(1) Department of Biomedical Engineering, Tufts University, Medford, MA 02155, USA
(2) Center for Molecular Medicine, Maine Medical Center Research Institute, Scarborough, ME 04074, USA
* Corresponding author Email: Ying.Chen@tufts.edu
The Snail gene family encodes evolutionarily conserved transcriptional repressor proteins that bind to E box sequences. In mammals, the SNAI1 and SNAI2 proteins are key regulators of the epithelial–mesenchymal transition both during embryonic development and tumour metastasis. Recent work has shown that SNAI1/SNAI2 gene and protein expression is regulated during tumourigenesis by a number of microRNAs. Furthermore, the existence of double-negative feedback loops that provide exquisite regulation of SNAI1/SNAI2 expression and the expression of microRNAs binding to the
Here, we review new studies highlighting the regulatory interplay of miRNA expression, regulation of SNAI1 and SNAI2 expression and the epithelial–mesenchymal transition in the progression and metastasis of ovarian and other epithelial tumours. We also summarise new data demonstrating a role for SNAI1 and SNAI2 expression in mediating acquired tumour drug resistance.
This review highlights the important roles played by the transcriptional repressors SNAI1 and SNAI2 during tumour progression and metastasis, as well as their regulatory interplay with the expression of multiple different microRNAs.
Snail gene family members encode highly conserved zinc finger transcriptional repressor proteins[1,2]. In mammals, the Snail family genes include
MicroRNAs (miRNAs) are a class of short (approximately 22 nucleotide) non-coding RNA molecules that play important regulatory roles in a wide variety of cellular processes, including cell proliferation, differentiation, apoptosis, senescence and tumourigenesis[3,4]. miRNAs negatively regulate gene expression post-transcriptionally by binding to the 3ʹ-untranslated region (UTR) of their target mRNAs to inhibit their translation or decrease mRNA stability. It has been estimated computationally that more than 60% of protein-coding genes may be miRNA targets.
EMT is a cellular programme during which epithelial cells lose epithelial characteristics and acquire mesenchymal traits. EMT occurs normally during embryonic development, but also occurs during pathological processes such as fibrosis and tumour metastasis[6,7]. During EMT, epithelial cells disassemble intercellular junctional complexes, lose apical-basal polarity and adherence to the basement membrane and gain a more fibroblast-like morphology with increased motility and invasiveness. Several transcription factors, including SNAI1, SNAI2, ZEB1 and ZEB2, are key regulators of the EMT. In addition, recent studies have established that various miRNAs also regulate EMT. In this review, we summarise new studies highlighting the regulatory interplay of miRNA expression, regulation of SNAI1/SNAI2 expression and the EMT in the progression and metastasis of ovarian and other epithelial tumours. We also summarise new data demonstrating a role for SNAI1 and SNAI2 expression in mediating acquired tumour drug resistance.
Ovarian cancer is the fifth leading cause of cancer death in women in the United States and is the most lethal gynaecologic malignancy. The Cancer Genome Atlas (TCGA) Research Network has catalogued mRNA expression, miRNA expression, promoter DNA methylation and DNA copy number alterations in a large set of different tumour types, including 489 high-grade serous ovarian adenocarcinomas. They used non-negative matrix factorisation consensus clustering to define four gene expression subtypes of high-grade serous ovarian cancer, termed immunoreactive, differentiated, proliferative and mesenchymal. Survival duration did not differ significantly among patients with tumours of these different gene expression subtypes. Similar cluster analysis of miRNA expression defined three expression subtypes. The miRNA subtype 1 partially overlapped the mRNA proliferative subtype and miRNA subtype 2 partially overlapped the mRNA mesenchymal subtype. Survival duration differed significantly between these miRNA subtypes. Patients with miRNA subtype 1 tumours survived significantly longer than patients with tumours of the other two subtypes. While patient survival did not correlate with any of the four gene expression subtypes, a 193-gene transcriptional signature predictive of overall survival could be defined using an integrated expression data set from 215 samples. Subsequent analyses of the TCGA data of miRNA expression of serous ovarian tumours demonstrated that expression of a set of 34 miRNAs also was predictive of overall patient survival.
The TCGA Research Network has recently described prognostically relevant gene signatures for high-grade serous ovarian cancer. Together, these subtype and survival gene expression signatures provide a prognostic model of high-grade serous ovarian cancer classification, which they term ‘Classification of Ovarian Cancer’ (CLOVAR). In this study, the descriptions of the previously observed mesenchymal, differentiated, proliferative and immunoreactive subtypes of high-grade serous ovarian cancer were substantially expanded. During analysis, these subtype signatures were combined with the CLOVAR survival signature, containing genes whose expression was either strongly correlated (good prognosis genes) or anti-correlated (poor prognosis genes) with survival. The detailed stratification permitted by the new classifications resulted in the identification of a subset of high-grade serous ovarian cancer, with only 23 months median survival and 63% platinum therapy resistance, compared with median survival of 46 months and a platinum resistance rate of 23% in all other cases. These worst outcome tumours were classified as both CLOVAR poor prognosis (for survival) and CLOVAR mesenchymal subtype, which was enriched for EMT-related genes.
