EDIL3 influenced the αvβ3‐FAK/MEK/ERK axis of endothelial cells in psoriasis

Abstract One of the earliest events in the development of psoriatic lesion is a vascular network expansion. The abnormal vascular network is associated with increased endothelial cells (ECs) survival, proliferation, adhesion, migration, angiogenesis and permeability in psoriatic lesion. Our previous study demonstrated that epidermal growth factor‐like repeats and discoidin I‐like domains 3 (EDIL3) derived from psoriatic dermal mesenchymal stem cells (DMSCs) promoted cell–cell adhesion, migration and angiogenesis of ECs, but the molecular mechanism of upstream or downstream has not been explored. So, this study aimed to explore the association between EDIL3 derived from DMSCs (DMSCs‐derived EDIL3) and psoriasis‐associated angiogenesis. We injected recombinant EDIL3 protein to mouse model of psoriasis to confirm the roles of EDIL3 in psoriasis. Besides, we employed both short‐interference RNA (si‐RNA) and lentiviral vectors to explore the molecular mechanism of EDIL3 promoting angiogenesis in psoriasis. In vivo, this research found that after injected recombination EDIL3 protein, the epidermis thickness and microvessel density were both elevated. EDIL3 accelerated the process of psoriasis in the IMQ‐induced psoriasis‐like mouse model. Additionally, we confirmed that in vitro DMSCs‐derived EDIL3 is involved in the tube formation of ECs via αvβ3‐FAK/MEK/ERK signal pathway. This suggested that DMSCs‐derived EDIL3 and αvβ3‐FAK/MEK/ERK signal pathway in ECs play an important role in the pathogenesis of psoriasis. And the modification of DMSCs, EDIL3 and αvβ3‐FAK/MEK/ERK signal pathway will provide a valuable therapeutic target to control the angiogenesis in psoriasis.

occurs before epidermal changes and persists after clearance of the clinical lesions. 4 The dilatation and tortuous of vascular network is associated with increased endothelial cells (ECs) proliferation, adhesion, migration, angiogenesis and permeability in psoriatic lesions. [5][6][7][8] Recently, researchers found that the dermal mesenchymal stem cells (DMSCs) in psoriatic lesions produced more angiogenic and proinflammatory mediators and involved in course of psoriasis. [9][10][11] Our previous study demonstrated that psoriatic DMSCs promoted cellcell adhesion, migration and angiogenesis of ECs, but the molecular mechanism of upstream or downstream has not been explored. 12,13 Epidermal growth factor-like repeats and discoidin I-like domains 3 (EDIL3) is an extracellular matrix protein, also named as developmental endothelial locus-1 (Del-1), composed of two discoidin I-like domains and three EGF-like repeats, the second of which contains an Arg-Gly-Asp (RGD) motif. 14 The RGD motif can bind with integrin, then effect ECs functions including adhesion, migration and angiogenesis. 7,9,14 During angiogenesis in early embryogenesis, 7 tumour 15 and ischemic tissue, 16 EDIL3 is an important molecular for mediating ECs functions. Our prior study demonstrated the mRNA and protein expression of EDIL3 in DMSCs was significantly upregulated in psoriasis. 9 In vitro, EDIL3 derived from DMSCs (DMSCs-derived EDIL3) in psoriasis promotes angiogenesis of ECs. 13 Based on previous research, we explored the mechanism of DMSCs promoting angiogenesis in psoriasis. First, we injected recombinant EDIL3 protein to mouse model of psoriasis to confirm the roles of EDIL3 in psoriasis. Besides, we employed both shortinterference RNA (si-RNA) and lentiviral vectors to explore the molecular mechanism. daily with a semi-quantitative scoring system including scaling and erythema. To evaluate microvessel densities during modelling, at on day 4, half of every group mice were sacrificed by cervical dislocation and skin samples were collected within 2 h for additional experiments. At the end of the experiment on day 7, remainder mice were sacrificed and skin samples were collected. Skin biopsies were processed for paraffin sections, which were then stained with haematoxylin and eosin (H and E). Moreover, skin biopsies were embedded in tissue-tek OCT compound (Thermo Scientific) and stored at −80°C.

