Temsirolimus (G); and Rapamycin (H).

Identification of the tumor immune microenvironment correlation with the prognostic signature

Growing evidence suggests that the tumor microenvironment has a critical role in tumor occurrence, development and prognosis22,23. Except for tumor cells, the other two main components in the tumor microenvironment are immune cells and stromal cells. Therefore, the estimate R package was applied to assess the immune score, ESTIMATE score and stromal score for ccRCC samples24. Immune scores were obviously different in the low-risk group and the high-risk group (Fig. 6A and C). The ssGSEA was applied for calculating the enrichment scores of immune cells, Immature dendritic cell, Eosinophil, Mast cell, and Neutrophil had higher enrichment scores in the low-risk group (Fig. 6D). CIBERSORT was employed for calculating the infiltration rate of 22 immune cells in ccRCC samples, and the infiltration of each ccRCC samples was shown by bar plot (see Supplementary Fig. S6 online). Then, we compared the differences of 22 immune cells between groups based on the risk score. The results suggested that Eosinophils, B cells naive, Dendritic cells resting, Macrophages M2, Mast cells resting, Dendritic cells activated T cells, CD4 memory resting, Monocytes and NK cells resting showed high infiltrate in the low-risk group; Macrophages M0, T cells CD4 memory activated, T cells follicular helper, Plasma cells, T cells regulatory (Tregs) and T cells CD8 showed high infiltrate in the high-risk group (Fig. 6E). In recent years, immune checkpoint inhibitor (ICI) therapy has been acknowledged as a potent treatment for various tumors such as melanoma, non-small cell lung cancer, and kidney cancer25. Therefore, we explored the differential expression of the 47 previously reported immune checkpoint genes in different risk groups (see Supplementary Table S1 online). The findings revealed that 34 immune checkpoint genes were differently expressed in different groups. With the exception of CD274, CD200, TNFSF18, TNFSF15, NRP1, and KIR3DL1, most immune checkpoint genes (28/34) were highly expressed in the high-risk score group (Fig. 6F).

Fig. 6.

Fig. 6

Open in a new tab

Characterization of the tumor immune microenvironment in high-risk and low-risk groups. (A-C) Comparison of immune score (A), stromal score (B) and ESTIMATE score (C) between the two groups. Cluster 1 represents the high-risk group and cluster 2 represents the low-risk group. (D) The infiltration abundance of immune cell subsets was evaluated by ssGSEA. (E) The differences of tumor infiltrating immune cells between the two groups were compared by CIBERPORT. (F) The difference of immune checkpoint genes expression between high-risk and low risk groups.

In addition, we further analyzed the correlation of risk score and 22 tumor infiltrating immune cells. The results suggested that B cells naive, NK cells resting, Dendritic cells activated, Eosinophils, Dendritic cells resting, Mast Cells resting, Macrophages M1, T cells CD4 memory resting, Monocytes and Macrophages M2 were negatively correlated with risk scores (Fig. 7A), while T cells follicular helper, T cells regulatory (Tregs), T cells CD4 memory activated and Macrophages M0 were positively correlated with risk scores (Fig. 7B). Above all, the risk signature based on HSP genes could effectively reflect the tumor immune microenvironment of ccRCC patients.

Fig. 7.

Fig. 7

Open in a new tab

Correlation analysis between risk score and tumor infiltrating immune cells. (A) Positively correlated with risk score. (B) Negative correlation with risk score.

