LncRNA HOTTIP facilitates the stemness of breast cancer via regulation of miR-148a-3p/WNT1 pathway

Abstract

Emerging evidence suggests that dysregulation of long non-coding RNA (lncRNA) plays a key role in tumorigenesis. The lncRNA, HOXA transcript at the distal tip (HOTTIP), has been reported to be up-regulated in multiple cancers, including breast cancer, and is involved in various biological processes, including the maintenance of stemness. However, the biological function and underlying modulatory mechanism of HOTTIP in breast cancer stem cells (BCSCs) remains unknown. In this study, we found that HOTTIP was markedly up-regulated in BCSCs and had a positive correlation with breast cancer progression. Functional studies revealed that overexpression of HOTTIP markedly promoted cell clonogenicity, increased the expression of the stem cell markers, OCT4 and SOX2, and decreased the expression of the differentiation markers, CK14 and CK18, in breast cancer cells. Knockdown of HOTTIP inhibited the CSC-like properties of BCSCs. Consistently, depletion of HOTTIP suppressed tumour growth in a humanized model of breast cancer. Mechanistic studies demonstrated that HOTTIP directly binds to miR-148a-3p and inhibits the mediation of WNT1, which leads to inactivation of the Wnt/β-catenin signalling pathway. Our study is the first to report that HOTTIP regulates the CSC-like properties of BCSCs by as a molecular sponge for miR-148a-3p to increase WNT1 expression, offering a new target for breast cancer therapy.

Keywords: HOXA transcript at the distal tip (HOTTIP); WNT1; breast cancer; miR-148a-3p; stemness.

Figure 1

The high expression of HOTTIP in breast cancer and BCSCs. A, The relationship between HOTTIP expression and the outcomes of breast cancer patients was analysed using the online tool, KM plotter (http://www.kmplot.com). B, The expression of HOTTIP by qRT‐PCR analysis in the MCF10A, MCF7 and T47D cells. C, F, The percentage of CD44⁺/CD24⁻ cells by Flow cytometry in the sphere cells of MCF7 and T47D, and their parental cells. D, E, Sphere formation capacities by sphere formation assay in the spheres cells and the parental cells. G, H, Western blot analysis showing the protein expression levels of OCT4, SOX2 and CK14, CK18 in the sphere cells and their parental cells. I, The relative expression of HOTTIP in sphere cells and parental cells was assessed by qRT‐PCR. Data are presented as mean ± SD. *P < .05, **P < .01, ***P < .001 compared to MCF‐10A or the parental cells

Figure 2

HOTTIP is involved in maintaining the stemness of BCSCs. A,B, The HOTTIP expression was measured in MCF7 and T47D sphere cells transfected with shHOTTIP or shControl plasmids, and in the parental MCF7 and T47D cell lines transfected with overexpression (OE)‐HOTTIP or control plasmids by qRT‐PCR analysis. C, The percentage of CD44⁺/CD24⁻ BCSCs subpopulations was detected by Flow cytometry in the HOTTIP‐silenced MCF7 and T47D sphere cells. D, Sphere formation assays showed the self‐renewal capacity of the HOTTIP‐silenced MCF7 and T47D sphere cells. E, Soft‐agar colony formation assays showed the clone formation ability of the HOTTIP‐overexpressed MCF7 and T47D cells. F, G, The expression of OCT4, SOX2 and CK14, CK18 in the HOTTIP‐silenced MCF7 and T47D sphere cells and in HOTTIP‐overexpressed parental cells by Western blot analysis. Data are presented as mean ± SD. **P < .01, ***P < .001 compared to Control or shControl group

Figure 3

miR‐148a‐3p negatively modulated the stemness of BCSCs. A, The relationship between hsa‐miR‐148a expression and the outcomes of breast cancer patients was analysed using the online tool, KM plotter (http://www.kmplot.com). B, Spearman's correlation coefficient analysis between miR‐148a‐3p expression and WNT1expression in 15 patients with BC. C, The relative expression of miR‐148a‐3p in the sphere cells of MCF7 and T47D, and their parental cells was assessed by qRT‐PCR. D, G, The relative expression of miR‐148a‐3p of the sphere cells transfected with miR‐148a‐3p mimic, and the parental cells transfected with miR‐148a‐3p inhibitor by qPCR analysis. E, The percentage of CD44+/CD24‐ BCSCs subpopulations determined in the miR‐148a‐3p mimic‐transfected sphere cells by flow cytometry. F, The self‐renewal capacity of the miR‐148a‐3p mimic‐transfected sphere cells was analysed by sphere formation assays. H, The clone formation ability of the miR‐148a‐3p inhibitor‐transfected parental cells was detected by soft‐agar colony formation. I, J, The protein levels of OCT4, SOX2 and CK14, CK18 in the miR‐148a‐3p mimic‐transfected sphere cells and the miR‐148a‐3p inhibitor‐transfected parental cells by Western blot. Data are presented as mean ± SD. * P < .05, ** P < .01 compared to control group

Figure 4

HOTTIP regulates miR‐148a‐3p by acting as a ceRNA. A, A putative binding site between HOTTIP and miR‐148a‐3p by DIANA TOOLS. B, C, D, Luciferase reporter assay in the HEK‐293T, MCF7 and T47D cell lines cotransfected with WT and MUT type HOTTIP reporters and miR‐148a‐3p mimics. E, F, qRT‐PCR was performed to test the expression of miR‐148a‐3p in the sh‐HOTTIP‐transfected sphere cells of MCF7 and T47D, and the OE‐HOTTIP‐transfected parental cells. Data are presented as mean ± SD. **P < .01, ***P < .001 compared to control or shControl group

Figure 5

HOTTIP modulates the stemness of BCSCs via miR‐148a‐3p/WNT1 signaling. A, C, The effects of miR‐148a‐3p inhibitor on the self‐renewal capacity of the shHOTTIP‐transfected MCF7 and T47D sphere cells by sphere formation assays. B, D, The effects of miR‐148a‐3p mimic on the self‐renewal capacity of the OE‐HOTTIP‐transfected MCF7 and T47D parental cells by soft‐agar colony formation. E, The effects of miR‐148a‐3p inhibitor on the WNT1 protein levels of the shHOTTIP‐transfected MCF7 and T47D sphere cells by Western blot assay. F, The effects of miR‐148a‐3p mimic on the WNT1 protein levels of the OE‐HOTTIP‐transfected MCF7 and T47D parental cells. G, The correlation between HOTTIP and miR‐148a‐3p was analysed by Spearman's correlation analysis. H, The relationship of HOTTIP and WNT1 expression was analysed by Spearman's correlation analysis. I, GEO dataset revealed a significant positive correlation between HOTTIP and WNT1 expression. Data are presented as mean ± SD. *P < .05, **P < .01, compare with Control or shControl group

Figure 6

Effect of HOTTIP knockdown on tumour growth. A, Image of tumour formation from HOTTIP‐silenced group and negative control group at the end of point. B, Tumour growth curves were plotted for HOTTIP‐silenced group and negative control group by the tumour volume. The tumour volume was calculated every three days. C, Comparison tumour weight between HOTTIP‐silenced group and negative control group. D, The expression of HOTTIP in HOTTIP‐silenced group and negative control group was determined by qRT‐PCR. E, OCT4, SOX2, CK14, CK18, and WNT1 expressions in HOTTIP‐silenced group and negative control group were detected by Western blot. Data are presented as mean ± SD. **P < .01, ***P < .001, compare with shControl group