Lycopene inhibits endothelial-to-mesenchymal transition of choroidal vascular endothelial cells in laser-induced mouse choroidal neovascularization

Abstract

Choroidal neovascularization (CNV), is a major cause of irreversible blindness among the elderly population in developed countries, which is resulted from subretinal fibrosis without effective therapeutic strategies. Endothelial-to-mesenchymal transition (EndMT) of choroidal vascular endothelial cells (CVECs) contributes to subretinal fibrosis. Lycopene (LYC), a non-pro-vitamin A carotenoid, plays an anti-fibrotic role. Herein, we explored the effect and mechanism of LYC on the EndMT of CVECs during CNV. Firstly, LYC inhibited EndMT in hypoxic human choroidal endothelial cells (HCVECs). Meanwhile, LYC inhibited proliferation, androgen receptor (AR) expression and nuclear localization in hypoxic HCVECs. Then LYC-inhibited AR promotes the activation of microphthalmia-associated transcription factor (MITF) in hypoxic HCVECs. In addition, LYC down-regulated AR and induced MITF up-regulated pigment epithelium-derived factor (PEDF) transcription and expression in hypoxic HCVECs. Moreover, LYC-induced PEDF bound to laminin receptor (LR), inhibiting EndMT of hypoxic HCVECs via down-regulating protein kinase B (AKT)/β-catenin pathway. In vivo, LYC alleviated mouse laser-induced subretinal fibrosis secondary to CNV via up-regulating PEDF without any ocular or systemic toxicity. These results indicate that LYC inhibits EndMT of CVECs via modulating AR/MITF/PEDF/LR/AKT/β-catenin pathway, showing LYC is a promising therapeutic agent for CNV.

Keywords: choroidal neovascularization; choroidal vascular endothelial cells; endothelial to mesenchymal transition; lycopene.

FIGURE 1

LYC inhibits EndMT in hypoxia‐exposed HCVECs. HCVECs were divided into normal and hypoxia groups. (A) Western blot (WB) is conducted to detect the protein levels of endothelial cell markers CD31 and VE‐cadherin, as well as mesenchymal cell markers α‐SMA and vimentin. (B–E) The relative expression of each molecule i analysed. HCVECs were divided into normal, hypoxia (48 h), hypoxia +0.1% DMSO, hypoxia +5 μM LYC, hypoxia +10 μM LYC, hypoxia +20 μM LYC and hypoxia +40 μM LYC groups. CD31 (F) and α‐SMA (G) mRNA levels were detected by qRT‐PCR. *p < 0.05, **p < 0.01, compared to normal group. ^# p < 0.05, ^## p < 0.01, compared to hypoxia group. n = 6 in each group.

FIGURE 2

LYC inhibits proliferation, AR expression and nuclear localization in hypoxia‐exposed HCVECs. HCVECs were divided into normal, hypoxia, hypoxia +0.1% DMSO, hypoxia +5 μM LYC, hypoxia +10 μM LYC, hypoxia +20 μM LYC, and hypoxia +40 μM LYC groups. (A) The viability of HCVECs was evaluated by means of CCK‐8 assay. (B) P‐AR and AR protein levels were detected by WB. (C, D) P‐AR and AR protein levels were analysed. (E) AR was detected in normal, hypoxia, hypoxia +0.1% DMSO and hypoxia +20 μM LYC groups by IF. DAPI (blue) was used for labelling the nucleus. (F) The co‐localization of AR as well as DAPI was analysed utilizing the Pearson coefficient (PCC). *p < 0.05, compared to the normal group. ^# p < 0.05, compared to the hypoxia group. n = 5 in each group.

FIGURE 3

LYC‐inhibited AR promotes the activation of MITF in HCVECs under hypoxic conditions. HCVECs were divided into normal, hypoxia, hypoxia +0.1% DMSO, hypoxia + LYC (20 μM), and hypoxia + LYC + TES groups. (A) P‐AR, AR, p‐MITF and MITF protein levels were detected utilizing WB. (B–E) The relative expression of each molecule was analysed. (F) Nuclear and cytoplasmic separation and WB were done to detect MITF translocation. (G, H) MITF relative expression in the nucleus and cytoplasm was analysed. *p < 0.05, compared to the normal group. ^# p < 0.05, compared to the hypoxia group. ^% p < 0.05, compared to the hypoxia + LYC group. n = 5 in each group.

