Genomic regulation of invasion by STAT3 in triple negative breast cancer

Abstract

Breast cancer is a heterogeneous disease comprised of four molecular subtypes defined by whether the tumor-originating cells are luminal or basal epithelial cells. Breast cancers arising from the luminal mammary duct often express estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth receptor 2 (HER2). Tumors expressing ER and/or PR are treated with anti-hormonal therapies, while tumors overexpressing HER2 are targeted with monoclonal antibodies. Immunohistochemical detection of ER, PR, and HER2 receptors/proteins is a critical step in breast cancer diagnosis and guided treatment. Breast tumors that do not express these proteins are known as "triple negative breast cancer" (TNBC) and are typically basal-like. TNBCs are the most aggressive subtype, with the highest mortality rates and no targeted therapy, so there is a pressing need to identify important TNBC tumor regulators. The signal transducer and activator of transcription 3 (STAT3) transcription factor has been previously implicated as a constitutively active oncogene in TNBC. However, its direct regulatory gene targets and tumorigenic properties have not been well characterized. By integrating RNA-seq and ChIP-seq data from 2 TNBC tumors and 5 cell lines, we discovered novel gene signatures directly regulated by STAT3 that were enriched for processes involving inflammation, immunity, and invasion in TNBC. Functional analysis revealed that STAT3 has a key role regulating invasion and metastasis, a characteristic often associated with TNBC. Our findings suggest therapies targeting STAT3 may be important for preventing TNBC metastasis.

Keywords: ChIP-seq; RNA-seq; STAT3; TNBC; invasion.

Figure 1. STAT3 binds in a subtype-specific manner across basal TNBC breast cancer

A. Heatmap of Spearman rank correlations between all pairwise comparisons of STAT3 binding sites across basal TNBC cell lines and TNBC tumors (blue). B. Venn diagram of replicate STAT3 binding sites in basal breast cancer cell line HCC70. 20,808 high-confidence binding sites were called in both replicates. C. Correlation plot of STAT3 binding sites in HCC70 replicates (Spearman rho=0.92) D. Canonical STAT3 motif enriched in binding sites in HCC70 and MDA-MB-231. E. Venn diagram of replicate STAT3 binding sites in basal breast cancer cell line MDA-MB-231. 7,767 high-confidence binding sites were called in both replicates. F. Correlation plot of STAT3 binding sites in MDA-MB-231 replicates (Spearman, rho=0.80).

Figure 2. STAT3 binds and regulates genes in basal TNBC cell lines

A. Heatmap of 737 differentially expressed gene transcripts in HCC70 in response to STAT3 knockdown (> 2.0-fold differences between siRNA treated cells and non-targeting vehicle controls) B. Cumulative distribution function plot of STAT3 binding sites near differentially expressed transcripts in HCC70. C. Heatmap of 548 differentially expressed gene transcripts in MDA-MB-231 in response to STAT3 knockdown (> 2.0-fold differences between siRNA treated cells and non-targeting vehicle controls) D. Cumulative distribution function plot of STAT3 binding sites near differentially expressed transcripts in MDA-MB-231.

Figure 3. STAT3 regulates a gene signature associated with an invasion phenotype in basal TNBC

(Data represents mean +/− SEM) A. Transwell invasion after STAT3 knockdown in HCC70 resulted in 1.6-fold reduced invasion (t-test, p-value<0.0001). B. Transwell invasion after STAT3 knockdown in MDA-MB-231 resulted in 1.9-fold reduced invasion (t-test, p-value < 0.0001). C. HCC70 proliferation measure after siRNA mediated STAT3 knockdown for 96 hrs. Knockdown of STAT3 did not result in reduced proliferation, but increased proliferation by 10% (t-test, p-value<0.0037). D. MDA-MB-231 proliferation measure after STAT3 knockdown for 96hrs. Knockdown of STAT3 resulted in 10% decrease in proliferation (t-test, p-value < 0.0040).