Colon Cancer-Upregulated Long Non-Coding RNA lincDUSP Regulates Cell Cycle Genes and Potentiates Resistance to Apoptosis

Abstract

Long non-coding RNAs (lncRNAs) are frequently dysregulated in many human cancers. We sought to identify candidate oncogenic lncRNAs in human colon tumors by utilizing RNA sequencing data from 22 colon tumors and 22 adjacent normal colon samples from The Cancer Genome Atlas (TCGA). The analysis led to the identification of ~200 differentially expressed lncRNAs. Validation in an independent cohort of normal colon and patient-derived colon cancer cell lines identified a novel lncRNA, lincDUSP, as a potential candidate oncogene. Knockdown of lincDUSP in patient-derived colon tumor cell lines resulted in significantly decreased cell proliferation and clonogenic potential, and increased susceptibility to apoptosis. The knockdown of lincDUSP affects the expression of ~800 genes, and NCI pathway analysis showed enrichment of DNA damage response and cell cycle control pathways. Further, identification of lincDUSP chromatin occupancy sites by ChIRP-Seq demonstrated association with genes involved in the replication-associated DNA damage response and cell cycle control. Consistent with these findings, lincDUSP knockdown in colon tumor cell lines increased both the accumulation of cells in early S-phase and γH2AX foci formation, indicating increased DNA damage response induction. Taken together, these results demonstrate a key role of lincDUSP in the regulation of important pathways in colon cancer.

Figure 1

lincDUSP is a candidate oncogenic lncRNA that is overexpressed in a subset of colon tumors. (A) Heatmap of differentially expressed lncRNAs (>2-fold change, p < 0.05) identified from 22 colon tumors and 22 matched normal colon tissue from TCGA RNA-Seq data. (B) Cluster graph of lincDUSP FPKM values from tumors vs. normal colon from TCGA RNA-Seq data. (C) Taqman qRT-PCR for lincDUSP in an independent cohort of patient-derived normal colon and colon tumor samples. Values were normalized to the normal colon sample ID number 4040169, and HPRT1 was used an endogenous control.

Figure 2

Knockdown of lincDUSP reduces proliferation and colony formation in vitro. (A,B) Verification of lincDUSP knockdown in V703 and V481 cells by LNA GapmeRs at 24, 48, and 72 hours. Error bars = SD (n = 2). (C,D) Proliferation assay in V703 and V481 cells treated with control GapmeRs vs. lincDUSP GapmeRs. The knockdown of lincDUSP significantly affects cell proliferations as compared to control cells. (E,F) Colony formation assay in V703 and V481 cells treated with control GapmeRs vs. lincDUSP GapmeRs. The depletion of lincDUSP has significant effects on the clonogenic capacity of colon cancer cells. Representative images shown below graph. Asterisks denote significant difference vs. negative control GapmeR by two-tailed t-test. **<0.05; ***<0.01.

Figure 3

Knockdown of lincDUSP potentiates apoptosis induction in vitro. (A,B) Caspase 3/7 GLO Assay was performed in V703 (Panel A) and V481 (Panel B) cells treated with 1.0 uM doxorubicin (Dox) for indicated timepoints 24 hours post-transfection with indicated GapmeRs. The knock down of lincDUSP was associated with increased Caspase 3/7 activity in both cell lines. P-value generated using student’s t-test; error bars = SE (n = 3). (C) Flow cytometry analysis of surface Annexin V in V481 cells transfected with control or lincDUSP-specific GapmeRs. At 48 hours post-transfection, cells were treated with 2.0 µM doxorubicin for 4 hours prior to staining. (D) Quantitation of Annexin V flow cytometry data. P-value generated using student’s t-test; error bars = SE (n = 4 from two independent experiments).

Figure 4

Knockdown of lincDUSP results in gene expression changes in cell cycle control pathways. (A) Verification of lincDUSP knock down by qRT-PCR prior to utilizing the RNA for RNA-seq. (B) Volcano plot of gene expression changes upon lincDUSP KD in V703 cells assayed by RNA-Seq. Genes with significantly decreased expression are shown in blue; genes with significantly increased expression are shown in red; LINC01605 (lincDUSP) shown in yellow. (C) Summary of significantly enriched pathways in V703 cells post lincDUSP KD (FDR < 0.05). The complete data is presented in Supporting Data File 3.

Figure 5

lincDUSP associates with chromatin within regulatory regions of differentially expressed genes. (A) The genome-wide occupancy of lincDUSP was identified by ChIRP-seq. The distribution of distances of lincDUSP ChIRP-Seq peaks to transcription start sites of protein-coding genes is shown. (B) Overlap between genes that are within 300 kb of a lincDUSP ChIRP-seq peak and differentially expressed genes identified in V703 cells post lincDUSP KD (RNA-Seq). (C) The distribution of distances of lincDUSP ChIRP-Seq peaks to transcription start sites of protein-coding genes that are also differentially expressed upon KD of lincDUSP.

Figure 6

lincDUSP knockdown affects cell cycle distribution and DNA damage response induction. (A) Cell cycle analysis of DAPI-stained V703 cells 24 hours post-transfection with indicated GapmeRs. Left panel: Negative control GapmeRs, Right panel: lincDUSP-specific GapmeRs. Black arrows indicate S-phase peak. Cell cycle models were fit using ModFit LT v4.0. (B) Representative images of Alexa 488-labeled γH2AX in V481 cells. Cells were transfected with indicated GapmeRs; time 0 represents cells 48 hours post-transfection. Cells were treated with 1 µM Doxorubicin (Dox) where indicated.