PNPLA6 regulates retinal homeostasis by choline through phospholipid turnover

Abstract

Although mutations in human patatin-like phospholipase PNPLA6 are associated with hereditary retinal degenerative diseases, its mechanistic action in the retina is poorly understood. Here, we uncover the molecular mechanism by which PNPLA6 dysfunction disturbs retinal homeostasis and visual function. PNPLA6, by acting as a phospholipase B, regulates choline mobilization from phosphatidylcholine and subsequent choline turnover for phosphatidylcholine regeneration in retinal pigment epithelial cells. PNPLA6-driven choline is supplied from retinal pigment epithelial cells to adjacent photoreceptor cells to support their survival. Inhibition of this pathway results in abnormal morphology, proliferation, metabolism, and functions of retinal pigment epithelial and photoreceptor cells, and mice with retina-specific PNPLA6 deletion exhibit retinitis pigmentosa-like retinal degeneration. Notably, these abnormalities are entirely rescued by choline supplementation. Thus, PNPLA6 plays an essential role in retinal homeostasis by controlling choline availability for phospholipid recycling and provide a framework for the development of an ophthalmic drug target for retinal degeneration.

Fig. 1. Histological features of the retina in tamoxifen-inducible systemic Pnpla6-deficient mice.

a HE staining and immunostaining for PNPLA6 (red, PNPLA6; blue, DAPI) in the RPE and sensory retina cells of healthy human retinas. Scale bar, 100 µm. Boxed areas in the left panels are magnified in the right panels. b qPCR analysis of Pnpla6 and the RPE marker Rpe65 in fractionated mouse eye tissue (n = 4 independent animals). c Tamoxifen administration protocol in Pnpla6^fl/flCreER(−) or Pnpla6^fl/flCreER(+) mice. d qPCR analysis of Pnpla6 in RPE cells isolated from Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice treated with tamoxifen, with Gapdh as a normalization control (n = 5 independent animals). e HE staining of retina sections from Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice. Brackets, ONL. Scale bar, 50 μm. f Quantification of retinal thickness (n = 6 from e). g Vertical number of nuclei in the ONL (n = 6 from e). h TEM images of tamoxifen-inducible Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) retinas. RPE, RPE and photoreceptors, outer segment of photoreceptors, and cell body of photoreceptors are shown, from left to right, respectively. White arrows, mitochondria of RPE; arrowheads, outer segment; red arrows, inner segment; yellow arrow, disc structure of photoreceptors; blue arrow, mitochondria of photoreceptors. Scale bar, 1 μm. i Thickness of the RPE layer (n = 5 from h). j Number of mitochondria in RPE (n = 5 from h). k Thickness of inner and outer photoreceptor segments (n = 5 from h). l Number of abnormal mitochondria in photoreceptors (n = 4 from h). m TEM images of RPE cells demonstrating phagocytosis of photoreceptors; arrowheads, microvilli. Scale bar, 1 μm. n Quantitative values of the number (n = 5 from m) and length ratio (n = 10 from m) of microvilli. Data are presented as mean ± SEM and representative of at least two independent experiments. P-values were derived from one-way ANOVA with Tukey’s multiple comparison test (b) and two-tailed unpaired-t test (d, f, g, i–l, and n). Source data are provided as a Source Data file.

Fig. 2. Visual function and histological features of the retina in ocular-selective Pnpla6-deficient mice.

