iMPs preparation

imMKCLs (cell line 7) were cultured as previously reported [15] in Iscove’s modified Dulbecco medium (IMDM; Sigma-Aldrich) containing 15% fetal bovine serum (FBS; Nichirei Bioscience), 1 µL/mL L-ascorbic acid (cat. #A4544, Sigma-Aldrich); 1 µL/mL 1-thioglycerol (cat. #M6145, Sigma- Aldrich); a 1% cocktail of 12,000 units/mL penicillin, 10 mg/mL of streptomycin, and 200 mM L-glutamine (cat. #10378-016, Thermo Fisher Scientific); and a 1% cocktail of 1 mg/mL human insulin, 0.55 mg/mL human transferrin, and 0.67 µg/mL sodium selenite (cat. #41400-045, Thermo Fisher Scientific) in the presence of 50 ng/mL human stem cell factor (SCF; cat. #193-15513, FUJIFILM Wako Pure Chemical) and 200 ng/mL TA316 (human thrombopoietin mimetic small compound; in-house synthesized) at 37℃ and 5% CO2. During the proliferation phase of 14 to 50 days, 100 µg/mL doxycycline (Dox; Clontech) was added to the medium to maintain c-MYC, BMI1, and BCL-XL overexpression.

In the differentiation induction phase, Dox was excluded from the medium, and 750 µM SR1 (aryl hydrocarbon receptor antagonist; NARD Institute), 10 mM Y27632 (Rho-associated protein kinase inhibitor; FUJIFILM Wako Pure Chemical), and 15 mM KP457 (inhibitor of GPIbα shedding on platelets; Kaken Pharmaceutical) were added. 1.0 × 105 imMKCLs/mL were incubated in 125-mL Erlenmeyer cell culture flasks shaken by Lab-Therm shakers (Kuhner) to generate shear stress. After incubating for 5 days, the culture medium was centrifuged at 700 x g for 15 min, and the cell pellet was suspended in phosphate-buffered saline (PBS) at 1% volume of the medium to collect iMPs.

Fig. 1

Preparation of iMPs and measurement of growth factors. A Schema of the preparation. Following 5 days of imMKCL differentiation with DOX removal, a mixture of mature megakaryocytes and platelets were collected as iMPs. B Concentration of growth factors in iMPs that have been previously reported to promote wound healing. Following the activation, PDGF-BB, TGF-β, EGF, VEGF, FGF2, and IGF1 in the supernatant were quantified using ELISA. n = 3. C Thrombin dependency of the growth factors released from iMPs. Their concentrations were determined with and without thrombin during the activation. n = 3. Data are presented as the mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. NS: not significant. ND: not detected

PRP preparation

PRP was extracted from healthy male donors aged 37 to 47 years using a commercial kit (Condensia, Kyocera) according to the instruction manual. Blood containing 5% sodium citrate (Citramin, Fuso Pharmaceutical Industries) was centrifuged at 600 x g for 7 min, and the plasma layer with buffy coat was collected. Following the second centrifugation at 2000 x g for 5 min, about 80% of the total volume of the supernatant was collected as platelet-poor plasma (PPP), and the remaining layer containing the cell pellet was stirred to acquire PRP. The blood cell densities of whole blood, PPP, and PRP were measured with an automated hematology analyzer (XQ-320, Sysmex). The platelet concentration of PRP was adjusted to five times that of whole blood by adding the calculated amount of PPP.

Measurement of cytokines

To activate platelets, 100 U/mL thrombin (Fuji Pharma) was added to the iMPs and PRP, and then both were incubated in a thermostatic chamber at 37 °C for 60 min. Following the activation, the iMPs and PRP were centrifuged at 2000 x g for 15 min to collect supernatant containing proteins secreted from the cells. The expression levels of cytokines, including PDGF-BB, transforming growth factor beta (TGF-β), epidermal growth factor (EGF), FGF2, VEGF, and insulin-like growth factor 1 (IGF1), were quantified by an enzyme-linked immuno sorbent assay (ELISA) kit (Quantikine, R&D Systems). All assays were conducted according to the manufacturer’s instructions.