| Immunofluorescent and H and E staining
For measuring epidermal thickness, paraffin sections were stained with H and E. and images were taken by inverted microscopy (Olympus). Epidermal thickness was measured under microscope.
The immunofluorescent (IF) staining of mice skin biopsies for CD31 was performed to assess the microvessel densities. The 5μm thick tissues were sectioned from frozen tissue. To inhibit nonspecific antigen-antibody reactions, tissues were blocked with goat serum (Boster Biological Technology) for 30 min and washed for 3 times with PBS buffer. Primary rabbit anti-CD31 (Abcam) was added and incubated at 4°C overnight. The next day, recovery temperature at 37°C, slides were washed in PBS for 3 times and incubated with secondary antibody (Goat anti-rabbit IgG; Zhongshanjinqiao) for 2 h at room temperature. Finally, slices were washed for three times with PBS, incubated with 4′, 6-diamidino-2-phenylindole (DAPI; Solarbio) for 10 min at room temperature. Sections were then imaged using LSCM (Olympus FV1200MPE).

| Samples collection
10HEK293 cells were kindly donated by Dr. Ruixia Hou (Department of Dermatology, Taiyuan Central Hospital of Shanxi Medical University). DMSCs from five patients with psoriasis and five healthy volunteers were isolated from skin tissues and the isolated and identified methods have been described in previous work. 17 The patients with psoriasis had been diagnosed both clinically and pathologically and received neither pharmaceutical nor physical therapy within least 12 weeks. Characteristics of psoriatic patients are shown in

| Cells isolation and identification
DMSCs were isolated and cultured as described previously. 17 Briefly, skin tissue was cut into pieces and subcutaneous fat was removed.
The dermis was separated from the epidermis by incubating in 0.25% dispase (Hyclone). Then, dermal pieces were chopped further and filtered. The suspension containing cells was collected. Finally, after centrifugation, cells pellets were seeded into a culture dish and cultured in DME/F12 supplemented with 10% FBS and 1% antibiotics. The non-adherent cells were removed after 48-72 hours.
The medium was refreshed every 5 days and the colonies of DMSCs appeared at 7 to 10 days. The phenotype of DMSCs was identified The expression of CD31 on cells surface was analysed by flow cytometry (Beckman Coulter).

| HUVECs co-culture with DMSCs
Transwell chamber with 0.4 μm pore filters was selected to create an indirect interaction microenvironment for co-culture. HUVECs were co-cultured with DMSCs in a 12-well transwell plate for 48 h at 37°C in 5% CO 2 . DMSCs (8 × 10 4 cells/well) were seeded into the upper chambers in the transwell plates followed by 500 μl medium (10% v/v FBS in DME/F12). Then HUVECs were seeded into the lower chambers at 1:1 ratio (DMSCs: HUVECs) followed by 1000 μl EBM medium supplemented with EGM-2. Before HUVECs adhesion, DMSC-seeded chambers were moved into new 12-well plates for 6 h to exclude the influence that HUVECs adhered to the well plate. After co-cultured, HUVECs were harvested for real-time quantitative polymerase chain reaction (RT-qPCR), western blot and HUVECs function analysis.
We divided HUVECs into five groups according to DMSCs cocultured with them. The details of groups were as follows: Control group: HUVECs were cultured in the lower chamber alone. C-DMSCs group: DMSCs from healthy volunteers were co-culture with HUVECs. C-DMSCs EDIL3-high group: DMSCs from healthy volunteers with overexpressed EDIL3 co-cultured with HUVECs. P-DMSCs group: DMSCs from psoriasis co-cultured with HUVECs. P-DMSCs EDIL3-low group: DMSCs from psoriasis with low-expressed EDIL3 co-cultured with HUVECs.

| Short-interference RNA transfection to P-DMSCs
For silencing EDIL3 of P-DMSCs, EDIL3 si-RNA (GenePharma) was transfected using HiperFect transfection reagent (QIAGEN) following the manufacturer's instructions. P-DMSCs were seeded into a 6-well plate until reached about 50% confluence and washed with DMEM/F12 for 3 times prior to transfection.

