Epigenetic re-wiring of breast cancer by pharmacological targeting of C-terminal binding protein

Abstract

The C-terminal binding protein (CtBP) is an NADH-dependent dimeric family of nuclear proteins that scaffold interactions between transcriptional regulators and chromatin-modifying complexes. Its association with poor survival in several cancers implicates CtBP as a promising target for pharmacological intervention. We employed computer-assisted drug design to search for CtBP inhibitors, using quantitative structure-activity relationship (QSAR) modeling and docking. Functional screening of these drugs identified 4 compounds with low toxicity and high water solubility. Micro molar concentrations of these CtBP inhibitors produces significant de-repression of epigenetically silenced pro-epithelial genes, preferentially in the triple-negative breast cancer cell line MDA-MB-231. This epigenetic reprogramming occurs through eviction of CtBP from gene promoters; disrupted recruitment of chromatin-modifying protein complexes containing LSD1, and HDAC1; and re-wiring of activating histone marks at targeted genes. In functional assays, CtBP inhibition disrupts CtBP dimerization, decreases cell migration, abolishes cellular invasion, and improves DNA repair. Combinatorial use of CtBP inhibitors with the LSD1 inhibitor pargyline has synergistic influence. Finally, integrated correlation of gene expression in breast cancer patients with nuclear levels of CtBP1 and LSD1, reveals new potential therapeutic vulnerabilities. These findings implicate a broad role for this class of compounds in strategies for epigenetically targeted therapeutic intervention.

Proteomic Analysis of Zeb1 Interactome in Breast Carcinoma Cells

“…Recent reports suggest that the C-terminal binding protein 2 (CtBP2) is a critical transcriptional co-repressor that orchestrates the EMT program both in Zeb1-dependent and -independent ways [ 34 , 35 ]. Importantly, CTBP2 may serve as an independent prognostic marker for several tumors including hepatocellular carcinoma and lung cancer [ 36 , 37 ].…”

TBX2 acts as a potent transcriptional silencer of tumour suppressor genes through interaction with the CoREST complex to sustain the proliferation of breast cancers

“…Interactions are also facilitated between the CoREST and PRC2 complexes by the HOX Transcript Antisense lncRNA (HOTAIR) to coordinate CoREST activity with the repressive H3K27 methylation mark ( 21 ). Components of the CoREST complex are known to be deregulated and/or over-expressed in human breast tumours, resulting in the repression of epithelial pro-differentiation genes such as p21 WAF1/CIP1 ( 22 ), E-cadherin ( 23 ), Grainyhead-Like 2 (GRHL2) and FOXA1 ( 24 ) to promote a malignant mesenchymal state. Unsurprisingly, the pharmacological targeting of CoREST in human cancer has become a subject of intense investigation, whereby strategies have been developed to target both individual members of the complex, or multiple cofactors simultaneously, in an attempt to recover tumour suppressor expression ( 25 , 26 ).…”

Protein expression of the gp78 E3 ligase predicts poor breast cancer outcome based on race

“…To determine the predictive value of gp78 expression in mammary malignancies, we analyzed a previously established racially diverse breast cancer cohort of patients ( n = 560) residing in a designated health disparities catchment area in Eastern North Carolina (median follow-up, 8.5 years) and arrayed in tissue microarray (TMA) format ( 6 , 11 , 49 , 50 ) to define the association between gp78 protein levels and breast cancer patient tumor characteristics and survival. The expression of gp78 was measured by IHC using a quantitative digital pathology platform in which pathologist-annotated regions of tumor were scored in terms of the IHC staining level assessed by an increasing pixel intensity stratification: 0, 1 + , 2 + , or 3 + ( 6 , 11 , 49 , 51 ). The percent of cells with these intensities were then used to assign protein expression values based on a continuous metric referred to as the H-score (0–300 scale) using the following equation: (H-score = 3 [%3 + ] + 2 [%2 + ] + 1 [%1 + ]) ( 6 , 11 , 49 ).…”

“…The expression of gp78 was measured by IHC using a quantitative digital pathology platform in which pathologist-annotated regions of tumor were scored in terms of the IHC staining level assessed by an increasing pixel intensity stratification: 0, 1 + , 2 + , or 3 + ( 6 , 11 , 49 , 51 ). The percent of cells with these intensities were then used to assign protein expression values based on a continuous metric referred to as the H-score (0–300 scale) using the following equation: (H-score = 3 [%3 + ] + 2 [%2 + ] + 1 [%1 + ]) ( 6 , 11 , 49 ). Using the Aperio platform, digital scoring for gp78 was strongly correlated with the manual score provided by a pathologist (adjusted Pearson’s R 2 = 0.93) ( Supplemental Figure 1A ; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.157465DS1 ).…”