PRMT1 and PRMT4 Regulate Oxidative Stress-Induced Retinal Pigment Epithelial Cell Damage in SIRT1-Dependent and SIRT1-Independent Manners

Oxidative stress-induced retinal pigment epithelial (RPE) cell damage is involved in the progression of diabetic retinopathy. Arginine methylation catalyzed by protein arginine methyltransferases (PRMTs) has emerged as an important histone modification involved in diverse diseases. Sirtuin (SIRT1) is a protein deacetylase implicated in the onset of metabolic diseases. Therefore, we examined the roles of type I PRMTs and their relationship with SIRT1 in human RPE cells under H2O2-induced oxidative stress. H2O2 treatment increased PRMT1 and PRMT4 expression but decreased SIRT1 expression. Similar to H2O2 treatment, PRMT1 or PRMT4 overexpression increased RPE cell damage. Moreover, the H2O2-induced RPE cell damage was attenuated by PRMT1 or PRMT4 knockdown and SIRT1 overexpression. In this study, we revealed that SIRT1 expression was regulated by PRMT1 but not by PRMT4. Finally, we found that PRMT1 and PRMT4 expression is increased in the RPE layer of streptozotocin-treated rats. Taken together, we demonstrated that oxidative stress induces apoptosis both via PRMT1 in a SIRT1-dependent manner and via PRMT4 in a SIRT1-independent manner. The inhibition of the expression of type I PRMTs, especially PRMT1 and PRMT4, and increased SIRT1 could be therapeutic approaches for diabetic retinopathy.


Introduction
Diabetic retinopathy is the leading cause of blindness. The breakdown of the blood-retinal barrier (BRB) mediated by oxidative stress is related to the progression of diabetic retinopathy [1,2]. Retinal pigment epithelial (RPE) cells are a vital component of the outer BRB and are vulnerable to oxidative stress [3]. However, the molecular mechanisms of oxidative stress-induced RPE cell damage are not fully understood.
Protein arginine methyltransferases (PRMTs) catalyse the methylation of the arginine residues of histone and nonhistone proteins. Mammals possess nine PRMTs, which are divided into three types according to their method of methylation. Type 1 PRMTs (PRMT1, PRMT2, PRMT3, PRMT4, PRMT6, and PRMT8) catalyse asymmetric dimethylation at arginine residues, whereas type II PRMTs (PRMT5 and PRMT9) catalyse symmetric dimethylation, and type III PRMTs (PRMT7) catalyse monomethylation [4]. PRMT1 is thought to be involved in diabetic retinopathy, as PRMT1 expression is increased via the generation of reactive oxygen species (ROS) in the retinas of streptozotocin-treated rats and high-glucose-treated bovine retinal capillary endothelial cells, which are a crucial component of the inner BRB [5]. However, the regulation of PRMTs by oxidative stress in RPE cells has not been elucidated.
Sirtuin (SIRT1), a mammalian ortholog of yeast Sir2 (Silent Information Regulator 2), is an NAD-dependent histone deacetylase that regulates diverse physiological and pathophysiological processes, such as senescence, circadian rhythms, autophagy, and apoptosis [6]. In RPE cells, decreased SIRT1 expression caused by ultraviolet light is related to RPE cell damage [7]. The treatment of RPE cells with resveratrol, which increases SIRT1 activity, suppresses inflammatory cytokine-induced vascular endothelial growth factor (VEGF) secretion, which is involved in age-related macular degeneration (AMD) [8]. These reports suggest that SIRT1 protects against RPE cell dysregulation. However, the mechanisms regulating SIRT1 in RPE cells have not been evaluated.
In this study, we evaluated type I PRMT expression and SIRT1 expression under hydrogen peroxide-(H 2 O 2 -) induced oxidative stress and demonstrated that oxidative stress-induced PRMT1 expression increases RPE cell apoptosis via SIRT1 downregulation, whereas PRMT4 does so independently of SIRT1 expression. After treatments, the cells were treated with 500 g/mL of 3-(4,5-dimethysssl-thiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT; Sigma) and incubated for 3 h in a CO 2 incubator. Cells with a functional mitochondrial succinate dehydrogenase can convert MTT to formazan. The formazan crystals formed were solubilized in DMSO (Sigma) and measured with an ELx808 microplate spectrophotometer reader at = 570 nm (BioTek).

Animal Experiments.
Hyperglycemia was induced in overnight fasted, 10-week-old male SD rats ( = 7) by intraperitoneal injection of streptozotocin (55 mg/kg) dissolved in cold and fresh citrate buffer (0.1 M and pH 4.5). Control rats ( = 7) were injected with citrate buffer. Three days after STZ injection, plasma glucose level was determined after overnight fasting with Accu-Chek Aviva (Roche, Swiss). Rats with a blood glucose level of 300 mg/dL or higher were considered as diabetes. After 2 weeks, for preparation of cryosections, the rats were anaesthetized and eyeballs were enucleated, and then they were killed by CO 2 inhalation. All animal experiments were performed in accordance with National Institutes of Health animal research standards. And protocols were approved by the Chonnam National University Laboratory Animal Research Center.

Immunohistochemistry (IHC) and Digital Image Analysis.
Eyeballs were fixed in 4% paraformaldehyde in phosphatebuffered saline (PBS, pH 7.4) for 2 hours at 4 ∘ C. Eyeballs were Oxidative Medicine and Cellular Longevity

PRMT1 and PRMT4 Expression Is Increased in the RPE Layer of Streptozotocin-Treated Rats.
To confirm the increase of PRMT1 and PRMT4 expression in vivo, we generated rats with streptozotocin-(STZ-) induced diabetes, which show severe hyperglycemia and are known to have induced diabetic retinopathy via oxidative stress [12,13]. As shown in

Discussion
Type I PRMTs are important pathophysiological regulators as they promote the production of asymmetric dimethylarginine (ADMA), a metabolic by-product that inhibits nitric oxide synthase (NOS), which is involved in cardiovascular disease, diabetes, and other metabolic disorders [14][15][16]. In addition to producing ADMA, type I PRMTs regulate various cellular processes, such as transcription, RNA splicing, and signal transduction, by catalysing the asymmetric dimethylation of histone or nonhistone proteins [17]. Recent studies have revealed the role of type I PRMTs in diabetic nephropathy [18,19]. However, their roles in diabetic retinopathy are rarely known. and GCL (b) Figure 5: PRMT1 and PRMT4 expression is increased in the RPE layer of streptozotocin-treated rats. Eyeballs were enucleated from vehicle-treated and STZ-treated rats and cryosections were prepared. (a, b) PRMT1 (a) and PRMT4 (b) expressions were measured by immunohistochemistry analysis (C: choroid, RPE: retinal pigment epithelium, PL: photoreceptor layer, OLM: outer limiting membrane, ONL: outer nuclear layer, OPL: outer plexiform layer, INL: inner nuclear layer, IPL: inner plexiform layer, and GCL: ganglion-cell layer). Representative images were from at least three independent experiments. To quantify the DAB signaling, semiautomated analysis protocol was performed as described in Section 2. stress increases PRMT3 expression, leading to increased ADMA generation in preglomerular vascular smooth muscle cells [20]. Treatment with human serum albumin, which induces oxidative stress in renal proximal tubular epithelial cells, increased PRMT1 expression [21]. In contrast, few studies have revealed the protective effects of type I PRMTs in oxidative stress. Very recently, Huang et al. reported that arsenic-induced oxidative stress recruits PRMT1 to the histone 4 arginine 3 and PRMT4 to the histone 3 arginine 17 for asymmetric dimethylation, which leads to increased ferritin transcription via the antioxidant responsive element in HaCaT cells (human keratinocytes) [22]. In the report by Huang et al., in contrast to our results, the expression of PRMT1 and PRMT4 was not changed and PRMT1 and PRMT4 increased ferritin to protect cells from oxidative stress. This discrepancy was likely due to the different cell types (retinal pigment epithelial cells versus keratinocyte) or the kind of oxidative stress (H 2 O 2 versus arsenic), as arsenic treatment induces superoxide anion (O 2 •− ) and hydroxyl radical ( • OH) production [23]. Moreover, PRMT1 and PRMT4 were examined within 6 h of arsenic treatment, while our experiment involved a relatively long duration of H 2 O 2 treatment. Wang et al. reported that treatment with lithium and valproate acid, which protect against H 2 O 2 -induced oxidative stress, increase PRMT4 expression in NSC34 cells [24]. However, they did not establish the function of PRMT4.
In this study, we also showed that H 2 O 2 -induced SIRT1 downregulation is involved in RPE cell damage. Several lines of evidence support our findings. Wu   Bhattacharya et al. reported that decreased SIRT1 expression in RPE cells induces p53 acetylation-mediated apoptosis, leading to the progression of age-related macular degeneration (AMD) [26]. Indeed, p53 and its target genes are closely involved in RPE cell apoptosis [27,28]. As a deacetylase, SIRT1 inhibits p53 activity via deacetylation at lysine 382 [29]. In our study, transfection with an enzymatic-dead mutant, SIRT1, did not inhibit the H 2 O 2 -induced RPE cell apoptosis. Here, we provide novel evidence that SIRT1 expression and its enzymatic activity are vital for RPE cell maintenance.
Interestingly, we found that SIRT1 expression is negatively regulated by PRMT1 but not by PRMT4. Scalera et al. reported that red wine decreased PRMT1 expression in a SIRT1-dependent manner in human endothelial cells [30]. However, in our study, SIRT1 overexpression did not alter PRMT1 expression. This may be due to a cell-type-specific response (endothelial versus epithelial cells). We speculated that PRMT4 regulates SIRT1 expression, as PRMT4 increases the stability of SIRT1 mRNA by methylating HuR protein at arginine 217 in stem cells and HeLa cells [31,32]. However, PRMT4 did not influence SIRT1 protein expression in ARPE-19 cells. To our knowledge, this is the first report on the relationship between PRMT and SIRT1 in cell function. We provide the first evidence that PRMT1 regulates SIRT1 under oxidative stress.
Signaling induced by high-glucose is highly associated with oxidative stress [33]. Very recently, we reported that PRMT4 expression is increased by high-glucose in RPE cells [34]. Consistent with previous results, PRMT4 expression was increased in the RPE and outer limiting membrane (OLM) layers of STZ rats. In addition, PRMT1 expression was increased in the RPE and overall layers of retina of STZ rats. It may be speculated that increased PRMT1 and PRMT4 expression in other layers by STZ treatment contributes to the progression of retinopathy. Further studies should be performed to reveal this speculation.
In this study, we demonstrated that oxidative stressinduced RPE cell damage is regulated by type I PRMTs (PRMT1 and PRMT4) and oxidative stress-induced SIRT1 downregulation is involved in the PRMT1-mediated RPE cell apoptosis pathway. Oxidative stress is a major cause of diabetic retinopathy. Taken together, our data suggest a model of signaling pathways involved in oxidative stress-induced RPE cell apoptosis ( Figure 6). Therefore, the inhibition of type I PRMT expression, especially PRMT1 and PRMT4, and the increase in SIRT1 expression could be therapeutic approaches for diabetic retinopathy.