| Real-time quantitative polymerase chain reaction (RT-qPCR)
Transfected P-DMSCs were harvested and mRNA was extracted  treat group-(Ct(target gene)-Ct(reference gene)) control group.

| Tube formation assay
In vitro, the tube formation of ECs was assessed using Matrigel matrix (mimics the natural basement membrane matrix of ECs

| Statistical analysis
Statistical analyses were performed using 17.0 SPSS software (Chicago, USA). Differences among groups were evaluated using repeated measures analysis of variance (anova). A two-tailed t-test was used when two groups were compared for statistical significance.

| EDIL3 accelerated the process of psoriasis in vivo
To further confirm the roles of EDIL3 in psoriasis. We intra-dermally  Researches showed that EDIL3 promoted tumours growth through promoted angiogenesis. 15 To explore whether EDIL3 accelerated the pathological progress of psoriasis is also through modulated angiogenesis, we conducted IF analysis. We used double IF with DAPI and CD31 to detect the microvascular density via two-photon confocal laser microscopy. We confirmed an upregulation of blood vessel density in psoriatic mice skin at day 4 and 7 ( Figure 1D,F). And the results indicated that EDIL3 significantly promoted the growth of microvasculature. In addition, compared to that at day 4, EDIL3injected skins exhibited more increased blood vessel density at day 7, and EDIL3 caused a time-dependent increase of angiogenesis.
Together, these data suggested that the upregulation of EDIL3 can promote microvascular formation in development of psoriasis.

| Identification and culture of DMSCs and HUVECs
DMSCs were incubated in DMEM/F12 supplied with 10% FBS and 1% antibiotic at 37°C in a humidified atmosphere containing 5% ( Figure S2E).

| DMSCs-derived EDIL3 stimulated the expression of α vβ 3 in HUVECs
To determine the optimized time range of lower expression of We also found compared with the controls, the expression of αvβ3 was upregulated in all co-cultured groups (p < 0.05) ( Figure 3A,B).
And in P-DMSCs group, the expression of αvβ3 was higher than C-DMSCs EDIL3-high (p < 0.05) ( Figure 3A,B). Together, these studies confirmed that not only C-DMSCs EDIL3-high upregulated integrin αvβ3 but also the increase of αvβ3 in P-DMSCs was more evident.  Figure 4A,B). In addition, in overexpressed EDIL3 group and P-DMSCs group, we found that p-FAK, p-MEK and p-ERK1/2 were upregulated in HUVECs ( Figure 4A). Both the P-DMSCs and C-DMSCs EDIL3-high groups stimulated the expression of p-MEK and p-ERK1/2, but no significant changes of the both groups ( Figure 4B).

| DMSCs-derived EDIL3 activated the FAK/MEK/ERK signal pathway in HUVECs
Then, the p-FAK in P-DMSCs group was significantly increased than C-DMSCs EDIL3-high ( Figure 4B). After knockdown of EDIL3 in P-DMSCs, we found the p-FAK, p-MEK and p-ERK1/2 proteins were all significantly decreased ( Figure 4B). All data suggested the activation of FAK/MEK/ERK signal pathway in HUVECs, which was positively regulated through EDIL3 and αvβ3 ( Figure 4C).

| DMSCs-derived EDIL3 through the α vβ 3-FAK/MEK/ERK axis induced tube formation of HUVECs
The process of tube formation is a result of dynamic reorganization of vascular system in vitro and is the characteristic trait of ECs.
HUVECs belong to ECs and have the ability of tube formation in vitro.
This study further confirmed that our protocols isolating HUVECs was successful and feasible ( Figure 5A). The number and stability of tube mesh formation was found to be increased in co-culture groups. After 48 h, the control no treatment with DMSCs displayed very few branches; however, the prominent branching networks were observed in DMSCs-treated groups ( Figure 5A-E) and the increase was significant (all p < 0.05) ( Figure 5F). Tube-forming ability in C-DMSCs, as measured by the number of meshes assembled, was significantly decreased compared to that in P-DMSCs (p < 0.001) ( Figure 5F). In EDIL3 overexpressed group, after co-culture, sprouting HUVECs were numerous and tubes were longer compared to these in C-DMSCs group (Figure 5 B-F) (p < 0.01). Knockdown of EDIL3 in P-DMSCs significantly inhibited the tube formation of HUVECs compared to that in P-DMSCs (p < 0.001) ( Figure 5F). A similar effect was observed in EDIL3 overexpressed C-DMSCs and P-DMSCs; however, the increased tube meshes in P-DMSCs were more significant (p < 0.05) ( Figure 5F). Combining all results, these indicated that DMSCs-derived EDIL3 through the αvβ3-FAK/MEK/ ERK axis induced tube formation of HUVECs.

