Coordinated reprogramming of renal cancer transcriptome, metabolome and secretome associates with immune tumor infiltration

Abstract

Background: Clear cell renal cell carcinoma (ccRCC) is the most common subtype of renal cancer. The molecules (proteins, metabolites) secreted by tumors affect their extracellular milieu to support cancer progression. If secreted in amounts detectable in plasma, these molecules can also serve as useful, minimal invasive biomarkers. The knowledge of ccRCC tumor microenvironment is fragmentary. In particular, the links between ccRCC transcriptome and the composition of extracellular milieu are weakly understood. In this study, we hypothesized that ccRCC transcriptome is reprogrammed to support alterations in tumor microenvironment. Therefore, we comprehensively analyzed ccRCC extracellular proteomes and metabolomes as well as transcriptomes of ccRCC cells to find molecules contributing to renal tumor microenvironment.

Methods: Proteomic and metabolomics analysis of conditioned media isolated from normal kidney cells as well as five ccRCC cell lines was performed using mass spectrometry, with the following ELISA validation. Transcriptomic analysis was done using microarray analysis and validated using real-time PCR. Independent transcriptomic and proteomic datasets of ccRCC tumors were used for the analysis of gene and protein expression as well as the level of the immune infiltration.

Results: Renal cancer secretome contained 85 proteins detectable in human plasma, consistently altered in all five tested ccRCC cell lines. The top upregulated extracellular proteins included SPARC, STC2, SERPINE1, TGFBI, while downregulated included transferrin and DPP7. The most affected extracellular metabolites were increased 4-hydroxy-proline, succinic acid, cysteine, lactic acid and downregulated glutamine. These changes were associated with altered expression of genes encoding the secreted proteins (SPARC, SERPINE1, STC2, DPP7), membrane transporters (SLC16A4, SLC6A20, ABCA12), and genes involved in protein trafficking and secretion (KIF20A, ANXA3, MIA2, PCSK5, SLC9A3R1, SYTL3, and WNTA7). Analogous expression changes were found in ccRCC tumors. The expression of SPARC predicted the infiltration of ccRCC tumors with endothelial cells. Analysis of the expression of the 85 secretome genes in > 12,000 tumors revealed that SPARC is a PanCancer indicator of cancer-associated fibroblasts' infiltration.

Conclusions: Transcriptomic reprogramming of ccRCC supports the changes in an extracellular milieu which are associated with immune infiltration. The proteins identified in our study represent valuable cancer biomarkers detectable in plasma.

Keywords: CAFs; Immune infiltration; Metabolome; Renal cancer; SPARC; Secretome.

Fig. 1

Proteomic analysis of RCC secretome. A Left: The scheme of the experiment. CM from all cell lines was subjected to proteomic and metabolomics analysis. Cells were collected and used for transcriptomic analysis. Right: nanoHPLC-MS/MS revealed 85 CM proteins that were consistently altered in all ccRCC cell lines the same direction (upregulated/downregulated) and verified as secretome components. Proteins selected for validation are shown in red font. Complete data are shown in Additional file 1: Table S4. The analysis was performed using protein extracts isolated from 3 independent biological experiments per each cell line. B Gene ontology analysis of the 85 consistently altered proteins. C Validation of top altered proteins using ELISA. The analysis was performed using conditioned media isolated from 3 independent biological experiments per each cell line. Statistical analysis: One-way ANOVA with Dunnett's Multiple Comparison Test. *p < 0.05, **p < 0.01, ***p < 0.001. D The expression of genes encoding altered CM proteins is disturbed in ccRCC cell lines. The table shows result of microarray analysis (complete data are shown in Additional file 1: Table S5). The plots show results of qPCR validation of the analyzed genes in RNA isolated from 3 independent biological experiments. Statistical analysis: One-way ANOVA with Dunnett's Multiple Comparison Test. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 2

A. The expression of genes encoding proteins of ccRCC secretome is altered in ccRCC tumors. A The plots show results of qPCR analysis performed in non-neoplastic kidney samples (N) and ccRCC tumor samples (T). All: analysis performed in all samples, without differentiation into TNM Stage/Fuhrman grade. N: n = 92; T: n = 92; TNM stage: tumor samples were classified into Stage 1 (N1: n = 49, T1: n = 49), Stage 2 (N2: n = 12, T2: n = 12), Stage 3 (N3: n = 27, T3: n = 27), Stage 4: N4: n = 4, T4: n = 4). Fuhrman grade: tumor samples were classified into Grade 1 (N1: n = 27, G1: n = 27), Grade 2 (N1: n = 51, G2: n = 51), Grade 3 (N3: n = 9, G3: n = 8), Grade 4 (N4: n = 1, G4: n = 1). Statistical analysis was performed using Wilcoxon matched-pairs signed rank test or paired t test, depending on data normality distribution. B The plots show results of UALCAN/CPTAC analysis on proteomic data from ccRCC tumors (n = 110) and normal kidney tissues (n = 84)

