MeHA synthesis

High molecular weight hyaluronic acid sodium salt (HAs, Mw = 1.5–1.8 106 Da from Streptococcus equi, Sigma Aldrich, Milan, Italy) was modified to graft photoactive polymerizable groups by reacting with methacrylic anhydride (Me, Sigma Aldrich Milan, Italy) as previously described [31, 65, 66]. Briefly, 1 g of HAs was dissolved in 10 mL of pure water (H2O, Carlo Erba, Cornaredo, Italy) and stirred at room temperature (RT) for complete dissolution. MeHA was obtained by reacting the primary hydroxyl groups (-OH) with Me at 4 °C, keeping the pH between 8 and 9 using a sodium hydroxide solution (NaOH, Sigma Aldrich, Milan, Italy). An excess of 30 mol% ME per (-OH) was used. The reaction was carried out for 12 h and it was stopped by precipitating MeHA into cold anhydrous ethyl alcohol (EtOH, Sigma Aldrich, Milan, Italy). The supernatant was recovered by vacuum filtration. The isolated MeHA polymer was solubilized in pure water then dialyzed against distilled water for 5 days and freeze-dried (LaboGene’s CoolSafe 55 − 4 PRO, Bjarkesvej, Denmark).

1 H nuclear magnetic resonance (NMR)

1 H Nuclear magnetic resonance (NMR, Bruker AVIII 400HD, Fällanden, Swiss) was employed to assess the success of the functionalization reaction. MeHA (5 mg/mL) was completely dissolved in deuterium oxide (D2O) by using a vortex mixer, and it was transferred into NMR tubes. The data were collected at a frequency of 400 MHz. Phase and baseline corrections were applied before obtaining the areas (integrals) of purely absorptive peaks. The presence of methacrylate moieties on the HAs backbone was confirmed by the peaks at 5.7 and 6.2 ppm.

3D printing of MeHA

Freeze-dried MeHA was dissolved in diH2O at a concentration of 4% (w/v) containing 0.1% (w/v) 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure 2959, Sigma Aldrich, Milan, Italy), as previously reported [31]. Briefly, 3D printing was performed on “Rokit Invivo 4D2” (Rokit Healthcare Inc., Seoul, Korea) using 1.80 firmware. The input printing model was sliced with a grid pattern using New Creator K 1.57.70. The printing speed was set at 6 mm/s. The dispenser temperature was set at 15 °C, while the bed was set at 0 °C. A needle of 0.6 mm, a layer thickness of 0.4 mm and a fill density of 50% were used to build porous patches measuring 50 mm × 50 mm × 3 mm. During printing, UV light (λ: 365 nm) was used to crosslink the biomaterial ink enhancing the mechanical properties and avoiding the collapse of the structures. After printing, 3D porous structures were also post-crosslinked in a UV cabinet (Analytik Jena UVP crosslinker, CL-1000, λ: 365 nm) for 10 min. Finally, the patches were freeze dried for 24 h and stored at -80 °C before using.

Dynamic mechanical analysis

A-Q800 (TA-Instrument, New Castle, DE, USA) was employed to assess the mechanical properties of 3D printed structures. The frequency was varied between 0.5 and 5 Hz, using an amplitude of 100 µm in compression, a preload of 0.001 N and a force track of 125%. The tests were performed in a closed chamber in a wet state at room temperature. Elastic modulus (E’) is reported as mean value ± SD, n = 5.

Swelling behavior

Dried 3D patches were weighted (w0) scaffolds and placed in 5 mL of sterile medium that was supplemented with antibiotics and allowed to swell under physiological conditions for up to 5 h (pH = 7.4, T = 37 °C). The swollen hydrogels were then removed at specific time intervals, immediately blotted on filter paper to remove the superficially absorbed water, the weight was measured (wt), and the samples were then put back into the solution. Equation (1) was used to calculate the swelling ratio (Q):

$$Q} = }\left( - } \right)/$$

(1)

Results have been reported in Fig. 5G as mean value ± SD, n = 5.

Scanning electron microscopy

The 3D printed MeHA patch was observed by scanning electron microscopy (SEM, FEI Quanta 200 FEG, Hillsboro, OR, United States). Before the analysis, 3D patches were prepared as described before, frozen, lyophilized for 48 h. The lyophilized structure was coated with an ultrathin layer of Au/Pt by using an ion sputter and then observed by SEM.

