Abstract Background The mechanisms driving the development of cardiovascular injury in type 2 diabetes (T2D) remain incompletely understood. We recently demonstrated that red blood cells (RBCs) from patients with T2D (T2D-RBCs) act as mediators of endothelial dysfunction, but the mechanisms underlying this interaction are not clarified. We found that RBCs via upregulation of arginase 1 attenuate nitric oxide bioavailability and endothelial function. It is increasingly clear that extracellular vesicles (EVs) are actively secreted by practically all cell types, including RBCs, and represent a novel mechanism of intercellular communication. Purpose This study aimed to determine whether EVs derived from T2D-RBCs are involved as mediators in vascular injury through the signalling of arginase 1. Methods T2D-RBCs and RBCs from age-matched healthy controls (H-RBCs) were isolated and incubated with Krebs-Henseleit buffer (20% haematocrit). Following 18h incubation, the conditioned medium was collected for EV isolation using sequential ultracentrifugation and membrane affinity column. EV concentration was measured by nanoparticle tracking analysis. Aortas isolated from wild-type mice were incubated with EVs derived from T2D-RBCs and H-RBCs for 18h. Endothelium-dependent and -independent relaxations (EDR and EIDR, respectively) of the aortae were evaluated in a wire myograph. The involvement of arginase was investigated by the addition of the arginase inhibitor 2(S)-amino-6-boronohexanoic acid (ABH) either to the 18h co-incubation of EVs with the aortic segments to selectively investigate the contribution of EV-derived arginase (ABH, 10 mM) or to the aortae following the 18h co-incubation to selectively target vascular arginase (ABH, 100 µM). All animal experiments were performed according to the principles of laboratory animal care (NIH Publication no. 85-23 revised 1985) and human procedures according to the declaration of Helsinki and approved by the Swedish Ethical Review Authority. Results T2D-RBCs had a ten times lower EV concentration compared to H-RBCs. T2D-RBC-derived EVs isolated using both methods significantly impaired EDR (Fig. 1A-C), whereas EIDR was not affected. This effect was observed irrespective of if the same volume or concentration of EVs were administered. Inhibition of arginase with ABH during the co-incubation with EVs derived from T2D-RBCs completely prevented the impairment of EDR induced by the EVs (Fig. 1D). Administration of ABH to the aortae following the co-incubation also attenuated the impairment of EDR. Immunohistochemical staining revealed upregulation of arginase 1 (Fig. 2A-B) but not arginase 2 (Fig. 2C-D) in the vasculature following incubation with EVs from T2D-RBCs. Conclusion EVs derived from T2D-RBCs induce endothelial dysfunction, and arginase 1 derived from these EVs mediates this endothelial dysfunction. These results shed new important light on the mechanism underlying vascular injury mediated by RBCs in T2D.