Analysis of long non-coding RNA expression profiles in clear cell renal cell carcinoma

Abstract

To investigate the expression patterns of long non-coding RNAs (lncRNAs) in clear cell renal cell carcinoma (ccRCC) and in metastatic renal cell carcinoma (RCC), the present study downloaded three human exon arrays available from the public Gene Expression Omnibus. The probes of the human exon arrays were re-annotated and the probes uniquely mapping to lncRNAs were retained at the gene level. Following the analysis of GSE53757 and GSE46699, which contained paired ccRCC cancer and normal adjacent tissue samples, 32 differentially expressed lncRNAs (adjusted P<0.01) in ccRCC were identified. Various lncRNAs, including ENSG00000177133, NR_024418, T-cell leukemia/lymphoma 6 (TCL6), growth arrest-specific transcript 5, deleted in lymphocytic leukemia 2, colorectal neoplasia differentially expressed (CRNDE) and MIR155HG, have been reported to be abnormally expressed in cancers. Of these genes, NR_24418 and TCL6 have been reported to be associated with ccRCC. Following analysis of GSE47352, which contained 4 primary metastatic and 5 non-metastatic tumor samples, the 50 top differentially expressed lncRNAs were identified in metastatic ccRCC (Mann-Whitney U test, P<0.05). Comparison with the ccRCC associated lncRNAs revealed that the lncRNA CRNDE demonstrated an increased expression in ccRCC and metastatic ccRCC samples, which suggested that CRNDE is important in the progression of ccRCC. The lncRNA ENSG00000244020 was decreased in ccRCC and metastatic ccRCC, suggesting that silencing of ENSG00000244020 may be important in ccRCC development. Overall, a set of lncRNAs was identified as differentially expressed in ccRCC and metastatic ccRCC, providing potential candidates for the discovery of novel cancer biomarkers and therapeutic targets to improve diagnosis and therapy in RCC.

Keywords: clear cell renal cell carcinoma; data mining; long non-coding RNA; microarray analysis.

Figure 1.

Clustering heatmap of 72 paired normal samples (green) and tumors (blue) based on the 85 differentially expressed lncRNAs in GSE53757. Each column represents one sample and each row represents one lncRNA. Gene expression levels are indicated as follows: Red, high expression; blue, low expression. LncRNAs, long non-coding RNAs; TCL6, T-cell leukemia/lymphoma 6; CRNDE, colorectal neoplasia differentially expressed.

Figure 2.

Clustering heatmap of 65 paired normal samples (green) and tumors (blue) based on the 32 differentially expressed lncRNAs in GSE46699. Each column represents one sample and each row represents one lncRNA. Gene expression levels are indicated as follows: red, high expression; blue, low expression. LncRNAs, long non-coding RNAs; TCL6, T-cell leukemia/lymphoma 6; CRNDE, colorectal neoplasia differentially expressed.

Figure 3.

Distribution of expression differentials between experimental data set GSE53757 and validation data set GSE46699. FC, fold change.

Figure 4.

Clustering heatmap of 4 primary metastatic (green) and 5 non-metastatic tumor (blue) samples based on the 50 differentially expressed lncRNAs in GSE47352. Gene expression levels are indicated as follows: Red, high expression; blue, low expression. LncRNAs, long non-coding RNAs; CRNDE, colorectal neoplasia differentially expressed.