This article is mentioned in:

Introduction

Acute lung injury (ALI)/acute respiratory distress syndrome (ARDS) is a critical condition characterized by neutrophil infiltration of the lungs, injury of alveolar epithelial cells and capillary endothelium, and protein-rich edema (1). ARDS is a severe lung condition developing from ALI, and epidemiological studies from Europe published in the past 20 years reported an ARDS incidence ranging from 5–8 cases/1,00,000 individuals (1). If not treated correctly or in a timely manner, ALI may lead to acute respiratory failure, including diffuse pulmonary interstitial and alveolar edema and severe hypoxic respiratory, insufficiency characterized by the progressive hypoxemia and respiratory distress (2,3). Despite the development of promising new therapies to treat ALI during the past decade, including venovenous extracorporeal membrane oxygenation and lung-protective ventilation, the mortality of ALI remains high (40–50%) (1). Therefore, it is important to investigate novel therapeutic methods for treating ALI.

Macrophages in lung tissue include alveolar macrophages, pulmonary interstitial macrophages and bronchial macrophages, and these are an important regulator of the local inflammatory microenvironment and immune response during ALI (4). When stimulated by external pathogenic factors, macrophages may be converted into a pro-inflammatory phenotype (M1) and release a large number of inflammatory mediators, such as tumor necrosis factor (TNF)-α, interleukin (IL)-1 and IL-6, recruiting various immune cells and initiating an inflammatory cascade (5). However, with an increase in the level of pro-inflammatory cytokines, some macrophages will polarize toward an immunosuppressive phenotype (M2) and in turn inhibit the inflammatory reaction (6). Moreover, the disruption of the balance between M1 and M2 macrophages is considered to be one of the most important causes of the uncontrolled inflammatory cascade in ALI (4,6).

Mesenchymal stromal cells (MSCs) are a type of fibroblast-like cells that have the ability of self-renewal and differentiation (7). MSCs have been detected in various tissues in the adult human body, such as the bone marrow, adipose tissue, umbilical cord blood and synovial fluid, from which they may be easily isolated (8). Although MSCs have a prominent ability to proliferate and differentiate, the fact that they die soon after being transplanted into the injured area suggests that their paracrine properties largely constitute the basis for their medical application (9,10). Furthermore, the clinical use of MSCs has been limited by the long waiting times for cell preparation and the potential tumorigenicity safety concerns (11). To solve these aforementioned issues, some researchers have collected the bioactive factors secreted by MSCs in conditioned media (CM) as an alternative for MSCs and have reported positive results on improving cardiac function, skin wound repair and bone regeneration (11–13). In addition, the advantages of secure storage, transportation and administration also indicate CM may be a promising alternative for MSCs.

CM from MSCs is composed of a broad repertoire of bioactive factors, including cytokines, growth factors, microRNAs, proteasomes and exosomes. The wide range of anti-inflammatory cytokines secreted by MSCs, including transforming growth factor (TGF)-β, IL-1 receptor antagonist (IL-1Ra) and IL-10, suggests that CM can regulate the inflammatory process via a variety of mechanisms. For example, TGF-β can induce M1 macrophage polarization towards the M2-like phenotype, inhibit the maturation of B cells and decrease the secretion of TNF-α and immunoglobulin E. Moreover, IL-1Ra can block IL-1α/IL-1β signaling pathways, inhibit the antigen-presenting ability of dendritic cells and increase the number of immunosuppressive T cells, while IL-10 is a potent anti-inflammatory cytokine that can markedly inhibit the production of interferon (IFN)-γ, IL-2 and TNF-α (14–17).

Factors of the host microenvironment serve a key role in MSC-mediated immunomodulation (18). For instance, exposure of MSCs to inflammatory signals such as TNF-α and IL-1, is the basis of the in vivo MSC-mediated immunosuppression (17,19). Given that the in vitro cell culture condition does not include inflammatory signals, preconditioning MSCs is a prerequisite for MSC-based therapy (20). MSCs derived from various tissues have been reported to express functional Toll-like receptors (TLRs), a type of single-pass transmembrane proteins (21,22). TLR ligations can activate the immune defensive function of MSCs and alter their secretome profiles. For example, TLR4 ligands can activate the NF-κB pathway in MSCs and induce MSCs to release a variety of molecules, including IL-6 and vascular endothelial growth factor (23). Therefore, priming MSCs with TLR ligations may be a promising method for enhancing their therapeutic effects.

Flagellin is a subunit protein of the flagellum, a whip-like appendage that enables bacterial motility, and it is a commonly used TLR5 ligand (24). The majority of previous studies have focused on preconditioning MSCs with TLR3 and TLR4 ligations (22,23,25). While Linard et al (21) reported the superior anti-inflammatory effect of flagellin-activated MSCs in an irradiation-induced proctitis model compared with normal MSCs, whether MSCs preconditioned with flagellin are effective in other inflammation-related diseases and the mechanism underlying the beneficial effects of flagellin preconditioning on the MSCs remain elusive.

The aim of the present study was to investigate the effects of CM from normal adipose derived-MSCs (ADSCs) and CM from flagellin-activated ADSCs (F-CM) in a mouse model of ALI induced by intraperitoneal injection of lipopolysaccharide (LPS). Experiments were performed both in vitro and in vivo. The effects of CM and F-CM on the polarization and expression levels of inflammation-related factors were first investigated in RAW264.7 cells. Furthermore, CM and F-CM were injected into mice with LPS-induced lung injury to assess their anti-inflammatory activity in vivo. The current study also aimed to determine whether flagellin preconditioning can increase the beneficial effects of CM from ADSCs against LPS-induced ALI.

Materials and methods

Culture of human ADSCs

Human-ADSCs were purchased from the Type Culture Collection of the Chinese Academy of Sciences. The eight donors (men, 4; women, 4) were aged from 21–42 years old (the sex and age range of each sample were provided by the Type Culture Collection of the Chinese Academy of Sciences). The cells were centrifuged at 500 × g for 20 min at room temperature, and the cell pellet was diluted with D10 media which was composed of 89% DMEM-low glucose (DMEM-LG; Hyclone; Cytiva), 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin (Invitrogen; Thermo Fisher Scientific, Inc.). Then, cells were cultured in T-75 flasks (BD Biosciences) at 37°C in a humidified incubator containing 5% CO2, and the culture media was changed twice a week. ADSCs were passaged when they were confluent, and the cells obtained at the end of the third passage were used for the experiments. ADSCs from different donors were separately cultured. ADSCs from all eight donors were used in all subsequent experiments, including the animal study.

