Preview

PDF, English Download (9MB) | Terms of use

Abstract

Despite overall improvements in breast cancer management, breast cancer continues to be a major health challenge worldwide, with high mortality rates. Following initially successful therapies, some cells are able to evade the treatment and remain in the body at undetectable levels (known as minimal residual disease) and could eventually lead to a lethal tumor recurrence. Understanding the nature of this elusive cell population is of a great importance in tumor eradication and relapse prevention. Nevertheless, the intricate mechanistic details about these processes and the nature of these cells remain poorly understood, mainly due to difficulties in obtaining patient material of MRD and being able to study them over the course of treatment. To characterize MRD in breast cancer, we employed an inducible TetO-MYC/TetO-Neu/MMTV-rtTA mouse model and primary 3D cultures of mammary organoids that yield a correlate of MRD upon oncogene silencing. A combination of immunofluorescence, RNA sequencing, lipidomics and metabolomics revealed the unique nature of the residual cells. Despite the phenotypic similarity to the normal population, the residual cells exhibited a distinct transcriptional profile. This profile was also different when compared to the tumor, highlighting the unique properties of the residual cell population. In addition, lipid profiles of the residual cells were also distinguished from both normal and tumor populations. Surprisingly, despite the inactivated oncogenes and seemingly normal phenotype of the residual structures, the residual cells bore a metabolic resemblance to tumor cells. They retained some tumor metabolic hallmarks, which persevered long after the oncogenes were silenced and tumors had regressed. This was demonstrated in our 3D cultures and verified in vivo on histological sections and fresh samples of the mammary glands following tumor regression. Enhanced glycolysis, the urea cycle and NOS2 activity were the most prominent features preserved in the residual cells. As shown in correlation with publicly available microarray datasets of patient samples following neoadjuvant treatment, these traits could be particularly important in MRD of the basal-like HER2 positive and HER2 negative breast cancer subtypes. In conclusion, our findings suggest vast and profound effects of the changes happening over the course of tumor progression, which result in an altered metabolic network even in the absence of oncogene signaling, implying the existence of a “metabolic memory” probably imprinted through the changes in the epigenetic landscape. This phenomenon, particularly its potential driver(s), remains to be investigated. Furthermore, enhanced glycolysis, urea cycle and NOS2 activity in the residual cells compared to the normal, could provide an opportunity to interfere with MRD, offering the potential of preventing tumor recurrences.