Malin restoration as proof of concept for gene therapy for Lafora disease

Abstract

Lafora disease is a fatal neurodegenerative childhood dementia caused by loss-of-function mutations in either the laforin or malin gene. The hallmark of the disease is the accumulation of abnormal glycogen aggregates known as Lafora bodies (LBs) in the brain and other tissues. These aggregates are responsible for the pathological features of the disease. As a monogenic disorder, Lafora disease is a good candidate for gene therapy-based approaches. However, most patients are diagnosed after the appearance of the first symptoms and thus when LBs are already present in the brain. In this context, it was not clear whether the restoration of a normal copy of the defective gene (either laforin or malin) would prove effective. Here we evaluated the effect of restoring malin in a malin-deficient mouse model of Lafora disease as a proof of concept for gene replacement therapy. To this end, we generated a malin-deficient mouse in which malin expression can be induced at a certain time. Our results reveal that malin restoration at an advanced stage of the disease arrests the accumulation of LBs in brain and muscle, induces the degradation of laforin and glycogen synthase bound to the aggregates, and ameliorates neuroinflammation. These results identify malin restoration as the first therapeutic strategy to show effectiveness when applied at advanced stages of Lafora disease.

Keywords: Lafora disease; gene therapy; glycogen; neurodegeneration; neuroinflammation.

Graphical Abstract

Figure 1

Malin restoration in young malin^KO mice. PAS staining in brain. Representative images from control, 4-, 5-, and 7-month-old malin^KO and malin^KO+OE mice after 1 or 3 months of malin restoration. The upper panels show a cortical region and the lower panels the CA2/CA3 hippocampal region. Scale bar: 100 µm. n = 4–5 mice.

Figure 2

LB component analysis of young malin^KO mice after malin restoration. (A) Representative immunofluorescence images from the cortical region obtained with anti-MGS (green), anti-laforin (green), or anti-p62 (red) stainings in combination with DAPI (blue). Control and 4-month-old malin^KO (malin^KO 4 m), malin^KO+OE[4+1], and malin^KO+OE[4+3] mice are shown. Scale bar: 100 µm. (B) Quantification of number of aggregates per area (mm²) detected by MGS, laforin, or p62 staining. ANOVA with Tukey’s multiple comparison test and unpaired t-test. MGS: control versus malin^KO 4 months: P = 0.01, malin^KO 4 months versus malin^KO+OE[4+1]: P = 0.48, malin^KO 4 months versus malin^KO+OE[4+3]: P = 0.0092, malin^KO+OE[4+1] versus malin^KO+OE[4+3]: P = 0.21. Laforin: control versus malin^KO 4 months: P = 0.0001, control versus malin^KO+OE[4+1]: P = 0.021, malin^KO 4 months versus malin^KO+OE[4+1]: P = 0.022, malin^KO 4 months versus malin^KO+OE[4+3]: P = 0.009. p62: control versus malin^KO 4 months: P = 0.01, malin^KO 4 months versus malin^KO+OE[4+1]: P = 0.23, malin^KO 4 months versus malin^KO+OE[4+3]: P = 0.69, malin^KO+OE[4+1] versus malin^KO+OE[4+3]: P = 0.51. P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***), unpaired t-test. (C) Magnification of representative LBs found in 4-month-old malin^KO mice (left) compared with malin^KO+OE[4+3] mice (right). MGS and laforin (green) were co-stained with p62 (red), the overlay image for each one is shown (merge). Fluorescent profile along a line across the LB through its geometrical centre is shown in each case. The size of the fluorescent particle is indicated in each graphic. (D) Fluorescent intensity of MGS and laforin in p62-positive LBs from malin^KO, malin^KO+OE[4+1], and malin^KO+OE[4+3] mice. ANOVA with Tukey’s multiple comparison test and unpaired t-test. MGS, malin^KO 4 months versus malin^KO+OE[4+1]: P ≤ 0.0001, malin^KO 4 months versus malin^KO+OE[4+3]: P ≤ 0.0001, malin^KO+OE[4+1] versus malin^KO+OE[4+3]: P = 0.9. Laforin: malin^KO 4 months versus malin^KO+OE[4+1]: P = 0.0006, malin^KO 4 months versus malin^KO+OE[4+3]: P = 0.013, malin^KO+OE[4+1] versus malin^KO+OE[4+3]: P = 0.77. p62: malin^KO 4 months versus malin^KO+OE[4+1]: P = 0.065, malin^KO 4 months versus malin^KO+OE[4+3]: P = 0.075, malin^KO+OE[4+1] versus malin^KO+OE[4+3]: P = 0.86. P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). Data are shown as mean ± SD. Each dot represents one mouse (n = 4–8 as indicated).

