Commercial cow milk contains physically stable extracellular vesicles expressing immunoregulatory TGF-β

Abstract

Scope: Extracellular vesicles, including exosomes, have been identified in all biological fluids and rediscovered as an important part of the intercellular communication. Breast milk also contains extracellular vesicles and the proposed biological function is to enhance the antimicrobial defense in newborns. It is, however, unknown whether extracellular vesicles are still present in commercial milk and, more importantly, whether they retained their bioactivity. Here, we characterize the extracellular vesicles present in semi-skimmed cow milk available for consumers and study their effect on T cells.

Methods and results: Extracellular vesicles from commercial milk were isolated and characterized. Milk-derived extracellular vesicles contained several immunomodulating miRNAs and membrane protein CD63, characteristics of exosomes. In contrast to RAW 267.4 derived extracellular vesicles the milk-derived extracellular vesicles were extremely stable under degrading conditions, including low pH, boiling and freezing. Milk-derived extracellular vesicles were easily taken up by murine macrophages in vitro. Furthermore, we found that they can facilitate T cell differentiation towards the pathogenic Th17 lineage. Using a (CAGA)12-luc reporter assay we showed that these extracellular vesicles carried bioactive TGF-β, and that anti-TGF-β antibodies blocked Th17 differentiation.

Conclusion: Our findings show that commercial milk contains stable extracellular vesicles, including exosomes, and carry immunoregulatory cargo. These data suggest that the extracellular vesicles present in commercial cow milk remains intact in the gastrointestinal tract and exert an immunoregulatory effect.

Fig 1. Characterization of bovine milk-derived extracellular vesicles.

(A) Size distribution of isolated EVs observed in a NanoSight LM12. Raw data was analyzed with NTA software, with a minimum expected particle size of 50nm. At least 200 tracks had to be analyzed per sample for inclusion in final analysis. Data presented is a representation of >15 samples in multiple experiments. (B) Electron microscopy of the ultracentrifugation pellet from milk showed both exosomes with spherical shapes (30–100nm) and larger EVs ranging from 100–200nm. (C) Exosome capture assay, exosomes were captured with an anti-CD63 antibody, prior to elution and validation by NTA. (D) Detection of bovine specific RNA in EVs. β -Casein, β-Lactoglobulin (β LG) and elongation factor-1α (EF1α) were detected by RT-qPCR in both colostrum and commercial milk. (E) Detection of immunoregulatory miRNAs in EVs, including miR-21, miR-30a, miR-92a, miR-99a and miR-223.

Fig 2. Milk-derived EVs are stable under degrading conditions.

Extracellular vesicles (100μl, 200μg/ml) were acidified (pH = 2), boiled (15 minutes at 105°C) and frozen (liquid nitrogen) to determine their stability. Stability was assessed using Nanoparticle Tracking Analysis, by comparing particle size and concentration to untreated vesicles. (A) Milk-derived EVs showed no significant differences in concentration. (B) As a control, EVs isolated from macrophage culture supernatant were analyzed. The concentration of these vesicles was significantly reduced after all treatments. Statistically significant differences were determined by Mann-Whitney test, *p<0.05. Error bars represent mean ± S.D. (N = 4).

Fig 3. Cellular uptake of EVs in murine cells.

Macrophages were incubated with PKH67-labeled milk-derived EVs (green) for various time points 37°C or 4°C as control for active uptake. (A) Images were obtained on a Leica fluorescent microscope and represent three separate experiments (magnification 400x). (B) Flow cytometric analysis for PKH67-staining (FITC wavelength) in macrophages was performed. The MFI of two separate experiments (performed in duplo) was averaged. (C) Flow cytometric analysis for PKH67-staining in RAW 264.7 macrophages and NIH 3T3 fibroblasts, washed with either PBS or citric acid to remove surface bound EVs and in primary adherent splenocytes. ΔMFI was corrected for unstained EVs in culture. (D) Confocal microscopy confirmed intracellular uptake of EVs, membranes were stained with F4/80 (red) (magnification 2000x). N/D means not done. Error bars represent mean ± S.D. (N = 3).

Fig 4. Active TGF-β is present on milk-derived extracellular vesicles.

Milk EVs were added to NIH 3T3-cells transduced with CAGA₁₂-luc construct. After 20h of culture, the reporter cells were lysed and the luciferase activity was measured. (A) A dose response of pSmad3/4-signaling was induced by the incubation with milk EVs. (B) Standard curve using recombinant rhTGF-β 1. (C) Blocking active TGF-β using an anti-TGF-β 1,2,3 antibody (2,5 μg/ml) abolished the induction of Smad-signaling by milk EVs. (D) Induction of Th17 differentiation, facilitated by TGF-β (1ng/ml) as positive control or milk EVs (400μg/ml), measured by the increased expression of ROR-γT and IL-17 mRNA compared to stimulation in absence of both N = 4 mice). (E) Blocking active TGF-β using an anti-TGF-β1,2,3 antibody (5μg/ml) inhibits Th17 differentiation (N = 4 groups, 3 mice in each group). Statistically significant differences were determined by Mann-Whitney test, *p<0.05, **p<0.01, ***p<0.005. Error bars represent mean ± S.D.