Emerging role of long non-coding RNA SOX2OT in SOX2 regulation in breast cancer

Abstract

The transcription factor SOX2 is essential for maintaining pluripotency in a variety of stem cells. It has important functions during embryonic development, is involved in cancer stem cell maintenance, and is often deregulated in cancer. The mechanism of SOX2 regulation has yet to be clarified, but the SOX2 gene lies in an intron of a long multi-exon non-coding RNA called SOX2 overlapping transcript (SOX2OT). Here, we show that the expression of SOX2 and SOX2OT is concordant in breast cancer, differentially expressed in estrogen receptor positive and negative breast cancer samples and that both are up-regulated in suspension culture conditions that favor growth of stem cell phenotypes. Importantly, ectopic expression of SOX2OT led to an almost 20-fold increase in SOX2 expression, together with a reduced proliferation and increased breast cancer cell anchorage-independent growth. We propose that SOX2OT plays a key role in the induction and/or maintenance of SOX2 expression in breast cancer.

Figure 1. Schematic genomic organization of SOX2 and SOX2OT and their expression patterns in breast cancer.

A) The genomic context of SOX2 and its upstream region. The gray boxes show the proposed promoter and enhancer regions , . B) The genomic context of SOX2 and SOX2OT, derived from the UCSC browser. SOX2 is located in an intron of SOX2OT. The triangles above each gene show the location of primers used in qRT-PCR. The SOX2 region is enlarged, and the direction of gene transcription shown with arrows. The phylogenetic conservation of each region is shown below the gene diagram (Mammal Cons). Vertical rectangular shading shows the conserved exonic regions of SOX2OT. The locations of these genes are adopted from UCSC genome browser March 2006 (NCBI36/hg18). C) Heat map showing the expression of SOX2 and SOX2OT in breast cancer samples analyzed by TCGA. It covers the expression of SOX2OT; chromosome 3: 182810845–182941699. SOX2 region is shown by an arrow. The samples are classified based on the estrogen receptor status. D and E) Expression of SOX2 and SOX2OT respectively in ER+ (n = 595) and ER− (n = 176) breast cancer samples (data derived from TCGA). Mann-Whitney rank sum test showed that these genes were expressed differently according to estrogen status: p values of <0.05 and <0.005 were calculated for D and E respectively.

Figure 2. Expression of SOX2 and SOX2OT in different breast cancer cell lines.

A) The expression of SOX2 and SOX2OT relative to HPRT and GAPDH in 18 breast cell lines measured by qRT-PCR. Scatter plot showing the expression of SOX2 and SOX2OT. Black and white circles represent ER− and ER+ cell lines. B) Box plot indicating expression of SOX2 and SOX2OT in ER+ and ER− breast cancer cell lines. * represents p value <0.05. C) Expression of SOX2 and SOX2OT in five MCF-7 sub-lines relative to the MCF-7 parental line was measured by qRT-PCR. Error bars represent standard deviations of three technical replicates.

Figure 3. Expression on SOX2 and SOX2OT in breast cancer cell lines grown in suspension culture.

A and B) Expression of SOX2 and SOX2OT in MDA-MB-231 (A) and MCF-7 (B) breast cancer cell lines cultured in suspension relative to the cells grown as monolayer was measured by qRT-PCR. Three consecutive passages of cells were grown in suspension as S-p1 to S-p3 respectively. M represents the expression in monolayer culture. Error bars represent the standard error of the mean of three biological replicates. C) Western blot analysis showing expression of SOX2 in MCF-7 and MDA-MB-231 cells grown as monolayers and in suspension. D) Expression of SOX2 and SOX2OT in TamC3 and TamR3 derivatives of MCF-7 grown in suspension relative to MCF-7 monolayer culture measured by qRT-PCR. Error bars represent standard deviations of three technical replicates. E) Expression of SOX2 and SOX2OT in monolayer MDA-MB-231-luc-D3H2LN cells (in vitro) and the same cells grown as orthotopic xenografts (in vivo) were measure by qRT-PCR. The error bars are standard error of the mean of six individual tumors. *, **, and *** represent p values of <0.05, 0.01 and 0.001 respectively.

Figure 4. Ectopic expression of SOX2OT in MDA-MB-231 cells.

A) Relative expression of SOX2 and SOX2OT measured by qRT-PCR in MDA-MB-231 cells transfected with control vector and plasmid containing a SOX2OT gene (NR_004053.3). The bar graphs show the fold change relative to control vector. ** and *** represent p values of <0.01 and 0.001 respectively. B) Western blot analysis showing the over-expression of SOX2 in cells transfected with vector containing the SOX2OT gene.

Figure 5. Effect of ectopic expression of SOX2OT on cell growth.

A) The proliferation rates of cells transfected with control vector and plasmid containing the SOX2OT gene were measured after 1, 3, 5 and 7 days of culture. B) Proliferation rates of cells transfected with control vector or plasmid containing the SOX2OT gene, in monolayer and suspension culture, as well as in soft agar. * and ** represent p values of <0.05 and 0.01 respectively. C) Image of MDA-MB-231 cells grown in soft agar for two weeks. The top panel represents the colony form in cells with ectopic expression of SOX2OT. The bottom panel represents cells containing the control vector. D) Effect of ectopic expression of SOX2OT on cell cycle distribution. The cells were treated with and without paclitaxel for 24 h and stained with PI. Following FACS analysis, DNA histograms were analyzed using ModFit LT. The data are representative examples for duplicate tests.