Table 2.

Further studies on circulating tumor EVs-DNAs (2014–2019).

Samples/Aims	Main Results	Reference
Three prostate cancer (PCa) cell lines. Plasma of human (PCa) patients (n = 4).	Different gDNA fragments in the subpopulations of EVs (Abs, MVs, and EXOs). EV-gDNA could harbor specific gDNA mutations of the parent cells. Plasma EVs also carry double-stranded gDNA with no differences in MVs/EXOs.	[17]
Glioblastoma, PC3 prostate cancer, or U87 cancer cell lines. Plasma of a PCa mouse model; human plasma of mCRPC patients (n = 40).	Large EVs (oncosomes) contain most of the circulating chromatinized DNA (up to 2 Mb). L-EVs from human mCRPC patients contained large-sized dsDNA, covering the entire tumor genome, with reported cancer-specific (MYC/PTEN) genomic alterations.	[18]
Whole blood samples of pancreatic cancer (PDAC) patients (n = 127) and controls.	KRAS mutations were more detectable in exoDNA than in cell-free DNA, but mutant KRAS was also detected in a substantial minority of healthy samples.	[19]
Serum from patients with (PDAC) pancreatic cancer or pancreatic disease and from healthy individuals.	The minimal exosomal DNA used for digital PCR analyses was 0.663 ng. Potential clinical utility of circulating exosomal DNA for identification of KRASG12D and TP53R273H mutations in patients with pancreas-associated pathologies.	[20]
Engineered exosomes from fibroblasts-like mesenchymal cells (iEXosomes).	Compared to liposomes, iExosomes facilitate therapeutic targeting of oncogenic KRAS in pancreatic cancer.	[21]
Bioreactor-based generation of clinical-grade iExosomes.	Large-scale production of clinical-grade iExosomes for targeting KRAS in pancreatic cancer.	[22]
Xenotransplant mouse model of human glioma-cancer stem cells featuring an intact blood–brain barrier (BBB).	The three types of glioma-derived EVs (ABs, MVs, and EXOs) contained gDNA sequences. Some sequences appeared in all EVs, whereas a few sequences appeared exclusively in one type of EVs. All tumor-derived EVs cross the intact BBB and can be detected in the peripheral blood.	[23]
Comparison of circulating cfDNA and EV-DNA, their origins, and their respective advantages and disadvantages for cancer diagnostic.	Mutated cfDNA, more tumor-specific and enriched in smaller fragments, is more efficient for prognosis of late tumor stages. Exosomal gDNA (between 2.5–10 kb) might be a better potential biomarker for early cancer diagnosis.	[24]
An immortalized HeLa cervical cancer (2D) cell culture and a three-dimensional (3D) organoid culture.	The EV secretion dynamics were significantly different for both culture types: 2D culture remains a valuable tool for the search of human cir-tEV-gDNA cancer biomarker, whereas the 3D culture seems more useful for searching cir-tEV-RNA.	[25]
Mechanisms of nuclear content loading to exosomes.	A link between micronuclei (MN) formation and the generation of some specific exosomal loading with gDNA was identified by inducing genomic instability.	[26]
Human mast (HMC-1) cell line and (TF-1) erythroleukemic cell line.	Exosome-enriched small extracellular vesicles (sEVs) were discriminated by a high resolution iodixanol density gradient into two novel heterogeneous EV subpopulations of low density (LD) and high density (HD) with different RNA/DNA EV cargoes.DNA was predominantly localized on the outside or surface of sEVs.	[27]
Human colon (DKO1) and glioblastoma (Gli36) cell lines; normal primary kidney epithelial cells and human plasma.	Necessary reassessment of the “classical” exosome composition and biogenesis: extracellular dsDNA is not associated with exosomes or any other types of small EVs, but with extracellular particles (EPs).	[28]