Cell reports · 6 authors, 5 centres
This summary was generated by AI from a single paper. It has not been reviewed by a clinician and is not clinical advice. Verify against the source before acting on it.
This study identifies PNPLA2 as a retinyl ester hydrolase in the retinal pigment epithelium that mobilizes retinyl esters from retinosomes, providing substrate for RPE65 to generate 11-cis-retinal. PNPLA2 knockout mice show impaired visual function, delayed dark adaptation, and reduced chromophore regeneration. These findings reveal a critical missing link in the visual cycle and suggest PNPLA2 dysfunction may contribute to vision disorders.
**Background:** The visual cycle regenerates 11-cis-retinal, the chromophore of visual pigments, after photobleaching. In the retinal pigment epithelium (RPE), retinyl esters (REs) are stored in lipid droplets called retinosomes. The key isomerase RPE65, which converts all-trans-RE to 11-cis-retinol, resides in the endoplasmic reticulum (ER), not in retinosomes. Therefore, an unknown retinyl ester hydrolase (REH) must mobilize REs from retinosomes to provide substrate for RPE65. Patatin-like phospholipase domain containing 2 (PNPLA2) is a known triglyceride lipase associated with lipid droplets in other tissues and has been shown to have REH activity in vitro, but its role in the visual cycle was unknown.
**Methods:** The authors used Pnpla2 global knockout (KO) mice and wild-type (WT) controls. They assessed PNPLA2 expression in ocular tissues by qRT-PCR, western blot, and immunohistochemistry. Retinal function was evaluated by scotopic and photopic electroretinography (ERG). Dark adaptation was measured by ERG a-wave recovery after photobleach. Retinoid levels (REs and 11-cis-retinal) were quantified by HPLC under dark-adapted, light-adapted, and post-bleach conditions. Lipid droplet accumulation in RPE was visualized by BODIPY staining and RE autofluorescence. Subcellular localization of PNPLA2, CGI-58, and RPE65 was examined by immunofluorescence and lipid droplet fractionation. REH activity of PNPLA2 was tested in HEK293A cells overexpressing PNPLA2 and/or CGI-58, and by siRNA knockdown of endogenous PNPLA2. A cell-free liposome-based REH assay was also performed. The effect of PNPLA2 on RPE65-mediated 11-cis-retinol production was measured in HEK293A-LRAT cells coexpressing RPE65 and PNPLA2.
**Key Results:** PNPLA2 was expressed in RPE and retina, and localized to lipid droplets/retinosomes, while RPE65 was in the ER and not on lipid droplets. Pnpla2 KO mice showed significantly lower scotopic a-wave and b-wave amplitudes at all tested light intensities (e.g., a-wave at 2.8 log cd·s/m²: WT ~400 μV, KO ~200 μV, p<0.001) and reduced photopic b-wave amplitudes compared to WT. Dark adaptation was delayed: after 60 min, WT a-wave recovered to 100% of baseline, while KO recovered only to 66%. RE levels were significantly higher in KO mice under all conditions (dark-adapted: KO ~800 pmol/eye, WT ~500 pmol/eye; light-adapted: KO ~1200 pmol/eye, WT ~800 pmol/eye; p<0.01). 11-cis-retinal levels were similar in dark- and light-adapted states but were significantly lower in KO after 15 min and 40 min of dark adaptation (15 min: KO ~200 pmol/eye, WT ~300 pmol/eye; 40 min: KO ~350 pmol/eye, WT ~450 pmol/eye; p<0.05). Lipid droplets accumulated markedly in RPE of KO mice. In cell assays, overexpression of PNPLA2 reduced RE levels by 31.3% (433 vs 629 pmol/mg protein), and coexpression with CGI-58 reduced RE by 66.7% (210 vs 629 pmol/mg protein). siRNA knockdown of PNPLA2 increased RE by 42% (221 vs 156 pmol/mg protein). In the cell-free liposome assay, PNPLA2 plus CGI-58 produced significant all-trans-retinol from retinyl palmitate. Coexpression of PNPLA2 with RPE65 increased 11-cis-retinol production by 21% (2008 vs 1660 pmol/mg protein) in cells and by 50% in the in vitro isomerase assay.
**Clinical Implications:** This study identifies PNPLA2 as the long-sought retinyl ester hydrolase that mobilizes REs from retinosomes, providing substrate for RPE65 and thereby supporting chromophore regeneration. PNPLA2 dysfunction may impair dark adaptation and contribute to vision loss in conditions such as age-related macular degeneration or other retinopathies where retinoid metabolism is disrupted. The findings suggest that enhancing PNPLA2 activity could be a therapeutic strategy to improve visual cycle efficiency. However, the global knockout model has limitations, including premature death at ~3 months from cardiac issues, and PNPLA2 is also expressed in neural retina, so its role there remains unclear. Future studies using conditional RPE-specific knockouts are needed to confirm the RPE-specific role.