Tri-lineage differentiation (osteocyte, chondrocyte and adipocyte) of the flagellin-activated ADSCs

For osteogenic and adipogenic differentiation, ADSCs were seeded on 6-well plates at the density of 1×105 cells/well. Osteogenic induction was performed using differentiation media consisting of DMEM-LG supplemented with 10% FBS, 50 mg/ml of ascorbic acid, 10 mM β-glycerophosphate and 100 nM dexamethasone (all Sigma-Aldrich; Merck KGaA). The media was changed every 3 days until day 21.

Adipogenic studies were performed by culturing the cells in differentiation media containing 1 mM dexamethasone, 50 mM 3-isobutyl-1-methylxanthine and 10 mg/ml insulin (all Sigma-Aldrich; Merck KGaA).

For chondrogenesis, 1×105 cells were centrifuged at 500 × g in a polypropylene tube (15 ml; BD Biosciences) for 10 min at 4°C. Aggregates were incubated in chondrogenic media consisting of DMEM supplemented with 10% FBS, 1% insulin-transferrin-selenium, 1 mM sodium pyruvate and 50 mM L-proline (all Sigma-Aldrich; Merck KGaA).

After the differentiation processes were complete, cells were fixed in 4% paraformaldehyde for 30 min at 37°C and stained with Alizarin Red (10 min at 37°C), Oil Red O (10 min at 37°C) and Toluidine Blue (30 min at 37°C; all Beyotime Institute of Biotechnology), respectively. Then, cells were observed with a light microscope (Optiphot; Nikon Corporation) at a high-power magnification of ×100.

Flow cytometry

After being incubated with D10 media containing flagellin (100 ng/ml; cat. no. P7388; Beyotime Institute of Biotechnology) at 37°C for 48 h, ADSCs were harvested and suspended in PBS containing 4% FBS. Non-specific detection of the Fc component of the CD antibodies were blocked with 5% BSA (Beyotime Institute of Biotechnology) for 30 min at room temperature. Then, cells were incubated with fluorescence-conjugated antibodies directed against mouse anti-human CD31 (cat. no. 555446), CD45 (cat. no. 560178), CD73 (cat. no. 550257), CD90 (cat. no. 555593) and CD105 (cat. no. 562380; all from BD Biosciences; 1:20) for 30 min in the dark at 4°C After being washed three times with PBS, the stained cells were sorted using a FACSCalibur flow cytometer (BD Biosciences). The result was analyzed using BD FACStation software (version 6.1X; BD Biosciences).

Preparation of CM and F-CM

When reaching 80–90% confluency, the culture media was changed to a mixture of D10 media and flagellin (100 ng/ml). After 48 h incubation at 37°C, flagellin-activated ADSCs were washed three times with warm PBS and re-supplemented with serum-free DMEM. After 48 h, the CM was collected, filtered with a 0.2-µm filter (EMD Millipore) and concentrated 50-fold using ultrafiltration units with a 3-kDa molecular weight cutoff (EMD Millipore) (26). The concentrated F-CM was stored at −80°C before being used for the following experiments. The CM was prepared using the same method, except no flagellin was added.

Effect of serum deprivation on the viability of ADSCs

Normal ADSCs and ADSCs pretreated with flagellin (100 ng/ml) for 48 h at 37°C were cultured in 96-well plates at the density of 2×104 cells/well with serum-free culture media. After 0, 24 and 48 h, the serum-free culture media was removed and 200 µl fresh culture media containing 10% Cell Counting Kit-8 reagent (CCK-8; Dojindo Molecular Technologies, Inc.) were added to each well according to the manufacturer's instructions. After 2 h incubation at 37°C, the absorbance was measured at 450 nm.

Culture and treatment of RAW264.7 cells and THP-1 cells

RAW264.7 cells were purchased from the Type Culture Collection of the Chinese Academy of Sciences. Cells were cultured in 6-well plates at the density of 1×105 cells/well, and randomly divided into the control (20 µl/ml PBS), LPS (1 µg/ml LPS; Beyotime Institute of Biotechnology), CM (1 µg/ml LPS + 20 µl/ml CM) and F-CM (1 µg/ml LPS + 20 µl/ml F-CM) groups. The cells were cultured at 37°C for 24 h and then tested.

THP-1 cells were purchased from the Type Culture Collection of the Chinese Academy of Sciences. Cells were cultured in RPMI-1640 medium (Hyclone; Cytiva) containing 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin (complete medium) and maintained at 37°C in a humidified, 5% CO2 environment with medium changed every 2 days. Cells were cultured in 6-well plates at the density of 1×105 cells/well, and randomly divided into the control (20 µl/ml PBS), LPS (1 µg/ml LPS), CM (1 µg/ml LPS + 20 µl/ml CM) and F-CM (1 µg/ml LPS + 20 µl/ml F-CM) groups. The cells were cultured at 37°C for 24 h and then tested.

Reverse transcription-quantitative PCR (RT-qPCR)

To extract RNA, 0.5 ml TRIzol® reagent (Beyotime Institute of Biotechnology) was added in each well of cells after 24 h of culture or small pieces of the mouse lung tissue. mRNA dissolved in the TRIzol® reagent was isolated via centrifugation (12,000 × g/min) at 4°C for 15 min. cDNA was synthesized using a RT kit (cat. no. 4368813; Thermo Fisher Scientific, Inc.), Mouse-specific or human-specific primers with SYBR Green Master Mix (cat. no. 309155; Thermo Fisher Scientific, Inc.) used in the study are presented in Table I. The thermocycling conditions were as follows: Initial denaturation at 95°C for 5 min, 40 cycles of denaturation at 95°C for 30 sec, annealing at 58°C for 30 sec and extension at 72°C for 45 sec. The relative mRNA expression was calculated using the 2−ΔΔCq method (27). The GAPDH gene was used as the internal control.

Table I. Primer sequences (5′→3′) for reverse transcription-quantitative PCR.

Table I.

Primer sequences (5′→3′) for reverse transcription-quantitative PCR.

