Retinal Phenotyping of a Murine Model of Lafora Disease

Abstract

Lafora disease (LD) is a progressive neurologic disorder caused by biallelic pathogenic variants in EPM2A or EPM2B, leading to tissue accumulation of polyglucosan aggregates termed Lafora bodies (LBs). This study aimed to characterize the retinal phenotype in Epm2a^-/- mice by examining knockout (KO; Epm2a^-/-) and control (WT) littermates at two time points (10 and 14 months, respectively). In vivo exams included electroretinogram (ERG) testing, optical coherence tomography (OCT) and retinal photography. Ex vivo retinal testing included Periodic acid Schiff Diastase (PASD) staining, followed by imaging to assess and quantify LB deposition. There was no significant difference in any dark-adapted or light-adapted ERG parameters between KO and WT mice. The total retinal thickness was comparable between the groups and the retinal appearance was normal in both groups. On PASD staining, LBs were observed in KO mice within the inner and outer plexiform layers and in the inner nuclear layer. The average number of LBs within the inner plexiform layer in KO mice were 1743 ± 533 and 2615 ± 915 per mm², at 10 and 14 months, respectively. This is the first study to characterize the retinal phenotype in an Epm2a^-/- mouse model, demonstrating significant LB deposition in the bipolar cell nuclear layer and its synapses. This finding may be used to monitor the efficacy of experimental treatments in mouse models.

Keywords: EPM2A; Lafora; Lafora body disorder; disease; electroretinography; gene; knock outs; optical coherence tomography; periodic acid Schiff reaction; phenotype.

Figure 1

Full-field electroretinogram (ERG) and optical coherence tomography (OCT) findings in Epm2a^−/− knock out (KO) mice and control (WT) littermates. (A) Dark-adapted (DA) ERG a-wave amplitude measures to different intensities of flash (ranging from 0.01 cd·s·m⁻² to 10 cd·s·m⁻²) showed no difference between KO and WT mice at both 10 months and 14 months of age. (B) The DA ERG b-wave amplitude measures to different intensities of flash (ranging from 0.0025 cd·s·m⁻² to 10 cd·s·m⁻²) showed no difference between KO and WT mice at both time points. (C) Light-adapted (LA) ERG amplitude measures to four different flicker frequencies (5–20 Hz) showed no difference between KO and WT mice at both time points. Notably, the LA flicker amplitudes are lower at higher stimulus frequencies (15 and 20 Hz) compared to lower stimulus frequencies (5 and 10 Hz) in both KO and WT mice. This is because the depolarizing bipolar cell contribution that predominates the LA ERG at lower flicker frequencies shows larger amplitudes compared to the hyperpolarizing bipolar cell contribution that predominates at high flicker frequencies. Similar findings have been consistently observed in multiple WT mouse models. (D) The average retinal thickness was comparable between KO and WT mice at both time points. (E) A representative horizontal line scan from KO and WT mice, respectively, showing how the average retinal thickness was calculated.

Figure 1

Full-field electroretinogram (ERG) and optical coherence tomography (OCT) findings in Epm2a^−/− knock out (KO) mice and control (WT) littermates. (A) Dark-adapted (DA) ERG a-wave amplitude measures to different intensities of flash (ranging from 0.01 cd·s·m⁻² to 10 cd·s·m⁻²) showed no difference between KO and WT mice at both 10 months and 14 months of age. (B) The DA ERG b-wave amplitude measures to different intensities of flash (ranging from 0.0025 cd·s·m⁻² to 10 cd·s·m⁻²) showed no difference between KO and WT mice at both time points. (C) Light-adapted (LA) ERG amplitude measures to four different flicker frequencies (5–20 Hz) showed no difference between KO and WT mice at both time points. Notably, the LA flicker amplitudes are lower at higher stimulus frequencies (15 and 20 Hz) compared to lower stimulus frequencies (5 and 10 Hz) in both KO and WT mice. This is because the depolarizing bipolar cell contribution that predominates the LA ERG at lower flicker frequencies shows larger amplitudes compared to the hyperpolarizing bipolar cell contribution that predominates at high flicker frequencies. Similar findings have been consistently observed in multiple WT mouse models. (D) The average retinal thickness was comparable between KO and WT mice at both time points. (E) A representative horizontal line scan from KO and WT mice, respectively, showing how the average retinal thickness was calculated.

Figure 2

Periodic Acid Schiff-Diastase) (PASD) staining in Epm2a^−/− knock out (KO) mice and control (WT) littermates. (A) PASD staining of KO mice shows Lafora body (LB) deposition in KO mice at both 10 and 14 months (top panel). The LBs (black arrows) were noted in the inner plexiform layer (IPL), inner nuclear layer (INL) and outer plexiform layer (OPL). The LBs were most numerous in the IPL. The LBs were not visualised in WT littermates (bottom panel). All the other retinal layers are also labelled: retinal pigmented epithelium (RPE), photoreceptor outer segment (OS), photoreceptor inner segment (IS), outer nuclear layer (ONL) and ganglion cell layer (GCL). (B) Automated quantification of the LBs within the IPL showed a significantly higher LB deposition in KO mice compared to WT mice at both time points (**** p < 0.0001). There is a significant increase in the number of LBs in KO mice at 14 months, in comparison to 10 months (p = 0.035).