Cell culture

Human skin fibroblasts derived from adult donors (cat. #CC-2511, Lonza) and HUVECs derived from normal neonate (cat. #C2517A, Lonza) were preliminarily cultured in the manufacturer’s recommended growth medium (LONZA) at 37℃ and 5% CO2. To evaluate the paracrine effects of iMPs and PRP in the wound environment, the horizontal co-culture system NICO-1 (Ginrei Lab) was used for in vitro assays. In this system, a filter of 0.6 μm (ICCP filter, Ginrei Lab) is inserted between two linked wells through which proteins, amino acids, and extracellular vesicles can pass [18]. 1 × 104 fibroblasts or HUVECs were seeded in one well of NICO-1 and incubated for 24 h. After rinsing with PBS, the culture medium was changed to basal medium. Serum-free Dulbecco’s modified Eagle medium (DMEM; Nacalai Tesque) was used as basal medium for the fibroblasts. To prevent excessive apoptosis, DMEM / Ham’s F-12 (FUJIFILM Wako Pure Chemical) containing 10% FBS (Life Science Group) was prepared as basal medium for HUVECs. Co-culture was initiated by filling the other well of NICO-1 with basal medium containing 10% iMPs, PRP, or PBS as control and 10 U/mL thrombin.

Fig. 2

Effects of iMPs on fibroblasts in vitro. A Shema of the assay using the NICO-1 co-culture system. This system has a filter of 0.6 μm between two linked wells, enabling evaluation of the paracrine effect in a biological wound environment. Primary cells were seeded in one well, while culture medium containing 10% PBS, PRP, or iMPs filled the other well. B Fibroblast proliferation. Following 72 h of co-culture, the number of fibroblasts per well was counted. n = 6. C Cell cycle of fibroblasts. Representative plots show 7AAD staining after 72 h of co-culture. The positive rate of 7AAD is indicated in the bar graph. n = 6. D Gene expressions of ACTA2, FGF2, and VEGF in fibroblasts. ACTA2 is a differentiation marker of myofibroblasts. FGF2 and VEGF are secreted from fibroblasts and contribute to cell-cell interactions in wound healing. n = 6. E Concentrations of FGF2 and VEGF released into the culture medium from fibroblasts following the co-culture. n = 6. Data are presented as the mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001

Analysis of proliferation and cell cycle

Following 72 h of co-culture, cells were detached with 0.25% Trypsin-EDTA (FUJIFILM Wako Pure Chemical) and collected for cell counting with a flow cytometer (FACS Canto II, Becton, Dickinson & Co.) and counting beads (cat. #C36950, Thermo Fisher Scientific). The number of fibroblasts in each group was determined after excluding propidium iodide (cat. #00-6990-42, Thermo Fisher Scientific) positive dying cells. A cell cycle assay was performed as previously reported [19]. Briefly, collected cells were fixed and permeabilized in 100 mL of BD Cytofix/Cytoperm (Becton, Dickinson & Co.) for 15 min on ice and then washed in BD Perm/Wash buffer (Becton, Dickinson & Co.). Next, the cells were treated with BD Cytoperm permeabilization buffer Plus (Becton, Dickinson & Co.) for 10 min on ice, washed in BD Perm/Wash buffer, and incubated in BD Cytofix/Cytoperm for 5 min on ice. After another wash with BD Perm/Wash, the cells were stained with Ki67-Alexa647 (cat. #558615, Becton, Dickinson & Co.) at 4 °C overnight (dilution 1:30 in BD Perm/Wash buffer). The cells were then washed with BD Perm/Wash and incubated with 7AAD (cat. #559925, Becton, Dickinson & Co.) at 1:40 dilution for 20 min at room temperature. Finally, the cells were analyzed with a flow cytometer within an hour.

