Endurance-dependent urinary extracellular vesicle signature: shape, metabolic miRNAs, and purine content distinguish triathletes from inactive people

Abstract

Extracellular vesicles (EVs) enriched with bioactive molecules have gained considerable attention in nanotechnology because they are critical to intercellular communication while maintaining low immunological impact. Among biological matrices, urine has emerged as a noninvasive source of extracellular-contained liquid biopsy, currently of interest as a readout for physiological adaptations. Therefore, we aimed to evaluate chronic adaptations of endurance sport practice in terms of urinary EV parameters and evaluated by food consumption assessment. Two balanced groups of 13 inactive controls vs. triathlon athletes were enrolled; their urinary EVs were obtained by differential ultracentrifugation and analyzed by dynamic light scattering and transmission electron and atomic force microscopy. The cargo was analyzed by means of purine and miRNA content through HPLC-UV and qRT-PCR. Specific urinary EV signatures differentiated inactive versus endurance-trained in terms of peculiar shape. Particularly, a spheroid shape, smaller size, and lower roughness characterize EVs from triathletes. Metabolic and regulatory miRNAs often associated with skeletal muscle (i.e., miR378a-5p, miR27a-3p, miR133a, and miR206) also accounted for a differential signature. These miRNAs and guanosine in urinary EVs can be used as a readout for metabolic status along with the shape and roughness of EVs, novel informative parameters that are rarely considered. The network models allow scholars to entangle nutritional and exercise factors related to EVs' miRNA and purine content to depict metabolic signatures. All in all, multiplex biophysical and molecular analyses of urinary EVs may serve as promising prospects for research in exercise physiology.

Keywords: EVs; Guanosine; Microscopy; Physical exercise; Urine; miRNA.

Fig. 1

Schematic steps for EV extraction from urine

Fig. 2

Results of urinalysis. ACR: albumin-to-creatinine ratio. One point of albumin and 8 points of protein data were under the detection level of 5 mg/L and 68 mg/L, respectively

Fig. 3

EV average size distribution: the figure reports the comparison between the average size distributions of extracellular vesicles collected by IN controls and TR. In both cases, a broad distribution has been detected by DLS measurements, with a significant difference in the distribution profile. Large vesicles (>100 nm of diameter) were predominant for IN people, where two peaks of comparable intensity can be observed at ~100 nm and ~200 nm. Instead, small vesicles (<40 nm of diameter) dominated the distribution for TR, with a tail extended above 100 nm

Fig. 4

EV shape morphology: TEM analysis showed different shapes of EVs comparing IN (upper panels) to TR (lower panels). The upper panel reports three images of EVs collected by inactive people, while the lower panel reports three images of EVs collected by triathletes. The number of analyzed macrovesicles is very low due to the low EV concentration of the solution used for the grid preparation. In addition, with any treatment aimed to increase the contrast of biologic materials, such as exosomes, only large vesicles can be detected

Fig. 5

The images derived from AFM scanning and are representative of EV topography of IN (A and A’) and TR (B and B’). Panels A and B represent the original scanning area while A’ and B’ represent scanning after applying the mask to isolate vesicles on which analysis of roughness have been performed and reported in Table 6

Fig. 6

Levels of muscle-specific miRNAs in uEVs. miRNA levels are expressed as mean ± standard deviation (SD) of threshold cycle (Ct) values obtained from 28 subjects (A). DeltaCt (ΔCt) values of microRNAs differentially expressed in EVs between IN males and TR male athletes (B). Bars represent mean ± SD values of ΔCt per sample category. Asterisks denote differential expression p values: (*) <0.05

Fig. 7

Muscle-specific miRNA expression: the violin plots show the median, mean (big dots), IQR, and all data points of 10 miRNAs analyzed (miR23a-3p has not been analyzed with this method due to the violation of assumptions of normality and homogeneity). The scale factor for the Bayesian test has been set as r = 0.707, assuming a Cauchy prior distribution; Bayes factors (BF) are reported on a logarithmic scale as log_e(BF₀₁).

Fig. 8

Network analysis: the figure shows the network analysis connecting food intake, anthropometric data, purines, and miRNA content of uEVs in the TR (triathlon) and IN (control) groups. It should be noted that in the triathlon group there were less edges and that some of them were related to negative correlations among variables