Impaired BDNF-TrkB trafficking and signalling in Down syndrome basal forebrain neurons

Abstract

Brain derived neurotrophic factor (BDNF) and its receptor tropomyosin-related kinase B (TrkB) play crucial roles in neuronal development, synaptic transmission, and neuroplasticity. Deficits in BDNF/TrkB signalling and trafficking have been identified in several neurodegenerative diseases, including Alzheimer's disease (AD). Individuals with Down syndrome (DS) are at an increased risk of developing AD compared to the general population. Basal forebrain neurons (BFNs) are among the first to degenerate in AD and DS, but the mechanisms underlying their vulnerability remain unclear. Using BFNs derived from the Dp1Tyb mouse model of DS, we investigated neurotrophic signalling and trafficking deficits in AD-DS. We found enlarged early endosomes and elevated levels of active Rab5, a GTPase critical for early endosome formation, in Dp1Tyb BFNs. These abnormalities were associated with impaired transport of internalised TrkB from axon terminals to the soma. Using microfluidic devices, we demonstrated that axonal BDNF stimulation enhanced signalling endosome dynamics in wild-type but not Dp1Tyb BFNs, which is likely due to impaired axonal ERK1/2 signalling. Our findings establish a link between Rab5 hyperactivation, endosomal dysfunction, and impaired ERK1/2 signalling, highlighting the interplay between trafficking and neurotrophic signalling, and underscore the importance of targeting endolysosomal and signalling pathways to mitigate neuronal dysfunction in AD-DS.

Fig. 1. Early endosomes are enlarged and display increased Rab5 activity in basal forebrain neurons isolated from Dp1Tyb embryos.

A Representative confocal images of EEA1 immunofluorescence in WT and Dp1Tyb BFNs at DIV14. B EEA1 positive puncta were significantly larger in Dp1Tyb neurons compared to WT (Student’s t test; ****p < 0.0001). C The size distribution of EEA1 positive puncta in Dp1Tyb BFNs showed a significant shift from smaller to larger binned areas compared to WT littermates (Student’s t test; *p = 0.0320). D Quantification of EEA1 puncta number in Dp1Tyb BFNs compared to WT (Student’s t-test; ***p = 0.0002). E Representative images of WT and Dp1Tyb neurons incubated with GST-Rab5BD to measure levels of GTP- Rab5 in the cell bodies (dashed lines) following 0, 5, 30 min BDNF treatment. F BDNF significantly increased GTP bound Rab5 in BFNs after 30 min treatment. GTP-Rab5 levels were significantly elevated in Dp1Tyb BFNs compared to WT at baseline and with BDNF treatment. (2-way ANOVA with Tukey’s multiple comparisons; BDNF; ****p < 0.0001, genotype; ****p < 0.0001. Multiple comparisons; ****p < 0.0001, ***p = 0.0009, **p = 0.0072, *p = 0.0368). Different coloured points represent cells from separate Ns. Bars represent mean ± SEM. N = 4 biological replicates. 15–25 cells were imaged per N. Scale bar = 10 µm.

Fig. 2. BDNF enhancement of signalling endosome transport is impaired in Dp1Tyb basal forebrain neurons.

A Schematic of the three-chamber MFC used to culture primary BFNs for axonal transport experiments. Axonal compartments were incubated with Alexafluor555-HcT (HcT-555) ± BDNF and time-lapse imaging was conducted within the MFC microgrooves. Endosomes were manually tracked using TrackMate. Pink arrows highlight a paused endosome, yellow/cyan arrows highlight retrogradely moving endosomes. Scale bar = 20 µm. B, C Average endosome speed in WT and Dp1Tyb BFNs in response to BDNF. Lines represent paired conditions within the same three-chambers MFC. BDNF significantly enhances signalling endosome speed in WT (paired Student’s t test; *p = 0.0372), but not in Dp1Tyb (paired Student’s t test; p = 0.1462) BFNs. D, E Percentage of pausing of HcT-555-positive signalling endosomes in response to BDNF. BDNF significantly reduces the percentage of pausing in WT (paired Student’s t test; *p = 0.0186), but not Dp1Tyb (paired Student’s t test; p = 0.4451) primary BFNs. F, G Speed distribution curves of signalling endosome transport in WT and Dp1Tyb neurons ± BDNF. Endosome pausing (speed <0.25 μm/s) was excluded from the curves). All graphs N = 6 biological replicates, 5–8 axons per condition.

