Update your browser to view this website correctly. Update my browser now
Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by memory impairment and gradual loss of cognitive abilities. Amyloid-β (Aβ) generated by pathological cleavage of APP (amyloid precursor protein) has neurotoxic effects that contribute to AD pathology. Early defects in synaptic function and neuronal loss are found in the hippocampal formation, starting in the entorhinal cortex. In transgenic mouse models of AD-like pathology, long-term potentiation (LTP) of synaptic transmission in hippocampal pathways is commonly impaired. However, the cause of LTP defects in AD mouse models has not been defined and mechanisms might differ between hippocampal sub-regions. In the entorhinal perforant path input to the dentate gyrus (DG), LTP consolidation depends on synaptic activity-induced protein synthesis that is regulated at the level of translation initiation. Here we asked whether transgenic APP/PS1 amyloidogenic mice have defects in the stimulus-induced formation of the translation initiation complex during DG LTP. High-frequency stimulation was applied to the perforant path and changes in field excitatory postsynaptic potentials (fEPSPs) were recorded in the DG in vivo. LTP of the fEPSP slope was induced in both genotypes, with no difference in magnitude at 50 min post-HFS. However, fEPSP responses in APP/PS1 mice declined to baseline by 2 h, whereas wild-type mice exhibited stable LTP. This selective defect in LTP consolidation in APP/PS1 mice was associated with loss of translation initiation complex formation, detected as binding of eukaryotic translation initiation factor 4G (eIF4G) to eIF4E in m7GTP cap-pulldown assays in DG tissue lysates. We conclude that DG LTP consolidation is severely impaired in APP/PS1 mice in vivo and associated with impaired translation initiation underlying protein synthesis-dependent synaptic plasticity.
Alzheimer’s disease (AD) is a progressive neurodegenerative disease associated with loss of memory and formation of Amyloid-β plaques in the extracellular space and neurofibrillary tangles inside neurons. The hippocampal formation is affected early in the disorder, with neurodegeneration starting in the entorhinal cortex (EC) before spreading to the dentate gyrus (DG) and hippocampal subfields. Moreover, selective overexpression of APP in the EC results in Aβ formation and altered synaptic function. The accumulation of Aβ deposits in the hippocampus is associated with functional impairment in synaptic transmission and plasticity and progressive loss of dendritic spines, as a pathological correlate of amnesia and cognitive dysfunction in AD. However, the impact of Aβ deposits on stimulus-evoked synaptic signaling and plasticity in the entorhinal projection to the DG in vivo is poorly understood.
Long-term synaptic plasticity, like memory, is generally considered to require new protein synthesis for consolidation. Long-term potentiation (LTP) of synaptic transmission provides a tractable experimental model system for analysis of synaptic function and plasticity in AD mouse models. LTP in the entorhinal perforant path projection to the DG of rodents requires the synaptic activity-induced formation of the translation initiation complex, which supports protein synthesis underlying stable (late phase) LTP. The formation of the translation initiation complex depends on the binding of the mRNA cap-binding protein, eukaryotic translation initiation factor 4E (eIF4E), to the scaffolding protein eIF4G. The eIF4E-eIF4G complex together with other factors results in the recruitment of ribosomes to mRNA, thus starting translation. Convergent evidence from human AD and mouse AD-like models suggests dysregulation of translation as a key component. However, regulation at the level of eIF4E-eIF4G complex formation, which is critical for synaptic plasticity in DG, has not been investigated.
Transgenic mice carrying mutations in human APP and presenilin 1 (APP/PS1 mice) provide an established model of AD-like β-amyloidosis. APP/PS1 mice are associated with enhanced amyloid deposition, plaque formation, neuronal loss, and cognitive decline. LTP is also impaired in different strains of APP/PS1 mice that differ in time of disease onset and severity of amyloidosis, in both the hippocampal CA1 region and DG. Here, we specifically asked whether translation initiation underlying LTP consolidation in the DG is affected in APP/PS1 mice. We report normal DG LTP induction, but failure in LTP consolidation that is correlated with a defect in the stimulus-evoked formation of the translation initiation complex.
To determine whether APP/PS1 mice have impaired LTP consolidation in the DG in vivo associated with the impaired formation of eIF4E-eIF4G translation initiation complex.