Fig. 3

The expression of genes involved in secretion and ECM regulation is altered in ccRCC. A GO analysis of 1497 genes commonly altered in Caki-1 and KIJ265T cells when compared with RPTEC. The genes are grouped by functional categories defined by high-level GO Cellular Component terms. The gene group categorization was analyzed using by ShinyGO 0.76 analysis (http://bioinformatics.sdstate.edu/go/) and the data was imported into GraphPad Prism to generate the plot. Complete data are shown in Additional file 1: Table S5. B The expression of genes regulating secretion and extracellular space. The plots show qPCR validation of microarray results in RNA isolated from 3 independent biological cell culture experiments. Statistical analysis: One-way ANOVA with Dunnett's Multiple Comparison Test. *p < 0.05, **p < 0.01, ***p < 0.001. C The expression of proteins regulating secretion and extracellular space in ccRCC tumors. The plots show results of UALCAN/CPTAC analysis. For SYTL3 only data for phosphopeptides were available (SYTL3 (1): NP_001229313.1:S185. SYTL3(2): NP_001229313.1:S203). N: normal kidney samples (n = 84), T: ccRCC tumors (n = 110). D The expression of SLC9A3R1 protein and its phosphorylated variants in ccRCC tumors. N: normal kidney samples (n = 84), T: ccRCC tumors (n = 110). The analysis was performed using UALCAN/CPTAC platform

Fig. 4

The expression of genes encoding key ccRCC secretome proteins correlates with immune infiltration in tumors. A The expressions of SPARC correlates with the presence of endothelial cells in ccRCC tumors. The plots show results of analysis performed with Timer platform (http://timer.comp-genomics.org/). B Top enriched terms for biological processes for genes positively correlating with SPARC in ccRCC tumors. Only genes with r ≥ 0.5 correlations were analyzed. The analysis was performed using ShinyGO 0.76 (http://bioinformatics.sdstate.edu/go/). C The expression of SPARC correlates with key angiogenic regulators in ccRCC tumors. The plots show results of analysis performed with UALCAN (http://ualcan.path.uab.edu/index.html) platform. D SPARC emerges as the top gene correlating with immune infiltration in PanCancer analysis. Volcano plot shows the results of correlation analysis between the expression of 85 genes encoding proteins of ccRCC secretome and the presence of immune cells infiltrating 40 tumor types. SPARC correlations (r ≥ 0.9) are shown with colorful dots. E Representative plots showing top SPARC-CAFs correlations in cancers of breast (BRCA-LumA, BRCA-Her2) and colon (COAD). For detailed data see Additional file 2: Table S10. EPIC, MCPCOUNTER, TIDE: different algorithms utilized by Timer for calculation of immune infiltration

Fig. 5

ccRCC CM metabolome changes. A Top altered CM metabolites analyzed by GC–MS (complete GC–MS data are shown in Additional file 2: Table S11). B Validation of GC–MS. C The expression of SLC16A4 lactate transporter in ccRCC cells. The plot shows results of qPCR validation of microarray data. The analyses were performed using conditioned media (A, B) or RNA (C) isolated from 3 independent biological experiments. Statistical analysis: One-way ANOVA with Dunnett's Multiple Comparison Test. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 6

Coordinated reprogramming of ccRCC transcriptome, metabolome and secretome. The expression of genes encoding SPARC, SERPINE1 and STC2 is upregulated while expression of DPP7 is downregulated in ccRCC cells, which reflects altered levels of the encoded protein in the extracellular milieu. The expression of genes encoding proteins involved in protein trafficking and secretion is reprogrammed to support changes in concentrations of extracellular proteins and metabolites: MIA2 (cTAGE5) a receptor of endoplasmic reticulum, localizing to the ER exit sites (ERES), a critical regulator of COPII assembly, protein trafficking and export [–36]. ANXA3 localizes to endocytic compartments in ccRCC cells [37] and contributes to the regulation of vesicles release [38]. KIF20A is a kinesin crucial for the fission of RAB6-positive vesicles and their exit from Golgi/TGN membranes [27]. It also promotes secretion of factors involved in proliferation of castration-resistant prostate cancer [39]. SYTL3 is a critical effector of RAB27B, enabling kinesin-microtubule-dependent movement of secretory granules towards plasma membrane [40]. ABCA12 is a transmembrane lipid transporter, required for the proper transcriptional programming of vesicle trafficking and cytoskeletal remodeling pathways, lipid raft composition, as well as formation and functioning of secretory granules in pancreatic cells [41]. PCSK5 is a proprotein convertase, which cleaves the target proproteins converting them into their active functional forms. Enhanced expression of SLC16A4 transporter ensures increased secretion of lactate. Changes in concentrations of extracellular proteins (STC2, SPARC, SERPINE1, TGFBI, DPP7, and TF) as well as metabolites affect the functioning of TME cells. TGN: trans-Golgi network; COPII: The Coat Protein Complex II vesicles; Pyr: pyruvate; αKG: α-keto glutarate; PM: plasma membrane; CAFs: cancer-associated fibroblasts; DCs: dendritic cells; ECs: endothelial cells. 4-OH-P: 4-hydroxyproline. Upregulation/downregulation is shown with red/blue font, respectively