a Protocol for ocular-specific Pnpla6 deletion. b Immunostaining for PNPLA6 in retinas from tamoxifen-treated Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice. Red, PNPLA6; blue, DAPI; arrow, RPE. Scale bar, 50 μm. c qPCR of Pnpla6 in isolated RPE or brain from Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice, with Gapdh or Hprt1 as a control (n = 3 independent animals). d HE staining of Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) retinas; brackets, ONL. Scale bar, 50 μm. e OCT analysis of Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) retinas. Brackets, ONL. f Quantified values in OCT analysis of Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) retinas (n = 4 from e). g TEM images of Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) retinas. Arrowheads, disc structure of photoreceptors. Scale bar, 1 μm. h ERG analysis of Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice under a dark condition (150.3 cd.s/m²). i Amplitudes of a-wave and b-wave (n = 4 from h). j Amplitudes of a-wave and b-wave of the ERG under dark conditions (n = 7 independent animals). k qPCR of Pnpla6 in isolated RPE or brain from 3-month-old Pnpla6^fl/fl and Best1-CrePnpla6^fl/fl mice (n = 3 independent animals). l HE staining of Pnpla6^fl/fl and Best1-CrePnpla6^fl/fl retinas. Scale bar, 50 μm. m TEM images of Pnpla6^fl/fl and Best1-CrePnpla6^fl/fl retinas. Arrows, microvilli of RPE; arrowheads, disc structure of photoreceptors. Scale bar, 1 μm. n OCT analysis of ONL thickness in Pnpla6^fl/fl and Best1-CrePnpla6^fl/fl mice (n = 3 independent animals). o TUNEL staining of Pnpla6^fl/fl and Best1-CrePnpla6^fl/fl retinas. Blue, DAPI; Red arrow, TUNEL-positive cells. Scale bar, 50 μm. p Amplitudes of a-wave and b-wave of the ERG under a dark condition (150.3 cd.s/m²) (n = 4 independent animals). Data are presented as mean ± SEM and representative of at least two independent experiments. P-values were derived from two-way ANOVA with Tukey’s multiple comparison test (f, j, and n) and two-tailed unpaired-t test (c, i, k, and p). *P < 0.05, **P < 0.01, and ***P < 0.001, versus control mice (f, j, and n). Source data are provided as a Source Data file.

Fig. 3. Changes in phospholipid metabolism in PNPLA6-knockdown ARPE-19 cells and in primary RPE cells isolated from tamoxifen-inducible, ocular-selective Pnpla6-deficient mice.

a qPCR of PNPLA6 in PNPLA6-knockdown (si PNPLA6) and control (scramble) ARPE-19 cells, with GAPDH as an internal control (n = 6 biological replicates). b PC and LPC levels in PNPLA6-knockdown ARPE-19 cells relative to control cells, with the levels in control cells as 1 (n = 3 biological replicates). c Lipidomics of PC and LPC molecular species in PNPLA6-knockdown ARPE-19 cells and control cells (n = 4 biological replicates). d GPC and choline levels in PNPLA6-knockdown ARPE-19 cells relative to control cells, with the levels in control cells as 1 (n = 4 biological replicates). e Lipidomics of PE and LPE molecular species in PNPLA6-knockdown ARPE-19 cells and control cells (n = 4 biological replicates). f Heatmap representation of the ratio of individual phospholipids and lysophospholipids in PNPLA6-knockdown ARPE-19 cells relative to control cells (n = 4 biological replicates). g PC and LPC levels in RPE cells isolated from tamoxifen-inducible, ocular-selective Pnpla6-deficient mice relative to those from control mice, with the levels in control mice as 1 (n = 3 biological replicates). h Lipidomics of PC and LPC molecular species in Pnpla6-deficient RPE cells and control cells (n = 3 biological replicates). i Choline levels in Pnpla6-deficient RPE cells relative to control cells, with the levels in control cells as 1 (n = 4 biological replicates). j Heatmap representation of the ratio of individual phospholipids and lysophospholipids in Pnpla6-deficient RPE cells relative to control cells (n = 4 biological replicates). Data are presented as mean ± SEM and representative of at least two independent experiments. P-values were derived from two-way ANOVA with Tukey’s multiple comparison test (c–e, h and i) and two-tailed unpaired-t test (a, b, and g). Source data are provided as a Source Data file.

Fig. 4. PNPLA6 knockdown alters cellular function and metabolism in ARPE-19 cells.

