Successive neuron loss in the thalamus and cortex in a mouse model of infantile neuronal ceroid lipofuscinosis

Abstract

Infantile neuronal ceroid lipofuscinosis (INCL) is caused by deficiency of the lysosomal enzyme, palmitoyl protein thioesterase 1 (PPT1). We have investigated the onset and progression of pathological changes in Ppt1 deficient mice (Ppt1-/-) and the development of their seizure phenotype. Surprisingly, cortical atrophy and neuron loss occurred only late in disease progression but were preceded by localized astrocytosis within individual thalamic nuclei and the progressive loss of thalamic neurons that relay different sensory modalities to the cortex. This thalamic neuron loss occurred first within the visual system and only subsequently in auditory and somatosensory relay nuclei or the inhibitory reticular thalamic nucleus. The loss of granule neurons and GABAergic interneurons followed in each corresponding cortical region, before the onset of seizure activity. These findings provide novel evidence for successive neuron loss within the thalamus and cortex in Ppt1-/- mice, revealing the thalamus as an important early focus of INCL pathogenesis.

Figure 1

Schematic diagram of the thalamic nuclei and cortical regions analyzed in Ppt1^−/− mice. Measurements of cortical thickness were made in primary motor (M1) and somatosensory barrelfield (S1BF) cortex, primary visual (V1) and primary auditory cortex (Au1) and the lateral entorhinal cortex (LEnt), together with optical fractionator estimates of the number of lamina IV granule neurons and parvalbumin or somatostatin positive interneurons in S1BF and V1. The same unbiased stereological method was used to investigate the survival of thalamic relay neurons in the ventral posterior nucleus (VPM/VPL), dorsal lateral geniculate nucleus (LGNd), medial geniculate nucleus (MGN), central medial (CM), and of parvalbumin positive inhibitory neurons in the reticular thalamic nucleus (Rt, darker shading) at different stages of disease progression.

Figure 2

Progressive cortical atrophy in Ppt1^−/− mice. Cortical thickness measurements reveal progressive thinning of the cortical mantle that occurs at different rates between subfields that serve different functions. Compared with age matched controls (+/+), significant thinning of the cortical mantle in Ppt1^−/− mice was first evident in the primary visual cortex (V1) at 5 months of age and was subsequently observed in the primary auditory cortex (Au1) and primary motor (M1) at 7 months of age (** p<0.01). A similar decline in the thickness of the somatosensory barrelfield (S1BF) cortex was also evident in Ppt1^−/− mice, but did not reach statistical significance. In contrast, cortical atrophy in the entorhinal (LEnt) cortex was less pronounced in Ppt1^−/− mice.

Figure 3

Progressive accumulation of autofluorescent storage material in Ppt1^−/− mice. Graphs of the relative amount of autofluorescent storage material over time reveal the progressive increase in storage material in the somatosensory barrelfield cortex (S1BF) and thalamus of Ppt1^−/− mice compared with age-matched controls (+/+). Direct comparison of storage material accumulation revealed the similar rates of its accumulation in both regions.

Figure 4

Progressive astrocytosis in Ppt1^−/− mice. (A) Immunohistochemical staining for glial fibrillary associated protein (GFAP) reveals the pronounced upregulation of this marker of astrocytosis with increased age in the primary motor cortex (M1), somatosensory barrelfield cortex (S1BF) and primary visual cortex (V1) of Ppt1^−/− mice compared to age matched controls (+/+). GFAP immunoreactive astrocytes were evident within all laminae of M1, S1BF and V1 of mutant mice from 3 months of age onwards, initially most prominently in deeper laminae, but spreading to involve all laminae and increasing in intensity with disease progression. Laminar boundaries are indicated by roman numerals. (B) At 3 months of age Ppt1^−/− mice exhibit pronounced astrocytosis that is confined to individual thalamic nuclei, with GFAP-positive astrocytes prominent in the reticular (Rt), ventral posterior (VPM/VPL), lateral dorsal geniculate (LGNd) and medial geniculate (MGN) nuclei. The boundaries of these nuclei are indicated by white dashed lines. (C) Progressive accumulation of GFAP staining in the thalamus of Ppt1^−/− mice compared to age matched controls at 1, 3, 5 and 7 months of age. Scale bar = 400 μm in A; 200 μm in B and C.