FIGURE 4

LYC‐down‐regulated AR and ‐induced MITF up‐regulate PEDF transcription and expression in HCVECs under hypoxic conditions. HCVECs have been classified into normal, hypoxia, hypoxia +0.1% DMSO, hypoxia + LYC (20 μM), hypoxia + LYC + TES, and hypoxia + LYC + ML329 groups. (A) PEDF mRNA levels were detected by qRT‐PCR. (B) PEDF protein levels were detected by WB. (C) PEDF protein levels were analysed. (D) The MITF binding site (5′‐CACGTG‐3′, from −672 to −667) in the human PEDF promoter is shown. (E) The activity of the human PEDF promoter was detected by luciferase reporter assay in HCVECs. (F) The direct binding of MITF to the PEDF promoter was detected by CHIP. *p < 0.05, compared to the normal group. ^# p < 0.05, in comparison to the hypoxia group. ^% p < 0.05, in comparison to the hypoxia + LYC group. n = 5 in each group.

FIGURE 5

LYC‐induced PEDF binds to LR on HCVECs under hypoxic conditions. HCVECs were divided into normal, hypoxia, hypoxia +0.1% DMSO, hypoxia + LYC (20 μM), hypoxia + LYC + control siRNA, hypoxia + LYC + PEDF siRNA and hypoxia + human PEDF recombinant protein groups. (A) PEDF‐R and LR protein levels were detected by WB. (B‐C) PEDF‐R and LR protein levels were analysed. (D) HCVECs were divided into normal, hypoxia, hypoxia +0.1% DMSO and hypoxia + LYC groups. PEDF and LR were labelled in CECs. (E) The co‐localization of PEDF and LR was analysed. *p < 0.05, in comparison to the normal group. ^# p < 0.05, in comparison to the hypoxia group. ^% p < 0.05, in comparison to the hypoxia + LYC group. n.s. denoted no significance. n = 5 in each group.

FIGURE 6

LYC‐induced PEDF inhibits EndMT of HCVECs via down‐regulating AKT/β‐catenin pathway under hypoxic conditions. HCVECs were divided into normal, hypoxia, hypoxia +0.1% DMSO, hypoxia + LYC (20 μM), hypoxia + LYC + control siRNA, hypoxia + LYC + PEDF siRNA, hypoxia + LYC + SC79, and hypoxia + LYC + SKL2001 groups. (A) P‐AKT, AKT, p‐β‐catenin and β‐catenin protein levels were detected by WB. (B, C) The relative expression of p‐AKT and p‐β‐catenin was analysed. (D) CD31, VE‐cadherin, α‐SMA and vimentin protein levels were detected by WB. (E–H) The relative expression of each molecule was analysed. *p < 0.05, in comparison to the normal group. ^# p < 0.05, in comparison to the hypoxia group. ^% p < 0.05, in comparison to the hypoxia + LYC group, n = 5 in each group.

FIGURE 7

LYC alleviates mouse laser‐induced subretinal fibrosis secondary to CNV via up‐regulating PEDF. The mice were divided into normal, CNV 14 days, CNV 14 days + 0.1% DMSO (oral gavage; 0.8 mg/kg/day for consecutive 14 days), CNV 14 days + LYC (oral gavage; 0.8 mg/kg/day for consecutive 14 days), CNV 14 days + LYC + control siRNA, CNV 14 days + LYC + PEDF siRNA, and CNV 14 days + CON groups. (A) Representative images of subretinal fibrosis lesions by marking α‐SMA in each group (except the normal group) at post‐laser 14 days were shown. Scale bar = 100 μm. (B) Quantitative analyses for the area of α‐SMA were done. Snail (C) and α‐SMA (D) mRNA levels were detected by qRT‐PCR. (E) Snail and α‐SMA protein levels were detected by Western blot. Snail (F) and α‐SMA (G) protein levels were analysed. *p < 0.05, in comparison to the normal group. ^# p < 0.05, in comparison to the hypoxia group. ^% p < 0.05, in comparison to the hypoxia + LYC group. n = 5 in each group.

FIGURE 8

LYC causes no ocular or systemic toxicity. The mice were divided into normal, 0.1% DMSO (oral gavage; 0.8 mg/kg/day for consecutive 14 days), and LYC (oral gavage; 0.8 mg/kg/day for consecutive 14 days) groups. (A) HE staining on retina‐RPE‐choroid cryosection was done. (B) The A/B ratio was analysed. (C) TUNEL staining on retina‐RPE‐choroid cryosection was done. (D) The body weight of each mouse was measured. (E) HE staining on mouse heart, lung, liver, kidney, brain and stomach tissue was done. n = 4 in each group.

FIGURE 9

The mechanism diagram of LYC on subretinal fibrosis during CNV is shown.