Evaluation of FGF2 function on HUVEC proliferation

To verify the effect of FGF2 contained in iMPs on HUVECs, 1.5 or 10 ng/mL recombinant FGF2 (cat. #ab9596, Abcam) was added to the basal medium instead of the co-culture. After 72 h of incubation, the number of HUVECs was determined with a flow cytometer, as described above. In addition, 2 nM of an FGFR inhibitor, Pemigatinib (cat. #S0088, Selleck) or Futibatinib (cat. #S8848, Selleck), was added to the basal medium at the initiation of the co-culture with iMPs. Dimethyl sulfoxide (DMSO) was used as the inhibitor solvent and control. The number of HUVECs was determined after 72 h.

RT-qPCR

Total RNA was extracted from fibroblasts or HUVECs using the RNeasy Micro Kit (Qiagen) and reverse transcribed using the ReverTra Ace qPCR RT Master Mix Kit (Toyobo). Synthesized cDNA was subjected to quantitative reverse transcription polymerase chain reaction (RT-qPCR) with TaqMan Gene Expression Master Mix (Thermo Fisher Scientific) using CFX Connect (Bio-Rad). Amplification was performed using specific primers (cat. #4331182, Thermo Fisher Scientific). The analyzed genes and their primer ID were GAPDH (Hs02786624_g1), ACTA2 (Hs00426835_g1), FGF2 (Hs00266645_m1), VEGF (Hs00900055_m1). Gene expressions were quantified as the relative mRNA level normalized to GAPDH.

Transwell migration assay

HUVEC migration was analyzed using a transwell with an 8.0-µm pore membrane insert (Corning). The lower chamber contained 2.5 mL basal medium with 10% iMPs, PRP, or PBS and 10 U/mL thrombin. 100 µL of serum-free DMEM including 5 × 104 HUVECs was added to the insert, which was placed on each well of a 24-well plate. After 24 h of incubation, remaining cells on the upper side of the membrane were wiped out with cotton swabs. Then, the cells that penetrated the lower side were fixed with 4% paraformaldehyde phosphate buffer solution (FUJIFILM Wako Pure Chemical) and stained with 1% crystal violet (bioWORLD) for 10 min. The center of the membrane was photographed with a digital phase contrast microscope (BZ-X700, Keyence), and the stained area was quantified using analysis software (BZ-X Analyzer, Keyence).

RNA-seq

After 72 h of co-culture, total RNA was extracted from HUVECs using the RNeasy Micro Kit. RNA-seq libraries were prepared from at least three biological replicates according to the manufacturer’s protocol. Briefly, ~ 10 ng total RNA was used as input for cDNA conversion using a SMART-Seq v4 Ultra Low Input RNA Kit for Sequencing (cat. #634890, Clontech, Takara). cDNA was fragmented using an S220 Focused-ultrasonicator (Covaris). The cDNA library was then amplified using a NEBNext Ultra DNA Library Prep Kit for Illumina (cat. #E7370L, New England Biolabs). Finally, the NEBnext library size was estimated using a bioanalyzer with an Agilent High Sensitivity DNA Kit. Sequencing was performed using a NextSeq500 System (Illumina) with a single-read sequencing length of 60 bp. TopHat (version 2.1.1; with default parameters) was used to map to the reference genome (UCSC/hg19) with annotation data from iGenomes (Illumina). Gene expression levels were quantified using Cuffdiff (Cufflinks version 2.2.1; with default parameters).

Knockdown of ANGPTL4

24 h after seeding HUVECs in NICO-1, siRNA was transfected using Lipofectamine RNAiMAX (cat. #13778030, Thermo Fisher Scientific). The cells were incubated for 6 h in serum-free Opti-MEM (Thermo Fisher Scientific) containing 10 nM of siRNA SMARTpool (cat. #M-007807-02, Dharmacon) or Non-Targeting Control siRNAs (cat. #D-001210-03, Dharmacon). After changing the medium to basal medium, the cells were co-cultured with iMPs or administered 1.5 ng/mL FGF2, as described above. The number of HUVECs in each group was counted after 72 h with a flow cytometer. The knockdown efficiency in HUVECs was evaluated by RT-qPCR 72 h after the transfection.

