ECE2023 Oral Communications Oral Communications 6: Endocrine-related Cancer (6 abstracts)
1Division of Endocrinology and Diabetes, University Hospital of Würzburg, Würzburg, Germany; 2School of Mathematics, University of Birmingham, Birmingham, United Kingdom; 3Irving Institute for Cancer Dynamics, Columbia University, New York, United States; 4Max Delbrück Center for Molecular Medicine, Berlin, Germany; 5Core Unit SysMed, University of Würzburg, Würzburg, Germany; 6Institute of Metabolism and System Research, University of Birmingham, Birmingham, United Kingdom; 7Centre for Endocrinology, Diabetes and Metabolism (CEDAM), Birmingham Health Partners, Birmingham, United Kingdom
Background: Molecular mechanisms of malignant adrenocortical tumorigenesis remain elusive despite previous comprehensive genomic studies. Nonetheless, cellular heterogeneity of primary, recurrent and metastatic adrenocortical carcinoma (ACC) havent been fully investigated.
Aim: To characterize the molecular profile of different cell subtypes in primary, recurrent and metastatic ACC by single-nuclei RNA sequencing (snRNA-Seq), using adult human normal adrenal glands (NAG, n=6) and adrenocortical adenomas (ACA, n=12) as reference.
Methods: Single nuclei were isolated from 14 ACC samples, including 6 primary tumours, 3 local recurrence and 5 ACC metastasis from 9 different patients. Genomic alterations were evaluated by targeted next-generation sequencing in 6 cases, whereas CTNNB1 mutation were excluded by Sanger sequencing in the remaining 3 cases. snRNA-Seq was performed using 1CellBio. Seurat R package was used for data analysis, integration and exploration. Integration of datasets was performed by anchor pairs between ACC and NAG. Identification of cell subpopulations was performed using unsupervised clustering followed by differential gene expression analysis, with cluster annotation based on marker genes and scores for hallmark gene sets or the gene signatures of previously identified ACA cell populations.
Results: The ACC tumour microenvironment was relatively devoid of immune cells compared to NAG, emphasising that ACC is an immunological cold tumour. The integrated analysis revealed the presence of 7 subpopulations of cortical-derived cells in ACC. Among these, ACC recurrence and metastasis were highly represented by a cell population overexpressing IGF2, DLK1, and the translationally controlled tumour protein TPT1, and significant enrichment of ribosomal biogenesis and proliferation-associated signalling pathways. A second major subpopulation, enriched within primary ACC samples, displayed a transcriptional signature of increased cholesterol homeostasis and steroid synthesis similar to that observed in our cortisol-producing ACA. The intra-tumour differential gene expression analysis revealed that a decrease in expression of steroid hormone production genes, an upregulated cellular proliferation, an increased expression of IGF2, DLK1, TPT1 as well as the silencing of MEG3 and MEG8 are distinguishing ACC from NAG/ACA cells. Cortical cells statistically indistinguishable from NAG made up the remaining bulk of the ACC cell population, while smaller subpopulations were highly enriched with cell cycle genes or characterized by ACSM3, KRTCAP2 or ZNF331.
Conclusion: Our analyses represent the first molecular characterization of ACC at single-nuclei level, allowing the investigation of the molecular heterogeneity of ACC. The transcriptomic signatures showed the presence of 7 different subpopulations, including a specific cell population overexpressing IGF2, DLK1 and TPT1 associated to recurrent and metastatic ACC.