Gene	Species	Forward primer	Reverse primer
TNF-α	mice	CCCTCACACTCAGATCATCTTCT	GCTACGACGTGGGCTACAG
IL-6	mice	GCC TTC CCT ACT TCA CAA	ACA ACTCTT TTC TCA TTT CCA C
GAPDH	mice	GGAGCGAGATCCCTCCAAAAT	GGCTGTTGTCATACTTCTCATGG
CD86	mice	TGTTTCCGTGGAGACGCAAG	TTGAGCCTTTGTAAATGGGCA
IL-10	mice	GCTCTTACTGACTGGCATGAG	CGCAGCTCTAGGAGCATGTG
IL-1	mice	CTATGTCTTGCCCGTGGAG	CATCATCCCACGAGTCACA
CD163	mice	ATGGGTGGACACAGAATGGTT	CAGGAGCGTTAGTGACAGCAG
TNF-α	human	CTCCCAGGTCCTCTTCAAGG	TTGATGGCAGAGAGGAGGTT
IL-6	human	CGGGAACGAAAGAGAAGCTC	ACAACAACAATCTGAGGTGC
GAPDH	human	CCACRCCRCCACCTTTGAC	ACCCTGTTGTTGCTGTAGCCA
CD86	human	TGCTCATCTATACACGGTTACC3	TGCATAACACCATCATACTCGA
IL-10	human	GCTGTCATCGATTTCTTCCC	TCAAACTCACTCATGGCTTTGT
IL-1	human	AATCTGTACCTGTCCTGCGTGTT	TGGGTAATTTTTGGGATCTACACTCT
CD163	human	CCAGTCCCAAACACTGTCCT	CACTCTCTATGCAGGCCACA
PPARγ	human	ACCAAAGTGCAATCAAAGTGG A	ATG AGG GAG TTG GAA GGC TCT
ALP	human	ACTGGGGCCTGAGATACCC	TCGTGTTGCACTGGTTAAAGC
RUNX2	human	TCAACGATCTGAGATTTGTGGG	GGGGAGGATTTGTGAAGACGG
SOX9	human	AGCGAACGCACATCAAGAC	CTGTAGGCGATCTGTTGGGG
COL2A1	human	TGGACGATCAGGCGAAACC	GCTGCGGATGCTCTCAATCT
Adiponectin	human	GGCTTTCCGGGAATCCAAGG	TGGGGATAGTAACGTAAGTCTCC

[i] TNF, tumor necrosis factor; IL, interleukin; RUNX2, RUNX family transcription factor 2; PPARγ, peroxisome proliferator activated receptor γ; ALP, alkaline phosphatase; COL2A1, collagen type II α 1 chain.

ELISA

Commercial ELISA kits for human [C-C chemokine ligand 5 (CCL5; cat. no. DRN00B), IFN-γ (cat. no. DIF50), IL-1β (cat. no. MLB00C), IL-1Ra (cat. no. DRA00B), IL-4 (cat. no. D4050), IL-6 (cat. no. D6050), IL-8 (cat. no. D8000C), IL-10 (cat. no. D1000B) and TGF-β (cat. no. DB100B)] and mouse [TNF-α (cat. no. MTA00B), IL-1 (cat. no. MLB00C), IL-6 (cat. no. DY406), IL-10 (cat. no. M1000B) and monocyte chemoattractant protein (MCP)-1 (cat. no. MJE00B)] cytokines (all from R&D Systems, Inc.) were used to determine the concentrations of inflammatory-associated factors in the bronchoalveolar lavage fluid (BALF) according to manufacturer's protocol.

Western blotting

To detect the expression of NF-κB pathway-related proteins, ADSCs treated with 100 ng/ml flagellin at 37°C for 0, 30 and 120 min were harvested to obtain the cytoplasmic and nuclear proteins using a Keygen Kit (Nanjing Keygen Biotech Co., Ltd.) according to manufacturer's protocol. Total protein was quantified using the bicinchoninic acid Protein assay kit (Beyotime Institute of Biotechnology). The same amount of protein (20 µg/lane) was placed into a 10% SDS-PAGE and transferred onto 0.22 µm PVDF membranes (EMD Millipore). Then, the transferred membranes were blocked in 5% BSA (Beyotime Institute of Biotechnology) at room temperature for 2 h and incubated with primary antibodies [p65, cat. no. AF5881, 1:1,000; IκBα, cat. no. AF5204, 1:1,000; phosphorylated (p)-IκBα, cat. no. AF5851, 1:1,000; β-actin, cat. no. AF5006, 1:2,000; histone deacetylase 1 (HDAC1), cat. no. AF5180 1:2,000; all from Beyotime Institute of Biotechnology] overnight at 4°C. β-actin and histone deacetylase 1 served as the internal controls. The membranes were washed with PBST containing 0.05% Tween (Shanghai Aladdin Bio-Chem Technology Co., Ltd.) three times followed by incubation with the corresponding horseradish peroxidase-conjugated secondary antibody (Beyotime Institute of Biotechnology) for 1 h at 37°C. Protein bands were visualized using ECL reagents (Roche Diagnostics GmbH) and then scanned with the Image Quant LAS4000 system (Cytiva). Protein expression levels were semi-quantified using Gel-Pro Analyzer software (version 4.0; Media Cybernetics, Inc.), with β-actin or HDAC1 as the loading control.

Animal model

All animal experiments were approved by the Animal Care and Use Committee of the Shanghai Children's Hospital. In total, 32 male BALB/c mice (age, 6–8 weeks; weight, 20–25 g) were purchased from Shanghai Lab Animal Research Center. Mice were maintained on a standard diet and water ad libitum (12 h light/dark cycle with humidity of 60±5% and temperature 22±3°C). All mice were randomly divided into the control group (n=8), the LPS group (n=8; ALI mice received PBS), the CM group (n=8; ALI mice received CM) and the F-CM group (n=8; ALI mice received F-CM). PBS, CM or F-CM (5 µl/g of body mass of the animals) were injected into the tail vein of mice in the LPS, CM and F-CM groups once a day for 2 days before the experiment. ALI was induced via intraperitoneal injection of LPS (20 µl/g) (28,29). Then, 24 h after injection, mice were euthanized with 0.9% saline containing sodium pentobarbital (50 mg/ml) via intraperitoneal injection at dose of 250 mg/kg of body mass of the animals.