Mutational landscape in high-risk group and low-risk group

We assessed differences in the distribution of somatic mutations between groups, with higher somatic mutation rates in the high-risk group (84.25%, 107/127) than in the low-risk group (77.56%, 159/205), and the 20 most frequently mutated genes were listed (Fig. 8A-B). We compared tumor mutational burden (TMB) between groups based on risk score and explored the relationship between TMB and risk score. Significantly higher TMB was observed in the high-risk group than in the low-risk group (Fig. 8C). Risk score was positively correlated with TMB (Fig. 8D). To explore the effect of TMB on prognosis in renal cancer patients, we used the survminer R package to determine the optimal cutoff value for TMB to group patients and observed that patients with low TMB had better prognosis than patients with high TMB (Fig. 8E). The collaborative impact of combining risk scores and TMB in prognostic stratification was assessed. It was found that TMB did not affect risk score-based prognostic prediction. Significant difference in OS was observed according to risk score in the high TMB subgroup and the low TMB subgroup (low TMB and low risk score vs. low TMB and high-risk score; high TMB and low risk score vs. high TMB and high-risk score; Fig. 8F). Also apparent from Fig. 8F the low TMB and low risk groups have the best OS, while the high TMB and high-risk groups have the worst OS. These results suggested that HSP gene-based risk score may be a potential predictor of survival independent of TMB.

Fig. 8.

Fig. 8

Open in a new tab

Association of risk signatures with tumor mutational burden (TMB). (A, B) The top 20 mutated genes in the high-risk group (A) and low-risk group (B). (C) The differences of TMB in high-risk and low-risk groups. (D) Risk score is positively correlated with TMB. (E) Survival analysis of patients in the high TMB group and low TMB group. (F) Survival analysis of patients in the screening cohort stratified by both TMB and risk score.

HSPA7 promotes the proliferation, invasion and metastasis of ccRCC cells

Our previous research indicated that HSP genes may impact the prognosis of ccRCC patients, with the most significant expression differences observed in HSPA7, HSPA8, and HSP90B1. While previous studies have demonstrated the role of HSPA8 in promoting colon cancer progression through Wnt/β-Catenin signaling pathways26, and the influence of HSP90B1 on malignant tumor prognosis27. However, there are few studies on the mechanism of HSPA7 affecting the progression of malignant tumors. In order to further explore the role of HSPA7 in the progression of ccRCC, we carried out cell experiments to preliminarily explore its mechanism. We detected the expression level of HSPA7 in several common ccRCC cell lines. Among the cell lines we detected, the expression of HSPA7 was the highest in A498 cells and the lowest in OSRC2 cells (Fig. 9A). Therefore, we chose these two cell lines for follow-up experiments. To investigate the function of HSPA7 in ccRCC cells, endogenous HSPA7 was stably knocked down by small interfering RNA (siRNA) in A498 cells (Fig. 9B) and overexpressed in OSRC2 cells (Fig. 9C). Consequently, these two cell lines were selected for follow-up experiments. CCK-8 assays were utilized to assess the effects of HSPA7 on the viability of human ccCRC cell lines. The findings demonstrated that knockdown of HSPA7 led to decreased proliferation in A498 cells (Fig. 9D), while overexpression of HSPA7 resulted in increased viability in OSRC2 cells (Fig. 9E). Endogenous knockdown of HSPA7 resulted in suppressed mesenchymal marker expression (Vimentin and N‐cadherin), upregulated epithelial marker expression (E‐cadherin), as well as decreased migratory and invasive ability in A498 cells (Fig. 9F-I and N-O). Conversely, overexpression of HSPA7 led to opposite effects on phenotype in OSRC2 cells (Fig. 9J-M and P-Q). Additionally, the impact of HSPA7 on the viability and cell cycle of ccRCC cells was also investigated. There was a significant reduced in the proportion of cells in S/G2 phase with knockdown HSPA7 in A498 cells (see Supplementary Fig. S7A online), whereas overexpression of HSPA7 in OSRC2 cells increased the proportion of cells in S/G2 phase (see Supplementary Fig. S7B online). The siRNA sequences of HSPA7 are shown in Supplementary Table S2.

Fig. 9.

