Cholinergic Basal Forebrain Lesion Decreases Neurotrophin Signaling without Affecting Tau Hyperphosphorylation in Genetically Susceptible Mice

Abstract

Alzheimer’s disease (AD) is a progressive, irreversible neurodegenerative disease that destroys memory and cognitive function. Aggregates of hyperphosphorylated tau protein are a prominent feature in the brain of patients with AD, and are a major contributor to neuronal toxicity and disease progression. However, the factors that initiate the toxic cascade that results in tau hyperphosphorylation in sporadic AD are unknown. Here we investigated whether degeneration of basal forebrain cholinergic neurons (BFCNs) and/or a resultant decrease in neurotrophin signaling cause aberrant tau hyperphosphorylation. Our results reveal that the loss of BFCNs in pre-symptomatic pR5 (P301L) tau transgenic mice results in a decrease in hippocampal brain-derived neurotrophic factor levels and reduced TrkB receptor activation. However, there was no exacerbation of the levels of phosphorylated tau or its aggregation in the hippocampus of susceptible mice. Furthermore the animals’ performance in a hippocampal-dependent learning and memory task was unaltered, and no changes in hippocampal synaptic markers were observed. This suggests that tau pathology is likely to be regulated independently of BFCN degeneration and the corresponding decrease in hippocampal neurotrophin levels, although these features may still contribute to disease etiology.

Keywords

Alzheimer’s disease basal forebrain brain-derived neurotrophic factor (BDNF)cholinergic neuron hippocampus lesion neurotrophin tau hyperphosphorylation

INTRODUCTION

Alzheimer’s disease (AD) is an irreversible neurodegenerative disease that results in progressive memory and cognitive dysfunction. Degeneration of the basal forebrain cholinergic neurons (BFCNs) is an early characteristic feature of AD, with pathological changes observed prior to clinical manifestation of the condition [1, 2]. BFCN loss also closely correlated with the burden of senile plaques composed of deposits of amyloid-β (Aβ) protein as well as with the presence of neurofibrillary tangles (NFTs), composed of hyperphosphorylated tau protein [3 –6], and likely directly underpins aspects of the observed cognitive decline [7 –9].

Genetically modified strains of mice have been developed which phenotype the major features of the human disease: increased production of Aβ protein, formation of plaques, tau hyperphosphorylation, and memory impairment [10 –14]. Use of these models has revealed that development of NFTs most likely occurs downstream of Aβ neurotoxicity [15]. However, neither Aβ- nor NFT-producing transgenic mouse models develop the early, marked loss of BFCNs characteristic of human AD. The immunotoxin p75-saporin has been used to selectively ablate BFCNs, and BFCN lesioning in mouse models of AD that overproduce Aβ results in increased Aβ deposition and memory impairment [16 –19]. Moreover, BFCN lesioning of mouse models genetically programmed to form NFTs as well as amyloid pathology also results in increased levels of hyperphosphorylated tau [19]. This raises the question of whether lesioning of BFCNs could accelerate tau hyperphosphorylation and NFT formation independently of Aβ generation.

One mechanism by which tau phosphorylation could be induced by BFCN loss is through changes in neurotrophic signaling. The levels of brain-derived neurotrophic factor (BDNF) mRNA and protein are selectively decreased in the hippocampus and cortex of AD patients compared to normal aged controls [20], and high BDNF levels predict slower cognitive decline in AD patients [21]. BDNF deprivation of cultured mouse hippocampal neurons can cause increased tau hyperphosphorylation due to increased Akt and glycogen synthase kinase-3β(GSK3β) activity [22], the latter being a major tau kinase and a target of BDNF-mediated signaling [23, 24]. Moreover, it has previously been reported that BFCN lesions reduce the hippocampal levels of BDNF [25], and increase the activity levels ofGSK3β [26].

Here we tested whether tau pathology could be induced or exacerbated by BFCN degeneration by performing BFCN lesions with the immunotoxin, p75-saporin, in pR5 tau transgenic mice. These tau transgenic mice express the human tau isoform containing the P301L familial frontotemporal dementia (FTD) mutation and develop tau hyperphosphorylation, NFTs and impaired spatial reference memory from 6 months of age [10 , 14].

MATERIALS AND METHODS

Animals

All animal experiments were performed within the guidelines of, and approved by, the institutional Animal Ethics Committee. The pR5 (P301L) tau transgenic mice were generated as previously described [12]. These mice harbor the longest human tau isoform with the hereditary FTD with parkinsonism (FTDP) P301L mutation under the control of the neuron-specific mouse Thy1.2 promoter [27]. Mice were housed on a 12-h dark/light cycle with water and standard chow available ad libitum. For our study, hemizygous transgenic males and females aged 3 and 7 months were used, with age- and gender-matched non-transgenic littermates used as controls. Results from males and females were quantified separately; however, as no significant differences were found between the sexes, data were pooled and further analyses were conducted as mixed gendercohorts.