In vivo wound healing assay

8-12-week-old female diabetic mice (BKS.Cg-m +/+ Leprdb/Jcl, CLEA Japan) were prepared for the animal experiments. Anesthesia was performed using isoflurane (Fujifilm Wako Pure Chemical) with an animal anesthetizer device (MK-AT210D, Muromachi Kikai). The concentration of isoflurane was 2% to provide an appropriate depth of anesthesia during the following surgical procedure. After hair removal with shaving and depilatory cream (Veet, Reckitt Japan), full thickness skin defects of 20 mm in diameter were introduced on the center of the back using a scalpel and scissors. The mice were randomly divided into three groups: iMPs, PRP, and PBS as control. 100 µL of each sample and 10 U of thrombin were applied onto the wound surface, and polyurethane film (Tegaderm, 3 M) was placed over the wound. This sample application procedure was also performed on day 3. Mice were placed in individual cages during the experiment. On days 0, 3, 6, 8, 10, 13, 15, 17, and 20, after removing the dressing, the wound was photographed with a digital camera (EOS 80D, Canon). Each non-epithelialized area was calculated using Adobe Photoshop software. At the end of the observation period, the mice were humanely euthanized with carbon dioxide in accordance with the institutional guidelines. All animal experiments were reported in line with the ARRIVE guidelines 2.0.

Histological analysis

On day 10 of the wound healing assay described above, mice from each group were euthanized with carbon dioxide for histological analysis. Skin samples, including wounds, were taken and fixed with 10% Formalin Neutral Buffer Solution (Fujifilm Wako Pure Chemical). The samples were embedded in paraffin and sectioned at each wound axially.

Masson’s trichrome staining was performed to evaluate collagen formation. Tissue sections were stained in Iron Hematoxylin Working Solution, 0.75% Orange G Staining Solution, Masson’s Staining Solution A, and Aniline Blue Solution. All solutions were purchased from Muto Pure Chemicals. Images of the whole section were acquired using a digital microscope (BZ-X700, Keyence) at low magnification. The area where the loss of hair follicles was observed was identified as the original wound, and the total area of the blue-stained area within the wound was measured using analysis software (BZ-X Analyzer, Keyence).

To quantify angiogenesis in the wound, immunohistochemical staining was performed. Tissue sections were de-paraffined with xylene. Antigen retrieval was performed by microwave treatment with EDTA Buffer (Genostaff). Endogenous peroxidase was blocked with 0.3% H2O2 in methanol, followed by incubation with G-Block (Genostaff) and an Avidin/Biotin Blocking Kit (Vector Laboratories). The sections were incubated with Anti-CD31 Rabbit Monoclonal Antibody (cat. #ab182981, Abcam) overnight. They were then incubated with Biotin-Conjugated Anti-Rabbit IgG (cat. #BA-1000, Vector Laboratories), followed by the addition of peroxidase conjugated streptavidin (Nichirei). Peroxidase activity was visualized using diaminobenzidine. The sections were counterstained with Mayer’s Hematoxylin (Muto Pure Chemicals) and mounted with Malinol (Muto Pure Chemicals). Images were acquired using a digital microscope (BZ-X700, Keyence) at high magnification. The average number of stained blood vessels in 10 random fields of view in each wound was calculated.

Validation of FD-iMPs

The iMPs lyophilization procedure was performed using VA-140 S (Taitec). iMPs collected as described above were transferred to a test tube and rotated in an ethanol bath at -60 °C for preliminary freezing. After 24 h of incubation at -80 °C, the tube was attached to a vacuum freeze-dryer to complete the process. Freeze-dried iMPs (FD-iMPs) were stored at 4 °C for 90 days to evaluate their stability and then applied for growth factor measurement by ELISA and for the in vivo wound healing assay. In these experiments, FD-iMPs were previously suspended in PBS to the same volume as before lyophilization.

Statistical analysis

All data are represented as means ± SEM and were analyzed using GraphPad Prism 8 (GraphPad Software). Pairwise comparisons were performed using a two-tailed t-test with assumed equal variances. When comparing three or more groups, one-way ANOVA and the Tukey-Kramer test were applied. Conventional statistical significance was set at p < 0.05.