Incorporation and release of MSC-sEVs from MeHA dressing

The 3D printed MeHA patches were placed into a 24-well tissue culture plate and loaded with 20 or 120 µg of MSC-sEVs resuspended in 100 µL PBS. Control condition (0 µg) was MeHA patches loaded with 100 µL PBS. The release profile was evaluated by placing the loaded dressing in a 24-well plate containing serum-free culture medium for up to 11 days. Then, samples were withdrawn at selected time points and stored at -80 °C until all samples were collected. The release of sEVs were quantified using the BCA Protein Assay Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Data reported as mean value ± SD, n = 3.

MSCs culture in bioreactor

Commercially available MSCs (Human Adipose-derived Mesenchymal Stem Cells; ScienCell Research Laboratories, Inc., CA, USA) were seeded at a density of 5 × 105 cells into handheld 3D bioreactors (VITVO®; Rigenerand Srl, Medolla, MO, Italy) [27] up to 14 days. The bioreactor is a closed system with two optical transparent membranes that allow gas exchange. Internally, a 400 μm matrix of biocompatible polyester separates two chambers. Each chamber has a port acting as input or output depending on the media flow. By input port, liquid can enter the first chamber, pass through the 3D matrix, fill the second chamber, and exit through the outlet port. Thus, the 3D matrix works as a filter that retains cell and as substrate for cell growth. MSCs at passage 3 were resuspended in 1.4 mL of culture medium (DMEM, high glucose, no glutamine, no phenol red; Thermo Fisher Scientific, Waltham, MA, USA) completed with exosome-depleted fetal bovine serum (Capricorn Scientific GmbH, Ebsdorfergrund, Germany) and injected into bioreactors by a syringe through one of the two ports provided. Then, bioreactors were placed in an incubator set at 37 °C and 5% CO2. Fresh medium (1.5 mL) was injected through both ports every 48 h and an equal volume of CM was collected. In 14 days, about 20 mL CM was collected for each bioreactor. At each harvest, CM was centrifuged at 300xg for 10 min at RT to eliminate dead cells, filtered with a 0.22 μm filter to remove cell debris.

Calcein AM staining

The growth and viability of MSCs into bioreactors were detected by staining with 1 µM Calcein AM (Life Technologies, Chagrin Falls, OH, USA) for 20 min at 37 °C and 5% CO2. After removing the excess probe, a z-stack (z-step 2.5 μm) was acquired with a confocal microscopy (Nikon A1 confocal microscope, Nikon Corporation, Tokyo, Japan) equipped with a 10X objective.

Cell flow cytometry

MSCs were incubated with the following fluorescent monoclonal mouse anti-human antibodies: CD44 FITC (BD Biosciences, San Jose, CA, USA); CD73 APC (eBioscienceTM, Thermo Fisher Scientific); CD90 PE (eBioscienceTM); CD105 PE (BD Biosciences); CD14 PE (eBioscienceTM); CD34 FITC (eBioscienceTM); CD45 APC (eBioscienceTM); HLA-DR FITC (eBioscienceTM). After two washes, fluorescent cells were detected with Attune NxT flow cytometer (Thermo Fisher Scientific). Data were analyzed using Attune NxT software (Thermo Fisher Scientific). Experiments were performed in triplicate.

sEV isolation

The clarified CM was loaded onto Amicon Ultra-15 100 kDa centrifugal devices (Amicon Ultra-15 centrifugal ultrafilters with Ultracel-PL PLGC membrane, 100 kDa; Millipore, MA, USA), previously sterilized with 70% ethyl alcohol (Sigma-Aldrich, Saint Louis, MA, USA), and centrifuged at 2000xg for 20 min at 4 °C. The filtrate was washed with sterile PBS (Euroclone, Milan, Italy) through a further centrifugation at 2000xg for 20 min at + 4 °C. The hMSC-SEVs were finally recovered from the filtering unit, quantified by Pierce™ BCA protein assay kit (Thermo Fisher Scientific), and stored in small aliquot at -80 °C immediately.

Nanoparticle tracking analysis

Nanoparticle tracking analysis (NTA) uses laser light scattering and Brownian motion to determine EVs size and concentration. Measurement of particle size and particle size distribution was performed with Nanosight NS300 (Malvern, UK) instrument equipped with a 488 nm laser. All samples were diluted in filtered PBS to a final volume of one mL. Ideal measurement concentrations were found by pre-testing the ideal particle per frame value (20–100 particles/frame). For each measurement, five 1-min videos were captured under temperature 25 °C and syringe pump speed 30. Data are represented as averaged finite track length adjustment (FTLA) concentration / size.