Another group recently described a set of integrated genomic analyses that further defined the mesenchymal gene expression subtype of serous ovarian cancer defined by the TCGA analysis and its relationship to the EMT. Integration of the TCGA mRNA, miRNA, DNA copy number and DNA methylation data revealed a miRNA-regulatory network that defined an integrated mesenchymal (iM) subtype associated with poor overall survival in 459 cases of serous ovarian cancer from the TCGA cohort and 560 cases from independent cohorts. In these analyses, 219 genes were predicted to be targets of 19 miRNAs, including genes encoding the EMT inducers SNAI2 and ZEB2. Clustering analyses based on the 219 miRNA-associated genes identified two clusters in both the TCGA cohort and in three independent patient cohorts. These clusters were termed the iM subtype and the integrated epithelial (iE) subtype. Patients with tumours of the iM subtype had significantly shorter overall survival than patients with tumours of the iE subtype. Histological features of the tumours also differed between the iM and iE subtypes. Eight key miRNAs, including miR-506, miR-141 and miR-200a, were predicted to regulate 89% of the targets in the iM network (Figure 1). The least studied of these miRNAs was miR-506. Functional assays in ovarian cancer cell lines demonstrated that miR-506 bound the 3’ UTR of the
miRNA–gene interaction network of eight key miRNAs and the EMT-related genes they are predicted to regulate in serous ovarian cancer cells. Reproduced with permission from Yang et al..
Another recent study supports the hypothesis that induction of the EMT programme is important for ovarian cancer progression and metastasis. Transcriptome and pathway analyses were performed on 14 matched sets of primary and metastatic serous ovarian cancer samples collected from seven patients with Stage III/IV cancer. Unsupervised hierarchical clustering revealed that the metastatic samples of five of the seven patients grouped closely with their respective primary samples, while the metastatic samples of the other two patients clustered most closely with one another, distant from their respective primary samples. Pathway analysis showed that 13 of the 20 most significantly enriched pathways were associated with the EMT. Further analyses focusing on EMT-associated genes clearly distinguished the primary from the metastatic samples in six of the seven patients.
A siRNA screen identified the homeobox transcription factor ALX1 as a novel regulator of EMT in ovarian cancer. siRNA-mediated attenuation of
Another recent report determined the effects of
Similar modes of regulation of SNAI1/SNAI2 expression are utilised in other cancer types. Breast cancer is the most commonly diagnosed cancer in women and is the second leading cause of death in women. MiR-124 expression was significantly suppressed in human breast cancer specimens, and expression levels were inversely correlated with the histological grade of the cancer. Ectopic expression of miR-124 in the MDA-MB-231 and BT-549 breast cancer cell lines strongly inhibited cell motility and invasive capacity, as well as the EMT programme. Luciferase reporter assays demonstrated that the
miRNA regulation of SNAI2 expression may be involved in the early progression to malignancy and metastasis during prostate cancer. miR-182 and miR-203 coordinately regulate SNAI2 expression in a premalignant prostate cell EMT model consisting of the prostate primary epithelial cell line EP156T and progeny mesenchymal EPT1 cells. miR-182 and miR-203 expression was completely repressed during EMT from EP156T cells to the mesenchymal EPT1 cells. The 3ʹ-UTR of the
A recently revealed aspect of the regulation of
In a mouse model of progressive prostate cancer, the
Integration of the miR-203/SNAI1 and miR-200/ZEB1/ZEB2 double-negative feedback loops. The top panel shows the core network integrating described interactions between miR-203, the miR-200 family, SNAI1, ZEB1, ZEB2 and E-cadherin (CDH1). The bottom panels show the stable epithelial (E) and mesenchymal (M) states obtained after dynamic analyses. Green indicates downregulated expression; red indicates upregulated expression. The red ‘‘lightning bolt’’ in the lower left panel indicates SNAI1 upregulation triggering the transition from the epithelial to the mesenchymal state. Reproduced with permission from Moes et al..
SNAI1 and SNAI2 expression may also play a role in a tumour’s acquisition of resistance to chemotherapy. Standard treatment for ovarian cancer involves tumour debulking surgery followed by platinum-taxane chemotherapy. Within 6 months, platinum-resistant tumours recur in approximately 25% of patients. Analysis of the A2780 ovarian adenocarcinoma cell line and a cisplatin-resistant variant revealed that the cisplatin-resistant cells exhibited a more mesenchymal morphology, with spindle-shaped cells that extended pseudopodia. The cisplatin-resistant cells exhibited increased motility and invasive behaviour, and siRNA-mediated knockdown of SNAI1 and/or SNAI2 expression in the cisplatin-resistant cells restored the epithelial morphology exhibited by the parental A2780 line, reduced cellular motility and invasion and increased sensitivity of the knockdown cells to cisplatin. Importantly, increased expression of the
SNAI1 and SNAI2 expression had previously been shown to mediate resistance to radiation and paclitaxel in the A4 ovarian cancer cell line. In these cells, this resistance was caused, at least in part, by SNAI1/SNAI2-mediated repression of the pro-apoptotic genes
This review highlights the important roles played by the transcriptional repressors SNAI1 and SNAI2 during tumour progression and metastasis, as well as their regulatory interplay with expression of multiple different miRNAs. The identification of double-negative feedback loops connecting transcription of various miRNAs and the EMT transcriptional regulators
CLOVAR, Classification of Ovarian Cancer; EMT, epithelial–mesenchymal transition; iE, integrated epithelial; iM, integrated mesenchymal; miRNA, microRNA; TCGA, The Cancer Genome Atlas; UTR, untranslated region.
Work in TG’s laboratory on Snail family genes was funded by NIH grant R01HD034883.
All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.