Surgery

2-month-old male and female pR5 (P301L) mice were anesthetized by intraperitoneal (i.p.) injection of ketamine (100 mg/kg) and the muscle relaxant xylazine (10 mg/kg). Each mouse was then placed in a stereotaxic frame (David Kopf Instruments). A single infusion of murine p75-saporin (0.4 mg/ml; Advanced Targeting Systems) or control rabbit IgG-saporin (0.4 mg/ml) was administered using a 30 G needle attached to a 5 ml Hamilton syringe and pump (World Precision Instruments). The needle was lowered into the medial septum (A-P 0.9 mm; M-L 0 mm; D-V 4.2 mm from Bregma) according to coordinates in a mouse brain atlas, and the toxin was infused at a rate of 0.400 μl/min (1.5 μl total volume). The needle was then left in place for 5 min to allow for diffusion. Immediately after surgery and 24-h post-surgery, mice were injected subcutaneously with the analgesic torbugesic (2 mg/kg), and the antibiotic Baytril (5 mg/kg).

Tissue preparation

Mice were sacrificed 1 and 5 months after lesion surgery by cervical dislocation, and the brain was divided into three portions: the anterior portion of the brain containing the basal forebrain, and the remainder of the two hemispheres containing the hippocampus. The anterior portion was cut in the coronal plane and kept in 4% paraformaldehyde (PFA) for 3 h at room temperature, then post-fixed overnight at 4°C. After 24 h in phosphate-buffered saline (PBS) containing 30% sucrose, it was embedded in Frozen Section Compound (FSC22; Leica). Basal forebrain sections were cut in the coronal plane (40 μm) using a sliding microtome (SM2000r, Leica) and quantified for the number of choline acetyltransferase- (ChAT) and parvalbumin-positive neurons.

The remaining two portions were cut in half sagitally and the hippocampus and cortical tissue from one portion were dissected and snap-frozen in liquid nitrogen for sequential protein extraction and immunoblotting and enzyme-linked immunosorbent assay (ELISA). The remaining portion was drop fixed in 4% PFA for 24 h at 4°C, washed in PBS then embedded in paraffin wax for serial sectioning in the coronal plane (7 μm) using a rotary microtome. All sections used for comparative analysis were processed, stained and analyzed together.

Immunohistochemistry

For visualization of cholinergic neurons in the basal forebrain, free-floating sections were immunostained using a goat anti-ChAT antibody (1 : 1000; Millipore), biotinylated donkey anti-goat IgG (1 : 1000; Jackson Immunoresearch Laboratories) and avidin/biotin complex (ABC) reagent (Vector Elite kit: Vector Laboratories). ChAT-labeled cytoplasm and dendrites were revealed by a nickel-intensified diaminobenzidine (Ni-DAB) reaction that stained neurons black. Mounted sections were dehydrated and coverslipped with DePeX (Sigma-Aldrich).

Parvalbumin-positive neurons were identified using fluorescence immunohistochemistry. Briefly, free-floating sections were incubated overnight at room temperature in mouse anti- parvalbumin antibody (1 : 200; Millipore) followed by donkey anti-mouse Alexa 647 (1 : 1000; Invitrogen). They were then coverslipped with mounting medium for fluorescence (Dako). Fluorescence and bright field microscopy and image acquisition were performed using a fluorescence slide scanner (Axio, Zeiss) and a bright field slide scanner (Axio, Zeiss), respectively.

Quantification of ChAT- and parvalbumin-positive neurons

The numbers of ChAT- and parvalbumin-positive neurons in the basal forebrain were counted for each animal, as per Boskovic et al. [28]. Every third section (10 sections per animal) was counted starting from the beginning of the medial septum (1.18 mm anterior to Bregma). Identification of the medial septum/diagonal band of Broca areas was made using a mouse brain atlas. All measurements and analyses were performed using Imaris 7.2.3 software (Bitplane).

ELISAs

Hippocampal neurotrophin levels were measured by ELISA (Biosensis) in supernatant prepared from hippocampal homogenates. Soluble proteins were extracted using an acid extraction protocol as per the manufacturer’s instructions. Briefly, hippocampi were suspended in 20 volume/weight extraction buffer (0.05 M sodium acetate, 1 M sodium chloride, 1% Triton-X, Roche complete inhibitor cocktail tablet) and homogenized. A bicinchoninic acid assay (BCA; Thermo Scientific) to measure total protein content was performed to ensure that soluble lysate concentrations were within the assay range and could be standardized. An ELISA for BDNF was then performed according to the manufacturer’s instructions (Biosensis). The resulting measurements (pg) were normalized per mg of total solubleprotein.