Transmission electron microscopy (TEM)

For TEM acquisition, the protocol described elsewhere was follow [67]. Briefly, the sEVs were fixed in a 2% glutaraldehyde solution in phosphate buffer (ratio 1:1). The sEVs were then deposited, rinsed, and stained with heavy metal compounds onto a gridded slide according to the standard protocols. The slide was visualized with a TEM Zeiss EM 910 instrument (Zeiss, Oberkochen, Germany).

Exosome antibody array

The immunoblotting analysis of sEV specific markers were performed using the commercial Exo-Check™ exosome antibody array (Systems Biosciences, USA) according to the manufacturer’s instructions. The array contains eight known sEV markers, including CD63, CD81, ALIX, FLOT1, ICAM1, EpCam, ANXA5, and TSG101; four controls, including two positive controls (HRP Detection), blank spot (background control) and GM130 cis-Golgi marker, which monitors for any cellular contamination. Briefly, 50 µg of MSC-sEVs was lysed and labeled for 30 min with constant mixing. The labeled samples were washed and blocked with the blocking buffer. Array membrane was incubated with labeled lysate/blocking buffer mixture at 4 °C overnight on a rocker. The next day, the membrane was washed and incubated with a detection buffer for 30 min at RT on a shaker. The membrane was washed and the chemiluminescence was developed with Clarity Western ECL substrate (Bio-Rad, USA). Membrane array was developed on the chemiluminescence imaging system (ChemiDoc, Bio-Rad). Experiments were performed in triplicate.

sEV flow cytometry

sEVs were harvested with exosome-Human CD81 Flow Detection (from cell culture) (Thermo Fisher Scientific), according to the manufacturer’s instructions. In detail, 50 µL of sEV suspension was added to a tube containing 20 µL of CD81 magnetic beads, previously washed with 500 µL of Assay Buffer containing 0.1% bovine serum albumin (BSA, Sigma-Aldrich) in PBS, and incubated at 4 °C overnight under stirring (650 rpm/min). After incubation, the bead-bound sEVs were isolated with the MagnaRack magnetic separator (Thermo Fisher Scientific) and washed twice with of Assay Buffer. Isolated CD81-positive sEVs were then labeled with mouse anti-human CD63 PE (eBioscience™), CD81 PE (BD Biosciences), CD73 APC (eBioscience™), or CD90 PE (eBioscience™). After 1-h incubation at RT protected from light on an orbital shaker (1000 rpm/min), the bead-bound sEVs were washed twice and suspended in Assay Buffer. Two controls were performed: PBS (vehicle) and ultrafiltrated exosome-depleted medium were stained instead of sEVs. Data were collected with Attune NxT flow cytometer (Thermo Fisher Scientific) and analyzed using Attune NxT Software v2.5 (Thermo Fisher Scientific). Experiments were performed in triplicate and the data represent the average.

sEV internalization

sEVs or PBS (negative control) were stained with PKH67 (PKH67 Green Fluorescent, Sigma-Aldrich) for 20 min at 37 °C, as previously described [18]. The excess unincorporated dye was removed from the labeled solutions by using Exosome Spin Columns (MW 3000) (Thermo Fisher Scientific), following the manufacturer’s instructions. Then, 1 × 104 dermal fibroblasts/cm2 (ATCC, MA, USA) or 1 × 104 endothelial cells/cm2 (HUVEC; ThermoFisher Scientific) were incubated with the labeled sEVs or PBS for 3, 6, and 24 h in DMEM (EuroClone) or in Medium 200 PRF (M200PRF, ThermoFisher Scientific) without supplements), respectively. After incubation, cells were washed, and the nuclei were stained with Hoechst 33,342 (ThermoFisher Scientific) for 10 min at RT. Finally, cells were fixed with 4% paraformaldehyde, mounted with ProLong™ Glass Antifade Mountant (Thermo Fisher Scientific), then observed with a laser scanning confocal microscopy system (Nikon A1 confocal microscope, Nikon Corporation, Tokyo, Japan) equipped with a 63X objective. Experiments were performed in triplicate.

