Type I collagen hydrogels as a delivery matrix for royal jelly derived extracellular vesicles

Abstract

Throughout the last decade, extracellular vesicles (EVs) have become increasingly popular in several areas of regenerative medicine. Recently, Apis mellifera royal jelly EVs (RJ EVs) were shown to display favorable wound healing properties such as stimulation of mesenchymal stem cell migration and inhibition of staphylococcal biofilms. However, the sustained and effective local delivery of EVs in non-systemic approaches - such as patches for chronic cutaneous wounds - remains an important challenge for the development of novel EV-based wound healing therapies. Therefore, the present study aimed to assess the suitability of type I collagen -a well-established biomaterial for wound healing - as a continuous delivery matrix. RJ EVs were integrated into collagen gels at different concentrations, where gels containing 2 mg/ml collagen were found to display the most stable release kinetics. Functionality of released RJ EVs was confirmed by assessing fibroblast EV uptake and migration in a wound healing assay. We could demonstrate reliable EV uptake into fibroblasts with a sustained pro-migratory effect for up to 7 d. Integrating fibroblasts into the RJ EV-containing collagen gel increased the contractile capacity of these cells, confirming availability of RJ EVs to fibroblasts within the collagen gel. Furthermore, EVs released from collagen gels were found to inhibit Staphylococcus aureus ATCC 29213 biofilm formation. Overall, our results suggest that type I collagen could be utilized as a reliable, reproducible release system to deliver functional RJ EVs for wound healing therapies.

Keywords: Apis mellifera; Wound healing; drug delivery; extracellular vesicle delivery; regenerative medicine.

Figure 1.

Characterization of RJ EVs; (A) Transmission electron micrograph of RJ EVs; scale bar = 200 nm; (B) representative NTA histogram of particle distribution; (C) NTA analysis of median particle size in nm; mean ± SD; n = 5.

Figure 2.

Release kinetics of type I collagen and RJ EVs; (A) NTA analysis of particles released into the surrounding medium from 1 mg/ml (light gray), 2 mg/ml (dark gray), and 3 mg/ml (black) collagen gels on day 1, 3, and 7; Concentrations displayed as particles/ml; (B) Cumulative RJ EVs released from 1 mg/ml, 2 mg/ml, and 3 mg/ml collagen gels after 7 d; data displayed as percentage of initial RJ EVs integrated into the gels; (C) representative NTA histogram of Ctrl (PBS incubated with collagen without RJ EVs) and RJ EVs released from collagen gels; (D) Median size of RJ EVs released from 1 mg/ml, 2 mg/ml, and 3 mg/ml collagen gels on day 1, 3, and 7 and prior to integration into collagen gels (control); A, B, D: n = 5; mean ± SD; statistics described in methods.

Figure 3.

AFM imaging of 2 mg/ml collagen gels with and without RJ EVs; 3D height (5 × 5 μm) and amplitude (2 × 2 μm; scale bar = 400 nm) images of control collagen and collagen loaded with (2.5 × 10⁹/ml) RJ EVs, imaged in AC mode; Arrows indicate RJ EVs released from the collagen gel.

Figure 4.

Integration of RJ EVs into 3T3 L1 cells; Confocal imaging of 3T3 L1 cells on day 1, 3, and 7 after colocalization with Transwells containing collagen gels with CFSE-ELVs (Col-EV); RJ EV displays control cells receiving CFSE RJ EVs 24 h before analysis; ctrl was not incubated with RJ EVs; nuclei stained with Hoechst; Arrows indicate EVs within cells; scale bar = 100 μm (upper panels per group) and 25 µm (lower panels per group).

Figure 5.

RJ EVs released from collagen gels promote HdnF migration and decrease proliferation; (A) Quantitative analysis of scratch closure (4, 8, 12, and 24 h); fibroblasts colocalized with collagen gels with and without RJ EVs compared to ctrl; data displayed as percentage scratch closure compared to 0 h; (B) Representative phase contrast micrographs at time point 0 h and 24 h; (C) Schematic representation of the experimental setup: fibroblasts were preconditioned for 24 h with RJ EVs released from collagen gels, subsequently a scratch was inflicted and cellular migration was assessed after 4, 8, 12, and 24 h; (D) Scratch closure of collagen gels with (dark gray) and without (light gray) RJ EVs after 24 h, as percentage of migration ctrl; (E) proliferation of HdnF evaluated with BrdU assay in presence of collagen gels with (dark gray) and without RJ EVs (light gray), measured at 450 nm; data displayed as percentage of control; A, C: n = 4; E: n = 5; mean ± SD; statistics described in methods.

Figure 6.

RJ EVs increase contractile behavior of HdnFs. Upper panel: Area of collagen gels containing 1.5 × 10⁶ HdnFs per ml, 24 h after gel polymerization. Negative control (neg. ctrl) was incubated with 0% FBS, positive control (pos. ctrl) in 20% FBS, control (ctrl), RJ EV preconditioned HdnFs (precon.) and RJ EV in collagen gel (RJ EV) in 5% FBS; n = 6; * statistical difference to ctrl; # statistical difference to neg. ctrl; and statistical difference to pos. ctrl; mean ± SD statistics described in methods; lower panel representative images of collagen gels after 24 h incubation;.

Figure 7.

RJ EVs released from collagen gel inhibit S. aureus biofilm formation. (A) Schematic representation of experimental setup. (B) Numeric results for crystal violet biofilm staining, expressed as percentage of positive control, for collagen gels with and without RJ EVs (n = 4, mean + SD; statistics described in methods. (C) Representative image of crystal violet S. aureus biofilm staining.