Extracellular vesicles (EVs) carry multiple bioactive molecules, including proteins and nucleic acids. Cell-free, extracellular DNA has been reported to be present in cell cultures and in multiple body fluids, but its relative location in relation to EVs has not been described. This study demonstrates that DNase-sensitive nucleic acids are present on the surface of EV isolates. Association of EV-DNA was revealed by increase of EV zeta potential and particle number upon DNase treatment. Additionally, cells exposed to EVs with associated DNA show the presence of cytoplasmic DNA traces intracellularly. In conclusion, we suggest that DNA can be associated to the surface of EVs and can be taken up by recipient cells. DNA on EV surfaces may influence their function in recipient cells.

Extracellular vesicles (EVs) are membrane bound structures released by cells, and found in all body fluids. The populations of EVs are diverse, and the nomenclature includes terms such as exosomes, ectosomes and microvesicles. Previously, it has been shown that the subpopulations of EVs called exosomes can carry both mRNA and microRNA, which can mediate function in recipient cells^[1]. Additionally, dsDNA has previously been proposed to be associated with EVs^{[2][3][4][5][6]}, but its relative location to/in EVs has not been well described. This study reports for the first time that cell-free DNA can be associated with the outside of EVs, which may influence aggregation of these EVs.

This work aims to determine the relative location of cell-free DNA in relation to extracellular vesicles.

On a nucleic acid-stained gel, clear DNA bands were visualized, and these were present not in samples pretreated with DNase-I but in samples treated with RNAse (Fig. 1A). The EVs fraction isolated from the density gradients had a charge of -70 mV, which was increased to -50 mV by DNase-I treatment (Fig. 1B). This result indicates that DNA is present on the outside of the EVs, and this DNA contributes to the negative charge of the vesicles. Further, DNase treatment increased the number of particles measured by nanoparticle tracking analysis, suggesting that the DNA to a certain degree contributes to aggregation of the isolated EVs (Fig. 1C). As a biological readout, isolated EVs associated with extracellular DNA were taken up by human mesenchymal stem cell in a time-dependent manner (Fig. 1D and E).

EVs carry multiple bioactive molecules, including proteins, various RNA species, and according to the present study, DNA. In our case, the DNA was observed in EV isolates from a human mast cell line. Specifically, our study argues that the DNase-sensitive nucleic acids are present on the outside of the EVs. Furthermore, the EV-associated DNA can be taken up by recipient cells, which could alter cellular responses.

It is known that EVs can mediate an array of biological messages to recipient cells- including surface-to-surface antigen presentation and receptors activation- and can deliver RNA (e.g. mRNA, miRNA) cargo to recipient cells^[1][7]. However, the presence and function of DNA as a cargo on the outside of EVs is less explored. EVs have previously been associated with cell-free DNA that carries retrotransposon elements and oncogenes^[3][4], but overall EV-associated DNA has been extensively characterized. A recent report^[5] emphasizes the presence of dsDNA inside of the EVs, whereas we find that a majority of DNA from the human mast cell line is associated with outer perimeter of EVs, since it is sensitive to DNase treatment without lysing the EVs. Our study also indicates, for the first time, that DNA covering floated EVs can lead to an increase in the net negative charge of the vesicles. This was confirmed by reduction of net negative charge from EVs by DNase treatment. We were also able to monitor the increase in particle numbers after DNase treatment, indirectly suggesting that EV-DNA may lead to aggregation of EVs. These results are in line with previous observation^[8][9] made in various electron micrographs, showing clustering of EVs, which could have occurred because of DNA on the surface of these vesicles. The aggregation of EVs may be secondary to nonspecific aggregation of EVs during ultracentrifugation, but our study suggests that EV-related DNA can contribute to this observation. As we observed that the extracellular DNA floated in the density gradient, we argue strongly that it is associated with the floating vesicles with relatively low density. Also, it is has been shown that exogenous plasmids DNA if associated with EVs are taken up more efficiently in recipient cells^[10] than free DNA, again arguing that association to EVs could be involved in DNA uptake. Similarly, we also observed a time-dependent increase in cytoplasmic DNA foci in EV treated recipient human mesenchymal stem cells. Overall, this study highlights the need to define the EV-associated DNA in delivering biological function in cells that take up these EVs.

We conclude that cell-free DNA can be associated with the outside of EVs, which can cause aggregation of these EVs, possibly influencing the effects of EVs in recipient cells.

This work is based on cell-line-produced EVs, which may not reflect a biological function in vivo.

It is not known whether the EVs leave the producing cells with DNA on their surface, or whether that occurs extracellularly. DNA from the cells or supernatant of the cell culture could have been sticking to the released EVs at an early stage of the different isolation steps, which in that case putatively could be considered to be a contamination to the EVs.

In the future, the detailed nature of the EV-associated DNA needs to be determined, in relation to sequences and intracellular origin. Also, the role of the EV-associated DNA in recipient cells may be studied.

EVs were isolated from conditional medium from HMC1.2 cells cultured in medium supplemented with exosomes-depleted FBS^[11]. Cells used had a viability exceeding 99%. We used differential centrifugations to remove cells (300 x g, 10 min), apoptotic bodies and microvesicles together (16,000 x g, 20 min), with a final pelleting of remaining EVs at 120,000 x g for 3 h.

^[12]The EVs were further separated from any other pelleted material using a density gradient (at the 20–24% iodixanol concentration). EVs were treated with DNase against double stranded DNA (DNase-I; Turbo-DNase) and RNase, without lysing the EVs. Later, DNA was isolated from the EVs using Qiamp DNA isolation kits, and visualized by nucleic acid stain after agarose gel electrophoresis. Zeta potential on EVs and total particle numbers were measured per unit of protein using a nanoparticle tracking analysis system (ZetaViewer®, Particle Metrix, Germany). Uptake of DNase treated EVs was visualized by DAPI in recipient human mesenchymal stem cells, using fluorescent microscopy.

This work was funded by the Swedish Cancer Foundation, the Swedish Research Council, Assar Gabrielssons Fond, the Swedish Heart and Lung Foundation as well as VBG-group Centre for allergy and asthma research.

Ethical approval: Not applicable.

No animal was used in the experiments. Therefore, no ethical approval is required.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.