Novel Strategy for Acquiring Metabolically-Tagged Nascent Extracellular Vesicles: Implications for Identifying Surface Protein Markers of Extracellular Vesicles From Neuroblastoma Cells Cultured With Native Serum

Authors

Markus HP

de Jong E

Makridakis M

Frantzi M

Koçer A

Doi

10.1002/jev2.70177

PMID: 22367537 · DOI: 10.1002/jev2.70177 · Journal: Journal of Extracellular Vesicles (2025)

TL;DR

This study developed a metabolic labelling strategy using the sugar analogue ManNAz and copper-catalysed click chemistry to selectively isolate tumour-derived extracellular vesicles (tdEVs) from neuroblastoma cell lines cultured in native serum, bypassing the need for EV-depleted or serum-free conditions. Mass spectrometry-based proteomic profiling of isolated EVs identified gap junction protein GJC1 (connexin 45) as a potential surface marker for neuroblastoma-derived EVs, with high GJC1 expression correlating with poor overall survival across three independent transcriptomic datasets of primary neuroblastoma tumours.

Cohort & data

  • Three neuroblastoma cell lines: SH-EP2 (mesenchymal, MYCN-non-amplified), SH-SY5Y (adrenergic, MYCN-non-amplified), and Kelly (adrenergic, MYCN-amplified).
  • Proteomic analysis by label-free LC-MS/MS of EVs isolated under three conditions: native serum with ManNAz (Nat-ManNAz), EV-depleted serum with ManNAz (EVd-ManNAz), and EV-depleted serum with DMSO (EVd-DMSO).
  • Survival analyses used three publicly available transcriptomic datasets: Kocak (GSE45547; n = 476), Versteeg (GSE16476; n = 88), and NREC (GSE85047; n = 220).
  • DepMap CRISPR knock-out viability screens across >1100 cancer cell lines (37 neuroblastoma lines) used for dependency analysis.

Key findings

  • ManNAz metabolic labelling did not substantially impair cell viability (p = 0.375, repeated measures ANOVA) and did not alter EV proteomic composition (94.2% overlap of 787 proteins between ManNAz and DMSO conditions; Pearson r = 0.998 for shared human-specific proteins).
  • Selective isolation of ManNAz-labelled EVs was confirmed by western blot for EV markers (CD63, CD9, ALIX, syntenin-1, HSPA8) with undetectable background in DMSO controls.
  • Proteomic analysis of EVs from native serum yielded 510 protein detections (69 human-specific), with strong enrichment for the mesenchymal (MES) transcriptional signature (19/69 proteins, Bonferroni p = 7.63 x 10^-15), consistent with the MES-state of the SH-EP2 cell line.
  • EV proteomes were enriched for proteins involved in cell migration and wound healing, consistent with the metastatic origin of SH-EP2.
  • GJC1 was identified as a potential neuroblastoma EV surface marker through a stepwise filtering procedure: 724 unique proteins narrowed to 16 candidates based on transmembrane domain annotation, absence from healthy plasma/serum EV proteomes, and absence from EV proteomes of major blood cell types.
  • High GJC1 expression correlated with poor overall survival in all three transcriptomic datasets (Bonferroni-corrected p-values shown in Kaplan-Meier analyses) and poor event-free survival.
  • CRISPR dependency screening showed neuroblastoma cells are selectively dependent on GJC1 for survival and growth compared to other cancer types (p = 1.62 x 10^-14).
  • GJC1 was detected in EVs from all three neuroblastoma cell lines (SH-EP2, SH-SY5Y, Kelly), regardless of transcriptional program (MES vs. adrenergic) or MYCN amplification status, but was not detected in EVs from blood plasma of three healthy adult donors.

Genes & alterations

  • GJC1 (connexin 45): Identified as a potential pan-neuroblastoma EV surface marker. Located on chromosome 17q21.31, a region gained in ~50% of neuroblastoma. High expression correlated with poor OS and EFS. Selectively essential for neuroblastoma cell survival in CRISPR screens.
  • MYCN: Cell lines studied included MYCN-amplified (Kelly) and MYCN-non-amplified (SH-EP2, SH-SY5Y) lines. GJC1 expression in EVs was independent of MYCN amplification status. GJC1 was overexpressed during tumour formation in a TH-MYCN mouse model.

Clinical implications

  • GJC1 may serve as a surface marker for immunocapture-based enrichment of circulating neuroblastoma-derived EVs in liquid biopsies, enabling non-invasive tumour monitoring.
  • The mesenchymal transcriptional program, associated with therapy resistance and enriched in post-treatment/recurrent tumours, was strongly reflected in the EV proteome, suggesting EV profiling could inform on tumour transcriptional state for patient stratification and treatment monitoring.
  • The metabolic labelling methodology provides a framework for identifying tumour-specific EV surface markers in other cancer types without requiring prior knowledge of biomarkers.

Limitations & open questions

  • Validation was performed in cell lines only; well-powered clinical studies with patient samples are needed to establish GJC1 utility as a biomarker.
  • Developing antibodies against GJC1 extracellular epitopes is challenging due to small extracellular loops and high sequence similarity across the 21 human connexin isoforms.
  • GJC1 expression is not exclusive to tumour cells; specificity in the EV context relies on its absence from healthy blood-circulating EVs.
  • The functional role of GJC1 in neuroblastoma development and progression remains unexplored.
  • Healthy donor validation was limited to three adult donors; paediatric controls were not included in the western blot validation.
  • Native serum culture yielded lower protein coverage than EV-depleted serum, possibly due to sample handling losses or differential EV secretion rates.

Citations from this paper used in the wiki

  • “We proposed the gap junction protein GJC1 (connexin 45) as a potential biomarker, the high expression of which correlates with poor overall survival, exhibits selective dependence for growth and survival in neuroblastoma, and is overexpressed during early neuroblastoma development.” (Discussion, p. 15-16)
  • “GJC1 was readily detected in EVs from three neuroblastoma cell lines with different transcriptional programs (adrenergic vs. mesenchymal) and MYCN status.” (Discussion, p. 16)
  • “The mesenchymal transcriptional program of the parental cells was strongly reflected in the EV proteome. Mesenchymal neuroblastoma cells are characterized by increased therapy resistance and were enriched in post-treatment and recurrent tumours.” (Discussion, p. 16)
  • “We detected GJC1 in EVs from all three cell lines, indicating that its expression and sorting into EVs is not exclusive to either epigenetic program (mesenchymal vs. adrenergic) or MYCN status.” (Results, section 3.9)

This page was processed by paper-compiler on 2026-05-06.