Cell counting and protein content in BALF

Each mouse was euthanized with sodium pentobarbital via intraperitoneal injection. Both lungs of each mouse were lavaged three times via a tracheal cannula with 0.5 ml PBS at 4°C to obtain BALF. BALF was centrifuged (300 × g; 10 min; 4°C) and the protein content of the BALF supernatant was determined using the Braford method (Beyotime Institute of Biotechnology), according to the manufacturer's instructions. The cell pellet was washed with red blood cell lysis solution (Beyotime Institute of Biotechnology) at 37°C for 1 min, centrifuged at 300 × g at room temperature for 10 min and resuspended in 200 µl PBS. The total cell count was determined using a hemocytometer and differential cell count was measured using Wright-Giemsa staining (incubated with Wright-Giemsa for 1 min at room temperature; Beyotime Institute of Biotechnology). Macrophages and neutrophils were quantified by counting 200 cells per slide at ×400 magnification using light microscope.

Hematoxylin and eosin (H&E) staining

Lung tissue was washed three times with PBS and immersed in 4% paraformaldehyde (Beyotime Institute of Biotechnology) overnight at room temperature. Then, lung tissue was dehydrated with an increasing gradient concentration of ethanol (50, 70, 80, 90 and 100%) and embedded in paraffin (Beyotime Institute of Biotechnology). Tissue slices with thickness of 4 µm were attached to the slides coated with polylysine (Beyotime Institute of Biotechnology) and were incubated overnight at 60°C. Then, the slices were immersed in hematoxylin purple solution (Beyotime Institute of Biotechnology) at 37°C for 5 min, washed three times with deionized water, washed in 1% hydrochloric acid alcohol (Shanghai Aladdin Bio-Chem Technology Co., Ltd.) for 5 sec and rinsed in deionized water for 10 min. Slices were then immersed in eosin dyeing solution at 37°C (Beyotime Institute of Biotechnology) for 30 sec, in 95% ethanol for 2 min, in 100% ethanol for 2 min, in 100% xylene (Beyotime Institute of Biotechnology) for 5 min, and in the mixture of xylene and ethanol (1:1), respectively.

Lung injury was evaluated by two blinded observers according to a previous method by Smith et al (30). The parameters, lung edema, alveolar and interstitial inflammation, alveolar and interstitial hemorrhage, atelectasis and hyaline membrane formation, were scored as followed: i) No injury = score of 0; ii) injury in 25% of the field = score of 1; iii) injury in 50% of the field = score of 2; iv) injury in 75% of the field = score of 3; and v) injury throughout the field = score of 4. The lung injury score was the sum of these five criteria. In total, 12 random light microscopic (magnification, ×400) fields from each slide were imaged and the injury area was calculated using ImageJ 1.8 software (National Institutes of Health).

Lung wet/dry weight ratio

After the mice were euthanized, their lungs were excised, rinsed briefly in PBS, blotted dry and weighed to obtain the wet weight. Then, the lungs were placed in an oven at 80°C for 48 h to obtain the dry weight. The ratio of wet lung weight to dry lung weight was calculated.

Statistical analysis

Data are presented as the mean ± standard deviation. Differences between or among groups were examined for statistical significance using a unpaired Student's t-test, one-way ANOVA with Tukey's post hoc or Kruskal-Wallis with Dunn's post hoc test using GraphPad Prism 6 (GraphPad Software, Inc.). P<0.05 was considered to indicate a statistically significant difference. Experiments were repeated ≥8 times.

Results

Characterization of ADSCs with flagellin preconditioning

Flagellin-activated ADSCs were positive for MSC markers, including CD73 (Fig. 1C), CD90 (Fig. 1D) and CD105 (Fig. 1E), but negative for the endothelial cell marker CD31 (Fig. 1A) and the hematopoietic cell marker CD45 (Fig. 1B). Moreover, flagellin-activated ADSCs maintained a fibroblast-like morphology (Fig. 1F). The osteogenic (Fig. 1G), adipogenic (Fig. 1H) and chondrogenic (Fig. 1I) differentiation of flagellin-activated ADSCs was identified using Alizarin Red S, Oil Red O and Toluidine Blue staining, respectively.

Figure 1. Characterization of ADSCs with flagellin preconditioning and effects of serum deprivation on the viability of ADSCs. Flow cytometric analyses demonstrated that ADSCs in passage 3 were negative for (A) CD31 and (B) CD45, but positive for (C) CD73, (D) CD90 and (E) CD105. (F) Flagellin-activated ADSCs maintained a fibroblast-like morphology. Magnification, ×100. (G) Osteogenic, (H) adipogenic and (I) chondrogenic differentiation of flagellin-activated ADSCs was assessed using Alizarin Red S, Oil Red O and Toluidine Blue staining, respectively. Magnification, ×100. Expression levels of (J) ALP and (K) RUNX2 are used to investigate osteogenic differentiation. Expression levels of (L) adiponectin and (M) PPARγ are used to investigate adipogenic differentiation. Expression levels of (N) COL2A1 and (O) SOX9 were used to investigate chondrogenic differentiation. (P) Cell Counting Kit-8 was used to detect the viability of ADSC and flagellin-activated ADSCs cultured in serum-free media. Differences between groups were examined for statistical significance using Student's t-test. *P<0.05 vs. ADSCs group; ##P<0.01 vs. control group. RUNX2, RUNX family transcription factor 2; PPARγ, peroxisome proliferator activated receptor γ; ALP, alkaline phosphatase; COL2A1, collagen type II α 1 chain; ADSC, adipose derived- mesenchymal stromal cells; OD, optical density.

The mRNA expression levels of the osteogenic genes alkaline phosphatase (2.77±0.46) and RUNX family transcription factor 2 (3.83±0.48), the adipogenic genes adiponectin (66.74±4.43) and peroxisome proliferator activated receptor γ (10.77±1.29) and the chondrogenic genes collagen type II α 1 chain (3.97±0.91) and SOX9 (2.15±0.17), were significantly increased after differentiation induction compared with the control (Fig. 1J-O).

Effect of serum deprivation on the viability of ADSCs

Both the viability of ADSCs (0 h, 1.059; 24 h, 0.84; and 48 h, 0.62) and flagellin-activated ADSCs (0 h, 1.013; 24 h, 0.89; and 48 h, 0.74) decreased over time (Fig. 1P). However, ADSCs retained 59% viability and flagellin-activated ADSCs retained 73% viability. No differences were found between ADSCs and flagellin-activated ADSCs groups after 24 h incubation. At 48 h, the optical density value of the flagellin-activated ADSCs group was significantly higher compared with the ADSCs group.

