Extracellular vesicle formation in Euryarchaeota is driven by a small GTPase

Abstract

Since their discovery, extracellular vesicles (EVs) have changed our view on how organisms interact with their extracellular world. EVs are able to traffic a diverse array of molecules across different species and even domains, facilitating numerous functions. In this study, we investigate EV production in Euryarchaeota, using the model organism Haloferax volcanii. We uncover that EVs enclose RNA, with specific transcripts preferentially enriched, including those with regulatory potential, and conclude that EVs can act as an RNA communication system between haloarchaea. We demonstrate the key role of an EV-associated small GTPase for EV formation in H. volcanii that is also present across other diverse evolutionary branches of Archaea. We propose the name, ArvA, for the identified family of archaeal vesiculating GTPases. Additionally, we show that two genes in the same operon with arvA (arvB and arvC) are also involved in EV formation. Both, arvB and arvC, are closely associated with arvA in the majority of other archaea encoding ArvA. Our work demonstrates that small GTPases involved in membrane deformation and vesiculation, ubiquitous in Eukaryotes, are also present in Archaea and are widely distributed across diverse archaeal phyla.

Keywords: Archaea; Haloferax volcanii; extracellular vesicles; small GTPase; small RNAs.

Fig. 1.

EV production in H. volcanii DS2. (A) Transmission electron micrograph of EVs. The size bar represents 100 nm. EVs were quantified from the supernatants of cultures (B) grown at different temperatures and (C) from different stages of growth at 28 °C. Each point indicates one biological replicate (n = 3). Error bars indicate the average of three biological replicates ± SD. Temperature-dependent EV production (B) was measured using relative fluorescence units (RFU) normalized to culture OD₆₀₀. Significance is indicated above the graph (NS. indicates “not significant”, * indicates “P ≤ 0.05”). Growth-dependent EV production (C) was quantified by immunodetection measuring the intensity of signals on spot blot (original spot blot in SI Appendix, Fig. S2C), normalized to OD₆₀₀. Growth of cultures indicated in blue follows Left axis, while EV production follows Right axis.

Fig. 2.

RNA composition of haloarchaeal EVs. (A) Analysis of the size distribution of RNA extracted from one replicate of purified EVs and whole cells of H. volcanii and Hbt. salinarum. (B) Expression levels of different RNA subpopulations calculated in percentage from total RNA expression (using TPM values) comparing cellular and EV-associated RNA for H. volcanii (average of three replicates) and Hbt. salinarum (one replicate).

Fig. 3.

EV-associated RNA and proteins. Volcano plots of RNA (A) and protein (B) abundance in EVs in comparison to cellular RNA abundance and protein abundances from cell membranes. Differential RNA abundancies and adjusted P-values were calculated using DESeq2, and only transcripts with TPM > 10 are represented in this plot. Differential protein abundancies and adjusted P-values were calculated with DEP (Methods). Raw data are presented in SI Appendix, Tables S6 and S10. Red asymptotes indicate thresholds for enrichment (P = 0.05 and |fold change| = 2).

Fig. 4.

Analysis of H. volcanii EV-associated GTPase, ArvA. (A) Quantification of EVs in the culture supernatant of the ArvA knockout strain and the respective parental strain. (B) Quantification of EVs in the culture supernatant of a strain overexpressing ArvA (pTA1852-ArvA) compared to control empty vector (pTA1852-empty). EVs were quantified by immunodetection and were averaged over three replicates with error bars denoting one SD from the average value. Original spot blots are presented in SI Appendix, Fig. S11 A and B. (C) Map of the ArvA operon in H. volcanii.

Fig. 5.

Transfer of radioactively labeled RNA by EVs. EVs were isolated from cells (H26 and H26 ΔArvA) that were incubated with radiolabeled uracil, resulting in EVs associated with radiolabeled RNA. EVs were then incubated with nonlabeled wild-type cells and the intracellular radioactivity in decays per minute (DPM) was measured 20 and 90 min postincubation, and normalized by subtracting background radiation (~15 DPM). Significance was calculated using a one-tailed t test (* indicates P < 0.05).

Fig. 6.

The new family of archaeal vesiculating GTPases, ArvA. (A) Unrooted phylogenetic tree of the identified ArvA homologs across the archaeal domain. The red arrow indicates position of H. volcanii ArvA. Blue dots represent branches with bootstrap value greater than 95. Structural prediction of tertiary structure of the ArvA dimer (monomer depicted in green) with (B) closed and (C) open conformations [AlphaFold v2 (41, 42)]. The modeled GDP ligand (displayed as balls), comes from the distant structural homolog EngA from Thermus thermophilus HB8 (rmsd of 3.29-Å out of 69 C-alphas, PDB 2DYK). Hydrophobic residues on N-terminal α-helix (displayed in yellow) are highlighted as balls.