Back

Developmental Plasticity and Stromal Co-option Shape a Pituitary Neuroendocrine Tumor Transcriptional Continuum

Osorio, R. C.; Oh, J. Y.; Jung, J.; Ehrenberg, A. J.; Saha, A.; Lad, M.; Khela, H.; Brennick, N.; Giannikopoulos, P.; Seeley, W. W.; Grinberg, L. T.; Diaz, A.; Aghi, M. K.

2025-06-24 cancer biology
10.1101/2025.06.18.659743 bioRxiv
Show abstract

Pituitary neuroendocrine tumors (PitNETs) are common intracranial neoplasms with complex biology underpinned by unresolved cellular origins, molecular heterogeneity, and microenvironment interactions. Here, we employ single-nuclei RNA-sequencing (snRNA-seq) of 419,874 cells from human normal pituitaries and PitNETs with spatial transcriptomics to resolve these challenges. We identify multi-hormonal neuroendocrine cells in both normal and tumor tissues, originating as early pseudotime intermediates from pituitary stem cells, revealing an inherent plasticity that blurs traditional lineage boundaries. PitNETs exhibit a transcriptional continuum across subtypes, challenging their classification into discrete categories. Trajectory analysis uncovers divergent cellular origins: silent gonadotroph adenomas (SGAs), prolactinomas, silent corticotroph adenomas (SCAs), and Cushings adenomas are closely linked to differentiated neuroendocrine cells, while somatotroph and null cell adenomas (NCAs) appear to derive more directly from adult stem cells. Tumor cells co-opt robust cell-cell communication networks found in normal adult neuroendocrine cells. Spatial profiling further demonstrates that perivascular niches enhance tumorigenicity through angiogenic and epithelial-mesenchymal transition programs. Our work redefines PitNETs as ecosystems shaped by developmental plasticity and microenvironmental crosstalk, offering a roadmap for future therapies targeting lineage fluidity and stromal dependencies.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.