Clinically relevant morphological structures in breast cancer represent transcriptionally distinct tumor cell populations with varied degrees of epithelial-mesenchymal transition and CD44+CD24- stemness.

Evgeny V Denisov,Jochen Wilhelm,Igor N Lebedev,Tatiana S Gerashchenko,Nadezhda V Cherdyntseva,Marina V Zavyalova,Sergey V Vtorushin,Mikhail A Buldakov,Lubov A Tashireva,Nikolay A Skryabin,Aleksei A Sleptcov,Vladimir M Perelmuter

doi:10.18632/oncotarget.18022

Abstract

Intratumor morphological heterogeneity in breast cancer is represented by different morphological structures (tubular, alveolar, solid, trabecular, and discrete) and contributes to poor prognosis; however, the mechanisms involved remain unclear. In this study, we performed 3D imaging, laser microdissection-assisted array comparative genomic hybridization and gene expression microarray analysis of different morphological structures and examined their association with the standard immunohistochemistry scorings and CD44+CD24- cancer stem cells. We found that the intratumor morphological heterogeneity is not associated with chromosomal aberrations. By contrast, morphological structures were characterized by specific gene expression profiles and signaling pathways and significantly differed in progesterone receptor and Ki-67 expression. Most importantly, we observed significant differences between structures in the number of expressed genes of the epithelial and mesenchymal phenotypes and the association with cancer invasion pathways. Tubular (tube-shaped) and alveolar (spheroid-shaped) structures were transcriptionally similar and demonstrated co-expression of epithelial and mesenchymal markers. Solid (large shapeless) structures retained epithelial features but demonstrated an increase in mesenchymal traits and collective cell migration hallmarks. Mesenchymal genes and cancer invasion pathways, as well as Ki-67 expression, were enriched in trabecular (one/two rows of tumor cells) and discrete groups (single cells and/or arrangements of 2-5 cells). Surprisingly, the number of CD44+CD24- cells was found to be the lowest in discrete groups and the highest in alveolar and solid structures. Overall, our findings indicate the association of intratumor morphological heterogeneity in breast cancer with the epithelial-mesenchymal transition and CD44+CD24- stemness and the appeal of this heterogeneity as a model for the study of cancer invasion.

Highlights

IntroductionIntratumor morphological heterogeneity is a common phenomenon for many cancers and is represented by various morphological structures (named as histological, invasive, and infiltrative patterns) that reflect the different architectural arrangements of tumor cells
Intratumor morphological heterogeneity is a common phenomenon for many cancers and is represented by various morphological structures that reflect the different architectural arrangements of tumor cells
Genetic variability tends to be a more popular subject in the field of intratumor heterogeneity, the importance of morphological diversity and its relationship with cancer prognosis is well-known for a long time [3, 5] and emphasized by recent studies [35,36,37]

Summary

Introduction

Intratumor morphological heterogeneity is a common phenomenon for many cancers and is represented by various morphological structures (named as histological, invasive, and infiltrative patterns) that reflect the different architectural arrangements of tumor cells. In dependence on the prevalence of a certain type of morphological structure, cancers are classified to distinct histological forms with a specific prognosis and response to therapy. The number of different types of morphological structures in breast tumors varies from case to case and depends on the molecular subtype of BC [9]. Patients either with alveolar/trabecular structures or with all types of morphological structures demonstrated a poor response to neoadjuvant chemotherapy [11, 13]. Alveolar and trabecular structures were associated with poor metastasis-free survival in BC patients [14]

Methods

Results

Conclusion