Sphingosine-1-phosphate derived from PRP-Exos promotes angiogenesis in diabetic wound healing via the S1PR1/AKT/FN1 signalling pathway

Abstract

Background: Sphingosine-1-phosphate (S1P), a key regulator of vascular homeostasis and angiogenesis, is enriched in exosomes derived from platelet-rich plasma (PRP-Exos). However, the potential role of PRP-Exos-S1P in diabetic wound healing remains unclear. In this study, we investigated the underlying mechanism of PRP-Exos-S1P in diabetic angiogenesis and wound repair.

Methods: Exosomes were isolated from PRP by ultracentrifugation and analysed by transmission electron microscopy, nanoparticle tracking analysis and western blotting. The concentration of S1P derived from PRP-Exos was measured by enzyme-linked immunosorbent assay. The expression level of S1P receptor1-3 (S1PR1-3) in diabetic skin was analysed by Q-PCR. Bioinformatics analysis and proteomic sequencing were conducted to explore the possible signalling pathway mediated by PRP-Exos-S1P. A diabetic mouse model was used to evaluate the effect of PRP-Exos on wound healing. Immunofluorescence for cluster of differentiation 31 (CD31) was used to assess angiogenesis in a diabetic wound model.

Results: In vitro, PRP-Exos significantly promoted cell proliferation, migration and tube formation. Furthermore, PRP-Exos accelerated the process of diabetic angiogenesis and wound closure in vivo. S1P derived from PRP-Exos was present at a high level, and S1PR1 expression was significantly elevated compared with S1PR2 and S1PR3 in the skin of diabetic patients and animals. However, cell migration and tube formation were not promoted by PRP-Exos-S1P in human umbilical vein endothelial cells treated with shS1PR1. In the diabetic mouse model, inhibition of S1PR1 expression at wounding sites decreased the formation of new blood vessels and delayed the process of wound closure. Bioinformatics analysis and proteomics indicated that fibronectin 1 (FN1) was closely related to S1PR1 due to its colocalization in the endothelial cells of human skin. Further study supported that FN1 plays an important role in the PRP-Exos-S1P-mediated S1PR1/protein kinase B signalling pathway.

Conclusions: PRP-Exos-S1P promotes angiogenesis in diabetic wound healing via the S1PR1/protein kinase B/FN1 signalling pathway. Our findings provide a preliminary theoretical foundation for the treatment of diabetic foot ulcers using PRP-Exos in the future.

Keywords: Diabetic foot ulcer; Exosomes; Platelet-rich plasma; Sphingosine-1-phosphate; Wound healing.

Figure 1

Identification of PRP-Exos. (a) The morphology of PRP-Exos was viewed by transmission electron microscope. (b) The particle size distribution of PRP-Exos was measured by nanoparticle tracking analysis; mean ± SD: 88.01 ± 21.01 nm. (c, d) Western blotting and quantitative analysis of PRP-Exos-specific markers CD63, flotillin and TSG101, and PRP-AS-specific markers CD41 and calnexin. ^***p< 0.001, ^**p < 0.01, ^*p < 0.05. PRP-Exos compared with PRP-AS. PRP platelet-rich plasma, PRP-AS activated supernatants of PRP, PRP-Exos exosomes derived from PRP

Figure 2

Evaluation of wound healing treated with PRP-Exos. (a, b) Representative image and mode pattern showing the process of wound closure in the NC, PRP-AS and PRP-Exos groups at days 0, 3, 7 and 11 after the operation. (c) Quantification of wound closure rates. ^****p < 0.0001, comparing PRP-Exos with NC; ^**p < 0.01, comparing PRP-Exos with PRP-AS. (d, e) Images of H&E staining and quantification of the degree of re-epithelization indicated by horizontal black lines. ^****p< 0.0001, comparing PRP-Exos with NC; ^**p< 0.01, comparing PRP-Exos with PRP-AS. (f, g) Images and quantification of immunofluorescence staining for CD31 labelled with Alexa Fluor 488; the cell nucleus was stained with DAPI. ^**p < 0.01, comparing PRP-Exos with NC; ^*p < 0.05, comparing PRP-Exos with PRP-AS. NC normal control, PRP platelet-rich plasm, PRP-AS activated supernatants of PRP, PRP-Exos exosomes derived from PRP

Figure 3

Measurement of S1P in PRP-Exos and S1PR in vivo. (a) Concentration of S1P in PRP-Exos, PRP-AS and NC groups was measured by ELISA. (b) Relative mRNA levels of S1PR1–3 in human skin were analysed. (c) Relative mRNA levels of S1PR1–3 in the mouse skin were analysed. (d) Data from the Human Protein Atlas verified that S1PR1 was rich in endothelial cells of human skin. (e, f) Western bloting and quantification showing the relative levels of S1PR1–3 in HUVECs; ^****p < 0.0001, ^**p < 0.01, ns no significant difference. S1P sphingosine-1-phosphate, S1PR sphingosine-1-phosphate receptor, NG normal glucose, HG high glucose, HP high permeation, NC normal control, PRP platelet-rich plasm, PRP-AS activated supernatants of PRP, PRP-Exos exosomes derived from PRP

