PRMT5 is a therapeutic target in choroidal neovascularization

Abstract

Ocular neovascular diseases including neovascular age-related macular degeneration (nvAMD) are widespread causes of blindness. Patients' non-responsiveness to currently used biologics that target vascular endothelial growth factor (VEGF) poses an unmet need for novel therapies. Here, we identify protein arginine methyltransferase 5 (PRMT5) as a novel therapeutic target for nvAMD. PRMT5 is a well-known epigenetic enzyme. We previously showed that PRMT5 methylates and activates a proangiogenic and proinflammatory transcription factor, the nuclear factor kappa B (NF-κB), which has a master role in tumor progression, notably in pancreatic ductal adenocarcinoma and colorectal cancer. We identified a potent and specific small molecule inhibitor of PRMT5, PR5-LL-CM01, that dampens the methylation and activation of NF-κB. Here for the first time, we assessed the antiangiogenic activity of PR5-LL-CM01 in ocular cells. Immunostaining of human nvAMD sections revealed that PRMT5 is highly expressed in the retinal pigment epithelium (RPE)/choroid where neovascularization occurs, while mouse eyes with laser induced choroidal neovascularization (L-CNV) showed PRMT5 is overexpressed in the retinal ganglion cell layer and in the RPE/choroid. Importantly, inhibition of PRMT5 by PR5-LL-CM01 or shRNA knockdown of PRMT5 in human retinal endothelial cells (HRECs) and induced pluripotent stem cell (iPSC)-derived choroidal endothelial cells (iCEC2) reduced NF-κB activity and the expression of its target genes, such as tumor necrosis factor α (TNF-α) and VEGF-A. In addition to inhibiting angiogenic properties of proliferation and tube formation, PR5-LL-CM01 blocked cell cycle progression at G₁/S-phase in a dose-dependent manner in these cells. Thus, we provide the first evidence that inhibition of PRMT5 impedes angiogenesis in ocular endothelial cells, suggesting PRMT5 as a potential therapeutic target to ameliorate ocular neovascularization.

Figure 1

PRMT5 is highly expressed in neovascular age-related macular degeneration (nvAMD). PRMT5 immunostaining on sections of eyes from human (a) nvAMD patient (68 year old female) and (b) control (81 year old female), where DAPI (blue) shows the nuclei of the cells and red indicates PRMT5 expression in different layers of the retina, and in the RPE/choroid complex. Higher expression of PRMT5 is observed in the retinal pigment epithelium (RPE)/choroid in the eye with nvAMD. Representative images shown from n = 3 nvAMD patients and controls (see Supplementary Fig. S1). (c) Quantification of PRMT5 mean fluorescence intensity (MFI) in RPE and retina of nvAMD vs controls. Mean ± SEM, n = 3. *p < 0.05, Student’s t-test with Welch’s correction. Scale bars = 20 µm. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; IS/OS photoreceptor inner/outer segments; CC, choriocapillaris.

Figure 2

PRMT5 expression in murine laser-induced choroidal neovascularization (L-CNV). (a) Flat mount staining of L-CNV choroids, showing the expression of PRMT5 (red) in and around the neovascular lesion in the choroid that underwent laser treatment compared to untouched control. (b) Immunoblot, showing PRMT5 is highly expressed in the retina and choroid of the L-CNV mouse eyes compared to the untouched control eyes. See also Supplementary Fig. S2. (c,d) Cryosections of the L-CNV (c) retina and (d) choroid, showing PRMT5 expression in different layers of the retina, including the ganglion cell layer (GCL), inner plexiform layer (IPL), outer plexiform layer (OPL), and in the inner and outer segments (IS/OS) of photoreceptors, and in the retinal pigment epithelium (RPE)-choroid complex where neovascularization is observed (isolectin B4 [IB4] staining, green). Higher expression is observed in the GCL and in the RPE-choroid in L-CNV compared to untouched.

Figure 3

PRMT5 overexpression promotes cell growth. (a) Immunoblot, showing FLAG-tagged wild-type (wt) PRMT5 successfully overexpressed (left panel) or knocked down (right panel) in HRECs. (b) Immunoblot, showing FLAG-tagged wtPRMT5 successfully overexpressed (left panel) or knocked down (right panel) in iCEC2 cells. See also Supplementary Fig. S3. (c–f) Effect of PRMT5 on cell proliferation. Overexpression of wtPRMT5 promoted cell growth in HRECs (c) and iCEC2 cells (d), while shPRMT5 knockdown had the opposite effect in HRECs (e) and iCEC2 cells (f). Mean ± SEM, n = 3–4 biological replicates. *p < 0.05 wtPRMT5 vs. pLV-empty; ^#p < 0.05, shPRMT5 vs. shScramble, unpaired Student’s t-tests with Welch’s correction.