Paraffin-embedded Ni-DAB immunohistochemistry for tau hyperphosphorylation

To visualize hyperphosphorylated tau epitopes in the hippocampus, paraffin-embedded sections were dewaxed through a series of xylene and ethanol incubations, then immunostained with various primary antibodies in a humid chamber: AT8 (1 : 500; Thermo Scientific), AT270 (1 : 500; Thermo Scientific), Tyr18 (1 : 500; MédiMabs), or Tau5 (1 : 500; Merck Millipore). Sections were then incubated in biotinylated donkey anti-mouse IgG (1 : 1000; Jackson Immunoresearch Laboratories) and ABC reagent (Vector Elite kit: Vector Laboratories). Tau- and phospho-tau-labeled cytoplasm and dendrites were revealed by a Ni-DAB reaction that stained neurons black. Mounted sections were dehydrated and coverslipped with DePeX (Sigma-Aldrich). Bright field microscopy and image acquisition were performed using a bright field slide scanner (Axio, Zeiss) and representative images were taken using Imaris 7.2.3 software (Bitplane).

Measurement of tau phosphorylation levels

To measure the levels of tau and hyperphosphorylated tau epitopes, soluble (RAB) and insoluble (RIPA) fractions of hippocampal homogenate were sequentially extracted as previously described [29]. Hippocampi were suspended in 10 volumes/weight ice-cold RAB buffer (0.01 M MES, 1 mM EGTA, 0.5 mM MgSO₄, 0.75 M NaCl, 0.02 M NaF, 1 mM Na₃VO₄, 1 mM PMSF) containing Complete EDTA-free Protease Inhibitor Cocktail (Roche) and PhosSTOP Phosphatase Inhibitor Cocktail (Roche), and homogenized. Samples were kept on ice for 30 min, and then centrifuged at 21,000 g for 90 min at 4°C. Supernatant was extracted and stored as the RAB fraction at –80°C. The remaining pellet was resuspended and homogenized in the same volume of ice-cold RIPA buffer (Cell Signaling) with 0.02 M NaF, 1 mM Na₃VO₄, 1 mM PMSF, Complete EDTA-free Protease Inhibitor Cocktail (Roche), and PhosSTOP Phosphatase Inhibitor Cocktail (Roche). Samples were allowed to stand on ice for 30 min before being centrifuged at 21,000 g for 90 min at 4°C. The supernatant was extracted and stored at –80°C as the RIPA fraction. A BCA assay (Thermo Scientific) to measure the total protein content for each sample from each fraction was performed to ensure that changes in tau were comparable across samples.

Total tau and hyperphosphorylated tau epitopes were quantified from the soluble and insoluble fractions of hippocampal homogenate by western blot analysis. Equal amounts of protein per mouse were separated on a NuPage Novex 4–12% Bis-Tris Protein Gel (Life Technologies), then transferred onto an Immobilon-FL transfer membrane (Millipore). In order to block non-specific binding sites, membranes were incubated in 5% bovine serum albumin in PBS with 0.1% Tween-20 (PBS-T) for 1 h at room temperature, after which they were incubated overnight at 4°C in the following antibodies: AT8 (1 : 1000; Thermo Scientific), AT100 (1 : 1000; Thermo Scientific), AT180 (1 : 1000; Thermo Scientific), AT270 (1 : 1000; Thermo Scientific), Tyr18 (1 : 5000; MédiMabs) or Tau5 (1 : 1000; Merck Millipore), and GAPDH (1 : 5000; Cell Signaling). Membranes were washed three times for 10 min in PBS-T then incubated with either anti-rabbit Alexa Fluor 680 (1 : 50,000; Invitrogen) or anti-mouse Alexa Fluor 800 (1 : 50,000; Invitrogen) antibodies for 2 h at room temperature. Membranes were washed thoroughly before protein bands were imaged using an Odyssey Imaging System (LI-COR Biosciences). Image Studio software (LI-COR Biosciences) was used for quantification of western blots.

