Cholinergic abnormalities, endosomal alterations and up-regulation of nerve growth factor signaling in Niemann-Pick type C disease

Abstract

Background: Neurotrophins and their receptors regulate several aspects of the developing and mature nervous system, including neuronal morphology and survival. Neurotrophin receptors are active in signaling endosomes, which are organelles that propagate neurotrophin signaling along neuronal processes. Defects in the Npc1 gene are associated with the accumulation of cholesterol and lipids in late endosomes and lysosomes, leading to neurodegeneration and Niemann-Pick type C (NPC) disease. The aim of this work was to assess whether the endosomal and lysosomal alterations observed in NPC disease disrupt neurotrophin signaling. As models, we used i) NPC1-deficient mice to evaluate the central cholinergic septo-hippocampal pathway and its response to nerve growth factor (NGF) after axotomy and ii) PC12 cells treated with U18666A, a pharmacological cellular model of NPC, stimulated with NGF.

Results: NPC1-deficient cholinergic cells respond to NGF after axotomy and exhibit increased levels of choline acetyl transferase (ChAT), whose gene is under the control of NGF signaling, compared to wild type cholinergic neurons. This finding was correlated with increased ChAT and phosphorylated Akt in basal forebrain homogenates. In addition, we found that cholinergic neurons from NPC1-deficient mice had disrupted neuronal morphology, suggesting early signs of neurodegeneration. Consistently, PC12 cells treated with U18666A presented a clear NPC cellular phenotype with a prominent endocytic dysfunction that includes an increased size of TrkA-containing endosomes and reduced recycling of the receptor. This result correlates with increased sensitivity to NGF, and, in particular, with up-regulation of the Akt and PLC-γ signaling pathways, increased neurite extension, increased phosphorylation of tau protein and cell death when PC12 cells are differentiated and treated with U18666A.

Conclusions: Our results suggest that the NPC cellular phenotype causes neuronal dysfunction through the abnormal up-regulation of survival pathways, which causes the perturbation of signaling cascades and anomalous phosphorylation of the cytoskeleton.

Figure 1

Characterization of NPC1^-/- septal cholinergic neurons. A. Brain sections at the level of the medial septum (MS) from 8-week-old WT and NPC1^-/- mice were stained for ChAT and visualized with secondary antibodies conjugated to HRP. The insets show a magnification of the cholinergic cells in the upper panels. WT cholinergic neurons are smaller and less rounded than NPC1^-/- cholinergic neurons, and there is a reduction in the number of cholinergic fibers surrounding the labeled NPC1^-/- cholinergic cells. B. The number of ChAT-labeled cholinergic neurons was quantified in the septal area of WT and NPC1^-/- brain sections. Four sections from six WT and six NPC1^-/- age-matched mice were used for quantification. There are no differences in the number of septal cholinergic neurons, p = 0.829. C and D. The morphology of septal cholinergic neurons was evaluated in ChAT-stained brain sections from WT and NPC1^-/- mice. NPC1^-/- cells have an increased area and exhibit a higher ratio between the minor and major axes of cholinergic neurons, * p < 0.006. A total of 600 cells from six WT and six NPC1^-/- age-matched mice were used for quantification. E. Quantification of p75-labeled cholinergic fibers in different septal regions without cholinergic neuronal somas, as shown in the insets of Additional file 2: Figure 2A, *p < 0.0001.