Selective impairment of late LTP maintenance in APP/PS1 transgenic mice
We examined LTP induction and maintenance at perforant path-granule cell synapses in the DG of wild-type and 7 to 8 month-old APPswe/PS1L166P mice. The maximum initial slope of the fEPSP was taken as a measure of synaptic transmission efficacy. Based on analysis of input-out curves of fEPSP and population spike responses, there was no difference in basal synaptic transmission between APP/PS1 mice and their wild-type littermates (Suppl. Fig. 1). Following HFS, wild-type mice exhibited non-decremental increases in the fEPSP over a 3 h recording period (Fig. 1A). In APP/PS1 mice, HFS induced an increase in the fEPSP slope which was equivalent in magnitude to wild-type control at 40-50 min post-HFS (Fig. 1A). However, fEPSP slope values declined precipitously to baseline by 120 min post-HFS and remained at baseline until the termination of recording at 3 h post-HFS (Fig. 1A). Typical field potential traces are shown in figures 1C and 1D. We conclude that LTP induction and early maintenance are intact in APP/PS1 mice, while late maintenance is severely impaired. Previous work in freely moving, 17-18 months old APP/PS1A246E mice, distinct from the APP/PS1 strain used here, similarly demonstrated normal induction with accelerated decay of DG LTP. In the latter study, fEPSP measurements were obtained up to 60 min post-HFS, and then again at 24 h, with no intervening recordings. Here, we show the rapid decay of LTP during the first hours of LTP maintenance, which captures the critical period of protein synthesis-dependent LTP consolidation.
Impairment of HFS-evoked translation initiation complex formation in APP/PS1 mice
To assess the regulation of the translation initiation complex, cap-pulldown assays were performed in DG tissue collected 3 h post-HFS. In the ipsilateral HFS-treated DG of wild-type mice, the cap-pulldown analysis showed enhanced association of eIF4G with eIF4E relative to the contralateral control DG, indicating enhanced initiation complex formation (Fig. 1E, 1F). In contrast, application of HFS in APP/PS1 mice did not significantly alter levels of eIF4G detected on cap-bound eIF4E (Fig. 1E, 1F). The ability of eIF4G to interact with eIF4E is determined by translational repressors, the eIF4E binding proteins, such as eIF4E-binding protein 2 (4E-BP2), and cytoplasmic fragile-X-interacting protein 1 (CYFIP1). These repressors prevent the attachment of eIF4G to eIF4G, while stimuli that cause the release of either repressor will facilitate eIF4E-eIF4G interaction and promote translation. Here we found that 4E-BP2 and CYFIP1 are released from eIF4E in the HFS-treated DG of wild-type mice (Fig. 1E, 1F), as previously reported. In contrast, HFS in APP/PS1 mice failed to trigger the removal of the repressors from eIF4E (Fig. 1G, 1H), which is consistent with failure to recruit eIF4G to eIF4E. Immunoblot analysis of whole DG lysate samples showed no change in expression eIF4E, eIF4G, 4E-BP2, or CYFIP1 in HFS-treated DG relative to contralateral DG in wild-type or APP/PS1 mice (Fig. 1G, 1H).
Taken together, the present results suggest that 7-8 month-old APP/PS1 mice have normal induction and early maintenance of DG LTP but are unable to form stable LTP, and this defect is associated with impaired loading of eIF4G onto eIF4E, indicating impaired translation initiation. The mechanism underlying the defect remains to be elucidated. Impairment in brain-derived neurotrophic factor (BDNF) function is a possibility, as Aβ accumulation disrupts BDNF-TrkB signalling, which in turn is important for protein synthesis-dependent LTP consolidation. It also remains to be seen whether LTP consolidation in the CA1 region and other parts of the brain are affected in the same way as the DG. For instance, another APPswe/PS1ΔE9 AD model shows stable LTP in CA1 and DG region in vivo 1 h after LTP induction, but the impacts on subsequent LTP maintenance have not been explored.
In conclusion, APP/PS1 mice have intact DG LTP induction and early maintenance, but the impaired formation of stable late-phase LTP, and this defect is associated with the impaired formation of the translation initiation complex.
Using cap-pulldown analysis in mouse DG lysates, this study identified a defect in translation initiation complex formation coupled to a defect in LTP consolidation in vivo. However, the impact on global and specific mRNA translation remains to be established. The small volume of mouse DG also limits the number of biochemical analyses that can be performed alongside the cap-pulldowns. The biochemical regulation shown here has not been localized to a specific cell type or subcellular compartment, which might be elucidated by analysis of pooled synaptoneurosomal fractions enriched in pinched-off dendritic spines.
A major goal is to uncover the mechanisms upstream of translation initiation that lead to the defect. Impairment of BDNF-TrkB signaling is a viable possibility. Further studies are needed to assess the developmental time course of the pathology and isolate the specific impacts of PS1 mutations and APP expression.
APPswe/PS1L166P mice with a C57/BL6 background were bred at the University animal facility. Transgenic APP/PS1 male mice were housed in the animal care facility for at least 2 weeks before starting the electrophysiological experiments. 7-8 month-old mice, weighing 28–35 g were used for the experiments. Animals were fed an autoclaved standard rodent diet (SDS, England; RMI-E), and maintained in a climate control animal facility with room temperature (22°C ±1), and relative humidity (46 ±5%). Mice had free access to food and water and were on a 12 h light/dark cycle.
Antibodies used for immunoblotting were as follows: CYFIP1 (1:1000, Upstate #07-531), FMRP (1:1000, Abcam #17722), 4E-BP2 (1:1000, Cell Signaling #2845), GAPDH (1:5000, Santa Cruz Biotechnology #32233), p-eIF4E (1:1000, Cell Signaling #9741), eIF4E (1:1000, Cell Signaling #9742), eIF4G (1:1000, Cell Signaling #2498).