Fig. 9

Open in a new tab

HSPA7 promotes the proliferation, invasion and metastasis of A498 and ORSC2 cells, A498 cells were transfected with HSPA7-specific siRNA(A498-hspa7-siRNA), a control siRNA (A498-siRNA-NC) or PBS(A498-WT); ORCS2 cells were transfected with a HSPA7-specific vector (ORSC2-hspa7-OE), a control vector(ORSC2- vector) or PBS(ORCS2-WT). (A) Expression of HSPA7 in ccRCC cells detected by Western blot. (B) HSPA7 expression in ORSC2 cells with different treatments. (C) HSPA7 expression in A498 were cells with different treatments. (D-E) ORSC2 and A498 cells viability was assessed using CCK8 assay; (F-G) Expression of E-cadherin, Vimentin and N‐cadherin in A498 cells were evaluated by Western blot; (H-I) Scratch healing assay of A498 were cells with different treatments. (J-K) Expression of E‐cadherin, Vimentin and N‐cadherin in ORCS2 cells were evaluated by Western blot. (L-M) Scratch healing assay of evaluated in the A498 cells. (N-O) Transwell migration and invasion were evaluated in the A498 cells (P-Q) Transwell migration and invasion were evaluated in the ORCS2 cells. Each experiment was conducted in triplicate. An asterisk (*) represents significant difference with P < 0.05; (**) represents P < 0.01; (***) represents P < 0.001. Error bars are indicative of means ± SD. n.s., not significant.

Discussion

RCC is a common cause of cancer-related deaths worldwide in the urinary system28. In addition to surgical treatment, targeted therapy and immunotherapy are the main treatment modalities for RCC. HSPs are one of the largest families of molecular chaperones, and various types of physiological and/or pathological stress regulate their expression29,30. Heat shock proteins (HSPs) play a multifaceted and pivotal role in the progression of cancer, contributing to its development through various mechanisms. They facilitate tumor cell survival and proliferation by stabilizing carcinogenic proteins, inhibiting apoptosis signaling pathways, and maintaining protein folding and degradation balance. For instance, HSP90 extends the lifespan of cancer cells by stabilizing telomerase activity and promotes invasion and metastasis through activation of signaling pathways such as HIF-1α and NF-κB16. Heat shock proteins play a significant role in shaping the tumor microenvironment, as they facilitate tumor growth and metastasis by modulating immune and inflammatory responses. Furthermore, HSP60 contributes to the provision of essential nutrients and oxygen to tumors through its promotion of angiogenesis and regulation of glucose metabolism31. Additionally, involved in signal transduction and metabolic reprogramming, HSPs regulate cancer cell growth, invasion, and metastasis by interacting with downstream signaling molecules such as EGFR, STAT3, Wnt-β-catenin32. They also influence cellular metabolic pathways like glycolysis and oxidative phosphorylation to support the energy requirements of tumor cells. Heat shock proteins could also be potential therapeutic targets. For example, HSP90 inhibitors have been used to enhance the sensitivity of chemotherapy drugs and have shown therapeutic potential against multiple cancer types in clinical trials33. Overexpression or abnormal function of HSPs may also lead to the occurrence and development of tumors, so in-depth study of their regulatory mechanisms is of great significance for cancer treatment.

Systematic understanding of the HSP in ccRCC will help us understand its pathogenesis, guide clinical treatment and judge the prognosis of ccRCC patients. We downloaded the expression, survival, mutation and CNVs data of ccRCC from the public database. These data were used to explore the expression patterns, prognostic value and impact on tumor microenvironment of HSP110, HSP90, HSP70 and HSP60 families in ccRCC. Differential expression of HSP genes was observed in ccRCC and normal tissue. Three clusters generated by the consensus clustering analysis showed significant differences in OS and clinicopathological features, indicated that the expression of HSP is associated with the prognosis of ccRCC patients.