Figure 2

In vivo NGF response in WT and NPC1-/- mice. A and B. WT and NPC1^-/- mice respond similarly to axotomy of the fimbria-fornix. Brain sections from 8-week-old axotomized (unilateral fimbria-fornix axotomy) WT and NPC1^-/- mice were stained with an antibody against ChAT (A). Quantification shows an evident and similar reduction in the number of neurons labeled with ChAT (B). The differences between the phenotypes are not significant, p = 0.755. C. Brain sections from 8-week-old axotomized WT and NPC1^-/- mice were immunostained for ChAT, and the images were acquired by confocal microscopy. In WT mice, axotomized septal neurons (IL, ipsilateral to the axotomy) exhibit lower levels of ChAT than contralateral neurons (CL). In NPC1^-/- mice, ChAT staining in the axotomized septal neurons is similar to that in the contralateral neurons. D. The quantification of ChAT intensity in septal neurons (90-92 neurons for contralateral side and 40-42 neurons for ipsilateral) from three WT and three NPC1^-/- mice demonstrates a significant reduction in the ChAT levels in WT axotomized neurons, while NPC1^-/- axotomized neurons appear protected, *p < 0.005, unpaired one-way ANOVA. The increased staining of ChAT-labeled cells in C and D is not due to increased cell size because the total fluorescence or immunostaining intensity of each cell was normalized by cell area. E. Upper panel: ChAT levels were analyzed by western blot using homogenates of the MS from three 8-week-old WT and three NPC1^-/- mice. Bands corresponding to ChAT from three different WT and NPC1^-/- mice are shown. Lower panel: quantification of the western blot indicates that ChAT levels are significantly higher in NPC1^-/- mice than in WT mice. *p < 0.01, unpaired Student's t-test. n = 3 animals per phenotype. ChAT was normalized against total Akt. F-H. Increased ChAT staining in response to NGF infusion in NPC1^-/- septal cholinergic neurons. Brain sections from 8-week-old axotomized WT and NPC1^-/- mice infused with NGF for one week (F). Black arrows indicate cholinergic neurons with increased ChAT staining. Quantification shows that NGF protected septal cholinergic neurons equally in WT and NPC1^-/- mice (G). Quantification of ChAT staining in cholinergic cells (as indicated by the arrows in F). AU, arbitrary units. The differences are statistically significant, *p < 0.0001. Two sections through the medial septum of four WT and four NPC1^-/- age-matched mice were used for quantification. A total of approximately 180 cells were considered for each phenotype (H).

Figure 3

Increased neurite length in NGF-treated PC12-U18666A compared to controls. A. PC12 cells were treated for 24 hrs with 2 μg/ml of U18666A and for another 48 hrs with the same dose of the drug in the presence of NGF (5 ng/ml). After the treatments, PC12 cells were fixed and stained for filipin. B. The number of differentiated cells (cells with a neurite at least twice the diameter of the cell body), the length of the neurites and the number of neurites per cell were quantified in control and PC12-U18666A cells. A total of 100 cells from three different experiments were quantified. The differences between treatments are significant, *p < 0.02. C. PC12 cells were treated for 24 hrs with NGF (5 ng/ml) and for an additional 24 hrs with NGF alone (48 hrs NGF) or in the presence of U18666A (2 μg/ml) (48 hrs NGF + 24 hrs U18). D. The number of differentiated cells, the length of neurites and the number of neurites per cell were quantified in the conditions defined in C. The differences between treatments are significant, *p < 0.02. E. PC12 cells were treated for 72 hrs with NGF (72 hrs NGF) or 24 hrs with NGF followed by 48 hrs with U18666A (72 hrs NGF + 48 hrs U18). Cells treated with the drug are stained, as shown in the representative images. F. PC12 cells were treated for 48 hrs with NGF (48 hrs NGF) or 24 hrs with NGF followed by 24 hrs with U18666A in the absence of NGF (24 hrs NGF + 24 hrs U18). PC12 cells treated with the drug in the absence of NGF retracted neuritis, which is evident compared with cells that were treated with NGF for 48 hrs. G. PC12 cells were treated for 48 hrs with NGF (48 hrs NGF) or for 24 hrs with NGF followed by 24 hrs with wortmannin (100 nM) in the presence of NGF. PC12 cells treated with wortmannin retracted neuritis, indicating that neurite elongation is at least partly regulated by IP3K signaling.

Figure 4

Increased Akt activation in NGF-treated PC12-U18666A and in the medial septum of NPC1^-/- mice compared to WT. A. PC12 cells were treated for 24 hrs with 2 μg/ml U18666A and with 5 ng/mL NGF for different durations (0-360 min). After NGF treatment, PC12 cells were homogenized, and the samples were processed for SDS-PAGE, blotted and immunolabeled for phosphorylated Akt (pAKT), total Akt (AKT), phosphorylated ERK1/2 (pERK), total ERK1/2 (ERK), phosphorylated PLC-γ (pPLC-γ) and total PLC-γ (PLC-γ). B. PC12 cells were treated for 24 hrs with 2 μg/ml U18666A (U18666A) or not treated (Control) and then were left untreated (T0) or incubated with NGF for 2 hrs (2 h). After treatment, PC12 cells were homogenized, followed by immunoprecipitation with a polyclonal antibody against Trk (C14, Santa Cruz) and immunoblotting for total Trk (upper panel, monoclonal B3, Santa Cruz) or with a monoclonal antibody against phosphotyrosine (pTyr). The graph shows the quantification of total TrkA in control and PC12-U18666A cells at time 0 (T0) or after 2 hrs of NGF treatment (2 h). The levels of TrkA are equivalent in both experimental groups. C. The graph shows the quantification of pAkt normalized to total Akt ± SEM from three different experiments. The differences between treatments are significant, *p < 0.001 (two-way ANOVA). D. The graph shows the quantification of activated ERK1/2 (pERK) normalized to total ERK ± SEM from five different experiments. The differences between treatments are not significant (two-way ANOVA). E. The graph shows the quantification of activated PLC-γ (pPLC-γ) normalized against total PLC-γ ± SEM from four different experiments. The differences between treatments are significant, **p < 0.05, two-way ANOVA. F. Levels of phosphorylated Akt (pAKT) and total Akt (AKT) were analyzed by western blot of homogenates of the MS of three 8-week-old WT and three 8-week-old NPC1^-/- mice. Bands corresponding to pAkt and Akt from three different WT and NPC1^-/- mice are shown. G. Quantification of the western blot shown in F indicates that pAkt levels are significantly higher in NPC1^-/- mice than in WT mice, *p < 0.01, unpaired Student's t-test. n = 3 animals per phenotype. pAkt levels were normalized against total Akt.

