SARI prevents ocular angiogenesis and inflammation in mice

Abstract SARI (Suppressor of AP‐1, regulated by IFN‐β) is known to play an important role in some systemic disease processes such an inflammatory conditions and cancer. We hypothesize that SARI may also play a role in ocular diseases involving inflammation and neovascularization. To explore our hypothesis, further, we investigated an endotoxin‐induced uveitis (EIU) and experimental argon laser‐induced choroidal neovascularization (CNV) model in SARI wild‐type (SARIWT) and SARI‐deficient (SARI−/−) mice. Through imaging, morphological and immunohistochemical (IHC) studies, we found that SARI deficiency exacerbated the growth of CNV. More VEGF‐positive cells were presented in the retina of SARI−/− mice with CNV. Compared to SARIWT mice, more inflammatory cells infiltrated the ocular anterior segment and posterior segments in SARI−/− mice with EIU. Collectively, the results point to a potential dual functional role of SARI in inflammatory ocular diseases, suggesting that SARI could be a potential therapy target for ocular inflammation and neovascularization.


| INTRODUC TI ON
Age-related macular degeneration (AMD) is a leading cause of irreversible vision loss among elderly people. 1 It is estimated that, globally, the number of people with AMD would be 196 million in 2020, going up to 288 million by 2040. More than 60% of these cases are in Asia. 2  and provides a reasonable proxy of the mechanisms underlying HLA-B27-associated uveitis in humans.
SARI (Suppressor of AP-1, regulated by INF-β), also called as BATF2, is a member of BATF (Basic leucine zipper (bZIP) transcription factor, ATF-like) family, which contains BATF (otherwise known as SFA2), BATF2 and BATF3 (otherwise known as JDP1 and p21SNFT). 10 Various studies have demonstrated the anti-tumour role of SARI in multiple cancers, including lung cancer, 11 prostate cancer, 12,13 B lymphoma 14 and colon cancer. 15,16 Previous studies by us indicated that SARI was down-regulated in colon cancer 15,17 and inhibits colon cancer growth through inhibiting the translational activity of HIF-1α/VEGF and tumour angiogenesis. 15 SARI is also involved in innate immunity and infection immunity. 18,19 SARI protects mice from Mycobacterium tuberculosis and Listeria monocytogenes mediated Type 1 and Type 2 diseases. 18 During T cruzi infection, SARI functions as a negative regulator of IL-23a in innate immune cells. 19 We and a previous study by Kayama also demonstrated the protective role of SARI in colitis through regulating macrophage infiltration. 20 Against this background, we speculated that SARI may have a role to play in preventing angiogenesis and inflammation responses in certain ocular diseases. To explore our speculation further, CNV and EIU models were induced in SARI wild-type (SARI WT ) and SARI deficiency (SARI −/− ) mice. The present study expands the understanding of SARI function and provides a novel therapy target for certain ocular diseases.

| Animals
SARI knockout (SARI −/− , catalogue no. 019085) and SARI wild-type (SARI WT , catalogue no. 002448) mice, 6 to 8 weeks old, were purchased from the Jackson Laboratory. Experiments were approved by the Animal Ethics Committee of Sichuan University. All animal care, husbandry and experiments were conducted in adherence with institutional guidelines for the use of animals in Sichuan University.
Mouse gene type was confirmed via the genotyping protocols supplied by the Jackson Laboratory.

| Laser-induced CNV model
Animals were anaesthetized with ketamine (75 mg/kg) and xylazine (5 mg/kg) by intraperitoneal injection after a drop of 0.2% tropicamide, and 1% phenylephrine (Santen) for pupil dilatation was administered to the right eye. A coverslip was lubricated with 2.5%

| Fundoscopy and funds fluorescence angiography (FFA)
Mice were deeply anaesthetized by intraperitoneal injection with a mixture of ketamine (75 mg/kg) and xylazine (5 mg/kg).
Hyaluronate was used to keep the ocular surface moist. The pupils were dilated using 0.2% tropicamide and 1% phenylephrine (Santen). Retinal pathology was assessed using a Micron Ⅳ fundoscopy system (Phoenix Research Laboratories). FFA was carried out 4 minutes after intraperitoneal injection of 25 mg/mL 150KD FITC conjugated dextran (Sigma-Aldrich). Digital images of eyes were captured for one minute. Angiograms were graded by two experienced ophthalmologists using a leakage score system ( Table 1). The area of CNV lesion was measured in a masked fashion using Image J.

| Choroidal flat mount
The entire ocular globes were enucleated and fixed in 4% para-

| Immunohistochemistry
Eyes were fixed in 4% paraformaldehyde freshly made in PBS overnight at 4°C and cryoprotected in 30% sucrose, and then embedded in paraffin. Retinal sections were sagitally cut through the corneaoptic nerve axis (3-μm thick), mounted on slides and dried. After deparaffinization with graded ethanol and xylene solutions, tissue samples were incubated with a blocking reagent, and then with primary antibody against VEGF (Abcam) and CCL 2 (Abcam) at 4°C overnight. After rinsing the slides with PBS, the specimens were treated with secondary antibodies (Zsbio). Secondary Abs were labelled with the HRP, which was detected by diaminobenzidine (DAB; Maixin), whereas the nuclei were stained with haematoxylin (Beyotime). All the sections were examined under an Olympus BX600 microscope and SPOT Flex camera. The VEGF and CCL 2 expressions in each frame were scored as 0, 1, 2 and 3 based on the percentage of positive cells. 21 Score 0, the percentage of positive cells < 5%; score 1, 5% < the percentage of positive cells < 15%; score 2, 15% < the percentage of positive cells < 25%; score 3, the percentage of positive cells > 25%.

| EIU mouse model
Mice were intraperitoneal anaesthetized with ketamine (75 mg/ kg) and xylazine (5 mg/kg). EIU was induced in SARI WT or SARI −/− mice by single intravitreal injection of 50 ng lipopolysaccharide (LPS; Sigma-Aldrich-Aldrich). The control animals received the same volume of PBS. Animals were divided into 4 experimental groups ( Table 2, n = 6 per group). After 24 hours, all animals from each group were killed when the inflammation was provoked and reached its peak. 22 Eyes were carefully enucleated and processed for evaluation.

| Inflammation evaluation in the anterior chamber
The clinical signs of inflammation in the anterior chamber (eg miosis, iris hyperaemia and hypopyon) were evaluated by two experi-

| Statistical analyses
Data were presented as mean and Standard deviation (SD) at least 3 independent experiments. Intergroup comparisons were made by one-way analysis of variance (ANOVA) testing, followed by Tukey's

| SARI protected the leakage of CNV in mice
To investigate the potential role of SARI in ocular angiogenesis, a CNV model was established in SARI WT and SARI −/− mice. Leakage was evaluated in vivo using fundoscopy and fluorescein angiography at 7 days after laser photocoagulation. As shown in Figure 1A,  Figure 1C). To confirm the protective potential of SARI, the size of the leakage was graded and measured among the different genotype groups.
When compared to the SARI WT group, the SARI −/− group showed increased area of neovascularization and leakage (Table 3). SARI deficiency was associated with decreased number of spots with weak dye staining (score 0 or 1) and increased number of spots with strong staining (score 2 or 3) ( Figure 1B). Animals in the KO group also had a significantly larger leakage area of CNV than that in WT group ( Figure 1D). The results suggest that SARI may play a protective role in CNV.

| SARI alleviates the severity of CNV in mice
To quantify the degree of CNV formation, we evaluated the extent of vascularization in choroidal flat -mounts stained against IB4. Small vessels were fully visible with fluorescence in normal groups, and the IB4-labelled CNV outgrowths were observed in the CNV group.
Such findings seem more prominent in KO mice compared to wildtype mice (Figure 2A) damage were more obvious in KO mice ( Figure 2C). Thus, our results suggest that SARI attenuated the severity of experimental CNV.

Many growth factors are involved in different stages of CNV. 24
Among these growth factors, VEGF is the most important. 25 A previous study by us has demonstrated the inhibition role of SARI on VEGF expression in colon cancer cells. 15 To further determine the underlying mechanism of SARI attenuating the severity of CNV, the choroidal tissues were collected for VEGF staining. As shown in Figure 3A,B few VEGF-positive cells were observed in the SARI WT and SARI −/− mice of normal group. Laser-induced VEGF expression was presented in both SARI WT and SARI −/− mice, but more VEGFpositive cells were observed in the choroidal tissue of SARI −/− mice, compared with SARI WT mice ( Figure 3A,B). Collectively, the results demonstrated the inhibition role of SARI in ocular VEGF expression.

| EIU is exacerbated in SARI-deficient mice
To investigate whether SARI plays a role in regulating inflamma- response was more severe in the KO + LPS group, as evidenced by exudation into the anterior chamber ( Figure 4A). The mice in the SARI −/− group had significant higher clinical score (mean, ~3.875) than that in the SARI WT group (mean, ~2.750) (P < .05, Figure 4B).
Some mice in the SARI −/− group even had hypopyon. There was no inflammation seen in the SARI −/− normal group or the SARI −/− normal group ( Figure 4B). As indicated by these results, SARI deficiency resulted in the extensive inflammation during EIU.

| Loss of SARI increased in filtrating inflammatory cells
In

| D ISCUSS I ON
Our study demonstrated two effects of SARI on two common oc- was inhibited by SARI, which resulted in less inflammation in mice with EIU. Thus, SARI appears to play a dual function, inhibiting ocular inflammation and neovascularization. Furthermore, given that there were more neovascular and inflammation changes in SARI −/− mice in the CNV and EIU models, basal SARI expression could potentially play a prognostic role in ocular diseases management. These observations contribute to the further understanding of SARI in ocular disease and provide a potential therapy target for such disease.
During CNV in wet AMD, VEGF expression is induced and plays a crucial role in mediating the process. Thus, targeting VEGF is an efficient strategy for treating wet AMD. 27 Inflammation is the major cause of uveitis. In ocular inflammation, macrophages are located mostly in the surrounding inflammatory tissue and are essential for the vascularization and damage of the inflammatory tissue. 33 Thus, targeting macrophage is an efficient treatment strategy for inflammatory ocular disease, 34 and for ocular diseases involving neovascularization. 35 According to the chemotaxis of CCL 2 on macrophage infiltration, CCL 2 inhibitor shows a good therapeutic effect on ocular-related diseases. 36 In a previous study, we have demonstrated the inhibitory role of SARI in CCL 2 expression in epithelial colon cells through promoting the degradation of STAT1 under colitis condition. 26 SARI inhibited the infiltration of macrophage into colon tissues and colitis development in mice. 26 In the present study, we also confirmed the protective role of SARI in EIU, which may contribute to the inhibition of CCR 2-positive inflammatory cells infiltration in the anterior segment and posterior segment.
In summary, the present study demonstrated the dual functional role of SARI in ocular disease, inhibiting ocular inflammation and neovascularization, which suggests that SARI could be a potential therapy target for certain ocular diseases. However, the present study only investigated the functional role and underlying mechanism of SARI in regulating wet AMD and EIU in SARI WT and SARI −/− mice. Further studies are needed to gain a deeper understanding of the SARI regulatory pathways in order to identify potential targets for intervention.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no potential conflict of interest.