Measurement of pTrkB/TrkB, pErk/Erk, pAkt/Akt, pGSK3β/GSK3β, synaptophysin, and PSD-95

Phosphorylated and total levels of TrkB, the signaling kinases Erk, Akt, and GSK3β, the presynaptic protein synaptophysin, and the postsynaptic density protein-95 (PSD-95) were quantified from the soluble fraction of hippocampal homogenates by western blot analysis as described above. The following primary antibodies were used: pTrkB (phospho S478; 1 : 1000; Biosensis), TrkB (1 : 1000; R&D Systems), pErk (1 : 1000; Cell Signaling), Erk (1 : 1000; Cell Signaling), pAkt (1 : 1000; Cell Signaling), Akt (1 : 1000; Cell Signaling), pGSK3β (Ser9) (1 : 1000; Cell Signaling), GSK3β (1 : 1000; Cell Signaling), synaptophysin (1 : 1000; Dako), PSD-95 (1 : 2000; NeuroMab), and the loading control GAPDH (1 : 4000; Cell Signaling). Imaging was performed as describedabove.

Morris water maze

Spatial memory was assessed using the Morris water maze as previously described [28]. 3-month-old pR5 tau transgenic mice (1 month post-BFCN lesion) were placed in a 100 cm diameter circular pool filled with opaque water. Visual cues were placed around the room for navigation and a small circular (10 cm diameter) platform was hidden 1.5 cm below the water surface. Mice were trained over a 5-day acquisition phase. On each day mice were released from three different start positions, with 20-min intervals between each training run. On the sixth day mice were subjected to a probe trial (in which the platform was removed) to test spatial memoryretention.

Latency was measured as the time from when the mouse was placed in the water until it had remained in the arena for a maximum of 60 s or until it had remained on the platform for a total of 10 s. Data were collected and analyzed using Ethovision XT (Noldus).

Statistics

All data are expressed as mean±SEM. Statistical tests, including appropriate post-hoc analysis, were performed for each result and are reported in the figure legends. The significance threshold was set at p < 0.05 and analyses were performed using Graphpad Prism 6.

RESULTS

p75-saporin injection selectively lesions basal forebrain cholinergic neurons

In order to induce cholinergic neuronal degeneration in 2-month-old pre-symptomatic tau pR5 (P301L) tau transgenic mice we used the toxin mu p75-saporin; p75^NTR expression is restricted to BFCNs and endocytosis of the saporin-receptor complex induces cell death. Control mice were injected with IgG-saporin. One month after injection, behavioral, histological, and biochemical analyses were performed. Stereotaxic injection of p75-saporin into the medial septal area of the basal forebrain resulted in a 60% decrease in the number of ChAT-positive neurons compared to that in the IgG-saporin injected mice (Fig. 1A, B). This lesion also caused denervation of the BFCN projections to the hippocampus (Fig. 1D). The number of parvalbumin-positive neurons in the basal forebrain was unchanged following p75-saporin injection compared to the control (Fig. 1C), confirming the specificity and extent of the lesion.

Decreased BDNF levels and signaling receptor phosphorylation in the hippocampus following cholinergic denervation

To determine whether cholinergic denervation caused a reduction in the level of BDNF in the hippocampus, we measured the amount of BDNF protein in hippocampal homogenates from p75-saporin- and IgG-saporin-injected pR5 tau transgenic mice, and their wildtype age- and gender-matched littermates. The IgG-saporin-injected animals had equivalent levels of BDNF protein levels to their wildtype littermates (Fig. 2A). In contrast, mice in which BFCNs were lesioned with p75-saporin had significantly reduced BDNF protein levels (Fig. 2A). This reduction was considered biologically significant as it was similar to that of mice with a heterozygous knockout of BDNF (BDNF^+/–) (Fig. 2A). In further support of this, the level of phosphorylated TrkB in the hippocampus of BFCN lesioned mice was significantly reduced compared to that of unlesioned transgenic controls. The levels of total TrkB protein remained unchanged (Fig. 2B, quantification in C).

No change in tau hyperphosphorylation in the hippocampus following cholinergic denervation

We next determined whether the cholinergic denervation precipitated the development of tau pathology. A range of tau phosphorylation epitopes that are considered to be markers of pathological hyperphosphorylation, and which are found in disease states, were used to immunostain the hippocampus of lesioned and control pR5 tau transgenic mice and probe western blots of hippocampal lysates. Although increased levels of phosphorylated tau were observed in the hippocampus of pR5 mice compared to wildtype mice (Fig. 3, column 1 compared to columns 2–5), no difference in the density or distribution patterns of immunostaining of hippocampal sections was observed when comparing 3-month-old lesioned and unlesioned pR5 mice using three different phosphorylated tau epitopes (Fig. 3, column 2 versus column 3). A similar profile was observed in the cortex (data not shown).

Moreover, no difference in the level of either total human tau or phosphorylated human tau at any of the phospho-tau epitopes was observed by western blotting, in either the soluble (Fig. 4A, quantified in C) or insoluble lysate fractions (Fig. 4B, quantified in D), between p75-saporin- and IgG-saporin-injected mice. This suggests that loss of cholinergic innervation to the hippocampus is not sufficient to induce (or reduce) hippocampal tau hyperphosphorylation, even in mice that are susceptible to developing pathogenic changes in tau.