Immunomodulatory effects of F-CM on murine macrophages

The immunomodulatory effect of F-CM was investigated in vitro using RAW264.7 cells. The effects of F-CM on the expression of the M1 marker C-C chemokine receptor (CCR)7 and the M2 marker CD206 were evaluated with flow cytometry. After a 48-h incubation, F-CM groups exhibited a significantly lower percentage of CCR7-positive cells (67.66±5.32%; Fig, 2B), but a significantly higher percentage of CD206-positive cells (41.75±4.48%; Fig. 2C) compared with the LPS group (CCR7, 83.64±9.28%; CD206, 11.37±3.65%; Fig. 2). The CM group had a significantly higher percentage of CD206-positive cells (22.79±5.87%) compared with the LPS group, but the percentage of CCR7-positive cells of the CM group (80.26±8.96%) was not significantly different compared with the LPS group (Fig. 2).

Figure 2. Effects of F-CM on surface markers of murine macrophages. (A) Flow cytometric analysis of (B) CCR7 positive cells represent M1 macrophages and (C) CD206 positive cells represent M2 macrophages. Differences among groups were examined for statistical significance using ANOVA. **P<0.01 vs. control group; #P<0.05 and ##P<0.01 vs. LPS group; ●●P<0.01 vs. CM group. CM, conditioned medium; F-CM, flagellin-conditioned medium; LPS, lipopolysaccharide; CCR, C-C chemokine receptor.

The effects of F-CM on the expression of inflammation-related genes were investigated using PCR analysis. Compared with PBS (the control group), LPS significantly increased the expression levels of pro-inflammatory factors, including TNF-α (7.11±0.91; Fig. 3A), IL-1 (5.60±0.83; Fig. 3B) and IL-6 (3.77±0.60; Fig. 3C), the expression of the M1 macrophage marker CD86 (4.58±0.56; Fig. 3E) and even the anti-inflammatory factor IL-10 (2.47±0.38; Fig. 3D), but did not affect the expression of the M2 macrophage marker CD163 (0.99±0.32; Fig. 3F). Furthermore, CM and F-CM significantly inhibited the expression levels of pro-inflammatory TNF-α (CM, 3.73±0.67; F-CM, 2.14±0.61; Fig. 3A), IL-1 (CM, 3.48±0.80; F-CM, 3.22±0.95; Fig. 3B), IL-6 (CM, 2.70±0.68; F-CM, 1.65±0.35; Fig. 3C) and CD86 (CM, 2.55±0.38; F-CM, 1.81±0.42; Fig. 3E), but further enhanced the expression levels of CD163 (CM, 2.69±0.40; F-CM, 3.65±0.46; Fig. 3F) and the anti-inflammatory factor IL-10 (CM, 3.60±0.58; F-CM, 4.42±0.68; Fig. 3D), compared with LPS.

Figure 3. Effects of F-CM on expressions of inflammatory-related genes of murine macrophages. F-CM significantly suppressed the expression levels of (A) TNF-α, (B) IL-1, (C) IL-6 and (E) CD86, but enhanced the expression levels of (D) IL-10 and (F) CD163 in flagellin-activated RAW264.7 macrophages. Differences among groups were examined for statistical significance using ANOVA. The results are represented as a relative ratio to the control group. **P<0.01 vs. control group; ##P<0.01 vs. LPS group; ●P<0.05 and ●●P<0.01 vs. CM group. CM, conditioned medium; F-CM, flagellin-conditioned medium; LPS, lipopolysaccharide; IL, interleukin; TNF, tumor necrosis factor.

Immunomodulatory effects of F-CM on human macrophages

The immunomodulatory effect of F-CM was also investigated in THP-1 cells (Fig. 4). Consistent with the aforementioned results, LPS significantly increased the percentage of CCR7-positive cells (79.23±7.40%), but CM and F-CM significantly decreased the percentage of CCR7-positive cells (CM, 65.03±9.01%; F-CM, 50.56±8.57%; Fig. 4B). Moreover, CM and F-CM groups also had a significantly higher ratio of CD206-positive cells (CM, 27.98±5.75%; F-CM, 35.27±6.67%) compared with the LPS group (6.19±2.40%; Fig. 4C).

Figure 4. Effects of F-CM on surface markers of human macrophages. (A) Flow cytometric analysis of (B) CCR7 positive cells represent M1 macrophages and (C) CD206 positive cells represent M2 macrophages. Differences among groups were examined for statistical significance using ANOVA. *P<0.05 and **P<0.01 vs. control group; ##P<0.01 vs. LPS group; ●P<0.05 and ●●P<0.01 vs. CM group. CM, conditioned medium; F-CM, flagellin-conditioned medium; LPS, lipopolysaccharide; CCR, C-C chemokine receptor.

The effects of F-CM on the expression levels of inflammation-related genes were examined using PCR analysis. Compared with PBS (the control group), LPS significantly increased the expression levels of TNF-α (4.81±0.54), IL-1 (6.20±0.79), IL-6 (2.71±0.59) and IL-10 (1.42±0.19), as well as the expression of the M1 macrophage marker CD86 (3.28±0.59), but markedly decreased the expression of the M2 macrophage marker CD163 (0.60±0.20) (Fig. 5A-F). In addition, CM and F-CM significantly inhibited the expression levels of TNF-α (CM, 2.80±0.79; F-CM, 1.98±0.45), IL-1 (CM, 4.74±0.49; F-CM, 3.22±0.95), IL-6 (CM, 2.19±0.34; F-CM, 1.64±0.29) and CD86 (CM, 2.56±0.46; F-CM, 1.97±0.35), but enhanced the expression levels of CD163 (CM, 2.70±0.41; F-CM, 3.82±0.43; Fig. 5E) and the anti-inflammatory factor IL-10 (CM, 2.21±0.28; F-CM, 2.89±0.43; Fig. 5D).

Figure 5. Effects of F-CM on expression levels of inflammatory-related genes of human macrophages. F-CM significantly suppressed the expression of (A) TNF-α, (B) IL-1, (C) IL-6 and (F) CD86, but enhanced the expression levels of (D) IL-10 and (E) CD163 in flagellin-activated THP-1 macrophages. Differences among groups were examined for statistical significance using ANOVA. The results are represented as a relative ratio to the control group. *P<0.05 and **P<0.01 vs. control group; ##P<0.01 vs. LPS group; ●P<0.05 and ●●P<0.01 vs. CM group. CM, conditioned medium; F-CM, flagellin-conditioned medium; LPS, lipopolysaccharide; IL, interleukin; TNF, tumor necrosis factor.

Changes of inflammatory cytokines in CM from flagellin-activated ADSCs

The levels of CCL5, IFN-γ, IL-1β, IL-1Ra, IL-4, IL-6, IL-8, IL-10 and TGF-β were assessed using ELISA (Fig. 6A). No significant differences in the levels of CCL5 (CM, 36.75±10.87 pg/ml; F-CM, 35.25±10.50 pg/ml), IFN-γ (CM, 150.3±22.60 pg/ml; F-CM, 146.8±23.94 pg/ml), IL-1β (CM, 58.75±6.85 pg/ml; F-CM, 45.50±13.48 pg/ml), IL-1Ra (CM, 45.50±13.67 pg/ml; F-CM, 59.25±18.48 pg/ml), IL-4 (CM, 19.25±12.61 pg/ml; F-CM, 19.75±10.24 pg/ml) or IL-10 (CM, 41.45±18.30 pg/ml; F-CM, 34.38±15.64 pg/ml) were observed between the two types of CM. However, flagellin preconditioning significantly increased the levels of IL-6 (CM, 827.3±293 pg/ml; F-CM, 6,797±1,128 pg/ml), IL-8 (CM, 6,786±2,182 pg/ml; F-CM, 40,605±7,560 pg/ml) and TGF-β (CM, 130.3±31.19 pg/ml; F-CM, 436.8±105.5 pg/ml) compared with CM.

Figure 6. Effects of flagellin preconditioning on the production of inflammatory-related cytokines and NF-κB signaling pathway in ADSCs. (A) Levels of inflammation-related cytokines of MSCs and flagellin-activated MSCs. Differences between groups were examined for statistical significance using Student's t-test. **P<0.01 vs. CM group. (B) p65, IκBα and p-IκBα protein expression levels in the cytoplasm and p65 protein expression in nucleus were examined using western blotting. (C) Changes of protein expressions (cytoplasm p65, cytoplasm lκBα, cytoplasm p-lκBα and nucleus p65) and (D) p-IκBα/ lκBα ratio during incubation. Differences among groups were examined for statistical significance using ANOVA. The results are represented as a relative ratio to the 0 min groups. *P<0.05 and **P<0.01 vs. 0 min group. CM, conditioned medium; F-CM, flagellin-conditioned medium; p-, phosphorylated; IκB, inhibitor of NF-κB; IL, interleukin; CCL, C-C chemokine ligand; IFN, interferon; TGF, transforming growth factor; HDAC1, histone deacetylase 1; IL-1Ra, IL-1 receptor antagonist.

Effects of flagellin on NF-κB signaling pathway in ADSCs

Western blotting results indicated that the expression level of p65 protein, a subunit of NF-κB, was significantly decreased in the cytoplasm (30 min, 0.80±0.08; 120 min, 0.56±0.17), while its presence in the nucleus increased significantly after ADSCs were treated with flagellin (30 min, 1.18±0.13; 120 min, 2.05±0.19; Fig. 6B and C). Moreover, p-IκBα expression (30 min, 1.38±0.12; 120 min, 2.29±0.25) was increased, while total IκBα expression decreased in the cytoplasm (30 min, 0.94±0.07; 120 min, 0.77±0.14; Fig. 6B and C). The p-IκBα/IκBα ratio was also significantly increased over time (0 min, 0.082±0.014; 30 min, 0.12±0.01; 120 min, 0.245±0.026; Fig. 6D). Thus, the results indicated that NF-κB signaling was activated in ADSCs when treated with 100 ng/ml flagellin.

Effects of F-CM on lung exudation and cell infiltration

The lung/body weight, lung wet/dry weight ratios and the protein concentration in the BALF were detected to evaluate the effects of F-CM on pulmonary edema and vascular permeability (Fig. 7). The mice in the LPS group exhibited significantly higher lung/body weight ratio (6.69±0.52 mg/g; Fig. 7A), lung wet/dry weight ratio (6.01±0.72 g/g; Fig. 7B) and protein concentrations (639.1±40.23 µg/ml; Fig. 7C) compared with the mice in the control group (5.70±0.36 mg/g; 4.13±0.42 g/g; and 362.1±32.36 µg/ml, respectively). Moreover, both CM and F-CM significantly decreased the lung/body weight ratio (CM, 6.02±0.46 mg/g; F-CM, 5.71±0.51 mg/g), lung wet/dry weight ratio (CM, 5.04±0.65 g/g; F-CM, 4.47±0.58 g/g) and protein concentrations (CM, 463.8±36.48 µg/ml; F-CM, 388.5±57.51 µg/ml) compared with LPS, but the effects of F-CM were more prominent.

Figure 7. Effects of F-CM on lung exudation and cell infiltration. A total of 24 h after the intratracheal instillation of LPS, (A) lung/body weight ratio and (B) lung wet/dry weight ratio were calculated. (C) Total protein concentration, (D) total cells, (E) neutrophils and (F) macrophages in the BALF were measured. Differences among groups were examined for statistical significance using ANOVA. **P<0.01 vs. control group; #P<0.05 and ##P<0.01 vs. LPS group; ●P<0.05 and ●●P<0.01 vs. CM group. CM, conditioned medium; F-CM, flagellin-conditioned medium; BALF, bronchoalveolar lavage fluid; LPS, lipopolysaccharide.

The effect of F-CM on inflammatory cell recruitment was investigated by counting the cell numbers in the BALF. LPS significantly increased not only the total cell number (42.67±4.35 105/l; Fig. 7D), but also the number of inflammatory cells, including macrophages (11.63±1.78 105/l; Fig. 7F) and neutrophils (22.78±2.33 105/l; Fig. 7E) compared with the control group. However, the increase in the numbers of total cells, macrophages and neutrophils was significantly inhibited by CM (24.92±3.75 105/l; 8.04±1.43 105/l; and 11.77±1.43 105/l, respectively) and F-CM (19.56±2.38 105/l; 5.99±1.16 105/l; and 8.39±1.62 105/l, respectively). In addition, the BALF of mice treated with F-CM contained the fewest macrophages and neutrophils among mice undergoing a LPS challenge.