a SEM images of PNPLA6-knockdown (si PNPLA6) and control (scramble) ARPE-19 cells. Scale bar, 1 μm. b Cell sizes of PNPLA6-knockdown and control cells, with control cells as 1 (n = 7 from a, biological replicates). c Microarray of PNPLA6-knockdown cells relative to control cells. GO analysis of genes altered in knockdown relative to control cells (left) and a heatmap of proliferation-related genes that were decreased in knockdown relative to control cells (right). d Proliferation in normal medium (containing 64.3 μM choline) or in medium containing 10-fold excess choline (643 μM), with the normal medium as 1 (n = 3 biological replicates). e Cellular choline levels (n = 3 biological replicates). f Cell cycle analysis of control cells (scramble), PNPLA6-knockdown cells (si PNPLA6), and PNPLA6-knockdown cells with excess choline (si PNPLA6 + choline) (n = 3 biological replicates). g TEER analysis (n = 8 biological replicates). h Immunostaining of ZO-1. Green, ZO-1; blue, DAPI. Scale bar, 10 μm. Arrows indicate areas where ZO-1 was disturbed in the cell periphery. i Phagocytosis of FITC-labeled (red) porcine POS. Scale bar, 10 μm. j Quantification of phagocytosis, with control cells as 1 (n = 3 biological replicates). k JC-1 staining. Scale bar, 10 μm. Bottom right, magnified image. l The red/green ratio of JC-1 staining (n = 5 biological replicates). m Oxygen consumption rate determined by an extracellular flux analyzer (n = 4 biological replicates). n Cellular ATP levels (n = 4 biological replicates). o Immunoblotting of mitochondrial proteins (n = 3 biological replicates). p Densitometry of mitochondrial proteins relative to β-actin, with control cells as 1 (n = 3 from o, biological replicates). q Heatmap of water-soluble metabolites (n = 4 biological replicates). Data are presented as mean ± SEM and compiled from or representative of at least two independent experiments. P-values were derived from two-way ANOVA (m) and one-way ANOVA with Tukey’s multiple comparison test (d–g and n) and two-tailed unpaired-t test (b, j, l, and p). *P < 0.05 versus control cells (n). Source data are provided as a Source Data file.

Fig. 5. PNPLA6 knockdown in primary human RPE cells disturbs choline turnover.

a qPCR analysis of PNPLA6 in PNPLA6-knockdown (si PNPLA6) and control (scramble) primary human RPE cells, with GAPDH as an internal control (n = 3 biological replicates). b, c Lipidomics of PC (n = 4 biological replicates) (b) and LPC (n = 3 biological replicates) (c) in PNPLA6-knockdown and control cells. d Changes in choline level in PNPLA6-knockdown cells relative to control cells, with the levels in control cells as 1 (n = 5 biological replicates). e Proliferation of PNPLA6-knockdown cells relative to control cells, with the levels in control cells as 1 (n = 3 biological replicates). f TERR analysis of PNPLA6-knockdown and control cells (n = 3 biological replicates). g Phagocytosis of FITC-labeled (red) porcine POS in PNPLA6-knockdown and control cells. Scale bar, 10 μm. h Quantitative values of phagocytosis, with control cells set as 1 (n = 4 biological replicates). i JC-1 staining of PNPLA6-knockdown and control cells. Scale bar, 10 μm. j Red/green ratio of JC-1 staining (n = 4 biological replicates). k, l TEM images of PNPLA6-knockdown and control cells. Representative images (left) with quantification of mitochondria number (n = 5 biological replicates) (right) (k) and microvilli (n = 5 biological replicates) (l) are shown. Arrows, mitochondria. Scale bar, 1 μm. (m) Effects of choline and CHKα inhibitor (MN58b) on the proliferation of PNPLA6-knockdown cells (n = 8 biological replicates). Data are presented as mean ± SEM and representative of at least two independent experiments. P-values were derived from one-way ANOVA with Tukey’s multiple comparison test (m) and two-tailed unpaired-t test (a–f, h, j, k, and l). Source data are provided as a Source Data file.

Fig. 6. PNPLA6-driven choline is secreted from RPE cells and supports photoreceptor cell homeostasis.