Loss of basal forebrain cholinergic neurons in young pre-symptomatic mice does not affect tau hyperphosphorylation later in life

To determine whether a longer period of cholinergic denervation might exacerbate tau pathology in pR5 mice, we lesioned BFCNs of pre-symptomatic mice at 2 months of age, and then assessed the development of pathology 5 months later (in 7-month-old animals). The level of phosphorylation at hyperphosphorylation epitopes was again measured by immunohistochemistry (Fig. 3) and western blot analysis (Fig. 5) of hippocampal homogenates from lesioned and control mice. Consistent with the mice being of an age at which they develop frank tau pathology, all aged transgenic mice displayed increased levels of total tau as well as increased levels of hyperphosphorylated tau, particularly insoluble tau, across a range of hyperphosphorylation epitopes, compared to the levels seen in younger pR5 mice. However, similar to the results obtained in younger mice, BFCN lesioning did not change the level of total tau or phospho-tau observed in hippocampal sections (Fig. 3) or in soluble or insoluble lysate fractions (Fig. 5).

Reduction in hippocampal BDNF and cholinergic innervation does not affect synaptic function or cognition

To determine whether more subtle synaptic changes resulted from BFCN lesion, we measured the levels of a range of synaptic and trophic proteins. No differences in phosphorylated Erk1/2 or Akt (a kinase upstream of GSK3β), or any changes in total Erk1/2 or Akt were found by western blot when comparing levels in in hippocampal lysates of BFCN-lesioned and -unlesioned pR5 mice (Fig. 6A-D). Similarly, and consistent with the lack of change in phosphorylated Akt and phospho-tau, western blot analysis revealed no change in either phospho-GSK3beta levels or total GSK3β in the hippocampus of BFCN-lesioned mice compared to unlesioned mice (Fig. 6E, F). Furthermore, no difference in total synaptophysin levels or PSD-95 levels in the hippocampus was found (Fig. 6G-J).

Finally, to test whether the loss of BFCNs was sufficient to induce a deficit in cognitive function, 3-month-old lesioned and control pR5 tau transgenic mice, together with their age- and gender-matched wildtype littermates, were tested in the Morris water maze hippocampal-dependent learning and memory paradigm. The older cohort were not behaviorally tested as both lesioned and unlesioned pR5 animals were predicted to have deficits by 6 months of age [14]. Animals underwent a 5-day learning schedule, followed by a probe trial on the sixth day. No significant differences in escape latency were observed between any of the groups over the 5 days (Fig. 7A). In the probe trial, when the hidden platform was removed from the arena, time spent in the target quadrant, latency to the platform quadrant, and platform crossing frequency were assessed. Consistent with the evidence of robust learning by both groups of mice, no differences between lesioned and control mice were observed for these measures (Fig. 7B-D). This demonstrates that, despite changes in cholinergic innervation and reduced neurotrophin signaling, loss of BFCNs does not result in obvious impairment in allocentric navigation learning and memory. Lesioned and control mice traveled equivalent distances in the maze arena and no mouse displayed obvious motor difficulties (Fig. 7E).

DISCUSSION

BFCN loss is considered an early feature of AD. The degeneration correlates with Aβ burden in humans and can induce Aβ production and accumulation in genetically susceptible mouse models. The relationship between the second histological feature of AD, tau pathology, and BFCN degeneration is less well studied; however there are direct signaling pathways whereby BFCN loss, resulting in the downregulation of BDNF signaling, might accelerate tau hyperphosphorylation. Our results revealed that loss of BFCNs resulted in a decrease in hippocampal BDNF protein and TrkB receptor activity. However, trophic signaling more broadly, tau hyperphosphorylation in the hippocampus, and hippocampal-dependent learning and memory and synaptic integrity were unchanged by the lesion.

Our BFCN lesion was achieved by a direct medial septal injection of the p75-saporin immunotoxin, resulting in selective degeneration of BFCNs, with the extent of the lesion and the subtype specificity of the affected neurons being consistent with previous reports in which the toxin was injected bilaterally into the ventricles [17 –19]. In a number of these previous reports, BFCN lesioning was sufficient to induce specific cognitive deficits and accelerate Aβ pathology in AD mouse models [18, 30]. In our case, the lesion also reduced cholinergic innervation to the hippocampus and induced flow-on effects to neurotrophicsignaling. We therefore conclude that this model is suitable for testing whether BFCN degeneration can affect tau pathology.

