Spatial characterization of RPE structure and lipids in the PEX1-p.Gly844Asp mouse model for Zellweger spectrum disorder

Abstract

Zellweger Spectrum Disorder (ZSD) is caused by defects in PEX genes, whose proteins are required for peroxisome assembly and function. Peroxisome dysfunction in ZSD causes multisystem effects, with progressive retinal degeneration (RD) among the most frequent clinical findings. However, much remains unknown about how peroxisome deficiency causes RD. To study RD pathophysiology in ZSD, we used the PEX1-p.Gly844Asp (G844D) mouse model, which represents the common human PEX1-p.Gly843Asp (G843D) variant. We previously reported diminished retinal function, diminished functional vision, and neural retina structural defects in this model. Here, we investigate the retinal pigment epithelium (RPE) phenotype, examining morphological, inflammatory, and lipid changes at 1, 3, and 6 months of age. We report that RPE cells exhibit evident degeneration by 3 months that worsens with time, starts in the dorsal pole, and is accompanied by subretinal inflammatory cell infiltration. We match these events with imaging mass spectrometry for regional analysis of lipids in the RPE. We identified 47 lipid alterations preceding structural changes, 9 of which localize to the dorsal pole. 29 of these persist to 3 months, with remodeling of the dorsal pole lipid signature. 13 new alterations occur concurrent with histological changes. Abnormalities in peroxisome-dependent lipids detected by LC/MS/MS are exacerbated over time. This study represents the first characterization of RPE in a ZSD model, and the first in situ lipid analysis in peroxisome-deficient tissue. Our findings uncover potential lipid drivers of RD progression in ZSD, and identify candidate biomarkers for retinopathy progression and response to therapy.

Keywords: PEX1; dyslipidemia; eye/retina; inflammation; lipidomics; lipids; mass spectrometry imaging; peroxisome disease; retinal degeneration; retinal pigment epithelium.

Fig. 1

Structure and MALDI MSI lipid analyses on whole RPE flatmounts from 1-month-old PEX1-G844D mice. A: Confocal imaging of RPE flatmounts from 1-month-old WT and PEX1-G844D mice stained with the F-actin marker TRITC-phalloidin. The shaded grey area represents whole RPE tissue prepared for flat mounting. Nine confocal microscope images arranged in dorso-ventral orientation were taken at 20x magnification to visualize tissue integrity. B: Magnified representative RPE images are shown to facilitate observation of cell morphology; (scale bar: 100 μm). Histogram shows the quantification of RPE cell numbers from WT and PEX1-G844D RPE flatmounts (N = 3 per genotype). C: Image of the coordinate acquisition grid (red dots) for RPE sampling. D: Examples of MSI localization of lipids with heterogeneous spatial distribution (signal intensity). White arrows indicate the pole with the greatest abundance. E: Quantification of lipid abundance (signal intensity) in the dorsal and ventral poles. All lipids with different abundance (heterogeneous distribution) between the poles are shown. (N = 3 per genotype). ∗P < 0.05, ∗∗P < 0.01. Color scale indicates the lowest abundance in dark blue and highest in red/white.

Fig. 2

Early lipid changes in the PEX1-G844D RPE at 1 month of age. A: The heatmap shows the log₂ fold change of lipid abundance in mutant samples compared to the WT average, with red indicating increased and blue decreased signal intensity. B: Localization by MALDI MSI of lipids in whole RPE flatmounts from WT and PEX1-G844D mice (6 RPE flatmounts per slide: 3 WT in V position and 3 PEX1-G844D in inverted V position). Examples MSI analyses show lipids with increased or reduced abundance in PEX1-G844D compared to controls, and homogeneous or heterogeneous distribution between the dorsal and ventral pole. The IMS scale indicates the highest lipid density in white/red and the lowest lipid density in dark blue. C: The heat map shows lipids dysregulated in PEX1-G844D with heterogeneous distribution between the dorsal compared to ventral pole (log₂ fold change of lipid abundance). Only lipid species with statistically significant changes between mutant and WT RPE (Supplemental Table S1) are included in the heatmaps (A and C). Each column (1, 2, and 3) represents the ratio of signal intensity measured by MSI (A) in whole PEX1-G844D RPE compared to the average intensity in WT, or (C) in the PEX1-G844D dorsal pole compared to the PEX1-G844D ventral pole. Each row corresponds to one lipid species. (N = 3 per genotype).