Figure 3

Malin restoration at an advanced stage of Lafora disease in malin^KO mice. (A) PAS staining in brain. Representative images of cortical and CA2/CA3 hippocampal region from control, 11-month-old malin^KO, 15-month-old malin^KO, and malin^KO+OE[11+4] mice are shown. ANOVA with Tukey’s multiple comparison test and unpaired t-test. Control versus malin^KO 11 months: P = 0.00003, control versus malin^KO 15 months: P = 0.0001, control versus malin^KO+OE[11+4]: P = 0.0094, malin^KO 11 months versus malin^KO 15 months: P = 0.0032, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.96, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0094. Scale bar: 100 µm. (B) Total brain glycogen was determined from the same groups (µg/mg of tissue). Each dot represents one mouse (n = 3–6).

Figure 4

Soluble and insoluble brain fractions analysis. (A) Representative western blots of total brain homogenate, soluble, and insoluble fractions are shown for MGS, laforin, and p62. The images shown are cropped to show the band of interest, for full images see Supplementary material. Three samples from the same group were run for these images from a total of 3–6 mice per group. Loading control: LICOR—revert staining. (B) Quantifications of protein detection by western blot. Relative optical density units were related to loading control and normalized respect to control mice. Data are shown as mean ± SD. ANOVA with Tukey’s multiple comparison test and unpaired t-test. MGS total homogenate: control versus malin^KO 11 months: P = 0.025, control versus malin^KO 15 months: P = 0.02, malin^KO 15 months versus malin^KO[11+4]: P = 0.017, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.033, control versus malin^KO+OE[11+4]: P = 0.028. MGS insoluble fraction: control versus malin^KO 11 months: P = 0.01, control versus malin^KO 15 months: P = 0.0001, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.015, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.026, control versus malin^KO+OE[11+4]: P = 0.029. Laforin total homogenate: control versus malin^KO 11 months: P = 0.0008, control versus malin^KO 15 months: P = 0.02, malin^KO 15months versus malin^KO+OE[11+4]: P = 0.0076, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.08, control versus malin^KO+OE[11+4]: P = 0.021. Laforin insoluble fraction: control versus malin^KO 11 months: P = 0.04, control versus malin^KO 15 months: P = 0.0094, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0054, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.085, control versus malin^KO+OE[11+4]: P = 0.0062. p62 total homogenate fraction: control versus malin^KO 11 months: P = 0.105, control versus malin^KO 15 months: P = 0.032, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.54, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0511. p62 insoluble fraction: control versus malin^KO 11 months: P = 0.04, control versus malin^KO 15 months: P = 0.010, malin^KO 11 months versus malin^KO 15 months: P = 0.24, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.67, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.14. P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). Each dot represents one mouse (n = 3–6).

Figure 5

MGS, laforin, and p62 analysis in the hippocampus of malin^KO mice after malin restoration at an advanced stage of Lafora disease. (A) Immunofluorescence images from the hippocampal region using anti-MGS (green), anti-laforin (red), or anti-p62 (red) antibodies combined with DAPI staining in each group. (B) Quantification of number of particles per area (mm²) in CA2/CA3 region stained with anti-MGS antibody. ANOVA with Tukey’s multiple comparison test. MGS total aggregates: control versus malin^KO 11 months: P = 0.00038, control versus malin^KO 15 months: P = 0.0008, control versus malin^KO+OE[11+4]: P = 0.064, malin^KO 11 months versus malin^KO 15 months: P = 0.0697, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.0287, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0006. nLBs: control versus malin^KO 11 months: P = 0.0006, control versus malin^KO 15 months: P = 0.0078, control versus malin^KO+OE[11+4]: P = 0.1, malin^KO 11 months versus malin^KO 15 months: P = 0.0009, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.06, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0175. CAL, control versus malin^KO 11 months: P = 0.0023, control versus malin^KO 15 months: P = 0.0008, control versus malin^KO+OE[11+4]: P = 0.081, malin^KO 11 months versus malin^KO 15 months: P = 0.12, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.02, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0011. (C) Quantification of nLBs and CALs from the hippocampus in each group. Quantification of number of particles per area (mm²) in CA2/CA3 hippocampal region stained with laforin (D) or p62 (E). ANOVA with Tukey’s multiple comparison test. Scale bar: 100 µm. In all graphics, data are shown as mean ± SD. P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). Each dot represents one mouse (n = 3–6 as indicated in the graphic).

