A risk progression breast epithelial 3D culture model reveals Cx43/hsa_circ_0077755/miR-182 as a biomarker axis for heightened risk of breast cancer initiation

Nataly Naser Al Deen, Nadia A. Lanman, Shirisha Chittiboyina, Sophie A. Lelièvre, Rihab Nasr, Farah Nassar, Heinrich zu Dohna, Mounir G. AbouHaidar, Rabih TalhoukScientific 2021 · DOI

TL;DR

This study introduces a novel 3D breast epithelial cell model that mimics early-stage breast cancer progression, revealing a specific molecular axis (Cx43/hsa_circ_0077755/miR-182) that acts as a biomarker for heightened risk of breast cancer initiation. This axis's dysregulation patterns may also predict prognosis, offering a new avenue for early detection and risk assessment.

What Problem Does This Paper Address?

Early-onset breast cancer is a growing global concern, particularly in certain populations where it presents at younger ages and with poorer prognoses. Current diagnostic methods often fail to identify individuals at high risk for initiating this disease. While the interplay of mRNA, circular RNAs (circRNAs), and microRNAs (miRNAs) is known to be involved in established breast cancer, their role as early risk assessment markers, especially in the context of initial tumor development and epithelial polarity disruption, remains underexplored. This paper aims to bridge this gap by identifying molecular signatures that predict heightened risk of breast cancer initiation.

Methods

The researchers utilized a sophisticated 3D breast epithelial cell culture model, HMT-3522 S1 (S1), which can be manipulated to mimic normal and pre-malignant states. Specifically, they silenced the gap junction protein Cx43 in S1 cells (creating Cx43-KO-S1 cells) to induce a loss of epithelial polarity and multilayering, mimicking pre-cancerous mammary epithelial morphology.

Key techniques employed include:

  • circRNA Microarrays: To identify circRNAs that are differentially expressed in response to Cx43 silencing.
  • miRNA Sequencing: To identify miRNAs that are differentially expressed in the same model.
  • RT-qPCR: For validation of differentially expressed circRNAs.
  • Bioinformatics Analysis: Including prediction of miRNA response elements (MREs) for circRNAs and gene co-expression network analysis using tools like Arraystar, TargetScan, Ingenuity Pathway Analysis (IPA), and Cytoscape.
  • Patient Cohort Validation: Utilizing existing microarray data from early-stage breast cancer patients to validate findings.
  • Survival Analysis: Employing the Kaplan-Meier Plotter with METABRIC and other datasets to assess the prognostic value of the identified axis.

Datasets

  • In vitro datasets: Generated from the HMT-3522 S1 (S1) and Cx43-knockout S1 (Cx43-KO-S1) 3D breast epithelial cell culture models. This includes circRNA microarray data and miRNA sequencing data.
  • Patient validation cohort: Microarray data from early-stage breast cancer patients (n=45 invasive ductal carcinoma vs. 17 normal adjacent breast tissues) from Lebanon, previously published by Nassar et al. (2017). Additional patient data from TCGA and METABRIC datasets were used for survival analysis.

Key Insights and Ideas

  • Cx43's Role in Post-Transcriptional Regulation: The study highlights that Cx43 is not only involved in cell-cell communication but also plays a crucial role in post-transcriptional regulatory pathways that influence breast cancer initiation.
  • 3D Model as a Pre-Malignant Mimic: The 3D culture system effectively recapitulates key features of pre-malignant mammary lesions, making it a powerful tool for studying early cancer development.
  • CircRNA-miRNA Axis as a Biomarker: The identification of the Cx43/hsa_circ_0077755/miR-182 axis as a potential biomarker for heightened risk of breast cancer initiation is a significant finding, as it moves beyond established cancer markers to focus on the very early stages of the disease.
  • Prognostic Potential: The dysregulation patterns of this axis appear to correlate with prognosis, suggesting its utility in predicting disease progression.

Key Results

  • Silencing Cx43 in the 3D culture model led to the significant dysregulation of 121 circRNAs and 65 miRNAs.
  • A comprehensive analysis identified and validated the Cx43/hsa_circ_0077755/miR-182 axis as the sole molecular signature that met all criteria for a risk-assessment biomarker.
  • This axis exhibits an inverse dysregulation pattern between the circRNA and miRNA in both the cell culture model and the patient cohort, consistent with a sponging mechanism.
  • Survival analysis indicated that up-regulation of miR-182 is associated with poor prognosis in patients with grade II breast tumors, while down-regulation of Cx43 also associates with poor prognosis in the same tumor grade.

Limitations and Future Work

  • Limitations:
    • The study acknowledges that circRNAs can sponge multiple miRNAs, and each miRNA can target numerous mRNAs, implying that other downstream pathways might be involved.
  • While the 3D model is effective, it is an in vitro system and may not fully capture the complexity of in vivo tumor development.
  • The validation of the Cx43/hsa_circ_0077755/miR-182 axis was primarily based on existing patient data and in vitro models; further clinical validation is needed.
  • Future Work:
    • Directly validate the interactions between Cx43, hsa_circ_0077755, and miR-182 through experimental manipulation (e.g., overexpression).
  • Investigate whether overexpressing candidate miRNAs, including miR-182/183 cluster, can recapitulate tumor-initiation phenotypes observed upon Cx43 loss.
  • Validate this axis in serum samples from breast cancer-free individuals at risk, as well as in early-stage and advanced-stage breast cancer patients, to assess its potential as a non-invasive biomarker.
  • Further explore the role of this axis in predicting prognosis across different tumor stages and subtypes.

Summary updated

Abstract

mRNA-circRNA-miRNAs axes have been characterized in breast cancer, but not as risk-assessment axes for tumor initiation in early-onset breast cancer that is increasing drastically worldwide. To address this gap, we performed circular RNA (circRNA) microarrays and microRNA (miRNA) sequencing on acini of HMT-3522 S1 (S1) breast epithelial risk-progression culture model in 3D and chose an early-stage population miRNome for a validation cohort. Nontumorigenic S1 cells form fully polarized epithelium while pretumorigenic counterparts silenced for gap junction Cx43 (Cx43-KO-S1) lose epithelial polarity, multilayer and mimic premalignant in vivo mammary epithelial morphology. Here, 121 circRNAs and 65 miRNAs were significantly dysregulated in response to Cx43 silencing in cultured epithelia and 15 miRNAs from the patient cohort were involved in epithelial polarity disruption. Focusing on the possible sponging activity of the validated circRNAs to their target miRNAs, we found all miRNAs to be highly enriched in cancer-related pathways and cross-compared their dysregulation to actual miRNA datasets from the cultured epithelia and the patient validation cohort. We present the involvement of gap junction in post-transcriptional axes and reveal Cx43/hsa_circ_0077755/miR-182 as a potential biomarker signature axis for heightened-risk of breast cancer initiation, and that its dysregulation patterns might predict prognosis along breast cancer initiation and progression.
BibTeX
@inproceedings{deen2021risk,
  author = {Nataly Naser Al Deen and Nadia A. Lanman and Shirisha Chittiboyina and Sophie A. Lelièvre and Rihab Nasr and Farah Nassar and Heinrich zu Dohna and Mounir G. AbouHaidar and Rabih Talhouk},
  title = {{A risk progression breast epithelial 3D culture model reveals Cx43/hsa\_circ\_0077755/miR-182 as a biomarker axis for heightened risk of breast cancer initiation}},
  booktitle = {Scientific},
  year = {2021},
  doi = {10.1038/s41598-021-82057-y},
}
Figure