Figure 4

Assessment of HUVEC function. (a) Endocytosis assay indicated that exosomes labelled with PKH26 were carried into HUVECs within 6 h; F-actin represents the cytoskeleton. (b) CCK-8 assays were conducted to find the optimal time for HUVECs treated with PRP-Exos; H0 refers to zero time. (c) EdU assays were conducted for evaluation of HUVEC proliferation. Positive rate = (cells labelled with Alexa Fluor 555/cells labelled with DAPI) x 100%. (d) HUVEC cell cycle were measured by flow cytometry. Percentage of S phase cells was determined. (e, f) Cell scratch assays were observed in living cells at 0, 10, 20, 30 and 40 h. (g, h) Transwell assays were conducted for evaluation of HUVEC migration. (i–l) Tube formation assays were conducted and the segments length was used for evaluation of HUVEC tube formation capacity. ^****p < 0.0001, ^**p < 0.01, ^*p < 0.05, ns no significant difference. S1P sphingosine-1-phosphate, shS1PR1 sphingosine-1-phosphate receptor 1 shRNA, NC normal control, PRP platelet-rich plasm, PRP-AS activated supernatants of PRP, PRP-Exos exosomes derived from PRP

Figure 5

FN1 and p-AKT are involved in PRP-Exos-S1P-induced angiogenesis. (a) Proteomic sequencing analysis showed that FN1 was significantly lower in diabetic skin compared with nondiabetic skin. (b) Bioinformatic data from the Human Protein Atlas confirmed that FN1 was rich in endothelial cells of human skin. (c–f) The relative levels of FN1, p-AKT and VEGF-A in HUVECs after different treatments were measured by western blotting in the PRP-Exos 50 μg/ml vs PRP-AS 50 μg/ml groups. (g–k) The relative levels of FN1, p-AKT, VEGF-A and S1PR1 in HUVECs after different treatments were measured by western blotting in the shS1PR1 vs vector, shS1PR1 + S1P vs vector + S1P, shS1PR1 + PRP-Exos vs vector + PRPExos groups. ^****p < 0.0001, ^***p < 0.001, ^**p < 0.01. FN1 fibronectin 1, p-AKT phosphorylated protein kinase B, t-AKT total protein kinase B, VEGF-A vascular endothelial growth factor A, S1P sphingosine-1-phosphate, S1PR1 sphingosine-1-phosphate receptor 1, NC normal control, PRP platelet-rich plasm, PRP-AS activated supernatants of PRP, PRP-Exos exosomes derived from PRP

Figure 6

Evaluation of wound healing treated with adenovirus-shS1PR1, PRP-Exos and S1P. (a, b) Representative images and mode pattern showing the process of wound closure treated with GFP, shS1PR1, GFP + S1P, shS1PR1 + S1P, GFP + PRP-Exos, shS1PR1 + PRP-Exos at days 0, 3, 7 and 11 after the operation. (c–e) Quantification of wound closure rates in the GFP vs shS1PR1, GFP + S1P vs shS1PR1 + S1P and GFP + PRP-Exos vs shS1PR1 + PRP-Exos groups. (f–h) Images and quantification of immunofluorescence staining for CD31 labelled with Alexa Fluor 488, and FN1 labelled with Cy3. The cell nucleus was stained with DAPI. ^****p< 0.0001, ^**p < 0.01, ^*p < 0.05. GFP green fluorescent protein, S1P sphingosine-1-phosphate, shS1PR1 sphingosine-1-phosphate receptor 1 shRNA, PRP platelet-rich plasma, PRP-Exos exosomes derived from PRP, FN1 fibronectin 1

Figure 7

PRP-Exos-S1P regulates FN1 via the AKT signalling pathway. (a–d) The relative levels of FN1, p-AKT and VEGF-A in HUVECs after different treatments were measured by western blotting in the LY294002 vs NC and LY294002 + PRP-Exos vs PRP-Exos groups. (e–h) The relative levels of FN1, p-AKT and VEGF-A in HUVECs after different treatments were measured by western blotting in the SC79 vs NC and SC79 + shS1PR1 vs shS1PR1 groups. (i, j) Immunofluorescence for FN1 was measured in the LY294002 vs NC and LY294002 + PRP-Exos vs PRP-Exos groups. (k, l) Immunofluorescence for FN1 was measured in the SC79 vs NC and SC79 + shS1PR1 vs shS1PR1 groups. ^***p < 0.001, ^**p < 0.01, ^*p < 0.05. LY294002 inhibitor of AKT phosphorylation, SC79 agonist of AKT phosphorylation, NC normal control, PRP platelet-rich plasm, PRPExos exosomes derived from PRP, FN1 fibronectin 1, p-AKT phosphorylated protein kinase B, t-AKT total protein kinase B, VEGF-A vascular endothelial growth factor A

Figure 8

FN1 directly regulates VEGF levels. (a, b) The relative levels of VEGF-A were measured by western blotting in the si-FN1 vs si-NC groups. (c) ELISA assays were conducted for measurement of VEGF-A levels in the HUVEC supernatants after different treatments. (d, e) Immunofluorescence for VEGF-A labelled with Alexa Fluor 488 was conducted. ^****p < 0.0001, ^**p < 0.01. si-NC negative control siRNA, si-FN1 fibronectin 1 siRNA, FN1 fibronectin 1, VEGF-A vascular endothelial growth factor A

Figure 9

The molecular mechanism pattern was produced by Biorender (www.biorender.com). VEGF Vascular endothelial growth factor, S1P sphingosine-1-phosphate, S1PR1 sphingosine-1-phosphate receptor 1, PRP platelet-rich plasma, PRP-AS activated supernatants of PRP, PRP-Exos exosomes derived from PRP, FN1 fibronectin 1, AKT protein kinase B