Ezrin regulates synovial angiogenesis in rheumatoid arthritis through YAP and Akt signalling

Abstract This study aimed to investigate the role and regulatory mechanisms of Ezrin in synovial vessels in rheumatoid arthritis (RA). Synovial tissues were obtained from people with osteoarthritis people and patients with RA patients. We also used an antigen‐induced arthritis (AIA) mice model by using Freund's adjuvant injections. Ezrin expression was analysed by immunofluorescence and immunohistochemical staining in synovial vessels of patients with RA and AIA mice. We investigated the role of Ezrin on vascular endothelial cells and its regulatory mechanism in vivo and in vitro by adenoviral transfection technology. Our results suggest a role for the Ezrin protein in proliferation, migration and angiogenesis of vascular endothelial cells in RA. We also demonstrate that Ezrin plays an important role in vascular endothelial cell migration and tube formation through regulation of the Hippo‐yes‐associated protein 1 (YAP) pathway. YAP, as a key protein, can further regulate the activity of PI3K/Akt signalling pathway in vascular endothelial cells. In AIA mice experiments, we observed that the inhibition of Ezrin or of its downstream YAP pathway can affect synovial angiogenesis and may lead to progression of RA. In conclusion, Ezrin plays an important role in angiogenesis in the RA synovium by regulating YAP nuclear translocation and interacting with the PI3K/Akt signalling pathway.

| 9379 CHEN Et al. and actin cytoskeleton. It plays an important role in cell morphology maintenance, cell movement, adhesion, signal transduction, apoptosis regulation, and metastasis promotion and prognosis of malignant tumours. 5,6 According to recent literature, Ezrin has been implicated several endothelial cell-related processes, such as endothelial cell proliferation and neovascularization, 7 regulation of the water balance 8 and maintenance of the endothelial cell pore size. 9 Angiogenesis is a physiological complex process through which new microvessels form from the pre-existing vascular system and involves endothelial cell activation, basement membrane rupture and migration, endothelial cell expansion, lumen formation and, finally, the formation of new blood vessels. Prior research literature has shown that the Hippo-YAP signalling pathway plays a key role in angiogenesis, by inducing endothelial cell proliferation and angiogenesis, and promoting vascular development. However, to the best of our knowledge, no robust data have shown whether the Hippo-YAP pathway regulates RA synovial angiogenesis.
In this study, we aimed to investigate the expression and role of

| Tissue Specimens
We included tissue specimens derived from ten individuals.
Specifically, we included knee synovial specimens of patients with RA (n = 5; age =67.4 ± 5.3 years; one male and four female), and knee joint synovial specimens from patients with osteoarthritis (n = 5; age =69.2 ± 7.1 years; two male and three female) as controls.
All synovial tissue samples were obtained from patients admitted to Zhujiang Hospital of Southern Medical University for total knee replacement. The diagnosis of rheumatoid joints complied with the revised standards of the American College of Rheumatology in 1987. 10 All subjects provided informed consent. The experimental protocol was approved in advance by the Ethics Committee of Zhujiang Hospital of Southern Medical University, China.

| Animals
Female C57/BL6 mice (6-7 weeks old, ~20 g body weight) were maintained on a standard chow pellet diet with had free access to water, on a 12 h light/dark cycle. Groups of five mice were used for each treatment to a total number of 40 animals. Animals were monitored daily for their health status. Experimental protocols were approved

| Adjuvant-induced arthritis model and treatment
Female C57/BL6 mice (6-7 weeks old, 20 g body weight) were ob- Freund's incomplete adjuvant (CFA, Sigma-Aldrich) was given into the knee joint at day 7. Treatments were started on day 8. Verteporfin (VP) was administered intravenously at a dose of 1 mg/kg. The control group was injected with the same amount of normal saline at the same time. Intraarticular injection of adenovirus (5 μl) was performed once a week. Experimental protocols were approved by the Southern Medical University Committee in accordance with the Institutional Animal Care and Use Committee guidelines. The present study was carried out in accordance with the Declaration of Helsinki.

| Assessment of arthritis severity
Histologic sections were scored based on a scale of 0-3 for each of the following five parameters: synovitis (defined as hypercellularity of the synovium), joint space exudate (defined as leukocytes in the joint space), soft tissue inflammation (defined as leukocyte infiltration of the infrapatellar fat pads, joint capsule and the area adjacent to the periosteal sheath), cartilage degradation (defined by loss of haematoxylin and eosin staining; 0 = full-stained cartilage, 3 = fully unstained cartilage) and bone damage (defined according to the extent and depth of subchondral bone damage). The total histologic score was the sum of the scored of the individual arthritis features (up to a maximum of 15). 11

| Immunohistochemistry
Synovial tissue specimens from patients with RA were fixed in paraformaldehyde (PFA) and embedded in paraffin. Specimens were embedded in paraffin and 5 μm serial sections were performed for immunohistochemical staining based on immunoperoxidase technique with diaminobenzidine as the chromogen. Following deparaffinization, microwave antigen retrieval was performed by incubating sections in 10 mM EDTA pH 7.5 at 93℃ for 10 minutes, after which sections were allowed to cool for at least 20 min.

| Immunofluorescence
For immunofluorescence analysis, PFA-fixed plated cells were incubated with blocking solution (10% normal goat serum, 0.1% Triton ×100 in PBS), for 1 h at room temperature. Samples were then incubated with the primary antibody overnight at 4℃. For quantification of Ezrin, Ki67, YAP and DAPI colocalization, nuclear ROIs were segmented and the ratio of CD31-positive area to the total area was quantified using Image J analysis software.

| MTT test
To determine the effects of VP on cell proliferation, the 3-(4,5-di methylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay was carried out. For the MTT assay, in brief, 1 × 10 4 HUVECs/well were plated in 96-well plates with 100 μl of culture medium overnight. 24 h later, the medium was replaced with fresh medium containing different concentrations of drugs. 20 μl (0.5 mg/ml) of MTT reagent was added to the cells at different time points and incubated for 2 h at 37℃. The medium was then removed, and 150 μl of DMSO was added to dissolve the formazan crystals formed. A BioTek plate reader was used to read absorbance at 570 nm.

| Cell migration
Confluent cell monolayers were scraped with a pipette tip to generate a wound. Cell debris was removed by washing with PBS several times. Cell monolayers were photographed with a digital camera straight after the scratch would and 12 h after, and the degree of wound closure was determined with the Image J analysis software.

| Tube formation
We assessed the drug effects on the ability of HUVECs to reorganize and differentiate into capillary-like networks using an in vitro Matrigel morphogenesis assay. HUVECs (1.5 × 10 5 cells/ml) were seeded in 48-well plates pre-coated with Matrigel (50 μl/well).
Matrigel was then polymerized for 40 min at 37℃. After cell incubation for 6 h in serum-containing media, images of tube morphology were captured using a microscope, and the extent of tube formation was quantified by counting the number of meshes with the Image J analysis software.

| Western blot analysis
HUVECs were grown in 100-mm dishes. HUVECs were trypsinized, pelleted at 300 g and then lysed with 150 μl RIPA lysis buffer and

| Adenovirus vector and transfection
Adenoviruses expressing or silencing Ezrin, YAP were purchased   (Table 1) were determined relative to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene expression. qPCRs were performed in triplicate from two independent experiments.

| Statistical analysis
All statistics were performed using GraphPad Prism (version 5.00 for Mac, GraphPad Software, San Diego, CA, USA). Data were tested for normality using a D'Agostino and Pearson omnibus normality test and subsequently assessed for homogeneity of variance. Data that passed both tests were further analysed by a two-tailed un-

| Ezrin expression and angiogenesis were increased in the synovial tissue of AIA-treated mice
To confirm our preliminary results obtained from synovial tissue of RA patients, we performed animal experiments. After inducing arthritis in mice by administration of complete/incomplete Freund's adjuvant, we observed different degrees of proliferation, infiltration of numerous inflammatory cells in the synovial membrane and increased formation of the synovial blood vessels the synovial tissues ( Figure 2A). Immunohistochemical staining demonstrated that the number of CD31 + cells in the synovial tissues of AIA mice was increased compared to control mice, which is due to increased proliferation of vascular endothelial cells. There were significantly more CD31 + cells in synovial tissues in the AIA group than in the control group

| Blocking Ezrin inhibits angiogenesis and delays arthritis progression in AIA mice
To explore the potential role of Ezrin in arthritis, Ezrin expression was depleted by siRNA-mediated knockdown and increased through oeRNA-mediated overexpression in HUVECs ( Figure 3A).