Previous reports indicate that the medial septal BFCNs, which directly innervate the hippocampus, are responsible for the regulation of normal hippocampal function and excitability, including regulating BDNF gene expression [31, 32]. Gil-Bea et al. [25] further suggest that cholinergic hypofunction leads to impaired muscarinic signaling in the hippocampus that in turn decreases Arc and proBDNF levels. Consistent with this, several groups have previously reported reduced BDNF mRNA or protein in the hippocampus following a complete BFCN lesion in rats [33, 34]. It was therefore not surprising that cholinergic denervation induced a significant reduction in the levels of BDNF protein and TrkB receptor signaling in the pR5 mice, confirming the link between BFCN activity and neurotrophin availability in the hippocampus. Furthermore, this reduction was considered functionally significant, as the level of BDNF was similar to that of mice with only one copy of the BDNF gene.

BFCN loss has been linked with impaired learning and memory, particularly through studies in rats [35]. However, we have previously reported that, although BFCN lesioning in wildtype mice impairs uncued (idiothetic) maze tasks, it does not affect cued (allocentric) Morris water maze performance [30]. Similarly BDNF^+/– mice show no deficits in the Morris water maze [36]. We therefore did not expect a deficit in allocentric navigation due to the lesion alone. Rather, we hypothesized that the BFCN lesion would cause impairment in allocentric hippocampal-dependent memory due to a reduction in BDNF levels in combination with tau dysfunction in the hippocampus. However, no deficits in learning or memory were apparent in the hippocampal-dependent spatial navigation task following loss of BFCNs in pR5 mice in our study. This finding is consistent with the fact that we did not observe an increase in tau hyperphosphorylation epitopes or changes to markers of trophic signaling and synaptic function.

In contrast to our findings in the pR5 strain, Ramos-Rodriguez et al. [19] reported that 3-month-old APP/PS1 transgenic mice with BFCN lesions and increased Aβ burden (but not controls) were mildly impaired in the acquisition phase of the Morris water maze task. This supports our finding that loss of BFCNs alone is insufficient to induce allocentric deficits, but highlights that the combination of cholinergic hypofunction and exacerbated Aβ pathology appears to synergistically cause impairment in spatial working memory performance.

Our main hypothesis was that, like Aβ pathology, tau pathology would be exacerbated by BFCN degeneration. However, we found no evidence that a loss of BFCN innervation and the resulting decrease in BDNF protein in the hippocampus either induced or exacerbated the extent of tau hyperphosphorylation at a variety of epitopes. This was the case even when mice harbored BFCN lesions for the majority of their life and were assessed at an age at which NFT pathology was already present. Consistent with this, we saw no change in the activation of tau or upstream kinases. In contrast, Hawkes et al. [26] have reported that BFCN lesioning of adult male wildtype rats results in a transient decrease in the phosphorylation (increase in the activation) of GSK3β and a minor increase in tau phosphorylation at the AT270 residue in the hippocampus between 4 and 14 days post-lesion. It is possible that, at an intermediate time point between lesioning and analysis, tau phosphorylation (and GSK3β activity) was transiently increased in our animals. However, we reasoned that in the pR5 mice, in which the mutant human tau protein has the propensity to aggregate, such a transient increase in tau hyperphosphorylation should be sufficient to trigger a cascade effect resulting in NFT formation. This idea is not supported by our data. Rather, the significant reduction in BDNF and cholinergic innervation to the hippocampus were not sufficient to exacerbate tau hyperphosphorylation or aggregation in these tau transgenic mice. This could be because tau pathology occurs independent of cholinergic dysfunction, or alternatively could be due to compensatory mechanisms or the need for an additional molecular trigger.

In contrast, comparable BFCN lesions in APP/PS1 mice result in an exacerbation in amyloid pathology [18, 19]. Furthermore, Ramos-Rodriguez et al., [19] have demonstrated that tau hyperphosphorylation (but not NFT formation) is induced in the cortex following BFCN lesioning of symptomatic APP/PS1 mice. Based on our results, we suggest that the reported increase in tau hyperphosphorylation is likely to be a downstream consequence of the increased Aβ in BFCN-lesioned APP/PS1 mice rather than a direct result of the lesion.