Then, a risk model of HSP prognostic signature was constructed in the training cohort by LASSO regression analysis. The risk scoring system containing 6 genes (HSPA8, HSP90B1, HSPA7, HSPA12B, HSPA4L, HSPA1L) was constructed, the ccRCC patients could effectively stratified into low-risk and high-risk groups. The performance of HSP prognostic signature was confirmed in the internal cohort. The OS rate and OS time of ccRCC patients in the high-risk group were significantly shorter than those in the low-risk group. Furthermore, the risk score increased with increasing malignancy of ccRCC. ROC curve and AUC could reflect the good accuracy of this signature in predicting the survival of patients with ccRCC. Univariate and multivariate analysis suggested that the signature was an independent risk factor affecting the prognosis of ccRCC. Previous studies showed that HSPA8 affected the function of tumor-associated genes, it could inhibit the degradation of mutant forms of p53 and p73, and binds to the non-phosphorylated tumor-suppressor retinoblastoma (Rb) protein to reduce its degradation6. Long non-coding RNA DLX6-AS1 could affect the expression of HSP90B1 through miR-223 to promote cell growth and invasiveness of bladder cancer34. N6-Methyladenosine-Modified HSPA7 is correlated with the tumor microenvironment of glioblastoma and could predict the response to immune checkpoint therapy, which is considered as a prognostic risk factor for glioblastoma patients35. In non-small cell lung cancer, HSPA12B overexpression is correlated with cisplatin resistance36. The tumor suppressor miR-497 induces apoptosis, inhibits proliferation and migration in nasopharyngeal carcinoma cells by targeting ANLN and HSPA4L37. HSPA1L is a potential activator of IGF1Rβ and can regulate the transcription of β-catenin to enhance cancer stem cell-like properties, so it may be a potential cancer therapy target38. These studies suggested that HSPs may play different roles in different types of cancer. How these genes interact in the development of ccRCC remains to be further investigated. Meanwhile, we also verified that HSPA7 promoted the proliferation, invasion and metastasis of ccRCC cells in vitro.

We found differences in the tumor microenvironment of renal cancer patients in different risk groups, which may be one of the reasons affecting the prognosis of the two groups. Immune cells are one of the main components of tumor microenvironment. In the low-risk group, higher infiltrated levels of Macrophages M2, B cells naive, Eosinophils, Dendritic cells activated, Mast cells resting, Dendritic cells resting, Monocytes, T cells CD4 memory resting and NK cells resting were observed. In the high-risk group, the infiltration levels of plasma cells, macrophage M0, T cell CD8, T cell regulatory factors (Tregs), T cell CD4 memory activation, and T cell follicular helper were higher. The results of this study’s correlation analysis showed that risk scores were positively associated with four types of immune cells, especially in Tregs cells. Tregs could suppress T-cell and B-cell responses, secrete a variety of immunosuppressive cytokines, lead to immune escape of tumor cells, and are associated with aggressive disease of tumors39,40. High level of T cell CD8 infiltration is a poor prognostic factor in ccRCC41. The poor prognosis in the high-risk group is consistent with previous studies. ICIs are the focus of cancer treatment and have become one of the main treatment modalities for ccRCC. By exploring the expression of 47 immune checkpoint genes in the two groups of ccRCC, it was found that 34 immune checkpoint genes were differentially expressed between the low-risk group and the high-risk group, and most of the immune checkpoint genes (28/34) were highly expressed in the high-risk group. Based on these results, we consider that the poor prognosis of ccRCC patients with high-risk score may be associated with the tumor immune microenvironment, and ccRCC patients with high risk score may benefit more from immunotherapy. In clinical treatment, sunitinib has been widely used as a first-line treatment for advanced ccRCC42. Nevertheless, not all patients respond well to sunitinib therapy, so selecting patients who are sensitive to sunitinib treatment could be beneficial to achieve better clinical therapeutic outcome. Our study found that low-risk patients were more sensitive to sunitinib. Therefore, HSPs could be served as a pharmacodynamic biomarker in ccRCC patients, and understanding the expression of HSPs before treatment might be useful for designing better individualized treatment regimens.

Conclusion

We systematically analyzed the prognostic value of HSP110, HSP90, HSP70, HSP60 families in ccRCC, constructed a prognostic model including 6 genes, and analyzed the differences in tumor microenvironment of different groups in ccRCC patients. Patients in the high-risk and low-risk groups had different prognosis and immune cell infiltration. Patients in the high-risk group might benefit more from immunotherapy. HSPs might be involved in the targeted therapy of ccRCC patients. Our findings might provide new insights into the role of HSPs in the tumor microenvironment of ccRCC patients. Of course, our study has limitations and further multicenter studies and experimental investigations are needed. However, our study proposes new prognostic genetic markers with potential clinical application value, which may improve the prognosis of ccRCC patients and even develop new therapeutic strategies.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1. (2.2MB, pdf)

Acknowledgements

We would like to acknowledge the TCGA database for the data provided.

