Fig 2: L1CAM+ EVs contain a distinct and interesting protein signature. Total EVs were isolated from plasma from four healthy volunteers and subsequently immunoprecipitated with L1CAM or EpCAM antibodies. The samples from the four subjects were then pooled to generate a sufficient amount of protein for five membrane antibody arrays [Human Obesity Antibody Array (A), Human Phospho-Kinase Antibody Array (B), MAPK Antibody Array (C), Human Kidney Biomarker Antibody Array (D), and Human Apoptosis Antibody Array (E)]. An equivalent amount of protein per sample type was loaded to each spot. The levels of different proteins were calculated by densitometric analysis and normalized to the reference proteins present on each membrane. Results are depicted as heatmaps ranging from zero to maximum value and scaled according to the color bars.

Fig 3: Neuronal EVs are found in the circulation. (A) Total EVs and L1CAM+ EVs were isolated from plasma of nine healthy volunteers and their size distribution was examined by NTA using Nanosight-NS500. Figure depicts concentration after 1:1,000 dilution for Total EVs and 1:200 dilution for L1CAM+ EVs; actual concentrations are depicted in (B). (B) The graph shows the actual concentration of total EVs and L1CAM+ and CD81+ plasma EVs immunoprecipitated from plasma of nine healthy volunteers after adjusting for dilution. The percentages in the graph represent the ratio over total EVs. (C) Western blot image shows enrichment of neuronal markers (L1CAM, MAP-2, N-enolase, p-Tau235, and TUJ) in L1CAM+ EVs when compared to total EVs from a single healthy control. CD9 is a common exosomal marker present in EVs but not in the mouse brain lysate used as a positive control. An equivalent amount of EVs was loaded on the gel by adjusting the dilution of the isolates according to the EV concentration determined by NTA. (D) GFP levels evaluated by FACS; L1CAM+ EVs were isolated from 300 μl plasma derived from Nestin-GFP transgenic or WT mice. The EVs were conjugated with the beads and the levels of GFP were evaluated by FACS analysis. The results show the percentages of beads-antibody-EV complex that contained GFP above the detection threshold. (E) GFP levels evaluated by fluorescence; comparison between the levels of GFP in EVs in the samples described in (D) were measured by plate reader at excitation of 485 nm and emission 515 nm.

Fig 4: L1CAM+ EVs offer a higher detection level for p-tau, BDNF and pro-BDNF over plasma, serum and total EVs. For p-tau comparisons, total EVs were isolated from four plasma and serum samples from healthy volunteers followed by L1CAM immunoprecipitation. The levels of p-tau-Thr181 (A) and p-tau-Thr231 (B) are presented in the graph in L1CAM+ EVs, total EVs, plasma, serum, and in comparison, to the background signal (blank). Column bars represent the mean of four samples, error bars represent SEM. For BDNF and proBDNF comparisons, total EVs were isolated from 20 plasma samples from healthy volunteers followed by L1CAM immunoprecipitation. BDNF levels (C) are different depending on the type of fluid tested [F(2, 57) = 6.868, p = 0.002]; its levels are higher in L1CAM + EVs compared to plasma (p = 0.001) and total EVs (p = 0.016), whereas its levels in total EVs were no different than plasma (p = 0.254). Similarly, proBDNF levels (D) are dependent on the type of fluid tested [F(2, 57) = 4.41, p = 0.017]; its levels are higher in L1CAM + EVs compared to plasma (p = 0.007) and total EVs (p = 0.026), whereas its levels in total EVs were no different than plasma (p = 0.628).

Fig 5: Immuno Electron Microscopy images of L1CAM+ and CD81+ EVs. L1CAM+ or CD81+EVs were incubated with primary human anti-L1CAM or CD81 antibody followed by a secondary antibody conjugated with 6 nm gold particle. Scale bar = 100 nm, microscope settings; 120 kV, magnification 200,000x except the L1CAM+ labeled with anti-L1CAM (top left) which used a magnification of 160000x.