Figure 5

Quantitative thresholding image analysis of GFAP (A) and F4/80 (B) immunoreactivity in the somatosensory barrelfield cortex (S1BF) demonstrates the progressive increase in these markers with disease progression in Ppt1^−/− mice compared with age matched controls (+/+). (A) GFAP immunoreactivity was first significantly elevated in Ppt1^−/− mice at 3 months of age and continued to increase during disease progression. (B) F4/80 immunoreactivity in Ppt1^−/− mice was first significantly elevated in Ppt1^−/− mice at 5 months of age and increased further at 7 months of age. (***, p<0.001, ANOVA with post-hoc Bonferroni analysis).

Figure 6

(A) Activation of microglia in 7 month old Ppt1^−/− mice. Immunohistochemical staining for the microglial marker F4/80 reveals activated microglia within all laminae of the primary motor cortex (M1), somatosensory cortex (S1BF) and primary visual cortex (V1) of Ppt1^−/− mice at 7 months of age compared to age matched control mice (+/+). Many F4/80 immunoreactive microglia with swollen soma and retracted processes were present in mutant mice, but were absent in controls. Laminar boundaries are indicated by roman numerals. (B) Localized microglial activation within individual thalamic nuclei of 7 month old Ppt1^−/− mice is also revealed by F4/80 immunoreactivity. In these mutant mice, intensely stained F4/80 positive microglia with brain macrophage like morphology were present in the ventral posterior (VPM/VPL), dorsal lateral geniculate (LGNd) and medial geniculate (MGN) nuclei, but were virtually absent from adjacent thalamic nuclei and were not present in age matched controls. The boundaries of these nuclei are indicated by white dashed lines. Scale bar = 300 μm in A, B; 100 μm in inserts in A, B.

Figure 7

Progressive loss of thalamic neurons in Ppt1^−/− mice. Histograms of unbiased optical fractionator estimates of the number of Nissl stained thalamic relay neurons in the dorsal lateral geniculate nucleus (LGNd), medial geniculate nucleus (MGN), ventral posterior nucleus (VPM/VPL) and parvalbumin stained inhibitory neurons in the reticular thalamic nucleus (Rt PV) of Ppt1^−/− mice and age-matched controls (+/+) at different stages of disease progression. The number of neurons in each of these thalamic nuclei declined in Ppt1^−/− mice with increased age, but this occurred at different rates between nuclei. Neuron loss in mutant mice first became significant within the visual system (LGNd) at 3 months of age, but was relatively delayed in somatosensory (VPM/VPL) or auditory (MGN) relay nuclei and for inhibitory neurons (Rt). (* p<0.05; ** p<0.01; *** p<0.001, ANOVA with post-hoc Bonferroni analysis).

Figure 8

Progressive loss of cortical neurons in Ppt1^−/− mice. Histograms of unbiased optical fractionator estimates of the number of Nissl stained lamina IV granule neurons and somatostatin- (SOM) and parvalbumin- (PV) positive neurons in the primary visual (V1) and somatosensory barrelfield (S1BF) cortex of Ppt1^−/− mice and age-matched controls (+/+) at different stages of disease progression. In V1 of mutant mice a significant loss of lamina IV granule neurons and SOM- or PV-positive interneurons was already evident at 5 months of age. SOM-positive interneurons were also significantly reduced in number at 5 months of age in S1BF of Ppt1^−/− mice, but the significant loss of PV-positive interneurons and granule neuron populations in S1BF did not occur until 7 months of age. (** p<0.01; *** p<0.001, ANOVA with post-hoc Bonferroni analysis).

Figure 9

Progressive development of the seizure phenotype of Ppt1^−/− mice. (A, B) electroencephalogram (EEG) recordings made via chronically implanted epidural screw electrodes revealed the onset and nature of spontaneous seizure activity in Ppt1^−/− mice with increased age. (A) Representative example of a seizure recorded in a Ppt1^−/− mouse by two channel EEG. (B) Seizures were not evident in Ppt1^−/− mice until 7 months of age and were present in all mutant mice by 7.5 months of age, but were not present in age matched controls (+/+). There was no significant difference between mutant mice in the frequency or duration of seizures at 7 and 7.5 months of age. Graded scoring of EEG recording traces also revealed a significant worsening of the interictal background EEG in mutant mice above 7 months of age, compared to mutant mice at 6 and 6.5 months of age and age-matched controls. (*** p<0.001, ANOVA with post-hoc Tukey-Kramer multiple comparisons).