GEOapplication/xmlftp://ftp.ncbi.nlm.nih.gov/geo/series/GSE293nnn/GSE293911/primaryOK200TranscriptomicsHomo sapiensExpression profiling by high throughput sequencinghttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE293911GEOGSEfalseA Unique System of Paired PDX Models to Investigate the Progression of Potentially Lethal to Lethal Prostate Tumors at a Multi-omic LevelProstate cancer (PCa) exhibits substantial heterogeneity in aggressiveness among patients, with significant lethality in its advanced metastatic stages. Defining the molecular mechanisms underlying rapid disease progression remains a pressing unmet need. Here, we present a comparative analysis of a longitudinally paired patient-derived xenograft model of aggressive PCa progression before and after therapeutic intervention. Using multi-omic analysis, we characterized the MDA PCa 177-B (androgen receptor [AR]-negative, basal) and MDA PCa 189-1 (AR-positive, luminal) models, which exhibit distinct transcriptomic profiles yet originate from a shared clonal lineage. Targeted and single-cell DNA sequencing revealed that these models share 71.5% of copy number variations, tracing back to a common ancestral cell population. Targeted bisulfite sequencing of both models revealed epigenetic marks unique for aggressive disease compared to normal/benign prostate. Next, through single-cell RNA sequencing, we identified a subset of cells exhibiting a low differentiation profile, present in both PDX models and characterized by a unique molecular signature associated with adaptive immune responses and metastasis-related pathways. Further integrative analysis using data from a broad spectrum of disease samples demonstrated that, within this pair of aggressive PDX models, these low-differentiation clusters are exclusive to stem-like tumor cell populations. These findings support the hypothesis that a continuum of epigenetic, transcriptomic, and phenotypic transitions occurs during the progression of PCa from its potentially lethal to lethal stages. We suggest that this continuum can be captured through well-characterized, longitudinal PDX models derived from the same patient. Our multi-omic approach underscores the value of such models in uncovering the dynamic and adaptive mechanisms of PCa aggressiveness.2026/04/05GSE293911GSM8893244GSM889324524676293911Homo sapiens