| DISCUSS ION
Psoriasis is not only inflammatory-dependent but also angiogenesisdependent disease. 5,19 Cytokines produced by immune cells, ke- lesions is required for circulation of nutrients, signalling molecules, gas exchange, waste removal and to provide an enlarged endothelial surface area for inflammatory cell trafficking. 19 The initiation or maintenance of the chronic inflammatory state in psoriasis may depend partly on mediators of the angiogenic pathway. 21 It is generally admitted that vascular expansion and angiogenesis in psoriasis

Mesenchymal stem cells (MSCs) can differentiate into adipo-
cytes, osteoblasts and chondrocytes, are self-renewing and expandable stem cells. 22 We isolated DMSCs from skins of psoriatic patients and healthy volunteers and the results showed its surface markers were same as MSCs. And DMSCs also differentiated into adipocytes, osteoblasts and chondrocytes. CD31 is a specific marker of ECs 23 and ECs can form capillary-like structures plated on the Matrigel matrix in vitro, which can be used to define ECs. 24 In present study, we successfully isolated DMSCs and HUVECs and co-cultured them to evaluate the effect of DMSCs-derived EDIL3 on ECs.
Our previous study demonstrated that psoriatic DMSCs promoted cell-cell adhesion and migration of ECs, but the molecular mechanism has not been explored. 12 Our microarray analysis (data deposited under NCBI GEO GSE42632) showed that DMSCs from psoriatic skin lesions displayed significantly high expression of EDIL3. 17 Previous studies have reported that EDIL3 is a recently cloned and characterized unique matrix protein that is expressed during early embryogenesis in ECs, 25 and its expression is downregulated in later developmental stages. In healthy adult tissue, EDIL3 becomes quiescent or may no longer be expressed. 17,26,27 However, EDIL3 expression can be re-initiated during ischemia and is upregulated in tumour vascular tissues. 25,26 These findings suggested that EDIL3 may have an important role in mediating aberrant vascular remodelling. We further analysed the mRNA and protein of EDIL3 in P-DMSCs using RT-qPCR and western blot.
The results showed mRNA and protein expression of EDIL3 in psoriatic DMSCs was markedly higher than healthy. 9 To further explore the role of EDIL3 in psoriasis, in present study, we de-  active Ras-GTP, and the latter acts to mitogen activated protein kinase (MAPK or MEK) and extracellular signal-regulated kinase-1/2 (ERK1/2). 32 FAK/MEK/ERK signal pathway is involved in many cellular functions including cell survival and migration and inhibition of this pathway has been shown to halt the migration of a variety of cell types. 33 In addition, ERK promotes the transcription of a variety of genes, many of which involved in motility. 33 In the present study, we demonstrated that DMSC-derived-EDIL3 involved in EC angiogenesis and the function acted via the αvβ3-FAK/MEK/ERK signal pathway.
In conclusion, EDIL3 was upregulated in P-DMSCs and was an

ACK N OWLED G EM ENTS
We are very grateful to Dr. Hou Ruixia for primary DMSCs and the department of obstetrics and gynaecology for the umbilical cord.
We also thank the National Natural Science Foundation of China for funding.

FU N D I N G I N FO R M ATI O N
This work was supported by grants from National Natural Science Foundation of People Republic of China (No. 81803146).

CO N FLI C T O F I NTE R E S T
All authors indicated no potential conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.