In vivo electrophysiology in mice
Mice were anesthetized with urethane (injected i.p. 1.2 g/kg body weight), which was supplemented throughout surgery and recording as required. Mice were placed in a stereotaxic frame and body temperature was maintained at 37°C. In one hemisphere only, a bipolar stimulation electrode (NE-200, 0.5 mm tip separation, Rhodes Medical Instruments, Wood hills, CA) was positioned ipsilaterally onto the perforant path (3.8 mm posterior to bregma, 2.5 mm lateral to the midline, and 1.6 mm from the brain surface), while an insulated tungsten recording electrode (0.075 mm; A-M Systems) was positioned in the hilus of the DG (2 mm caudal to bregma, 1.5 mm lateral to the midline, and 1.5–1.7 mm from the brain surface). The recording electrode was lowered in 0.1 mm increments while monitoring changes in the stimulus-evoked waveform until the maximum, positive-going field excitatory postsynaptic potential (fEPSP) slope was obtained in the dentate gyrus. To generate input/output (I/O) curves, 5 stimulus intensities were used: 1) PS threshold, 2) 10% of maximum PS, 3) 30% of maximum PS, 4) 60% of maximum PS, and 5) maximum PS. The stimulus intensity ranged from approximately 80 µA (below PS threshold) to 400 µA (maximum PS). 4 responses were collected at each stimulus intensity and averaged. After generating an I/O curve, a stable 20 min baseline of evoked fEPSPs was recorded before delivery of high-frequency stimulation (HFS) to induce LTP. Stimulus pulses were biphasic, 0.1 ms pulse-width, delivered every 30 s (0.033 Hz) for baseline and post-HFS recording. The test pulse stimulation intensity was set to elicit a population spike amplitude of 30% of the maximum response. The HFS protocol consisted of 4 trains of stimuli with an interval of 10 s; each train consisted of 15 pulses at 200 Hz, at a stimulus intensity twice that used for test pulse. Evoked responses were recorded for 3 h post-HFS, and changes in the fEPSP slope were expressed in percent of baseline (average of 20 min baseline period). The maximum initial slope of the rising phase of the fEPSP was measured as an index of synaptic transmission efficacy. After completing the recording, mice were decapitated and the dentate gyri were micro-dissected and immediately frozen on dry ice for later use.
Tissue dissection and sample preparation
The ipsilateral (stimulated) and contralateral (control) DG were rapidly dissected on ice and homogenized in buffer containing 50 mM HEPEs, 100 mM NaCl, 1 mM EDTA, NP-40 0.5%, 1 mM dithiothreitol, 1 mM Na3VO4, 50 mM NaF, and 1× protease inhibitor cocktail from Roche #11836170001. Homogenization was performed manually with 10–12 gentle strokes in a tissue grinder and the homogenate was centrifuged 10 min at 14000 ×g at 4°C. Protein concentration was measured using BCA protein assay (Pierce, #23227). Homogenates were stored at -80°C until use.
Analysis of m7GTP binding proteins
For the m7GTP pull-down assay, 250-300 μg of protein lysate together with 30 μl of 7-methyl GTP-agarose beads (Jena bioscience #AC-141) were incubated for 90 min at 4°C. Beads were washed three times with m7GTP lysis buffer and bound proteins were separated to an SDS-PAGE (10% gels or 4–15% gradient gels). Immunoblotting was carried out as described below.
SDS–PAGE and immunoblotting
Samples from m7GTP pull-down assays and lysates were boiled at 70°C for 10 min in Laemmli sample buffer (Bio-Rad) and resolved in 10% SDS/PAGE gels. Proteins were transferred to nitrocellulose membranes (Biorad, #162-0112) which were then blocked with 5% BSA, probed with the respective antibodies, and developed using chemiluminescence reagents (Biorad, #1705061). The blots were scanned using Gel DOC XRS+ (BIO-RAD) and densitometric analyses were performed with Image J software (NIH, Bethesda, MD). In DG lysate (input) samples, densitometric values were expressed per unit of protein (GAPDH) applied to the gel lane. In the cap-pulldown analysis, protein expression was normalized to eIF4E. Values from the HFS-treated DG are expressed as percent change from the contralateral internal control DG of the same mouse.
Group values are reported as mean ±SEM. Statistical comparisons were calculated with the Two-way ANOVA with Bonferroni correction or multiple t-test using GraphPad Prism 8.02. The significance level was set at P <0.05.
This work was funded by the EU Joint Program–Neurodegenerative Disease Research (JPND) project CIRCPROT jointly funded by the RCN (to CB), the BMBF (to VL), and by EU Horizon 2020 co-funding (project no. 643417).
All experimental procedures approved by Norwegian National Research Ethics Committee in compliance with EU Directive 2010/63/EU, ARRIVE guidelines. Persons involved in animal experiments have Federation of Laboratory and Animal Science Associations (FELASA) C course certificates and training.