RNA isolation, cDNA synthesis and real-time PCR

Total RNA was isolated from MSC-sEVs-treated cells with the RNeasy Mini Kit (Qiagen, Hilden, Germany). Total RNA from skin biopsies with RNeasy Fibrous Tissue Mini Kit (Qiagen, Hilden, Germany). The RNA quality and concentration of the samples was measured with the NanoDrop™ 2000 (Thermo Fisher Scientific). For the first-strand cDNA synthesis, 500 ng of RNA were reverse-transcribed using the QuantiNova™ Reverse Transcription Kit (Qiagen) in a SimpliAmp™ Thermal Cycler (Thermo Fisher Scientific). Real-time PCR was carried out using the designed primers (Table S1 and Table S2) at a concentration of 700 nM and QuantiNova SYBR Green PCR (Qiagen) on a StepOnePlus™ Real-Time PCR System (Thermo Fisher Scientific). Data analysis was performed using the 2ΔΔCt method [68], and presented as mean fold change of six measurements.

Animal experiments

Genetically diabetic male mice db/db (strain C57BL/KsJ-m+/+Leprdb, Charles River Laboratories -Calco-Lecco) were housed under standard conditions. Blood glucose was measured (Contour XT, Bayer, Basel, Switzerland), and only mice with blood glucose ≥ 250 mg/dL were included in the study. Under anesthesia (isoflurane 3% plus 2 L/min O2), pressure ulcers were induced in the dorsal back using two magnetic disks of 12 mm diameter (anisotropic ferrite) and a thickness of 5.0 mm, with an average weight of 2.4 g and 1000 G magnetic force (Algamagnetic, Italy), as previously described [33, 69]. Briefly, a skin fold was raised and placed between the two magnets to generate a compressive pressure of 50 mm Hg [34]. Three ischemia-reperfusion (I/R) cycles were used, each single I/R cycle consisting of a period of 12 h of magnet placement followed by a rest period of 12 h without magnet. A surgical wound curettage was performed on day 3 to remove the ischemic skin and eschar. After curettage, mice were randomly assigned to the treatment groups (N = 5 in each group). Topical application of the medications began after curettage, weekly renewed and dressed with Tegaderm (3 M Health Care; Tegaderm Roll, St Paul, MN, USA). After curettage and before each dressing renewal, both wounds were photographed and wound areas measured using Nis-Elements AR 3.2 software (Nikon Corporation, Tokyo).

Histology, immunofluorescence and image analysis

Mice were sacrificed at 7, 14 or 21 days according to the experimental target. Skin samples were quickly dissected and fixed in 4% (v/v) paraformaldehyde solution and picric acid–saturated aqueous solution in 0.1 M Sörensen’s phosphate buffer (pH 7.4). Right wounds were embedded in paraffin, sectioned at 4 μm, and stained with hematoxylin and eosin (H&E). For immunofluorescence, left wounds were fixed as above for 24 h, washed for 48 h in 0.1 M phosphate buffer 5.0% sucrose and quickly frozen. Cryostat Sect. (14 μm thick, HM550 Microm, Bio-Optica) were incubated overnight with primary antibody anti-laminin (rabbit, 1:200 dilution; SIGMA Aldrich) and anti PGP-9.5 (rabbit, 1:350 dilution; Proteintech) at 4 °C in a humid chamber. After rinsing, sections were incubated with secondary antiserum Cy2 Donkey anti-Rabbit IgG (Jackson Immunoresearch), rinsed and mounted in glycerol containing 1,4-phenylendiamine (0.1 g/L). Immunofluorescence images were taken by a Nikon Eclipse E600 microscope equipped with the Q Imaging Retiga-2000RV digital CCD camera (Q Imaging, Surrey, BC, Canada) and a motorized z-axis stage. Analysis was performed using the Nis-Elements AR 3.2 software, by applying the same procedure to all images under comparison. The immunoreactive area was calculated as area/fraction (percentage of immunoreactivity over 400 × 300 μm area). For morphological analysis, five images and two levels/animal were sampled at the center of the repaired ulcer. All analyses were performed blindly. Epidermal thickness was determined at the equator of the lesioned area by H&E staining on histologic sections in the same area. The mean value of five measurements/section and three sections per animal was used for the statistical analysis.

Statistical analysis

The results were expressed as mean ± SD and analyzed by GraphPad Prism software. One-way analysis of variance (ANOVA) and Student’s t-test were used to evaluate the statistical significance (p < 0.05).