miRNA-141 attenuates UV-induced oxidative stress via activating Keap1-Nrf2 signaling in human retinal pigment epithelium cells and retinal ganglion cells

Activation of NF-E2-related factor 2 (Nrf2) signaling could protect cells from ultra violet (UV) radiation. We aim to provoke Nrf2 activation via downregulating its inhibitor Keap1 by microRNA-141 (“miR-141”). In both human retinal pigment epithelium cells (RPEs) and retinal ganglion cells (RGCs), forced-expression of miR-141 downregulated Keap1, causing Nrf2 stabilization, accumulation and nuclear translocation, which led to transcription of multiple antioxidant-responsive element (ARE) genes (HO1, NOQ1 and GCLC). Further, UV-induced reactive oxygen species (ROS) production and cell death were significantly attenuated in miR-141-expressing RPEs and RGCs. On the other hand, depletion of miR-141 via expressing its inhibitor antagomiR-141 led to Keap1 upregulation and Nrf2 degradation, which aggravated UV-induced death of RPEs and RGCs. Significantly, Nrf2 shRNA knockdown almost abolished miR-141-mediated cytoprotection against UV in RPEs. These results demonstrate that miR-141 targets Keap1 to activate Nrf2 signaling, which protects RPEs and RGCs from UV radiation.

Growth evidences have shown that microRNAs are important in almost all cellular biological processes including oxidative and anti-oxidant responses [14][15][16]. microRNA deregulation is associated with a number of diseases [14][15][16]. In the current study, we show that microRNA-141 ("miR- 141") activates Nrf2 signaling via selectively targeting and silencing Keap1, which inhibits UV-induced oxidative stress and apoptosis in RPEs and RGCs.

miR-141 expression protects RPEs and RGCs from UV
Our previous studies have demonstrated that Nrf2 activation could protect RPEs from UV [8,17,20]. MTT assay results in ( Figure 3A) demonstrated that, as compared to the control cells, the two lines of RPEs with miR-141 over-expression were largely protected from UV. UVinduced a much weaker viability reduction in miR-141-expressing RPEs ( Figure 3A). Meanwhile, UV-induced RPE cell death ( Figure 3B) and apoptosis ( Figure 3C and 3D) were also largely attenuated with miR-141 expression. In line with our previous studies [8,17,20], cell apoptosis was tested by caspase-9 activity assay ( Figure 3C) and Histone DNA apoptosis ELISA assay ( Figure 3D). miR-141-induced cytoprotection against UV was also observed in RGCs, where miR-141 expression dramatically attenuated UV-induced viability reduction ( Figure 3E) and apoptosis activation ( Figure 3F). Collectively, miR-141 expression significantly protects RPEs and RGC from UV.

miR-141 expression inhibits UV-induced ROS production in RPEs and RGCs
Nrf2 is a well-established anti-oxidant signaling [11,21]. Above results demonstrated that miR-141 activated Nrf2 signaling in RPEs and RGCs, next we analyzed its effect on UV-induced ROS production. As demonstrated, the intracellular ROS content was significantly increased following UV radiation in both RPEs ( Figure 4A) and RGCs ( Figure 4B). Remarkably, forced-expression of miR-141 significantly attenuated ROS production in UV-treated RPEs ( Figure 4A) and RGCs ( Figure 4B). The ROS content in miR-141-expressing cells following UV radiation was comparable to the untreated control level ( Figure 4A and 4B), indicating the potent anti-oxidant activity by miR-141 expression. Notably, the basal ROS content was also lower with miR-141 expression in both RPEs and RGCs ( Figure  4A and 4B). Together, miR-141 expression largely inhibits UV-induced ROS production in RPEs and RGCs.

DISCUSSION
Nrf2-inducing agents were shown to result in decreased oxidative stress and significant cytoprotection [19]. For example, our previous study has shown that
In this study, we showed that miR-141 expression activated Nrf2 signaling in both RPEs and RGCs. First, miR-141 targeted and degradated Nrf2 inhibitor Keap1 in the above eye cells. Second, Nrf2 was stabilized in the miR-141-expressing cells, which was evidenced by upregulation of Nrf2 protein, but not mRNA. Third, stabilized Nrf2 was shown to translocate to cell nuclei, leading to transcription of several ARE-genes (HO1, NOQ1 and GCLC). Importantly, activation of Nrf2 by miR-141 expression significantly ameliorated UV-induced oxidative stress and subsequent death/apoptosis of RPEs and RGCs. shRNA-mediated knockdown of Nrf2 in these cells, on the other hand, almost abolished miR-141cytoproteciton against UV. We therefore conclude that miR-141 protects RPEs and RGCs from UV radiation via activating Nrf2-Keap1 signaling.
To our best knowledge, this is the first report showing potential biological functions of miR-141 in eye cells. There are at least twenty-seven potential target genes of miR-141 have been characterized thus far, which were predicted by the software including PicTar, TargetScanS, and miRanda [24]. One of these targets is Keap1 [18,19]. Our results showing miR-141 activates Nrf2 signaling via targeting Keap1 in RPEs and RGCs were consistent with other studies [18,19]. For example, van Jaarsveld et al., showed that miR-141 depletes Keap1 to decrease cisplatin sensitivity in ovarian cancer cells [18]. Similarly, miR-141 activates Nrf2-dependent antioxidant pathway via silencing keap1 to confer resistance to 5-FU [19].
UV radiation and oxidative stress induce significant damages to RPEs and RGCs. These are the underlying pathological mechanism of AMD and other retinal degenerative diseases [3][4][5]. The results of the current study showing that miR-141 potently activates Nrf2 signaling, which attenuates UV-induced oxidative stress and RPEs/RGCs damages.

Primary culture of human RGCs
Fresh donor human eyes were obtained from ocular trauma patients administrated at the authors institutions. The human retina was separated very carefully from the globe and dissected to give a flat retinal preparation, which was then maintained in the DMEM/HamF12 medium plus 50 μg/mL gentamicin in the 35mm culture dish (Corning, Suzhou, China). Retinal cells were then isolated by enzymatic dissociation [27], and were cultured in F12 medium with 12% serum, plus nerve growth factor (NGF, Sigma) and basic fibroblast growth factor (bFGF, Sigma). RGCs were positive for neurofilaments and Thy-1, but were negative for glial fibrillary acidic protein (GFAP) [27]. Primary RGCs at passage 3-7 were utilized for experiments. Written-informed consent was obtained from each patient. Experiments and the protocols requiring clinical samples were approved by the Ethics Review Board (ERB) of all authors' institutions. All investigations were conducted according to the principles expressed in the Declaration of Helsinki as well as national/ international regulations.

Forced miR-141 expression
The pre-has-miR-141 was obtained from Applied Biosystem (Shanghai, China), which was sub-cloned into pSuper-puro-GFP vector (OligoEngine, Seattle, WA) to construct miR-141 expression vector. The miR-141 construct or the empty vector was transfected to the RPEs and RGCs through Lipofectamine 2000 reagents (Invitrogen, Shanghai, China). The transfection took 48 hours. Afterwards, puromycin (1.0 μg/mL) was added to the complete medium to select stable cells, which lasted 3-4 passages, until at least 95% of cells were GFP positive. Control cells were transfected with non-sense scramble microRNA-control ("miR-C") (Applied Biosystem). Mature has-miR-141-3p expression in the stable cells was tested by RT-qPCR assay.

Real-time quantitative PCR analysis
After applied treatment, cellular RNA was extracted via the Trizol reagents (Invitrogen), which was then utilized to perform the reverse transcription assay [17,20]. The PCR reaction mixture contained 1× SYBR Master Mix (Applied Biosystem), 500 ng RNA together with indicated primers. Real-Time quantitative PCR ("RT-qPCR") was performed by the ABI Prism 7300 Fast system (Shanghai, China). The ΔΔ Ct method was utilized to quantify mRNA expression, and GAPDH was tested as an internal control. Primers were described in our previous studies [8,20] and in published literatures [28]. Expression of mature hsa-miR-141-3p was examined by the same ABI Prism 7300 Fast system. Primers for hsa-miR-141-3p were described previously [29]. All the primers were synthesized by Genepharm (Shanghai, China).

Cell viability and cell death assays
Following treatment, the survival of cells was tested via the routine MTT assay; Cell death was examined by trypan blue staining assay. The detailed protocols were described in our previous publications [17,20,30].

Western blot assay
Western blot assay was performed as described [7,20,31]. Each band was quantified (in total gray) by ImageJ software, and was normalized to the indicated loading control [7]. Each lane was loaded with exact same amount of protein lysates (30 μg per sample). Same set of lysate samples were run in sister gels to test different proteins. For detection of nuclear proteins, cell nuclei were isolated by the nuclei isolation kit from Sigma [20].

Nrf2 shRNA knockdown
The two non-overlapping lentiviral Nrf2 shRNAs, Nrf2 shRNA-A (sc-37030-V) and Nrf2 shRNA-B (sc-44332-V), along with lentiviral scramble control shRNA, were purchased from Santa Cruz Biotech (Santa Cruz); The lentiviral shRNA (10 μL/mL) was added directly to cultured cells for 36 hours. Cells were then subjected to puromycin (1.0 μg/mL) selection for another 6 days. Nrf2 expression in the stable cells was tested by Western blot and RT-qPCR assays.

Statistical analysis
Quantitative results were normalized to the control values of each assay, and were presented as mean ± standard deviation (SD). Data were analyzed by one-way ANOVA. Significance was chosen as p < 0.05.