Self-reinforcing effects of mTOR hyperactive neurons on dendritic growth

Abstract

Loss of the mTOR pathway negative regulator PTEN from hippocampal dentate granule cells leads to neuronal hypertrophy, increased dendritic branching and aberrant basal dendrite formation in animal models. Similar changes are evident in humans with mTOR pathway mutations. These genetic conditions are associated with autism, cognitive dysfunction and epilepsy. Interestingly, humans with mTOR pathway mutations often present with mosaic disruptions of gene function, producing lesions that range from focal cortical dysplasia to hemimegalanecephaly. Whether mTOR-mediated neuronal dysmorphogenesis is impacted by the number of affected cells, however, is not known. mTOR mutations can produce secondary comorbidities, including brain hypertrophy and seizures, which could exacerbate dysmorphogenesis among mutant cells. To determine whether the percentage or "load" of PTEN knockout granule cells impacts the morphological development of these same cells, we generated two groups of PTEN knockout mice. In the first, PTEN deletion rates were held constant, at about 5%, and knockout cell growth over time was assessed. Knockout cells exhibited significant dendritic growth between 7 and 18 weeks, demonstrating that aberrant dendritic growth continues even after the cells reach maturity. In the second group of mice, PTEN was deleted from 2 to 37% of granule cells to determine whether deletion rate was a factor in driving this continued growth. Multivariate analysis revealed that both age and knockout cell load contributed to knockout cell dendritic growth. Although the mechanism remains to be determined, these findings demonstrate that large numbers of mutant neurons can produce self-reinforcing effects on their own growth.

Figure 1:

Confocal maximum projections and representative Neurolucida reconstructions of brainbow-expressing cells in 7 and 18 weeks-old control (A, C) and KO (B, D) mice. The inset (green box) shows an expansion of outer molecular layer dendrites with processes growing parallel to the hippocampal fissure. Scale bar=100 μm (200 μm for the inset).

Figure 2:

Progression of morphological changes between 7 and 18 weeks for brainbowexpressing KO (red) and control (blue) cells. Each point in the graphs represents an individual granule cell, while bars depict means ± SEM. A: KO cell somas were larger than controls at both 7 and 18 weeks. B: Total apical dendrite length was greater for KO cells relative to controls at both 7 and 18 weeks, and both groups exhibited dendritic growth between 7 and 18 weeks. C: KO cell apical dendrites had more branches than controls at both 7 and 18 weeks. Between 7 and 18 weeks, dendritic branching increased among KO cells only. D: Dendrite length in the outer molecular layer was greater for KO cells relative to controls at both 7 and 18 weeks, and both groups exhibited dendritic growth between 7 and 18 weeks. E: KO cells had more branches than controls at both 7 and 18 weeks in the outer molecular layer. Between 7 and 18 weeks, dendritic branching increased among KO cells. F: The schematic shows the dentate granule cell layer (DGCL) and the three regions of the molecular layer (inner, IML; middle, MML and outer, OML) used for quantitative morphological analysis. *, p<0.0125. **, p<0.0025. ***, p<0.00025 (Bonferroni-adjusted p-values).

Figure 3:

Titrating the “load” of PTEN KO cells to achieve variable deletion rates. Images are confocal optical sections of the dentate gyrus showing PTEN-expressing cells (green) and Nissl staining (purple) in control mice (row A) and mice with 2% (row B) and 14% (row C) deletion rates. PTEN KO cells appear purple in the merged images (white arrows). Note the concentration of KO cells within the inner third of the dentate granule cell body layer; the typical positioning of postnatally-generated neurons. Scale bar=50 μm.

Figure 4:

Neuronal reconstructions of biocytin-filled cells from mice with variable rates of PTEN deletion from the dentate gyrus (10–33%). Cells from animals with higher deletion rates show greater dendritic length and complexity. Scale bar=100 μm.

Figure 5:

Three dimensional plots of multiple linear regression analyses utilizing the percentage of KO cells (% KO) and animal age in weeks (wks) as independent variables. The dependent variables are shown on the y-axis for each graph. Soma area (A) was predicted by age only, while apical dendrite length (B) was predicted by both age and the percentage of KO cells. When dendrite length was examined by subregion, dendritic length in the inner molecular layer (IML, D) was predicted by %KO only, while both variables predicted length in the middle (MML, E) and outer (OML, F) molecular layers. P-values for each independent variable are given on the appropriate axis.

Figure 6:

Basal dendrite length is predicted by animal age. A: Images show neuronal reconstructions of PTEN KO granule cells from mice at 10–24 weeks (wks)-of-age. Basal dendrites are shown projecting upwards. Scale bar = 50 μm. B: The three-dimensional plot depicts the results of a multiple linear regression analysis with the percentage of KO cells (%KO) and animal age (wks) as independent variables and basal dendrite length as the dependent variable. Basal dendrite length was predicted by animal age. P values are provided on the appropriate axis.

Figure 7:

The present study tested two competing hypotheses. The simplest hypothesis (H1, black arrow) predicts that KO granule cell morphology is entirely driven by cell intrinsic effects of PTEN loss. The key prediction of this model is that there should be no relationship between the number of KO cells in an animal and the morphology of individual KO cells. Our data supports the alternate hypothesis (H2, blue arrows), which predicts that the cell intrinsic effects of PTEN loss act in concert with secondary changes induced by the abnormal cells to exacerbate aberrant morphological changes among KO cells.