a, b Proliferation (a; n = 6 biological replicates) and intracellular choline levels (b; n = 3 biological replicates) in 661 W cells cultured in choline-sufficient (+) or choline-deficient (−) medium. c TUNEL staining of 661 W cells. d Quantification of dead cells per field of view (n = 4 biological replicates). Scale bar, 50 μm. e, f Microarray analysis of 661 W cells. Scattered plot (e) and heatmap of apoptosis-related genes (f) in choline-deficient relative to choline-sufficient cells. g TEM images of 661 W cells. Arrowheads, mitochondria. Scale bar, 1 μm. h The number of mitochondria in choline-deficient (n = 6 biological replicates) and -sufficient (n = 4 biological replicates) cells per image in (g). i Choline concentrations in medium of PNPLA6-knockdown (si PNPLA6; n = 3 biological replicates) and control (scramble; n = 4 biological replicates) ARPE-19 cells cultured in choline-deficient medium. j Photos of 661 W cells cultured in a conditioned medium from PNPLA6-knockdown or control ARPE-19 cells. Scale bar, 50 μm. k Proliferation of 661 W cells (n = 3 biological replicates). l Proliferation of 661 W cells cultured in a conditioned medium from PNPLA6-knockdown or control ARPE-19 cells with or without 10-fold excess choline (n = 3 biological replicates). m qPCR of Slc44a1 in Slc44a1-knockdown or control 661 W cells (n = 4 biological replicates). n Proliferation of Slc44a1-knockdown or control 661 W cells, with control cells as 1 (n = 5 biological replicates). o Proliferation of Slc44a1-knockdown or control 661 W cells cultured with conditioned medium from PNPLA6-knockdown or control ARPE-19 cells (n = 3 biological replicates). p Proliferation of 661 W cells cultured in conditioned medium from PNPLA6-knockdown or control ARPE-19 cells with or without 1 μM MN58b or 10-fold excess choline (n = 3 biological replicates). Data are presented as mean ± SEM and representative of at least two independent experiments. P-values were derived from one-way ANOVA with Tukey’s multiple comparison test (l, o, and p) and two-tailed unpaired-t test (a, b, d, h, i, k, m, and n). Source data are provided as a Source Data file.

Fig. 7. Topical application of choline prevents retinal degeneration in Pnpla6-deficient mice.

a Protocol for topical choline application to the eyes of ocular-selective Pnpla6-deficient mice. b HE staining of retinas from tamoxifen-treated Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice with or without choline eyedrops. Scale bar, 50 µm. c OCT analysis of ONL thickness of retinas in Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice with or without choline eyedrops (n = 6 independent animals). d TEM analysis of retinas in Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice. Arrowheads, disc structure of photoreceptors. Scale bar, 1 µm. e ERG analysis of Pnpla6^fl/flCreER(−) and Pnpla6^fl/flCreER(+) mice with or without choline eyedrops. Amplitudes of a-wave (left) and b-wave (right) under scotopic conditions are shown (n = 3 independent animals). f Schematic presentation of the function of PNPLA6 in the retina. In RPE cells, mobilization of choline from PC by PNPLA6 is crucial for RPE homeostasis. PNPLA6-produced choline in RPE cells is transported into photoreceptor cells to maintain photoreceptor homeostasis. PNPLA6 deficiency in RPE cells hampers both RPE and photoreceptor functions, leading to retinal degeneration. Data are presented as mean ± SEM and representative of at least two independent experiments. P-values were derived from two-way ANOVA with Tukey’s multiple comparison test (c and e). *P < 0.05, **P < 0.01, and **** P < 0.0001 versus control mice (c, e). Source data are provided as a Source Data file.

Fig. 8. Mutations in PNPLA6-related disorders impairs enzymatic function of PNPLA6.

a PNPLA6 point mutations reported in human retinal degeneration. We generated three-point mutants, G726R (found in Laurence-Moon syndrome), T1058I (found in Boucher-Neuhäuser syndrome), and S1014A (the catalytic center), of human PNPLA6. b–d qPCR of PNPLA6 relative to GAPDH (b), intracellular choline levels (c), and cell proliferation (with the value of mock cells as 1) (d) in ARPE-19 cells transfected with WT or mutant PNPLA6 (n = 3 biological replicates). e 3D structural model of PlpD of Pseudomonas aeruginosa used a template. Amino acid sequences are shown in ribbons, and serine (S) and aspartate (D) in the active site are indicated by space-filling models (red, oxygen; yellow, carbon). f 3D structural model of residues 977−1260 including the patatin domain of PNPLA6, constructed by homology remodeling using SWISS-MODEL. Green, lid structure; light blue, α-helix; blue, β-strand. g The surface of the patatin domain of PNPLA6. Left, lid is pictured as ribbon; Right, all components are pictured as superficial structure. Pink, S1014; yellow, hydrophobic residues around the channel; green, lid. h Structure of the vicinity of the active center in the patatin domain of WT PNPLA6. Left, full image; Right, magnified image of the lid. i Structure of the vicinity of the active center in the patatin domain of T1058I mutant. Left, full image; Right, magnified image of the lid; dotted lines, hydrophobic interaction. Data are presented as mean ± SEM and representative of at least two independent experiments. P-values were derived from one-way ANOVA with Tukey multiple comparison test (b, c, and d). Source data are provided as a Source Data file.