Figure 6

Effect of late malin restoration on LB composition. (A) Representative immunofluorescent images from 15-month-old malin^KO mice and malin^KO+OE[11+4] mice. Co-stainings: MGS (green)/p62 (red) and laforin (green)/p62 (red), both combined with DAPI staining (blue), each channel is shown separately and as merge. Scale bar: 100 µm. (B) Two representative individual LBs and their immunofluorescent profile along a line crossing the geometrical centre are shown for each group. The profiles correspond to MGS (green)/p62 (red) or laforin (green)/p62 (red) combinations. Each channel and merge images are shown. The size of the p62-positive particle is indicated in each graphic. (C) Quantification of the percentage of double-positive MGS/p62 and laforin/p62 aggregates in each group. ANOVA with Tukey's multiple comparison test and unpaired t-test. MGS + p62: malin^KO 11 months versus malin^KO 15 months: P = 0.68, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.037, malin^KO 15 months versus malin^KO+OE[11+4]: 0.0146. Laforin + p62: malin^KO 11 months versus malin^KO 15 months: P = 0.002, malin^KO 11 months versus malin^KO+OE[11+4]: P ≤ 0.0001, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.02. (D) Fluorescent intensity of MGS and laforin in p62-positive aggregates. ANOVA with Tukey’s multiple comparison test and unpaired t-test. MGS mean I_max: malin^KO 11 months versus malin^KO 15 months: P = 0.76, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.0135, malin^KO 15 months versus malin^KO+OE[11+4]: 0.0273. Laforin mean I_max: malin^KO 11 months versus malin^KO 15 months: P = 0.1178, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.0004, malin^KO 15 months versus malin^KO+OE[11+4]: 0.0003. p62 mean I_max: malin^KO 11 months versus malin^KO 15 months: P = 0.37, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.311, malin^KO 15 months versus malin^KO+OE[11+4]: 0.67. P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). In all graphics, the data are shown as mean ± SD. Each dot represents one mouse.

Figure 7

Inflammatory response after malin restoration at advanced stage of Lafora disease. (A) Astrogliosis and microgliosis were determined by immunostaining using anti-GFAP (red, combined with DAPI, blue) or anti-CD11b antibodies, respectively. Representative images from the hippocampus are shown. Scale bar: 500 µm. (B) Quantification of GFAP and CD11b stainings represented as percentage of positive pixel in the hippocampal region. N = 3 sections per mice, 3–6 mice. (C) RNA from total brain homogenates was analysed for quantification of the expression of pro-inflammatory genes by qPCR: IL-6, CXCL10. Graphics show the relative expression levels (2^−ΔΔCt). Unpaired t-test. P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). CD11b: control versus malin^KO[11+4] 11 months: P = 0.028, control versus malin^KO 15 months: P = 0.0049, control versus malin^KO+OE[11+4]: P = 0.0073, malin^KO 11 months versus malin^KO 15 months: P = 0.0037, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.18, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.064. GFAP: control versus malin^KO 11 months: P = 0.002, control versus malin^KO 15 months: P = 0.005, control versus malin^KO+OE[11+4]: P = 0.035, malin^KO 11 months versus malin^KO 15 months: P = 0.75, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.82, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.86. IL6: control versus malin^KO 11 months: P = 0.03, control versus malin^KO 15 months: P = 0.004, control versus malin^KO+OE[11+4]: P = 0.14, malin^KO 11months versus malin^KO 15 months: P = 0.018, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.8, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.018. CXCL10: control versus malin^KO 11 months: P = 0.001, control versus malin^KO 15 months: P = 0.0025, control versus malin^KO+OE[11+4]: P = 0.1156, malin^KO 11months versus malin^KO 15 months: P = 0.0794, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.8, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.047. Each dot represents one mouse (n = 3–6 mice). Data are shown as mean ± SD.

Figure 8

Effect of late malin expression on skeletal muscle. (A) PAS staining and immunofluorescent images from quadriceps muscles are shown for each group. Co-stainings of MGS/p62 were performed. For simplicity, MGS and p62 are shown independently. MGS (magenta, middle panels) and p62 (magenta, lower panels) were combined with agglutinin (WGA) to visualize the fibres (green). Scale bar: 100 µm. (B) The amount of MGS and p62 in the tissue was quantified as percentage of positive pixel in each group and represented as mean ± SD. Unpaired t-test, MGS % positive pixel: malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0022; p62% positive pixel: malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.88. P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). Unpaired t-test, P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). (C) Quantification of the percentage of double-positive MGS/p62 individual aggregates. ANOVA with Tukey’s multiple comparison test and unpaired t-test; malin^KO 11 months versus malin^KO 15 months: P = 0.41, malin^KO 11months versus malin^KO+OE[11+4]: P = 0.0045, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.0277. (D) Representation of the MGS fluorescence intensity in p62-positive individual aggregates in each group. ANOVA with Tukey’s multiple comparison test and unpaired t-test. MGS: malin^KO 11 months versus malin^KO 15 months: P = 0.83, malin^KO 11months versus malin^KO+OE[11+4]: P = 0.0045, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.03. p62: malin^KO 11 months versus malin^KO 15 months: P = 0.048, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.0041, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.019. (E) Total amount of glycogen in muscle was determined (µg/g of tissue) in each group. Data are shown as mean ± SD, each dot represents a mouse. ANOVA with Tukey's multiple comparison test and unpaired t-test: control versus malin^KO 11 months: P = 0.017, control versus malin^KO 15 months: P = 0.0002, control versus malin^KO+OE[11+4]: P = 0.0012, malin^KO 11months versus malin^KO 15 months: P = 0.011, malin^KO 11 months versus malin^KO+OE[11+4]: P = 0.3, malin^KO 15 months versus malin^KO+OE[11+4]: P = 0.046. P-value < 0.01 (**), and P-value < 0.001 (***). Unpaired t-test, P-value ≤ 0.05(*), P-value < 0.01 (**), and P-value < 0.001 (***). n = 3–6 mice.