Dataset Information

Microarray expression profiling data for sorted prostate epithelial subpopulations based on LY6D

ABSTRACT: In the normal prostate, most basal and some luminal cells are castration-resistant (CR). The identity of these CR cells and their relation to CR prostate cancer are unresolved. We compared single-cell expression profiles of prostate cells sorted from hormonally naïve (HN) and castrated mice. We found both basal and luminal-localized cells, particularly the latter, were molecularly heterogeneous. CR luminal cells and a subset of HN luminal cells exhibited a similar “intermediate” expression pattern, including high-level expression of multiple prostate stem/progenitor marker genes and androgen receptor gene. We validated LY6D as a marker linking CR luminal cells to luminal progenitors. LY6D+ prostate cells, including LY6D+ luminal cells, were enriched for organoid-forming potential regardless of the presence or absence of androgen. Krt8-based lineage-tracing revealed that LY6D+ CR luminal cells produced LY6D- normal luminal cells upon regeneration, but LY6D+ luminal cancer cells under PTEN-deficiency. Furthermore, prostate cancers originating from CR luminal cells (LY6D+) exhibited a more advanced phenotype than those from HN luminal cells (LY6D+ or LY6D-). Lastly, LY6D amplification/upregulation appear associated with advanced prostate cancer in patient samples. Together, our studies demonstrate LY6D as a novel progenitor marker predictive of lethal CR disease.

ORGANISM(S): Mus musculus

PROVIDER: GSE92473 | GEO | 2018/12/21

REPOSITORIES: GEO

ACCESS DATA

Similar Datasets

Project description:Castration-resistant prostate cancer is a lethal disease. The cell type(s) that survive androgen-deprivation remain poorly described despite global efforts to understand the various mechanisms of therapy resistance. We recently identified in wild type mouse prostates a rare population of luminal progenitor cells that we called LSCmed according to their FACS profile (Lin?/Sca-1+/CD49fmed). Here we investigated the prevalence and castration resistance of LSCmed in various mouse models of prostate tumorigenesis. In intact mice, we show that LSCmed prevalence remains low (5-10% of epithelial cells) when prostatic androgen receptor signaling unaltered (malignant Hi-Myc mice) but significantly increases in models exhibiting reduced prostatic androgen receptor signaling, rising up to 30% in premalignant tumors (Pb-PRL mice) and to >80% in castration-resistant prostate tumors driven by Pten loss (Ptenpc-/- mice). LSCmed tolerance to androgen deprivation was demonstrated by their persistence (Ptenpc-/-) or further enrichment (Pb-PRL) 2-3 weeks after castration as evidenced by FACS analysis. Transcriptomic analysis revealed that LSCmed represent a unique cell entity as their gene-expression profile is different from luminal and basal/stem cells, but shares markers of each. Their intrinsic androgen signaling is markedly decreased, which explains why LSCmed tolerate androgen-deprivation. This also enlightens why Ptenpc-/- tumors are castration-resistant since LSCmed represent the most prevalent cell type in this model. We validated CK4 as a specific marker for LSCmed on sorted cells and prostate tissues by immunostaining, allowing for the detection of LSCmed in various mouse prostate specimens. In castrated Ptenpc-/- prostates, BrdU staining revealed massive proliferation of CK4+ cells, further demonstrating their key role in castration-resistant prostate cancer progression. In all, this study identifies LSCmed as a probable source of prostate cancer relapse after androgen deprivation and as a new therapeutic target for the prevention of castrate-resistant prostate cancer.

Dataset Information

Microarray expression profiling data for sorted prostate epithelial subpopulations based on LY6D

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure