Human adipose-derived stem cells promote seawater-immersed wound healing via proangiogenic effects

The angiogenic effect of hADSCs in seawater-immersed wounds

The level of wound vascularization is an important indicator for evaluating the effects of wound healing. We previously reported that seawater immersion significantly delays wound healing and the treatment of hADSCs can promote wound healing. Therefore, we further analyzed the effects of seawater and hADSCs on wound angiogenesis. Fluorescence staining of frozen sections (Figure 1A) showed that hADSCs could differentiate into vascular endothelial cells in damaged wounds to promote wound healing. Subsequently, CD31 immunohistochemical staining was performed on the wound sections to evaluate repaired wound vascularization. The results (Figure 1B–1D) showed that the numbers of new blood vessels in the control and the SW + hADSC groups were significantly higher than those in the SW group on days 14 and 21 (P < 0.001). These results suggest that seawater immersion can significantly delay the neovascularization of wounds, while hADSCs reduce the damage caused by seawater immersion and promote wound vascularization.

Identification of DEGs in hADSCs

To explore the potential vascularization mechanism of hADSCs to promote the healing of seawater immersion wounds, a gene expression microarray dataset (GSE48228) was downloaded. The data were normalized using the limma R package, and the mRNA expression levels of the two groups were compared (Figure 2A). The screening criteria were set to a p value of < 0.05, and |log FC| > 2, and the results showed 4,642 upregulated and 5,541 downregulated genes, and volcano maps were used to show the distribution of all DEGs (Figure 2B). The top 10 DEGs were selected and are listed in Table 1. DEGs were sorted according to the p value; DCN, XIST, and COX7A1 were the top three upregulated DEGs, and LCN2, CALB1, and COX7B2 were the top three downregulated DEGs.

Protein–protein interaction (PPI) network and enrichment analysis of the DEGs

PPI network and enrichment analyses are important for understanding the biological functions and signaling pathways of genes and the functional connections between proteins. The STRING database was used to perform PPI network analysis on the top 2,000 most upregulated and downregulated DEGs. Then, the degree algorithm was used to obtain the hub genes with the highest degree of the first 20 nodes (Table 2) and these were visualized using Cytoscape software (Figure 2C, 2D). TP53, FN1, CTNNB1, STAT3, and COL1A1 as well as CDK1, CCNB1, CCNA2, AURKB, and CDC20 were the top five hub genes for the upregulated and downregulated DEGs, respectively.

Furthermore, the database for annotation, visualization, and integrated discovery (DAVID) was used for gene ontology (GO) annotation and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis to further reveal the enrichment status of DEGs in the molecular functions (MFs), biological processes (BPs), cell components (CCs), and pathways. Regarding BPs (Figure 3A, 3E), the upregulated DEGs were mainly involved in GO:0030198 (extracellular matrix (ECM) organization), GO:0007155 (cell adhesion), and GO:0001525 (angiogenesis), while the downregulated DEGs were mainly involved in GO:0051301 (cell division), GO:0007067 (mitotic nuclear division), and GO:0007062 (sister chromatid cohesion). Regarding CCs (Figure 3B, 3F), the majority of upregulated DEGs were components of GO:0031012 (ECM), GO:0005578 (proteinaceous ECM), and GO:0005925 (focal adhesion), whereas the downregulated DEGs were mainly components of GO:0005654 (nucleoplasm), GO:0000777 (condensed chromosome kinetochore), and GO:0000775 (chromosome, centromeric region). For the MFs (Figure 3C, 3G), the upregulated DEGs were mainly involved in GO:0005178 (integrin binding), GO:0050840 (ECM binding), and GO:0008201 (heparin binding), whereas the downregulated DEGs were mainly involved in GO:0005515 (protein binding), GO:0044822 (poly(A) RNA binding), and GO:0043142 (single-stranded DNA-dependent ATPase activity). KEGG pathway analysis was used to define the pathways and functions of the DEGs (Figure 3D, 3H). Most upregulated DEGs were significantly enriched in hsa04512 (ECM-receptor interaction), hsa04151 (phosphoinositide 3 kinase (PI3K)-Akt signaling pathway), and hsa05205 (proteoglycans in cancer), whereas the downregulated DEGs were mainly enriched in hsa03030 (DNA replication), hsa04110 (cell cycle), and hsa00240 (pyrimidine metabolism).

Relationships between DEGs and angiogenesis, necroptosis, pyroptosis, ferroptosis, and anti-inflammatory activities

The top three most upregulated or downregulated DEGs with protein-coding functions and the top five hub genes were selected, and the correlations between the DEGs and hub genes and the markers of angiogenesis, necroptosis, pyroptosis, ferroptosis, and anti-inflammatory activities in skin tissue were further analyzed using the GEPIA database. Notably, the vascularization correlation results of the upregulated DEGs and hub genes (Figure 4 and Supplementary Figure 1) suggested that almost all upregulated DEGs and hub genes were strongly positively correlated with vascularization markers (P < 0.001, R > 0). Regarding the downregulated DEGs (Figure 5), most were not significantly correlated with vascular markers. However, LNC2 was positively correlated with VEGFA expression (P ≤ 0.01, R = 0.13). CALB1was negatively correlated with VEGFA (P < 0.01, R = –0.12) and positively correlated with PECAM1 (P < 0.001, R = 0.17). Regarding the downregulated hub genes, almost all were positively correlated with VEGFA (P < 0.05, R > 0; Supplementary Figure 2A–2E). Interestingly, all downregulated hub genes were strongly negatively correlated with VWF and PECAM1 (P < 0.001, R < 0; Supplementary Figure 2F–2O).

For the correlations of the upregulated DEGs/hub genes and the markers of necroptosis, pyroptosis, ferroptosis, and anti-inflammatory activities (Table 3 and Supplementary Table 1), the results indicated that most of the genes were strongly positively correlated with the markers of necroptosis, ferroptosis, and anti-inflammatory activities (P < 0.05, R > 0), and DCN, COX7A1, and FNA1 were highly negatively correlated with pyroptosis markers (P < 0.01, R < 0), whereas COL3A1 and most other upregulated hub genes were positively correlated with the pyroptosis markers (P < 0.05, R > 0). For correlations of the downregulated DEGs/hub genes and the markers of necroptosis, pyroptosis, ferroptosis, and anti-inflammatory activities (Table 4 and Supplementary Table 2), the results showed that CALB1 and most of the hub genes were highly positively correlated with the necroptosis markers of caspase-8 (Casp8; P < 0.001, R > 0), whereas LCN2, CCNB1, and CDC20 were negatively correlated with the necroptosis markers of receptor-interacting protein kinase (RIPK1; P < 0.05, R < 0). LCN2, COX7B2, and all downregulated hub genes were positively correlated with pyroptosis markers (P < 0.05, R > 0), while CALB1 was negatively correlated (P < 0.01, R < 0). LCN2 and CALB1 were significantly positively correlated with the ferroptosis markers (P < 0.001, R > 0), while COX7B2 and all downregulated hub genes were not significant (P > 0.05). The downregulated hub genes of CCNB1, AURKB, and CDC20 were strongly negatively correlated with the anti-inflammatory markers of interleukin (IL)-10 (P < 0.01, R < 0), while all downregulated DEGs and other downregulated hub genes were not significantly correlated with the anti-inflammatory markers (P > 0.05).

Effects of seawater and hADSCs on human umbilical vein endothelial cells (HUVECs)

The proliferation and migration ability of vascular endothelial cells play a vital role in wound vascularization. Therefore, the effects of the coculture of seawater and hADSCs on the proliferation and migration of HUVECs were investigated. The results of the cell counting kit-8 (CCK8) assay (Figure 6A) showed that cell proliferation activity was significantly inhibited at 10% and 20% seawater concentrations (P < 0.01), and the inhibitory effect continued to exist as culture time increased. Therefore, 10% seawater concentration was used as the experimental concentration for subsequent experiments.

The expression of VEGFR is closely related to the vascularization of skin wounds thus, the effects of seawater and hADSCs on the protein and mRNA expression levels of VEGFR in HUVECs were detected (Figure 6C) [10, 11]. Expression levels of VEGFR were significantly lower in the SW group than those in the control group but were significantly higher in the SW + hADSC group (P < 0.01).

5-Ethynyl-2′-deoxyuridine (EdU) and cell scratch assays were used to evaluate the effects of seawater and hADSCs on HUVEC proliferation and migration, respectively. The EdU results (Figure 6D, 6E) indicated that the proliferation efficiency of HUVECs was significantly inhibited in the simulated seawater environment, while co-culture with hADSCs significantly improved the proliferation activity (P < 0.01). The cell scratch test results (Figure 6F, 6G) showed that compared with the control group, HUVEC scratches in the SW group had a significant delay in closure at 12 h, and co-culture with hADSCs relieved the effect of seawater on cell migration and significantly promoted cell migration ability (P < 0.01). Moreover, HUVEC scratches in the control and SW + hADSC groups were mainly healed at 36 h, while part of the unhealed area remained in the SW group.

Immunofluorescence observations

To observe the differentiation effects of hADSCs, the cells used for injection therapy were labeled with red fluorescence with CM-DiL fluorescent dye (Invitrogen, Thermo Scientific, USA) in advance. On the day 14 after injection treatment, wounded skin tissues of the SW + hADSCs group were immediately excised and frozen sections of 10 μM were prepared. Frozen sections were dried for 15 min at room temperature, then soaked for 10 min in PBS solution to remove tissue glue and incubated with CD31 antibody (10 μg/mL, R&D Systems, USA) at 4°C overnight. FITC-labeled goat anti-mouse IgG (H + L) cross-adsorbed secondary antibody (Beyotime Biotechnology Company, Jiangsu, China) and 1X Hoechst33342 (Ribo, Guangzhou, China) were incubated with the slices for 4 h and 30 min, respectively. The slices were then washed with PBS and mounted with antifade mountant (ProLong, Thermo Scientific, USA). Images were acquired using a Zeiss fluorescence microscope (HLA100, Shanghai, China).

Immunohistochemistry observations

To observe the vascularization effect of hADSCs in repairing seawater-immersed wounds, wound skin tissues on days 14 and 21 were used to prepare stained sections. The wound tissues were initially made into paraffin sections through formaldehyde treatment, dehydration, paraffin embedding, and sectioning. Paraffin sections were dewaxed and antigens were repaired, blocked, and incubated with CD31 primary antibody (10 μg/mL, R&D Systems, USA) at 4°C overnight. Then, the sections were incubated for 4 h with secondary antibodies (Cell Signaling Technology, MA, USA) and washed with PBS. The sections were incubated with DAB chromogen and counterstained with hematoxylin for immunohistochemical analysis. Sections were photographed using an optical microscope (Olympus, Japan) and quantitative analysis was performed using Image-Pro Plus software (Medical Cybernetics, USA).

Cell proliferation assay

HUVECs were obtained from the American Type Culture Collection (Manassas, VA, USA) and cultured in high-glucose Dulbecco’s modified Eagle’s medium (HyClone, Utah, USA) mixed with 10% fetal bovine serum (Gibco, Grand Island, USA), and the medium was changed every two-to-three days.

To explore the effect of the co-culture of seawater and hADSCs on cell proliferation. HUVECs were seeded into 96-well plates at a density of 2000 cells/well. After the cells were cultured for 12 h, the medium was replaced with a high-glucose culture medium containing different proportions of seawater (volume fractions of 5%, 10%, and 20%) for 24, 48, or 72 h. Artificial seawater was configured as previously reported. At each observation time point, 10 μL of CCK8 reagent (Beyotime Biotechnology Company, Shanghai, China) was added to each well and incubated for 2 h. At the end of incubation, absorbance was measured at 450 nm using an enzyme-labeled instrument (Tecan, Shanghai, China).

First, 0.4-μM Transwell polycarbonate membranes (Corning, New York, USA) were used to realize the co-culture of hADSCs and HUVECs (Figure 6B), and the proliferation rate of HUVECs was detected using an EdU kit (Ribo, Guangzhou, China). HUVECs were seeded into 24-well plates at 2 × 10⁵ cells/well. When confluency reached 70–80%, the cells were divided into three groups; control, SW, and SW + hADSC. Cells of the control group were cultured in complete medium, and cells of the SW and SW + hADSC groups were cultured in complete medium mixed with seawater. After 48 h of incubation, the cell proliferation ratio of each group was determined using the EdU kit. Fluorescence staining was performed according to the manufacturer's protocol. The results were photographed with a Zeiss fluorescence microscope and analyzed using ImageJ software (Bethesda, MD, USA).

Cell migration assay

A cell scratch assay was used to explore the effect of the co-culture of seawater and hADSCs on cell migration. HUVECs were inoculated in six-well plates with 1 × 10⁶ cells/well and cultured to 80–90% confluency. The groupings and treatments were the same as those used in the EdU assay. The cell monolayer was scratched with a 10 μL sterile pipette tip quickly and washed two-to-three times with PBS. At 0, 12, 24, and 36 h observation time points, the scratch area was photographed using a Zeiss inverted phase contrast microscope (Observer D1, Shanghai, China) and measured using ImageJ software. The following formula was used to assess cell migration ability: $$\text{Migration area (\%)}=\frac{\text{S}0-\text{Sx}}{\text{S}0}\times 100\%$$ where S0 represents the initial scratch area (t = 0 h) and Sx represents the residual scratch area at the time of measurement (t = n h).

Quantitative real-time polymerase chain reaction (qRT-PCR)

HUVEC groupings and treatments were the same as those in the EdU assay. HUVECs were co-cultured with seawater and hADSCs for 72 h, then total RNA from each group was extracted with Trizol reagent (Vazyme, Jaingsu, China). Reverse transcription of 2 μg of total RNA into cDNA was used as a template for qRT-PCR. Using GAPDH as an internal reference gene, the normalized relative expression level of VEGFR was calculated using the 2^-ΔΔCt method. GAPDH forward primer sequence: 5'-TGTGGGCATCAATGGATTTGG-3'; reverse primer sequence: 5'-ACACCATGTATTCCGGGTCAAT-3'. VEGFR forward primer sequence: 5'-AGCAAAGGGTGGAGGTG-3'; reverse primer sequence: 5'-ACATAAATGACCGAGGC-3'. The reaction conditions were 95°C pre-denaturation for 30 s followed by 40 cycles of 95°C for 10 s, 60°C for 10 s, and 72°C for 10 s.

Western blotting

The proteins for western blotting from each sample were extracted and separated by PAGD and transferred to a PVDF membrane (Cell Signaling Technology, MA, USA) at 300 mA for 2 h. The membrane was initially incubated in blocking buffer (5% bovine serum albumin) for 2 h at room temperature then incubated overnight at 4°C with primary GAPDH (1:1000, Cell Signaling Technology, MA, USA) or VEGF-A antibody (1:1000, Abcam, Cambridge, MA, UK) followed by 2 h of incubation with the appropriate horseradish peroxidase-conjugated secondary antibodies (1:3000, Beyotime, China). Protein expression was visualized by chemiluminescence using an Alpha Imager scanner (Tecan, Thermo Fisher Scientific, USA).

Differential analysis of gene expression in hADSCs

The gene expression microarray dataset (GSE48228), comprised of samples from six hADSCs, one fibroblast, and one endothelial cell, was downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo). The dataset was based on the GPL6480 platform (Agilent-014850 Whole Human Genome Microarray), and the mRNA expression data of hADSCs and fibroblast control were downloaded and used for this study. The platform data were converted using R language software and standardized using the limma R package’s array function (http://www.bioconductor.org/). The differential analysis of mRNAs in hADSCs compared with fibroblasts was performed using the GEO2R analysis tool. An mRNA with a p value of < 0.05 and a base-2 log FC value greater than ± 2 was defined as a DEG, and the integrated DEG lists of upregulated and downregulated genes were saved for subsequent analysis.

Integration analysis of the DEGs

The top 2000 most upregulated and downregulated DEGs were selected, and the PPI networks were analyzed using the STRING database (https://string-db.org/) and visualized using Cytoscape (version 3.7.2) software [43]. Then, the degree analysis method in the CytoHubba plug-in was used to calculate the grade nodes in the PPI network, in which highly connected nodes were defined as hub genes, and the top 20 hub genes were selected and visualized using Cytoscape.

The DAVID (https://david.ncifcrf.gov/home.jsp) provides a comprehensive set of functional annotation tools to understand the biological meaning behind a large list of genes [44]. In this study, the DAVID was used for GO annotation and KEGG pathway enrichment analysis of the integrated DEGs. GO annotation analysis of DEGs involved the BP, CC, and MF.

Correlations between DEGs and gene markers of angiogenesis, necroptosis, pyroptosis, ferroptosis, and anti-inflammatory activities in GEPIA

The relationship between DEGs and gene markers of proangiogenic activity, necroptosis, pyroptosis, ferroptosis, and anti-inflammatory activity was determined using the GEPIA database (http://gepia.cancer-pku.cn/index.html) [45]. The gene markers, including VEGFA, VWF, PECAM1, Casp8, RIPK1, IL-1α, Casp1, Glutathione peroxidase-4 (GPX4), nuclear receptor coactivator-4 (NCOA4), IL-4, and IL-10, were selected from previous studies [35, 36, 46, 47]. The GEPIA dataset, an interactive online platform with normal sample information from the Genotype Tissue Expression project, was used for the correlation analysis of the DEGs and gene markers in normal skin tissues. The gene expression level was adjusted with log2 TPM, and Spearman’s method was used to determine the correlation coefficient. DEGs are plotted on the x-axis, while the gene markers of interest are plotted on the y-axis.

Statistical analysis

Data were analyzed with SPSS 17.0 and are presented as means ± standard deviation. Significance was determined by ANOVA, and the correlation analysis of gene expression was evaluated using Spearman’s rank test. A p value of < 0.05 was considered significant.