Abbreviations

HSPs

Heat shock proteins

ccRCC

Clear cell renal cell carcinoma

RCC

Renal cell carcinoma

LASSO

Least absolute shrinkage and selection operator

TCGA

The Cancer Genome Atlas

CNV

Copy number variation

ROC

Receiver operating characteristic curve

AUC

Area under curve

OS

Overall survival

GDSC

Genomics of Drug Sensitivity in Cancer

IC50

Inhibitory concentration

TMB

Tumor mutational burden

Author contributions

This study was conducted collaboratively by all authors. Mao Huang and Wenjing Liao performed the data collection and analysis. Qin Tang, Junwu Li and Xiaoyi DU analyzed the results and performed the experiments. Mao Huang, Liangdan Tang and Qin Tang drafted and reviewed the manuscript. All authors approved the final manuscript.

Funding

This research received no external funding.

Data availability

The data is available at the TCGA database (https://portal.gdc.cancer.gov/.)

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Wenjing Liao and Mao Huang contributed equally to this work.

References

  • 1.Rossi, S. H. et al. Epidemiology and screening for renal cancer. World J. Urol.36 (9), 1341–1353. 10.1007/s00345-018-2286-7 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Hsieh, J. J. et al. Genomic classifications of renal cell carcinoma: A critical step towards the future application of personalized kidney cancer care with pan-omics precision. J. Pathol.244 (5), 525–537. 10.1002/path.5022 (2018). [DOI] [PubMed] [Google Scholar]
  • 3.Wallis, C. J. D. et al. First-line systemic therapy for metastatic renal cell carcinoma: A systematic review and network meta-analysis. Eur. Urol.74 (3), 309–321. 10.1016/j.eururo.2018.03.036 (2018). [DOI] [PubMed] [Google Scholar]
  • 4.Rini, B. I., Campbell, S. C. & Escudier, B. Ren. cell. Carcinoma Lancet, 373(9669): 1119–1132. DOI: 10.1016/s0140-6736(09)60229-4. (2009). [DOI] [PubMed] [Google Scholar]
  • 5.Signoretti, S. et al. Renal cell carcinoma in the era of Precision Medicine: from Molecular Pathology to tissue-based biomarkers. J. Clin. Oncol.36 (36). 10.1200/jco.2018.79.2259 (2018). Jco2018792259. [DOI] [PMC free article] [PubMed]
  • 6.Liu, T., Daniels, C. K. & Cao, S. Comprehensive review on the HSC70 functions, interactions with related molecules and involvement in clinical diseases and therapeutic potential. Pharmacol. Ther.136 (3), 354–374. 10.1016/j.pharmthera.2012.08.014 (2012). [DOI] [PubMed] [Google Scholar]
  • 7.Sural, S. et al. HSB-1/HSF-1 pathway modulates histone H4 in mitochondria to control mtDNA transcription and longevity. Sci. Adv.6 (43). 10.1126/sciadv.aaz4452 (2020). [DOI] [PMC free article] [PubMed]
  • 8.Macario, A. J. & Conway Macario de, E. Molecular chaperones: Multiple functions, pathologies, and potential applications. Front. Biosci. 12, 2588–2600 (2007). 10.2741/2257 [DOI] [PubMed] [Google Scholar]
  • 9.Kampinga, H. H. et al. Guidelines for the nomenclature of the human heat shock proteins. Cell. Stress Chaperones. 14 (1), 105–111. 10.1007/s12192-008-0068-7 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Wu, J. et al. Heat shock proteins and Cancer. Trends Pharmacol. Sci.38 (3), 226–256. 10.1016/j.tips.2016.11.009 (2017). [DOI] [PubMed] [Google Scholar]
  • 11.Jee, H. Size dependent classification of heat shock proteins: A mini-review. J. Exerc. Rehabil. 12 (4), 255–259. 10.12965/jer.1632642.321 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Calderwood, S. K. Heat shock proteins in breast cancer progression–a suitable case for treatment? Int. J. Hyperth.26 (7), 681–685. 10.3109/02656736.2010.490254 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Ciocca, D. R. et al. Heat shock proteins in prostate cancer: From tumorigenesis to the clinic. Int. J. Hyperth.26 (8), 737–747. 10.3109/02656731003776968 (2010). [DOI] [PubMed] [Google Scholar]