Figure 5

Transferrin-positive endosome abnormalities and reduced recycling of the TrkA receptor in PC12-U18666A cells. A. PC12 cells were transfected with a Rab5-EGFP plasmid (which labels early endosomes in green) and after 24 hrs were treated (U18666A) or not treated (Control) with U18666A (2 μg/ml) for another 24 hrs. To load the cells with transferrin (labeling recycling endosomes in red), PC12 cells were serum-starved for 1 hr and treated with transferrin Alexa Fluor 555 (2 μg/ml) for 2 hrs at 37°C. Cells were fixed with paraformaldehyde and visualized with a confocal microscope. Scale bar, 5 μm. B. The fluorescence from Rab5-EGFP (n = 16-24 cells, from two different experiments) and transferrin (n = 57-60 cells, from two different experiments) associated with the perinuclear region of PC12 cells was measured and plotted. C, control PC12 cells. U18, PC12 cells treated for 24 hrs with U18666A (2 μg/ml), *p < 0.0001. C. Immunoendocytosis (2 hrs at 37°C) of endogenous TrkA (TrkA endogenous) and TrkA-Flag in PC12 cells not treated (Control) or treated (U18666A) for two days with the drug (2 μg/ml). The arrows indicate the presence of larger endosomes in U18666A-treated PC12 cells than in non-treated cells. D. Volume quantification of TrkA-positive endosomes of PC12 cells non-treated (C) or treated (U18) with U18666A (2 μg/ml). For endogenous TrkA-labeled endosomes (TrkA endo), a total of 63-69 vesicles were considered (from 6-7 representative cells from two different experiments). *p < 0.0001. For TrkA-Flag-labeled endosomes, a total of 15-19 endosomes were considered (from 2 representative cells). *p < 0.0003. E. Immunoendocytosis of TrkA-Flag was allowed to proceed for 2 hrs at 37°C in the presence of NGF (2 hrs inter). The Alexa Fluor 488-conjugated anti-Flag monoclonal antibody was then washed from PC12 cells by treatment with cooled EDTA. PC12 cells were fixed and mounted. For recycling experiments, immunoendocytosis of TrkA-Flag was allowed to proceed for 60 min, and the Alexa Fluor 488-conjugated anti-Flag monoclonal antibody was then washed from PC12 cells by treatment with cooled EDTA and further incubated with an anti-Alexa Fluor 488 antibody for 60 min at 37°C (60 min recycling), fixed and mounted. F. The fluorescence associated with PC12 cells after the treatments indicated in E was quantified and plotted. 2 hrs inter, 2 hrs internalization. 60 min rec, 60 min recycling. To measure the fluorescence associated with PC12 cells after 2 hrs of internalization of the Alexa Fluor 488-conjugated anti-Flag monoclonal antibody, 61-67 cells were considered from two different experiments. To measure the fluorescence associated with PC12 cells after 60 min recycling, 23-45 cells were considered from two different experiments. *p < 0.0023. G. To visualize the co-localization of TrkA-Flag with transferrin, PC12 cells not treated (control) or treated (U18666A) with the drug were incubated with the Alexa Fluor 488-conjugated anti-Flag monoclonal antibody and Alexa Fluor 555-conjugated transferrin (60 μg/ml) for 2 hrs in the presence of NGF and fixed. H. To visualize the co-localization of TrkA-Flag with LAMP2, PC12 cells not treated (control) or treated (U18666A) with U18666A were incubated with the Alexa Fluor 488-conjugated anti-Flag monoclonal antibody for 2 hrs in the presence of NGF, fixed and immunostained with a polyclonal antibody against LAMP2 (lysosomal marker).