| Ezrin regulates YAP expression
To explore the signalling pathways downstream of Ezrin, we hypoth- Western blotting on both the nuclear and cytosolic fractions indicated that nuclear YAP was significantly increased upon Ezrin overexpression and reduced upon Ezrin knockdown ( Figure 4A).
Immunostaining indicated that YAP is primarily localized in the nuclei cells. Ezrin siRNA cells significantly impaired YAP nuclear translocation, whereas Ezrin overexpression significantly promoted YAP nuclear translocation ( Figure 4B). In addition, immunohistochemical  Figure 4E).

| Blocking the Hippo-YAP pathway inhibits angiogenesis and delays progression of arthritis in AIA mice
In order to investigate the effect of activation of the Hippo-YAP pathway on the synovitis and synovial vessels of AIA mice, we selected VP, an inhibitor of Hippo-YAP. First, CCK8 was used to detect the effect of VP in inhibiting cell proliferation. A proliferation inhibition rate >50% was accordingly selected, and 1, 2, and 4 μM were selected as the experimental concentrations for subsequent experiments ( Figure 5A). Western blotting revealed that VP inhibited YAP expression ( Figure 5B), tube formation ( Figure 5C), and migration

| Hippo-YAP pathway and PI3K/Akt pathway interaction in vascular endothelial cells
Prior research showed that the PI3K/Akt signalling pathway is a classical proliferation pathway and may be regulated by YAP. We found that the expression of p-Akt was increased in different AIA groups, which suggested activation of the PI3K/Akt signalling pathway.
However, the expression of p-Akt in AIA mice treated with VP was found to be decreased compared to control mice ( Figure 6A). We set out to validate these results and found that, when VP was added to HUVECs, YAP expression was inhibited with the increase in VP concentration, and the ratio of p-S6/S6 and p-Akt/Akt was decreased ( Figure 6B). We next constructed an adenovirus to silence and overexpress YAP. Following transfection, we found that YAP silencing increased PTEN expression, although the expression of p-S6 and p-Akt and the ratio of p-S6/S6 and p-Akt/Akt were both decreased.
The reverse was observed following YAP overexpression. Previous studies suggested interactions between the PI3K/Akt signalling pathway and the Hippo pathway. Indeed, we observed PI3K/Akt pathway activation during HUVEC cell proliferation ( Figure 6C). We  Figure 6D). Our results confirmed interaction between the Hippo-YAP and the PI3K / Akt signalling pathways ( Figure 6E and 6F). Previous studies have shown that Ezrin can participate in cell processes in health and disease relate to development, and metastasis of tumours, and cell adhesion, movement and angiogenesis through a variety of signalling pathways. [16][17][18][19] In this study, we observed that the proliferation activity of synovial vessels was enhanced and the Ezrin expression in the synovial vessels was also higher in specimens via cell-to-cell contact. 25 In addition, VE cadherin-mediated PI3K/ AKT signalling pathway regulates the YAP subcellular localization.

| DISCUSS ION
Simultaneously, YAP in the nucleus can induce the expression of angiogenin-2 (Ang-2) and thereby promote angiogenesis germination and vascular remodelling. 26 In addition, a number of studies have demonstrated that YAP promotes smooth muscle proliferation and migration through the inhibition of the expression of smooth muscle-specific genes. 27 In the process of smooth muscle development, YAP also promotes proliferation and inhibiting the differentiation of smooth muscle cells. YAP conditional knockout mice exhibit severe vascular abnormalities. 28  Our study also has several limitations. The adjuvant-induced arthritis models used in this study may not fully mimic the natural process of RA in patients. Therefore, our findings may not be generalizable to RA in humans.

| CON CLUS ION
In conclusion, the results of the present study demonstrated that Ezrin expression was increased in RA synovial vessels. In addition, we also demonstrated that the PI3K/Akt pathway was activated by the Hippo-YAP pathway. Ezrin was found to promote the activation and nuclear translocation of YAP, thereby playing a potential role in transcription. Our results suggest that activated YAP can further activate the PI3K/Akt pathway, which in turn promotes proliferation and vascular activities of vascular endothelial cells. Ezrin or Hippo-YAP pathway inhibition was shown to inhibit the proliferation of RA synovial vessels as well as inhibit inflammatory processes of RA joints in mice (Figure 7). The results may provide a new direction for the treatment of RA synovial angiogenesis.

CO N FLI C T S O F I NTE R E S T
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data presented in this study are available on request from the corresponding author.