Fig. 3. Reduced somatic accumulation of TrkB in Dp1Tyb basal forebrain neurons following axonal stimulation with BDNF.

A Schematic representation of the TrkB retrograde accumulation assay in three-chamber MFCs. B Representative confocal microscopy images of DiI and DiO retrograde tracers, and TrkB accumulation in the somatic compartment ± axonal BDNF treatment. C Representative images of TrkB accumulation in the soma following axonal incubation with anti-TrkB antibodies and BDNF in WT and Dp1Tyb neurons. D, E A significant decrease in the accumulation of TrkB in Dp1Tyb BFNs per cell (Student’s t test; ***p < 0.0001) and per N (Student’s t test; **p = 0.0073). Different coloured points represent individual Ns. Bars represent mean ± SEM. N = 3 biological replicates. 10–15 cells per biological replicate. Scale bar = 10 µm.

Fig. 4. Selective impairment of ERK1/2 signalling following BDNF treatment in Dp1Tyb basal forebrain neurons.

A Representative confocal images of pERK1/2 staining in WT and Dp1Tyb neurons. B Quantification of pERK1/2 immunostaining. There was a significant effect of both BDNF treatment and genotype on ERK1/2 activation (2-way ANOVA with Tukey’s multiple comparisons; BDNF treatment time; ****p < 0.0001, genotype; ****p < 0.0001 and interaction; ****p < 0.0001. Multiple comparisons; ****p < 0.0001, *p = 0.0195). N = 5 biological replicates. Bars represent mean ± SEM. Coloured points represent individual Ns. C Representative confocal images of pERK1/2 and SMI-31 staining in WT and Dp1Tyb axons ± BDNF treatment. D pERK1/2 intensity per axon in WT and Dp1Tyb BFNs ± 5 min BDNF treatment (2-way ANOVA with Tukey’s multiple comparisons; BDNF treatment time; ***p = 0.0004; genotype; ***p = 0.0006. Multiple comparisons; **p = 0.0024, ****p < 0.0001)). N = 4 biological replicates. Bars represent mean ± SEM. Coloured points represent individual Ns. E, F BDNF treatment significantly increased the phosphorylation of ERK1/2 in WT BFN axons (paired Student’s t test; **p = 0.0089), but not in Dp1Tyb BFN axons (paired Student’s t test; p = 0.0915). Scale bar = 10 µm.

Fig. 5. The axonal retrograde transport of signalling endosomes in Dp1Tyb basal forebrain neurons is not affected by ERK1/2 inhibition.

A Schematic depicting HcT-555 axonal transport experiment ± ERK inhibitor U0126. B Representative kymographs of HcT-555-positive signalling endosome transport in WT and Dp1Tyb BFNs ± U0126. C, D Average endosome speed per N in WT and Dp1Tyb BFNs in response to BDNF ± U0126. Lines represent paired conditions within the same three-chambers MFC. U0126 significantly reduces signalling endosome speed in WT (paired Student’s t test; *p = 0.0135), but not in Dp1Tyb (paired Student’s t test; p = 0.7136) BFNs. E, F Percentage of pausing of HcT-555-positive signalling endosomes in response to U0126 per N. Lines represent paired conditions within the same three-chamber MFC. U0126 significantly increases the percentage of pausing in WT (paired Student’s t test; **p = 0.0094) but not Dp1Tyb (paired Student’s t test; p = 0.4571) BFNs. G, H Endosome frame-frame speed distribution curve for signalling endosomes in WT and Dp1Tyb primary BFNs ± U0126 treatment. For all graphs N = 4 biological replicates, 5–8 axons imaged per condition. Endosome pausing (speed <0.25 μm/s) was excluded from the curves.