Figure 4

PR5-LL-CM01 is antiangiogenic in vitro. (a) Structure of PRMT5 inhibitor PR5-LL-CM01. (b,c) PR5-LL-CM01 is antiproliferative at 48 h of treatment on HRECs (b) and on iCEC2 cells (c). Mean ± SEM, n = 3 technical replicates. Representative data from three biological replicates. PR5-LL-CM01 dose-dependently decreases cells in S-phase and increases cells in G₀/G₁ after 24 h of treatment in (d,f) HRECs, and in (e,g) iCEC2 cells. Mean ± SEM of percentage of cells, n = 3 biological replicates.

Figure 5

Treatment with PR5-LL-CM01 decreases p65me2, NF-κB activity, and NF-κB target gene expression in HRECs and iCEC2 cells. (a,b) Immunoblots and quantification, indicating inhibition of p65-R30me2 in a stepwise manner with increasing concentrations of PR5-LL-CM01 in HRECs (a) or iCEC2 cells (b) overexpressing FLAG-tagged wt-p65. Lower panels show ImageJ quantification of p65-R30me2 relative to loading control (β-actin) for three independent immunoblots. *p < 0.05 vs. 0 µM PR5-LL-CM01 group. See also Supplementary Fig. S4. (c,d) NF-κB luciferase assay. wtPRMT5 overexpressing HRECs (c) or iCEC2 cells (d) were stimulated with 10 ng/ml IL-1β ± 3 µM PR5-LL-CM01 for 4 h. PRMT5 overexpression augmented NF-κB induction upon IL-1β treatment, while PR5-LL-CM01 inhibitor treatment reduced this effect. *p < 0.05 vs. IL-1β untreated group; ^#p < 0.05 vs. IL-1β-induced group; ^$p < 0.05 vs. pLV-empty + IL-1β-treated group. (e,f) Both shScramble and shPRMT5 HRECs (e) or iCEC2 cells (f) were stimulated with 10 ng/ml IL-1β for 4 h. shPRMT5 inhibited NF-κB activity upon IL-1β treatment. *p < 0.05 vs. IL-1β untreated group; ^$p < 0.05 vs. shScramble + IL-1β-treated group. n = 3–4 technical replicates. (g,i) HREC vector cells or (h,j) iCEC2 vector cells were stimulated with 10 ng/ml IL-1β ± 3 µM PR5-LL-CM01 for 4 h. PR5-LL-CM01 significantly decreased the expression of TNFA (g,h) and VEGFA (i,j) mRNA. Likewise, shPRMT5 reduced TNFA and VEGFA levels. *p < 0.05 vs. IL-1β untreated and PR5-LL-CM01 untreated group; ^#p < 0.05 vs. IL-1β-induced group, and PR5-LL-CM01 untreated group; ^$p < 0.05 vs. shScramble group. Mean ± SD, n = 3–4 biological replicates. Unpaired Student’s t-test with Welch’s correction was used for comparing two means, and one-way ANOVA with Dunnett’s post hoc tests was used when comparing more than two means.

Figure 6

PRMT5 inhibition by shRNA knockdown or PR5-LL-CM01 treatment reduces tube formation in HRECs and iCEC2 cells. (a–d) Tube formation assay in HRECs (a,c) or iCEC2 cells (b,d) transduced with shScramble vector or shPRMT5 reveals that knockdown of PRMT5 reduces tube formation ability. ***p < 0.0001 vs. shScramble control, unpaired Student’s t-test with Welch’s correction. (e–h) Quantitative analysis of tube formation in HRECs (e,g) and in iCEC2 cells (f,h) demonstrates that PR5-LL-CM01 blocks tube formation in a dose-dependent manner. Mean ± SEM, n = 6–12 images. **p < 0.01; ***p < 0.001 vs. DMSO control, one-way ANOVA with Dunnett’s post hoc test. Representative data from three biological replicates. Scale bars = 500 µm. See also Supplementary Fig. S6.

Figure 7

Hypothetical model. PRMT5 methylates and activates NF-κB. This results in the induction of NF-κB downstream genes, including cytokines, angiogenesis factors, chemokines, and antiapoptotic genes, whose functions are critical for inflammation and angiogenesis. Thus, using PR5-LL-CM01 to block the activity of PRMT5 will inhibit neovascularization-associated eye diseases.