Aβ oligomers are toxic [37, 38], and can cause mis-sorting of tau into dendrites, increased tau phosphorylation, and destabilization and damage of microtubules and spines [39]. A lack of neurotrophic activity to compensate for, or reverse, this effect could therefore trigger a disease cycle. Consequently,hyperphosphorylation of tau in the above APP models, and humans, may require two “hits”: BFCN/BDNF loss and the presence of Aβ oligomers. As rodent Aβ protein is structurally dissimilar to the human protein and does not form toxic aggregates, our pR5 transgenic model may not possess the required amyloid pathology for the ‘second hit’ to induce tau phosphorylation. This idea is supported by in vitro data where neurotrophin starvation of hippocampal neurons increased tau phosphorylation in a manner that was temporally related to increased Aβ generation [22]. We therefore suggest that, although BFCN degeneration and lowered BDNF levels are risk factors for exacerbated Aβ pathology in genetically susceptible animals [19, 40], and for cognitive decline in humans [4 , 41], AD-associated tau pathology is unlikely to be mediated directly by BFCN degeneration and/or the resultant decrease in neurotrophin levels, albeit that these features may be essential contributing factors to disease etiology.

Footnotes

ACKNOWLEDGMENTS

This work was supported by the National Health and Medical Research Council of Australia (Project Grant 1049236 to E.J.C). M.T.T. is the recipient of an Australian Postgraduate Award and Alzheimer’s Australia Dementia Research Foundation Top-Up scholarship. pR5 transgenic mice and antibodies were kindly provided by Prof. Jürgen Götz. Imaging work was performed in the Queensland Brain Institute’s Advanced Microscopy Facility and generously supported by the Australian Research Council Linkage Infrastructure, Equipment and Facilities Grant (ARC LIEF) LE100100074. We thank the staff of the University of Queensland Biological Resources Facility for breeding and maintaining the animals used in this study, Prof. Götz and members of the Coulson laboratory for helpful discussions, and Rowan Tweedale for editorial assistance.

Authors’ disclosures available online ().

References

Ginsberg

, Che

, Wuu

, Counts

, Mufson

(2006) Down regulation of trk but not p75^NTR gene expression in single cholinergic basal forebrain neurons mark the progression of Alzheimer’s disease. J Neurochem 97, 475–487.

Grothe

, Heinsen

, Teipel

(2012) Atrophy of the cholinergic basal forebrain over the adult age range and in early stages of Alzheimer’s disease. Biol Psychiatry 71, 805–813.

Beach

, Kuo

, Spiegel

, Emmerling

, Sue

, Kokjohn

, Roher

(2000) The cholinergic deficit coincides with Aβ deposition at the earliest histopathologic stages of Alzheimer disease. J Neuropathol Exp Neurol 59, 308–313.

Kerbler

, Fripp

, Rowe

, Villemagne

, Salvado

, Rose

, Coulson

, Alzheimer’s Disease Neuroimaging Initiative (2015) Basal forebrain atrophy correlates with amyloid b burden in Alzheimer’s disease. Neuroimage Clin 7, 105–113.

Potter

, Rauschkolb

, Pandya

, Sue

, Sabbagh

, Walker

, Beach

(2011) Pre- and post-synaptic cortical cholinergic deficits are proportional to amyloid plaque presence and density at preclinical stages of Alzheimer’s disease. Acta Neuropathol 122, 49–60.

Teipel

, Heinsen

, Amaro

Jr , Grinberg

, Krause

, Grothe

, Alzheimer’s Disease Neuroimaging Initiative (2014) Cholinergic basal forebrain atrophy predicts amyloid burden in Alzheimer’s disease. Neurobiol Aging 35, 482–491.

Contestabile

(2011) The history of the cholinergic hypothesis. Behav Brain Res 221, 334–340.

DeKosky

, Harbaugh

, Schmitt

, Bakay

, Chui

, Knopman

, Reeder

, Shetter

, Senter

, Markesbery

(1992) Cortical biopsy in Alzheimer’s disease: Diagnostic accuracy and neurochemical, neuropathological, and cognitive correlations. Intraventricular Bethanecol Study Group. Ann Neurol 32, 625–632.

Mesulam

(2004) The cholinergic lesion of Alzheimer’s disease: Pivotal factor or side show? Learn Mem 11, 43–49.

10.

Deters

, Ittner

, Gotz

(2008) Divergent phosphorylation pattern of tau in P301L tau transgenic mice. Eur J Neurosci 28, 137–147.

11.

Garcia-Alloza

, Robbins

, Zhang-Nunes

, Purcell

, Betensky

, Raju

, Prada

, Greenberg

, Bacskai

, Frosch

(2006) Characterization of amyloid deposition in the APPswe/PS1dE9 mouse model of Alzheimer disease. Neurobiol Dis 24, 516–524.

12.

Gotz

, Chen

, Barmettler

, Nitsch

(2001) Tau filament formation in transgenic mice expressing P301L tau. J Biol Chem 276, 529–534.

13.

Jankowsky

, Fadale

, Anderson

, Xu

, Gonzales

, Jenkins

, Copeland

, Lee

, Younkin

, Wagner

, Younkin

, Borchelt

(2004) Mutant presenilins specifically elevate the levels of the 42 residue β-amyloid peptide in vivo: Evidence for augmentation of a 42-specific g secretase. Hum Mol Genet 13, 159–170.