  • 14.Zhao, M. et al. Increased expression of heat shock protein 27 correlates with peritoneal metastasis in epithelial ovarian cancer. Reprod. Sci.19 (7), 748–753. 10.1177/1933719111432875 (2012). [DOI] [PubMed] [Google Scholar]
  • 15.Calderwood, S. K. et al. Heat shock proteins in cancer: Chaperones of tumorigenesis. Trends Biochem. Sci. 31 (3), 164–172 (2006). 10.1016/j.tibs.2006.01.006 [DOI] [PubMed] [Google Scholar]
  • 16.Yun, C. W. et al. Heat shock proteins: Agents of cancer development and therapeutic targets in anti-cancer therapy. Cells9 (1). 10.3390/cells9010060 (2019). [DOI] [PMC free article] [PubMed]
  • 17.Santos, T. G., Martins, V. R. & Hajj, G. N. M. Unconventional secretion of heat shock proteins in Cancer. Int. J. Mol. Sci.18 (5). 10.3390/ijms18050946 (2017). [DOI] [PMC free article] [PubMed]
  • 18.Tsan, M. F. & Gao, B. Cytokine function of heat shock proteins. Am. J. Physiol. Cell. Physiol.286 (4), C739–C744. 10.1152/ajpcell.00364.2003 (2004). [DOI] [PubMed] [Google Scholar]
  • 19.He, L., Deng, T. & Luo, H. S. Heat shock protein 70 gene polymorphisms and cancer risk: A meta-analysis. Sci. World J. 2014, 540309. (2014). 10.1155/2014/540309 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Zou, T. et al. The Association between Heat-shock protein polymorphisms and prognosis in lung cancer patients treated with platinum-based chemotherapy. Front. Pharmacol.11, 1029. 10.3389/fphar.2020.01029 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Balachandran, V. P. et al. Nomograms in oncology: More than meets the eye. Lancet Oncol.16 (4), e173–e180. 10.1016/s1470-2045(14)71116-7 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Spill, F. et al. Impact of the physical microenvironment on tumor progression and metastasis. Curr. Opin. Biotechnol.40, 41–48. 10.1016/j.copbio.2016.02.007 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Bremnes, R. M. et al. The role of tumor-infiltrating immune cells and chronic inflammation at the tumor site on cancer development, progression, and prognosis: Emphasis on non-small cell lung cancer. J. Thorac. Oncol.6 (4), 824–833. 10.1097/JTO.0b013e3182037b76 (2011). [DOI] [PubMed] [Google Scholar]
  • 24.Yoshihara, K. et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat. Commun.4, 2612. 10.1038/ncomms3612 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Sanmamed, M. F. & Chen, L. A paradigm shift in cancer immunotherapy: From enhancement to normalization. Cell175 (2), 313–326. 10.1016/j.cell.2018.09.035 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Li, B. et al. HSPA8 activates Wnt/β-Catenin signaling to facilitate BRAF V600E colorectal cancer progression by CMA-Mediated CAV1 degradation. Adv. Sci. (Weinh). 11 (3), e2306535. 10.1002/advs.202306535 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Huang, X. et al. Identification of HSP90B1 in pan-cancer hallmarks to aid development of a potential therapeutic target. Mol. Cancer. 23 (1), 19. 10.1186/s12943-023-01920-w (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Bray, F. et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin.68 (6), 394–424. 10.3322/caac.21492 (2018). [DOI] [PubMed] [Google Scholar]