Fig. 3

PEX1-G844D retinal structure at 3 months of age. A and B: Confocal imaging of RPE flatmounts from 3-month-old (A) WT and (B) PEX1-G844D mice immunostained with anti-IBA1 antibody, a microglia/macrophage marker (green) and counterstained with TRITC-phalloidin (red). Top panels: The shaded grey area represents whole RPE tissue prepared for flat mounting. Nine confocal microscope images arranged in dorso-ventral orientation were taken at 20x magnification to visualize tissue integrity. Bottom panels: Magnified representative images are shown to facilitate observation of cell morphology at dorsal and ventral poles. White arrows indicate blebs within RPE cells. C: Quantification of RPE cell number per mm² in whole RPE tissue (left graph) and in dorso-ventral areas (right graph); N = 3 per genotype, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. D: Quantification of subretinal IBA1 positive cells per mm² on the whole RPE (left graph) and in dorso-ventral areas (right graph); N = 3 per genotype, ∗∗P < 0.01. E: Neuroretina flatmounts oriented with photoreceptors face up counterstained with the cone marker peanut agglutinin (PNA, green). F: Confocal imaging of dorsal and ventral poles of retinal cryosections, immunostained with anti-GFAP antibody (red) counterstained with DAPI (blue). GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer. Scale bar: 100 μm.

Fig. 4

Progression of the spatial lipid changes in the PEX1-G844D RPE at 3 months. A: The heatmap shows the log₂ fold change of signal intensity in mutant samples compared to the average intensity in WT samples, with red indicating increased intensity and blue indicating decreased intensity. B: Localization by MALDI MSI of lipids in whole RPE flatmounts from WT and PEX1-G844D mice (6 RPE flatmounts from 3-month-old mice: 3 WT in V position and 3 PEX1-G844D in inverted V position). Examples of MSI analyses showing 10 lipids identified with different abundance (homogeneous or heterogeneous) in PEX1-G844D RPE compared to WT control. C: The heat map shows the log₂ fold change of lipid intensity in the PEX1-G844D dorsal compared to ventral pole. Only lipid species with significant difference between mutant and WT RPE (from Supplemental Table S4) are included in the heatmaps (A and C). Each column (1, 2, and 3) represents the ratio of signal intensity measured by IMS in (A) PEX1-G844D RPE samples compared to WT, or (C) PEX1-G844D dorsal compared to ventral pole. Each row corresponds to a lipid species. (N = 3 per genotype).The heat map color scales indicate highest lipid density in dark red and lowest lipid density in dark blue. D: The Venn diagram shows single lipid species significantly affected in PEX1-G844D mouse RPE at 1 and 3 months of age.

Fig. 5

LC/MS/MS analyses of peroxisome metabolites at 3 months. Histograms present the levels of (A) very long chain-lysophosphatidylcholines (VLC-LPCs), and (B) phosphatidylethanolamine (PE), and phosphatidylcholine (PC) plasmalogens (PL) measured by LC/MS/MS in RPE tissue from WT and PEX1-G844D mice. (Each point represents 1 RPE; N = 8 per genotype, ∗∗P < 0.01).

Fig. 6

Later stage PEX1-G844D retinopathy phenotype at 6 months. A: Confocal imaging of dorsal and ventral areas of 6 month-old WT and PEX1-G844D RPE flatmounts immunostained with anti-IBA1 antibody, a microglia/macrophage marker (green) and counterstained with TRITC-phalloidin (red). B: Quantification of RPE cell number per mm² in whole RPE tissue (top graph) and quantification of subretinal IBA1 positive cells per mm² on the whole RPE (bottom graph, each point represents 1 RPE) (N = 3 per genotype, ∗P < 0.05, ∗∗∗∗P < 0.0001). C: Neuroretina flatmounts oriented with photoreceptors face up counterstained with peanut agglutinin (PNA), a cone marker (green). D: Confocal imaging of retinal cryosections from dorsal and ventral poles, immunostained with anti-GFAP antibody (red) counterstained with DAPI (blue). GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer. Scale bar: 100 μm. E, F: Histograms presenting the levels of (E) VLC-LPCs and (F) PE, PC plasmalogens (PL) measured by LC/MS/MS in RPE. (N = 4 per genotype, each point represents 1 RPE ∗∗P < 0.01, ∗∗∗P < 0.001).

Fig. 7

Comparison of common dysregulated lipids in PEX1-G844D RPE and other RD models. Venn diagrams comparing the overlap of lipid species affected in the PEX1-G844D RPE and in retinal tissue from other genetic RD mouse models with lipid dysregulation (Ascl6^−/−, Lpaat3^−/−, Mfsd2a^−/−and Mfp2^−/−). The diagram includes common lipids increased (left), and reduced (right) per model.