Effects of F-CM on reducing inflammation caused by LPS in an animal model

H&E staining was performed to determine the histological changes in the lung. In the control group, the pulmonary architecture was intact and the alveoli were uniform in size (Fig. 8A). Compared with the control group, the LPS group exhibited marked inflammatory changes, characterized by large areas of red serum, swelling of the alveolar walls and numerous infiltrating inflammatory cells (Fig. 8B). However, the histological changes caused by LPS administration were reversed by CM and F-CM (Fig. 8C and D). Moreover, lung injury score evaluation also identified the treatment effects of CM and F-CM (Fig. 8E). The LPS group (mean, 12.5) had significantly higher lung score compared with the control group (mean, 1.38), but the lung injury score in F-CM group (mean, 5.13) was significantly lower compared with the LPS group. However, no significant differences were found between the CM (mean, 7.75) and LPS groups.

Figure 8. Effects of F-CM on reducing pathological damages caused by LPS. Pathological examination of the lung tissues of the (A) control, (B) LPS, (C) CM and (D) F-CM groups 24 h after LPS challenge. (E) Lung injury score was analyzed using the Kruskal-Wallis test. Differences among groups were examined for statistical significance by Kruskal-Wallis with Dunn's post hoc test. Red arrows indicate infiltrated immune cells, and blue arrows indicate edema. Br indicates bronchia, and V indicates vessels. Scale bar, 100 µm. **P<0.01 vs. control group; #P<0.05 vs. LPS group. CM, conditioned medium; F-CM, flagellin-conditioned medium; LPS, lipopolysaccharide.

ELISA was performed to investigate the levels of inflammatory mediators involved in ALI. The mice in the LPS group had significantly higher levels of TNF-α (704.6±87.1 pg/ml), IL-1 (399.8±61.90 pg/ml), IL-6 (249.4±54.71 pg/ml), MCP-1 (370.8±67.46 pg/ml) and IL-10 (82.50±15.32 pg/ml) compared with the mice in the control group (117.5±52.7; 42.75±26.68; 61.25±24.06; 44.25±21.02 pg/ml; and 15.75±10.18, respectively; Fig. 9A-E). Consistent with the results from the in vitro study, CM and F-CM significantly reduced the production of TNF-α (CM, 413.3±96.05 pg/ml; F-CM, 263.4±0.58 pg/ml), IL-1 (CM, 195.0±42.38 pg/ml; F-CM, 133.0±30.56 pg/ml), IL-6 (CM, 187.4±44.4 pg/ml; F-CM, 138.0±33.79 pg/ml) and MCP-1 (CM, 239.8±44.54 pg/ml; F-CM, 171.0±45.91 pg/ml) compared with PBS. However, no differences in IL-10 (CM, 74.13±13.20 pg/ml; F-CM, 73.88±15.49 pg/ml) levels were observed among the LPS, CM and F-CM groups. It was also demonstrated that the F-CM group had significantly lower TNF-α, IL-1 and MCP-1 levels compared with the CM group. In addition, the F-CM group (3.57±0.93) had a significantly lower TNF-α/IL-10 ratio compared with the LPS group (8.54±1.06) and the CM group (5.58±1.30; Fig. 9F). However, no significant difference in TNF-α/IL-10 ratio was found between the Control (7.21±2.14) and LPS group.

Figure 9. Effects of F-CM on inflammation-related factors. Expression levels of (A) TNF-α, (B) IL-1, (C) IL-6, (D) IL-10 and (E) MCP-1 in the lung were determined using ELISA 24 h after LPS administration. (F) TNF-α/IL-10 ratio was also calculated. Differences among groups were examined for statistical significance using ANOVA. **P<0.01 vs. control group; #P<0.05 and ##P<0.01 vs. LPS group; ●P<0.05 and ●●P<0.01 vs. CM group. CM, conditioned medium; F-CM, flagellin-conditioned medium; LPS, lipopolysaccharide; TNF, tumor necrosis factor; IL, interleukin; MCP, monocyte chemoattractant protein.

Discussion

The causes of ALI include a disruption in the balance of pro- and anti-inflammatory factors, an imbalance between oxidation and anti-oxidation, disorders of the coagulation system and abnormal apoptosis (31). In particular, LPS, which comprises part of the wall of Gram-negative bacteria, can increase the expression of inflammatory factors, promote inflammatory cells to enter lung tissue and increase alveolar-capillary barrier permeability (28,32). Thus, the focus of treating ALI is to inhibit the LPS-induced inflammatory reaction (29). Despite their effectiveness in regulating inflammation, the heavy use of steroids is associated with severe adverse effects and may increase the risk of infection (32). Moreover, other therapeutic methods, such as mechanical ventilation, neuromuscular blocking agents and antimicrobial therapy, are only moderately effective (7).

Previous clinical studies reported that the use of MSCs is safe for patients with ARDS, with a superior anti-inflammatory effect (7,33). In addition, as the principal beneficial effects of MSCs are mediated via paracrine mechanisms, CM from MSCs may be a promising therapeutic approach to ALI (26). However, the concentration of biofactors in CM is considered insufficient for therapy; thus, CM must be concentrated for treatment. Furthermore, external stimuli can alter the composition of paracrine factors from MSCs, making them more useful in the treatment of particular diseases. For instance, exposing MSCs to pro-inflammatory cytokines for a short time, which mimics the acute inflammatory microenvironment, significantly enhances their anti-inflammatory effects (13,19). Flagellin is a potent activator of a broad range of cell types involved in innate and adaptive immunity. Following stimulation of TLR5-expressing cells with flagellin, a signaling cascade is triggered, which involves phosphorylation of IL-1 receptor-associated kinase 1, leading to activation of mitogen-activated protein kinase kinases and inhibitor of NF-κB kinases, ultimately activating inflammatory protein production via NF-κB and p38 mitogen-activated protein kinase (24). However, most previous studies have focused on the beneficial effects of ADSCs activated by cytokines such as TNF-α, IL-1β and IFN-γ (17,19,34). Thus, limited information is available regarding the immunomodulatory properties of flagellin-exposed ADSCs.