  • 29.Ciocca, D. R. & Calderwood, S. K. Heat shock proteins in cancer: Diagnostic, prognostic, predictive, and treatment implications. Cell. Stress Chaperones. 10 (2), 86–103. 10.1379/csc-99r.1 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Young, J. C. et al. Pathways of chaperone-mediated protein folding in the cytosol. Nat. Rev. Mol. Cell. Biol.5 (10), 781–791. 10.1038/nrm1492 (2004). [DOI] [PubMed] [Google Scholar]
  • 31.Hoter, A. et al. Heat shock protein 60 in hepatocellular carcinoma: Insights and perspectives. Front. Mol. Biosci.7, 60. 10.3389/fmolb.2020.00060 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Parma, B. et al. Harnessing mitochondrial metabolism and drug resistance in non-small cell lung cancer and beyond by blocking heat-shock proteins. Drug Resist. Updat. 65, 100888. 10.1016/j.drup.2022.100888 (2022). [DOI] [PubMed] [Google Scholar]
  • 33.Yan, L. R. et al. Comprehensive pan-cancer analysis of Heat shock protein 110, 90, 70, and 60 families. Front. Mol. Biosci.8, 726244. 10.3389/fmolb.2021.726244 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Fang, C. et al. Long noncoding RNA DLX6-AS1 promotes cell growth and invasiveness in bladder cancer via modulating the mir-223-HSP90B1 axis. Cell. Cycle. 18 (23), 3288–3299. 10.1080/15384101.2019.1673633 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zhao, R. et al. The N(6)-Methyladenosine-modified pseudogene HSPA7 correlates with the tumor microenvironment and predicts the response to immune checkpoint therapy in glioblastoma. Front. Immunol.12, 653711. 10.3389/fimmu.2021.653711 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Chen, W. et al. HSPA12B overexpression induces cisplatin resistance in non-small-cell lung cancer by regulating the PI3K/Akt/NF-κB signaling pathway. Oncol. Lett.15 (3), 3883–3889. 10.3892/ol.2018.7800 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Wang, S. et al. The potent tumor suppressor miR-497 inhibits cancer phenotypes in nasopharyngeal carcinoma by targeting ANLN and HSPA4L. Oncotarget6 (34), 35893–35907. 10.18632/oncotarget.5651 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Choi, S. I. et al. HSPA1L enhances Cancer Stem Cell-Like Properties by activating IGF1Rβ and regulating β-Catenin transcription. Int. J. Mol. Sci.21 (18). 10.3390/ijms21186957 (2020). [DOI] [PMC free article] [PubMed]
  • 39.Nakamura, K., Kitani, A. & Strober, W. Cell contact-dependent immunosuppression by CD4(+)CD25(+) regulatory T cells is mediated by cell surface-bound transforming growth factor beta. J. Exp. Med.194 (5), 629–644. 10.1084/jem.194.5.629 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Bignone, P. A. & Banham, A. H. FOXP3 + regulatory T cells as biomarkers in human malignancies. Expert Opin. Biol. Ther.8 (12), 1897–1920. (2008). 10.1517/14712590802494022 [DOI] [PubMed] [Google Scholar]
  • 41.Nakano, O. et al. Proliferative activity of intratumoral CD8(+) T-lymphocytes as a prognostic factor in human renal cell carcinoma: clinicopathologic demonstration of antitumor immunity. Cancer Res.61 (13), 5132–5136 (2001). [PubMed] [Google Scholar]
  • 42.Serzan, M. T. & Atkins, M. B. Current and emerging therapies for first line treatment of metastatic clear cell renal cell carcinoma. J. Cancer Metastasis Treat.710.20517/2394-4722.2021.76 (2021). [DOI] [PMC free article] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1. (2.2MB, pdf)

Data Availability Statement

The data is available at the TCGA database (https://portal.gdc.cancer.gov/.)

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES