NFκB-Mediated Expression of Phosphoinositide 3-Kinase δ Is Critical for Mesenchymal Transition in Retinal Pigment Epithelial Cells

Epithelial mesenchymal transition (EMT) plays a vital role in a variety of human diseases including proliferative vitreoretinopathy (PVR), in which retinal pigment epithelial (RPE) cells play a key part. Transcriptomic analysis showed that the phosphoinositide 3-kinase (PI3K)/Akt signaling pathway was up-regulated in human RPE cells upon treatment with transforming growth factor (TGF)-β2, a multifunctional cytokine associated with clinical PVR. Stimulation of human RPE cells with TGF-β2 induced expression of p110δ (the catalytic subunit of PI3Kδ) and activation of NFκB/p65. CRISPR-Cas9-mediated depletion of p110δ or NFκB/p65 suppressed TGF-β2-induced fibronectin expression and activation of Akt as well as migration of these cells. Intriguingly, abrogating expression of NFκB/p65 also blocked TGF-β2-induced expression of p110δ, and luciferase reporter assay indicated that TGF-β2 induced NFκB/p65 binding to the promoter of the PIK3CD that encodes p110δ. These data reveal that NFκB/p65-mediated expression of PI3Kδ is essential in human RPE cells for TGF-β2-induced EMT, uncovering hindrance of TGF-β2-induced expression of p110δ as a novel approach to inhibit PVR.


Introduction
Epithelial mesenchymal transition (EMT), a process by which epithelial cells lose their cell polarity and gain mesenchymal properties, is essential for animal development and wound healing. EMT also plays a critical role in a variety of human diseases including organ fibrosis, cancer metastasis and proliferative vitreoretinopathy (PVR). PVR, a fibrotic eye disease, develops in 8-10% of patients who undergo retinal surgery to correct rhegmatogenous retinal detachment (RRD) [1][2][3][4][5] and in 40-60% of those with open-globe injury [6][7][8][9][10][11][12][13]. There are an estimated 55,000 individuals experiencing RRD in the United States annually [14] and approximately 203,000 cases of open-globe injury worldwide each year [6,10,15]. Currently, repeat surgery is the only option for treating PVR [16][17][18]; however, the outcome of the surgery to restore vision is poor due to the retinal damage resulting from recurrent detachment and the PVR process itself [19]. In addition, adjuvant

RNA Sequencing
When cells had grown to 90% confluence in 10 cm dishes, they were treated with 10 ng/mL TGF-β2 for 48 h (h). Subsequently, total RNA was extracted using TriZol reagent (Thermo Fisher Scientific) according to the manufacturer's procedure. Differential gene expression analysis was performed by DESeq2 software (www.bioconductor.org (accessed on 24 November 2021)) between two different groups (and by edge R between two samples). The genes with the parameter of false discovery rate (FDR) below 0.05 and absolute log2 fold change ≥1.1 were considered differentially expressed genes. Differentially expressed genes were then subjected to enrichment analysis of GO functions and KEGG pathways.
To express SpGuides in the targeted cells, the top oligonucleotides 5 -CACCG-20-nt (target RELA DNA sequence p65 (ACTACGACCTGAATGCTGTG) or the lacZ sgRNA sequence)-3 and the bottom oligonucleotides 5 -AAAC-20-nt (20-nt: complementary target RELA DNA sequences or lacZ sgRNA sequence)-C-3 were annealed and cloned into the lentiCRISPR v2 vector by BsmBI (New England Biolabs, Boston, MA, USA), respectively. All clones were confirmed by DNA sequencing using primer 5 -GGACTATCATATGCTTACCG-3 from a sequence of the U6 promoter that drives expression of sgRNAs. DNA synthesis and sequencing were performed by Massachusetts General Hospital DNA Core Facility (Cambridge, MA, USA).

Lentivirus Production
Lentivirus was produced as described previously [34]. Briefly, the lentiCRISPRv2 vector inserted with sgRNA, the packaging plasmid psPAX2 (Addgene, Catalog #12260) and the envelope plasmid VSV-G (Addgene, Catalog #8454) were mixed together and then added to a mixture of Lipofectamine 3000 (Thermo Fisher Scientific) with Opti-MEM (Thermo Fisher Scientific). This transfection mix was kept at room temperature for 30 min (min) and then carefully added into HEK-293T cells in a 60 mm cell culture dish. After 18 h (37 • C, 5% CO 2 ), the medium was replaced with growth medium supplemented with 20% FBS, and lentiviruses were harvested at 24 h after changing the medium and then daily for 2 days. The virus-containing media were pooled and centrifuged at 800× g for 5 min to remove cell debris. The supernatant was used to infect APRE-19 cells supplemented with 8 µg/mL polybrene (Sigma). The infected cells were selected in media containing puromycin (Sigma) (4 µg/mL), and the resulting cells examined by Western blotting analysis [34,35].

Immunofluorescence
Immunofluorescence was performed as described previously [34,35]. Briefly, following starvation and treatment with TGF-β2, cells were fixed in 3.7% formaldehyde/PBS for 10 min at room temperature, followed by blocking with 5% normal goat serum in 0.3% Triton X-100/PBS for 30 min and overnight incubation at 4 • C with primary antibodies (1:200 dilution). After thorough washes with 0.3% Triton X-100/PBS to remove non-specific binding, samples were incubated with fluorescent-labeled secondary antibodies Dylight 549 or 488 (anti-rabbit IgG) (1:300 dilution in a blocking buffer) for 1 h, followed by washing with 0.3% Triton X-100/PBS, mounting in mount medium with 4 ,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories) and photographing using a fluorescent microscope. Experiments were repeated at least 3 times.

Cell Migration Assay
This assay was conducted according to procedures in our published literature [34]. Briefly, confluent monolayers of cells in 24-well plates were scratched using autoclaved 200 µL pipet tips across the wells and detached cells aspirated away using PBS. The remaining cells were treated with TGF-β2 (10 ng/mL). Scratched areas were photographed once after scratching for the initial width and again 18 h later. The data were analyzed using Image J and Adobe Photoshop CS4 software as described previously [34]. At least 3 independent experiments were performed.

Luciferase Assay-PIK3CD Reporter
This assay was conducted according to the Promega Luciferase Assay Systems. Cells seeded into 96-well plates were used at around 90% confluence. PGL3-PIK3CD (the p110δ gene) reporter vectors p-GL3-R1, p-GL3-GN43 (or p-GL3 control) and pRL-RK were transfected overnight into ARPE-19 cells with lipofectamine 3000 (Life Technologies), followed by treatment with TGF-β2 for 24 h. After washing with PBS twice, cells were lysed with 1× passive lysis buffer, gently rocked for 15 min at room temperature, followed by addition of 50 µL LAR (luciferase reporter assay) to a 12 mm × 50 mm tube (Part #E2371, Promega. Disposable Cubettes), and the lysates were transferred from a well into the LAR substrate. These mixtures were first read for firefly luciferase activity in a TD-20/20 luminometer (Turner Designer, Promega). Finally, 20 µL 1× Stop and Glo substrate was added into the tubes, mixed and read again to obtain Renilla luciferase activity [28]. At least 3 independent experiments were performed.

Chromatin Immunoprecipitation Assay
Chromatin immunoprecipitation (ChIP) was conducted as described previously (35). In brief, cross-linking of protein-DNA complexes was performed by adding 37% paraformaldehyde diluted to a 1% final concentration and incubation of cells at room temperature for Cells 2023, 12, 207 5 of 15 15 min, followed by addition of glycine (125 mM) to quench the fixation. In total, 500 µL of lysis buffer (10 µg/mL leupeptin, 10 µg/mL of aprotinin and 1 mM PMSF were added per 5 × 10 6 cells for resuspension. Cell lysates were sonicated to shear chromatin to an average length of~1 kb, followed by supernatant collected after centrifugation at 12,000× g. Agarose beads were incubated with 5 µg of an anti-NFκB or non-immune rabbit IgG at 4 • C with a rotation for 2 h, followed by adding samples to the beads and incubation at 4 • C overnight. The following day, beads were collected by centrifugation and washed 4 times. In total, 100 µL of Tris-EDTA buffer was added to the sample and boiled for 10 min. Finally, samples were centrifuged for 1 min at 12,000× g, and the supernatant was collected into a clean tube. The ChIP samples were amplified by PCR using the following primers: forward: 5 -AAGGAGGGAGAGATGGGA-3 , reverse: 5 -ATACACGCGCTCGCTCTT-3 for PIK3CD-2e promoter, and they were subjected to a gel analysis to detect NFκBbound DNA.

Statistics
Data from 3 independent experiments were analyzed using an unpaired t-test between two groups and one-way analysis of variance (ANOVA) followed by the Tukey honest significant difference (HSD) post hoc test among more than two groups as described preciously [34]. A p value less than 0.05 was considered significant difference.

RNA Sequencing Analysis Uncovers TGF-β2-Induced PI3K/Akt Signaling in RPE Cells
The levels of TGF-β2 in the vitreous are known to be associated with PVR [36], with TGF-β2 being considered to be the primary factor to induce EMT [5,25,36]. In order to identify molecules involved in EMT induced by TGF-β2 in a human RPE cell line (referred to as ARPE-19), we performed transcriptomic sequencing analysis. The results showed 4740 differentially regulated genes, of which 2405 were up-regulated and 2335 downregulated ( Figure 1A).
Gene ontology (GO) analysis of these genes sets indicated a broad distribution of these TGF-β2 targets in the intracellular organelle and membranes ( Figure 1B), with biological processes connected with cellular responses such as cell adhesion, immune system process and cell proliferation. Molecular function analysis indicated that TGF-β2 induced binding, structural and transcription regulator activities ( Figure 1B).
Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis predicted that TGF-β2 regulated processes related to focal adhesion and the actin cytoskeleton as well as the activated PI3K/AKT signaling pathway ( Figure 1C), as well as immune system processes and transcription regulator activities. Based on these observations, we speculated that the nuclear transcription factor (NFκB), which is key in inflammatory and immune responses [37], also played a role in experimental PVR [27][28][29].

Depletion of NFκB/p65 in RPE Cells Attenuates TGF-β2-Induced Akt Activation, Expression of p110δ as Well as EMT
It is not clear at present which of the PI3K family members plays a predominant role in TGF-β2-induced EMT in RPE cells. We previously reported that PI3Kδ inactivation prevents vitreous-induced activation of Akt in human RPE cells [34], and other researchers showed that Akt could be activated by TGF-β2 [38,39]. We therefore speculated that PI3Kδ

Depletion of NFκB/p65 in RPE Cells
Attenuates TGF-β2-Induced Akt Activation, Expression of p110δ as Well as EMT It is not clear at present which of the PI3K family members plays a predominant role in TGF-β2-induced EMT in RPE cells. We previously reported that PI3Kδ inactivation prevents vitreous-induced activation of Akt in human RPE cells [34], and other researchers showed that Akt could be activated by TGF-β2 [38,39]. We therefore speculated that PI3Kδ could play a role in TGF-β2-stimulated Akt activation in RPE cells. To gain insight into this, we performed a time course of TGF-β2-induced Akt activation. As shown in Figure 2A,E, TGF-β2 stimulated activation of Akt, as measured by phosphorylation on Ser473 (8.84 ± 1.14 fold) within 4 h, but it decreased at 48 h; in addition, the expressions of snail and fibronectin, markers of EMT, were increased (snail: 44.79 ± 26.93 fold; fibronectin: 8.43 ± 0.21 fold) after 24 h upon TGF-β2 treatment (Figure 2A,H,I).
Based on the notion that TNFα induces NFκB/p65-dependent expression of p110δ [30], we explored whether TGF-β2 could also stimulate p110δ transcription via NFκB/p65. Firstly, we examined the cellular localization of p65 as a marker of NFκB activation after TGF-β2 treatment. Immunofluorescence analyses confirmed that TGF-β2 treatment could promote the nuclear transfer of NFκB ( Figure 2B-D). Expression of NFκB-dependent genes (e.g., IL-8, IL-1β and IL-6) was also up-regulated by TGF-β2 treatment ( Figure 2J-L), suggesting that TGF-β2 is able to activate NFκB.  D). Co-localization analysis of NFκB with nucleus was performed by Image J [40]. (E-I). Quantitation of Western blotting band intensity in A. The graphs are mean ± standard deviation (SD) of three independent experiments. **p < 0.01, *** p < 0.001. (J-L). mRNA from ARPE-19 cells treated with TGF-β2 (10 ng/mL) for 48 h was extracted and subjected to reverse qPCR. The graphs are mean ± SD of 4 independent experiments. The data were analyzed using one-way ANOVA followed by the Tukey HSD post hoc test. ** p < 0.01, *** p < 0.001.
To determine if TGF-β2-induced expression of p110δ was dependent on NFκB, we established an ARPE-19 cell line with NFκB/p65 depleted by CRISPR/Cas9 using a sgRNA targeting RELA encoding NFκB/p65 ( Figure 3A). As shown in Figure 3B,C, there was a more than 90% reduction in NFκB/p65 expression in the NFκB/p65-depleted cells compared with those expressing lacZ-sgRNA as a control. Intriguingly, upon on the depletion of NFκB/p65, there was concomitance with an attenuation of TGF-β2-induced Akt activation and p110δ expression ( Figure 3D-F). These results encouraged us to investigate if depletion of NFκB/p65 could suppress TGF-β2-induced EMT of ARPE-19 cells. Both the Western blot and immunofluorescence analyses showed that depletion of NFκB/p65 blocked TGF-β2-induced expression of fibronectin and snail markers for EMT ( Figure 3G-L).
Cells 2023, 11, x FOR PEER REVIEW 8 of 16 more than 90% reduction in NFκB/p65 expression in the NFκB/p65-depleted cells compared with those expressing lacZ-sgRNA as a control. Intriguingly, upon on the depletion of NFκB/p65, there was concomitance with an attenuation of TGF-β2-induced Akt activation and p110δ expression ( Figure 3D-F). These results encouraged us to investigate if depletion of NFκB/p65 could suppress TGF-β2-induced EMT of ARPE-19 cells. Both the Western blot and immunofluorescence analyses showed that depletion of NFκB/p65 blocked TGF-β2-induced expression of fibronectin and snail markers for EMT ( Figure 3G-L). The data were analyzed using one-way ANOVA followed by the Tukey HSD post hoc test. (J). ARPE-19 cells expressing SpCas9 with LacZ or RELA (encoding NF-κB/p65) sgRNA treated with TGF-β2 (10 ng/mL) for 48 h were stained with antibodies to fibronectin (rabbit) and pan-keratin (mouse), followed by incubation with fluorescently labeled secondary antibodies. The slides were Shown is a representative of three independent experiments. Scale bar: 20 µm. (K,L). Integrated density of the fluorescence was analyzed by Image J. The graphs are mean ± SD of 3 independent experiments. The data were analyzed using one-way ANOVA followed by the Tukey HSD post hoc test. * p < 0.05, ** p < 0.01, *** p < 0.001.

Depletion of PI3Kδ in RPE Cells Prevents TGF-β2-Induced EMT
ARPE-19 cells undergo EMT upon stimulation with TGF-β2 [35], and these cells can also induce experimental PVR [43]. Hence, we next examined if PI3Kδ depletion in ARPE-19 cells could block TGF-β2-induced expression of fibronectin, a protein marker of EMT. As previously reported [35], TGF-β2 induced expression of fibronectin in ARPE-19 cells, with this expression being prevented (73.8 ± 2.3%) by CRISPR/Cas9-mediated depletion of p110δ as shown by analyses of Western blotting ( Figure 4F,G) and immunofluorescence ( Figure 4H-L). Taken together, depletion of PI3Kδ in ARPE-19 cells is able to block TGF-β2-induced EMT.   A,B). ARPE-19 cells with lacZ-sgRNA/SpCas9 or PK2-sgRNA/SpCas9 were examined by Western blotting using indicated antibodies. Shown is a representative of at least 3 independent experiments (A); Quantitation of Western blotting band intensity in A (B). (C-G). Serum-starved ARPE-19 expressing SpCas9 with LacZ or PK2 sgRNA were treated with TGF-β2 (10 ng/mL) for 48 h. Their lysates were subjected to Western blot analyses using indicated antibodies. Shown is representative of at least 3 independent experiments (C,F). Quantitation of Western blotting band intensity in (C-G). (H). ARPE-19 cells expressing SpCas9 with LacZ or PK2 sgRNA treated with TGF-β2 (10 ng/mL) for 48 h were first stained with antibodies against fibronectin (rabbit), pan-keratin (mouse), N-cadherin (rabbit) and β-catenin (rabbit), followed by incubation with fluorescently labeled secondary antibodies and mounting of the slides in mounting medium containing DAPI (blue). Red signals indicate proteins expression. Shown is a representative of 3 independent experiments. Scale bar: 20 µm. (I-L). Integrated density of the fluorescence was analyzed by Image J. The bar graphs are mean ± SD of 3 independent experiments. * p < 0.05, ** p < 0.01, *** p < 0.001.

NFκB/p65 Binding to the PIK3CD-2e Promoter Is Required for the TGF-β2-Induced p110δ Expression in RPE Cells
To gain additional insight into the mechanism of p110δ expression, we next evaluated if the PIK3CD promoter element responsible for the TGF-β2 responsiveness was the same locus as previously shown for TNFα [30]. To this end, we transfected ARPE-19 cells with a pGL3-basic luciferase reporter plasmid or a pGL3 plasmid containing the PIK3CD (-2e) promoter element, which contains an NFκB/p65 binding region that is responsive to TNFα stimulation in human ECs [30] and then treated these cells with TGF-β2 for 24 h ( Figure 5A). A luciferase reporter assay showed that the basal activity of the exon-2e promoter [30] was comparable in cells with or without stimulation, but its activity was enhanced upon TGF-β2 simulation; furthermore, depletion of NFκB/p65 diminished the TGF-β2-induced binding of this TF to the exon-2e promoter in ARPE-19 cells ( Figure 5B). Furthermore, ChIP using an antibody against NFκB/p65 showed that NFκB could bind to the PIK3CD-2e promoter (primers designed based on the sequences of PIK3CD-2e promoter) in the ARPE-19 cells treated with TGF-β2 ( Figure 5C). Taken together, these results demonstrate that NFκB/p65 binding to the PIK3CD-2e promoter is required for TGF-β2-induced expression of p110δ in ARPE-19 cells, similar to as was shown for response to TNFα stimulation in vascular ECs.  . (A,B). After cells had attached to a 96-well plate, they were transfected overnight with p-GL3-R1 (or p-GL3 control) or pRL-RK using lipofectamine 3000. Cells were then treated with TGF-β2 (10 ng/mL) for the next 24 h. Firefly luciferase activity and Renilla luciferase activity were read in a TD-20/20 luminometer. Data were calculated as firefly luciferase activity/Renilla luciferase activity. The mean ± SD of 3 independent experiments is shown. Data were analyzed using one-way ANOVA followed by the Tukey HSD post hoc test. * Denotes p < 0.05. (C). NFκB/p65 Figure 5. NFκB/p65 binding to the PIK3CD-2e promoter is required for TGF-β2-induced expression of p110δ in ARPE-19 cells. (A,B). After cells had attached to a 96-well plate, they were transfected overnight with p-GL3-R1 (or p-GL3 control) or pRL-RK using lipofectamine 3000. Cells were then treated with TGF-β2 (10 ng/mL) for the next 24 h. Firefly luciferase activity and Renilla luciferase activity were read in a TD-20/20 luminometer. Data were calculated as firefly luciferase activity/Renilla luciferase activity. The mean ± SD of 3 independent experiments is shown. Data were analyzed using one-way ANOVA followed by the Tukey HSD post hoc test. * Denotes p < 0.05. (C). NFκB/p65 binding to the PIK3CD-2e promoter determined by a ChIP assay. DNA from ChIP was subjected to PCR and gel analysis. Non-immune IgG served as a negative control.

Depletion of PI3Kδ Prevents TGF-β2-Induced Migration of RPE Cells
We next used a scratch-wound assay to examine if PI3Kδ depletion could prevent TGF-β2-induced migration of ARPE-19 cells because this cellular event is critical for the formation of epiretinal membranes, whose contraction causes retinal detachment [44]. As shown in Figure 6A,B, TGF-β2 promoted migration of ARPE-19 cells with lacZ-sgRNA/SpCas9 at 16 h, but depletion of p110δ with PK2-sgRNA/SpCas9 significantly attenuated TGF-β2induced migration of these cells at this time point, demonstrating that inactivation of PI3Kδ could prevent PVR pathogenesis. . Wells of confluence cells were scratched using 200 μL pipet tips, washed and photographed at the initial width, followed by treatment with TGF-β2 (10 ng/mL) for 24 h. Cell images were then taken, and wound areas were analyzed by image J and Adobe Photoshop CS4 software. The data of bar graphs are the mean ± SD of 3 independent experiments. Representative raw data of 3 independent experiments are shown underneath the bar graphs. (B). *** denotes p < 0.001 using one-way ANOVA followed by the Tukey HSD post hoc test. (C). Diagram of a loop pathway of PI3Kδ/Akt/NFκB/PI3Kδ induced by TGF-β2. Growth factors (GFs) and cytokines (CKs) in the vitreous [24,45] activate the PI3K/Akt signaling pathway, resulting in Mdm2 phosphorylation and a decline in p53 levels [28,44]. Receptorregulated PI3Ks consist of PI3Kα, PI3Kβ and PI3Kδ [31]. In particular, TGF-β2 in the vitreous [46] triggers heightened expression of Mdm2 [47], resulting in elevated levels of NFκB/p65 [48] and p110δ [30], the catalytic subunit of PI3Kδ [31]. These biochemical events (e.g., a decrease in p53 [49][50][51] and an increase in NFκB/p65 [52,53]) consequently promote cellular responses (e.g., proliferation, EMT, migration and contraction), driving PVR pathogenesis.

Discussion
In the present study, to explore the molecular mechanism by which EMT develops in RPE cells, RNA sequencing was employed to profile TGF-β2-induced changes in cultured . Wells of confluence cells were scratched using 200 µL pipet tips, washed and photographed at the initial width, followed by treatment with TGF-β2 (10 ng/mL) for 24 h. Cell images were then taken, and wound areas were analyzed by image J and Adobe Photoshop CS4 software. The data of bar graphs are the mean ± SD of 3 independent experiments. Representative raw data of 3 independent experiments are shown underneath the bar graphs. (B). *** denotes p < 0.001 using one-way ANOVA followed by the Tukey HSD post hoc test. (C). Diagram of a loop pathway of PI3Kδ/Akt/NFκB/PI3Kδ induced by TGF-β2. Growth factors (GFs) and cytokines (CKs) in the vitreous [24,45] activate the PI3K/Akt signaling pathway, resulting in Mdm2 phosphorylation and a decline in p53 levels [28,44]. Receptorregulated PI3Ks consist of PI3Kα, PI3Kβ and PI3Kδ [31]. In particular, TGF-β2 in the vitreous [46] triggers heightened expression of Mdm2 [47], resulting in elevated levels of NFκB/p65 [48] and p110δ [30], the catalytic subunit of PI3Kδ [31]. These biochemical events (e.g., a decrease in p53 [49][50][51] and an increase in NFκB/p65 [52,53]) consequently promote cellular responses (e.g., proliferation, EMT, migration and contraction), driving PVR pathogenesis.