In the present study, the MSC markers CD73, CD90 and CD105 were highly expressed in flagellin-activated ADSCs, while the expression levels of the endothelial cell marker CD31 and the hematopoietic cell marker CD45 were low, suggesting that flagellin preconditioning at 100 ng/ml for 2 days did not affect the characteristics of ADSCs. Since FBS may induce severe immune rejection, serum-free DMEM was used to collect paracrine factors from MSCs. Although serum-free incubation decreased the viability of ADSCs, flagellin-activated ADSCs retained 73% viability after 48 h. Previous studies have reported that LPS, a TLR5 ligand, could protect MSCs from serum deprivation by stabilizing cell membrane and activating ERK and PI3K/Akt signaling pathways (35,36). Given the similarities between TLR4 and TLR5, flagellin may exert its beneficial effects in similar ways (23). Therefore, ultrafiltration units with a 3-kDa molecular weight cutoff were used to concentrate CM, to retain the highest number of inflammatory regulators as possible.

Alveolar macrophages include tissue-resident and recruited macrophages, and account for 80% of permanent alveolar cells (6). Alveolar macrophages phagocytize foreign bodies, recognize antigens and secrete a variety of inflammatory factors, forming the first line of the host immune defense (6). Macrophages exhibit high functional plasticity and may be induced into different phenotypes when subjected to different biosignals (37). In general, macrophages may be divided into classically activated (M1) and alternatively activated (M2) macrophages. M1 macrophages usually serve a pro-inflammatory role and can secrete several pro-inflammatory factors, such as TNF-α, IL-1, and IL-6, which contribute to the clearance of pathogens (38). In contrast, M2 macrophages secrete anti-inflammatory factors, such as IL-10, and promote angiogenesis, wound healing and tissue remodeling (39). The present in vitro study demonstrated that CM from flagellin-activated ADSCs significantly enhanced the expression of the anti-inflammatory M2 macrophage marker CD163 and decreased the expression of the inflammatory M1 macrophage marker CD86 compared with CM from the regular ADSC model. These results indicated the superior ability of F-CM to regulate macrophage characteristics and immunosuppression.

To investigate the mechanism underlying the superior anti-inflammatory effects of F-CM, the cytokines responsible for the immunomodulatory properties of the CM were investigated. Since the composition of CM is complex, flagellin preconditioning significantly increased the levels of both pro-inflammatory (IL-6 and IL-8) and anti-inflammatory cytokines (TGF-β). Although IL-6 is considered as a pro-inflammatory cytokine, it may exert both pro-inflammatory and anti-inflammatory roles (37). Previous studies revealed that IL-6 can induce macrophages into the M2 phenotype in an animal model (37,40). Furthermore, Liu et al (15) and Chen et al (41) reported that MSCs reprogramed macrophages into the M2 phenotype mainly by secreting TGF-β. Therefore, the increased IL-6 and TGF-β levels may partly explain the beneficial effects of flagellin preconditioning. In the current study, western blotting results indicated that NF-κB signaling was activated in ADSCs after treatment with 100 ng/ml flagellin for 30 min, which may partly explain the reason for the altered cytokines levels.

An LPS-induced ALI mouse model was used in the present study, as its reliability and reproducibility have been previously confirmed (42). The results of the examination of the lung exudate, H&E staining and cell counting in the BALF identified that both CM and F-CM significantly prevented LPS-induced lung injury in mice, and the effects of F-CM treatment were superior. Furthermore, the levels of inflammation-related factors in the BALF were detected using ELISA. TNF-α is an inflammatory factor that increases rapidly in the early stages of inflammation (43). TNF-α can mobilize bone marrow leukocytes into the blood circulation, aggregate monocytes in the lung tissue and promote the release of inflammatory factors, such as IL-1, IL-6 and colony-stimulating factor (44). The role of IL-1 has numerous similarities with that of TNF-α in ALI caused by inflammation (45). For example, IL-1 can initiate inflammatory reactions together with TNF-α, and enhance lung injury induced by TNF-α, but it does not itself cause lung injury (46). IL-6 can further promote macrophage aggregation, enhance phagocytosis by macrophages and promote the expression of various inflammatory factors (47). MCP-1 is another inflammatory chemokine that regulates the chemotactic response of leukocytes to inflammatory stimuli (48). Moreover, IL-10 is a general immunosuppressive cytokine, which has been shown to promote the conversion of M1 macrophages into M2 macrophages, as well as counteract the pro-inflammatory effects of TNF-α (47).

In the present study, LPS challenge not only increased the levels of pro-inflammatory cytokines, but also those of the anti-inflammatory cytokine IL-10 in mice, suggesting that the anti-inflammation process was activated. F-CM significantly decreased the production of TNF-α, IL-1, IL-6 and MCP-1; however, there were no significant differences in IL-10 among the LPS, CM and F-CM groups, possibly because the alleviated inflammation reaction was associated with lowered production of anti-inflammatory factors. Previous studies have reported that drug interventions, such as Jaceosidin and Ginsenoside Rh2, greatly decreased the levels of pro-inflammatory cytokines, but do not alter the level of IL-10 (28,49,50). Thus, the TNF-α/IL-10 ratio may be a more suitable indicator of the anti-inflammatory effect. However, IL-10 can be produced by almost all leukocytes, including macrophages, dendritic cells, neutrophils, natural killer cells, B cells and T cells (43). Currently, the effects of CM on cells other than macrophages remain unknown. In addition, the reasons for the absence of an increase in IL-10 require further investigation.

Despite the promising results, there were several limitations to the present study. Due to the limited time and the complexity of CM (including exosomes, cytokines and microRNAs), only the changes in nine important cytokines in the CM following flagellin preconditioning were examined. Although it was found that three cytokines changed significantly after ADCSs were preconditioned with flagellin, which one serves the key role in regulating the polarization of macrophages and inhibiting inflammation is yet to be elucidated. Furthermore, only one F-CM administration mode was evaluated in the present study, and additional experiments are required to determine the optimal dosage and timing of administration.

In conclusion, the results of the present study suggested that flagellin preconditioning significantly enhanced the beneficial effects of CM from ADSCs against LPS-induced lung injury in mice. These findings may indicate a promising novel approach to the treatment of inflammation-induced ALI.

Acknowledgements

Not applicable.

Funding

This study was supported by the Shanghai International Science and Technology Cooperation Fund (grant no. 18410721300).

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

XD and RL collaborated to design the study. RL was responsible for experiments. YL analyzed the data and prepared the manuscript. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable (MSCs were purchased from the Type Culture Collection of the Chinese Academy of Sciences).

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References