Journal: bioRxiv
Article Title: Pancreatic cancer extracellular vesicles carry a time-of-day-regulated miRNA cargo that disrupts the skeletal muscle clock and bioenergetics
doi: 10.64898/2026.05.03.722338
Figure Lengend Snippet: (A) Clinical and molecular characterization of six pancreatic cancer patients (n=6) stratified by unsupervised hierarchical clustering of serum sEV miRNA expression profiles. Three clusters were resolved: Cluster 1 (P1, P3, P5; red), Cluster 2 (P4; blue), and Cluster 3 (P2, P6; green). Diagnoses are indicated above each patient identifier: PDAC, pancreatic ductal adenocarcinoma; NET, neuroendocrine tumor; Mesoth., malignant mesothelioma of the pancreas. Clinical parameters shown include age at diagnosis (dot plot, years), number of prior treatment lines (bar chart, top right), and comorbidity burden (bar chart, bottom right). The Jaccard similarity index between within-cluster patient pairs (CPM>0) is plotted for Clusters 1 and 3, demonstrating high intra-cluster miRNA-profile similarity (range 0.588–0.644). A Venn diagram illustrates the overlap between miRNAs detected in patient serum sEVs (T=428) and in PANC-1 sEVs across 9 time-points (T=1,426); 359 miRNAs are shared between the two datasets. Bar charts indicate the percentage of cluster-specific patient miRNAs detectable in the PANC-1 sEV secretome: Cluster 1, 84.6%; Cluster 2, 84.2%; Cluster 3, 83.9%. ( B ) UpSet plot (top) displaying the combinatorial overlap of miRNAs detected in each patient’s serum sEVs. Vertical bars indicate the number of miRNAs shared by each intersecting patient combination defined in the connected dot matrix below; horizontal bars on the right show the total number of miRNAs detected per patient. Below the UpSet plot, a heatmap shows expression levels (log2(CPM+1)) of the 11 candidate miRNAs functionally validated in this study across all six patients. Rows are ordered by hierarchical clustering of miRNA expression profiles; columns are arranged by patient cluster assignment (Cluster 1, red; Cluster 2, blue; Cluster 3, green). Color scale ranges from blue (low expression) to dark red (high expression). ( C ) Frequency of the 11 candidate miRNAs among the top 50 most highly expressed miRNAs in pancreatic cancer tumor specimens. For each of 495 tumor samples from the GDC data portal (TCGA-PAAD, CPTAC-3, and HCMI-CMDC), miRNAs were ranked by RPM, and a given miRNA was classified as highly expressed if any of its corresponding precursors ranked within the top 50. Bars show the percentage of samples meeting this criterion; absolute sample counts are given in parentheses. ( D ) Proposed mechanistic model. Within the PDAC tumor cell, a dysregulated clock (BMAL1↓, altered rhythmic output) drives miRNA sorting into EVs during biogenesis; miR-27b-3p is uniquely (among those tested) and rhythmically packaged, while other sEV miRNAs are loaded constitutively. Secreted sEVs exhibit variable particle number with stable size, and these miRNAs are detectable in patient serum (84– 85% overlap with PANC-1 sEV miRNAs). In the recipient skeletal muscle cell, the full sEV miRNA cargo collectively disrupts the circadian clock, altering period, phase, and amplitude. In parallel, individual miRNAs reprogram bioenergetics along four non-redundant trajectories: Energetic (miR-27b-3p + others; ↑OCR, aerobic/oxidative, ↑spare respiratory capacity); High metabolic (miR-191-5p + others; ↑OCR, ↑ECAR, high respiratory capacity); Quiescent (others; ↓OCR, ↓ECAR, low metabolic capacity); and Glycolytic (miR-183-5p + others; ↓OCR, ↑ECAR, glycolytic drift). The integration of circadian disruption, bioenergetic reprogramming, and proteostatic dysregulation collectively drives muscle-cell atrophy in vitro ; whether and how these molecularly distinct insults produce a systemic cachexia phenotype in vivo remains to be determined (gray box).
Article Snippet: The human pancreatic cancer cell line PANC-1, the murine fibroblast cell line NIH3T3, and the murine myoblast cell line C2C12 were purchased from American Type Culture Collection (ATCC, Manassas, VA).
Techniques: Expressing, Biomarker Discovery, Disruption, In Vitro, In Vivo