Discussion
In the present study, to explore the molecular mechanism by which EMT develops in RPE cells, RNA sequencing was employed to profile TGF-β2-induced changes in cultured RPE cells. These studies identified the PI3K/AKT signaling pathway to be significantly upregulated upon the TGF-β2 treatment. We previously showed that PI3Kδ plays an essential role in vitreous-induced Akt activation, and another research team demonstrated that TGF-β2 levels in vitreous from patients with PVR are associated with clinical PVR [28,54]. In the current study, we found that PI3Kδ is critical for TGF-β2-induced Akt and cellular responses including EMT and migration of RPE cells. At this stage, it remains elusive whether other PI3K isoforms are involved in PVR pathogenesis and how TGF-β2 may selectively use PI3Kδ to activate Akt. Indeed, other PI3K isoforms are also present in RPE cells [34], and it is not clear why they can apparently not compensate for PI3Kδ inactivation. PI3K isoforms have been shown to have isoform-selective functions at the cellular and organismal level, which may relate to their relative expression levels and isoform-selective coupling to upstream activating signals such as small GTPases [55]. PI3Kδ is highly expressed in human REP cells (34), and thereby the existence of a TGF-β2induced PI3Kδ/Akt/NFκB/PI3Kδ feedback loop ( Figure 6C) to enhance PVR pathogenesis may contribute to such a PI3K isoform-specific function.
We recently showed that PI3Kδ is also important in vitreous-induced activation of Mdm2 [34]. Other studies revealed that patients who harbor the MDM2 single nucleotide polymorphism (SNP) 309 G allele (MDM2 SNP309G ) are more likely to develop PVR [56] and that primary human fetal RPEs containing MDM2 SNP309G have greater potential to induce PVR in an animal model [21,57]. The MDM2 gene contains two promoters designated P1 and P2 [58,59], with P1 being considered as a housekeeping promoter, while P2 at the intron between exons 1 and 2 can be activated by a variety of transcription factors including the small mothers against decapentaplegic (Smad)3/4 complex [47] and specificity protein (Sp)1 in response to various stimuli such as TGF-β2 [58]. Blockade of the P2-driven MDM2 expression prevents vitreous-induced p53 degradation (49] or TGF-β2-induced EMT [35]. In addition, Mdm2 also executes activities independently of p53, including induction of nuclear factor (NF)κB subunit p65 (NFκB/p65) transcription [60]. Thus, the major function of the TGF-β2-induced loop of PI3Kδ/Akt/NFκB/PI3Kδ might be to suppress the levels of p53 via Mdm2 because p53 is a gatekeeper of retinal detachment [44]. As a matter of fact, both a decrease in p53 [34,44] and an increase in NFκB/p65 [46] can promote inflammation contributing to PVR pathogenesis (e.g., EMT).
In this study, we also found that depletion of p110δ significantly suppressed NFκB/p65 expression ( Figure 4A). Based on that, Mdm2 is able to transcriptionally induce NFκB/p65 expression [60], and we hypothesized that PI3Kδ affected NFκB abundance via Mdm2, which was activated by TGF-β2-induced PI3Kδ/Akt signaling. Thereby, we proposed that a circuit of PI3Kδ/Akt/Mdm2/NFκB/PI3Kδ triggered by TGF-β2 might exist in RPE cells. This feedback loop is currently under investigation.
PVR is still a major obstacle to successfully correct RD despite gradual improvements in surgical success rates over the past decades; in particular, there are over 75% of postsurgical re-detachments [61]. However, there is still no effective treatment for this blinding disease even though numerous clinically approved anti-proliferative medicines have been evaluated [61,62]. Nonetheless, novel approaches including CRISPR/Cas9 have shed light in clinical trials on the treatment of human diseases, especially eye diseases (62). In summary, based on our discovery of the molecular mechanism of PVR pathogenesis (e.g., EMT) [34,63], a genome-editing approach to targeting the PI3Kδ/Akt/NFκB/PI3Kδ feedback loop ( Figure 6C) has great potential for therapeutic treatment of EMT-related diseases, including PVR, addressing a currently unmet clinical need [64].