14.

Pennanen

, Wolfer

, Nitsch

, Gotz

(2006) Impaired spatial reference memory and increased exploratory behavior in P301L tau transgenic mice. Genes Brain Behav 5, 369–379.

15.

Nisbet

, Polanco

, Ittner

, Gotz

(2015) Tau aggregation and its interplay with amyloid-β. Acta Neuropathol 129, 207–220.

16.

Gil-Bea

, Gerenu

, Aisa

, Kirazov

, Schliebs

, Ramirez

(2012) Cholinergic denervation exacerbates amyloid pathology and induces hippocampal atrophy in Tg2576 mice. Neurobiol Dis 48, 439–446.

17.

Hartig

, Saul

, Kacza

, Grosche

, Goldhammer

, Michalski

, Wirths

(2014) Immunolesion-induced loss of cholinergic projection neurones promotes β-amyloidosis and tau hyperphosphorylation in the hippocampus of triple-transgenic mice. Neuropathol Appl Neurobiol 40, 106–120.

18.

Laursen

, Mork

, Plath

, Kristiansen

, Bastlund

(2013) Cholinergic degeneration is associated with increased plaque deposition and cognitive impairment in APPswe/PS1dE9 mice. Behav Brain Res 240, 146–152.

19.

Ramos-Rodriguez

, Pacheco-Herrero

, Thyssen

, Murillo-Carretero

, Berrocoso

, Spires-Jones

, Bacskai

, Garcia-Alloza

(2013) Rapid β-amyloid deposition and cognitive impairment after cholinergic denervation in APP/PS1 mice. J Neuropathol Exp Neurol 72, 272–285.

20.

Phillips

, Hains

, Armanini

, Laramee

, Johnson

, Winslow

(1991) BDNF mRNA is decreased in the hippocampus of individuals with Alzheimer’s disease. Neuron 7, 695–702.

21.

Laske

, Stellos

, Hoffmann

, Stransky

, Straten

, Eschweiler

, Leyhe

(2011) Higher BDNF serum levels predict slower cognitive decline in Alzheimer’s disease patients. Int J Neuropsychopharmacol 14, 399–404.

22.

Amadoro

, Corsetti

, Ciotti

, Florenzano

, Capsoni

, Amato

, Calissano

(2011) Endogenous Aβ causes cell death via early tau hyperphosphorylation. Neurobiol Aging 32, 969–990.

23.

Mai

, Jope

, Li

(2002) BDNF-mediated signal transduction is modulated by GSK3β and mood stabilizing agents. J Neurochem 82, 75–83.

24.

Sperber

, Leight

, Goedert

, Lee

(1995) Glycogen synthase kinase-3 β phosphorylates tau protein at multiple sites in intact cells. Neurosci Lett 197, 149–153.

25.

Gil-Bea

, Solas

, Mateos

, Winblad

, Ramirez

, Cedazo-Minguez

(2011) Cholinergic hypofunction impairs memory acquisition possibly through hippocampal Arc and BDNF downregulation. Hippocampus 21, 999–1009.

26.

Hawkes

, Jhamandas

, Kar

(2005) Selective loss of basal forebrain cholinergic neurons by 192 IgG-saporin is associated with decreased phosphorylation of Ser glycogen synthase kinase-3beta. J Neurochem 95, 263–272.

27.

Luthi

, Van der Putten

, Botteri

, Mansuy

, Meins

, Frey

, Sansig

, Portet

, Schmutz

, Schroder

, Nitsch

, Laurent

, Monard

(1997) Endogenous serine protease inhibitor modulates epileptic activity and hippocampal long-term potentiation. J Neurosci 17, 4688–4699.

28.

Boskovic

, Alfonsi

, Rumballe

, Fonseka

, Windels

, Coulson

(2014) The role of p75^NTR in cholinergic basal forebrain structure and function. J Neurosci 34, 13033–13038.

29.

Probst

, Gotz

, Wiederhold

, Tolnay

, Mistl

, Jaton

, Hong

, Ishihara

, Lee

, Trojanowski

, Jakes

, Crowther

, Spillantini

, Burki

, Goedert

(2000) Axonopathy and amyotrophy in mice transgenic for human four-repeat tau protein. Acta Neuropathol 99, 469–481.

30.

Hamlin

, Windels

, Boskovic

, Sah

, Coulson

(2013) Lesions of the basal forebrain cholinergic system in mice disrupt idiothetic navigation. PLoS One 8, e53472.

31.

Lindefors

, Ernfors

, Falkenberg

, Persson

(1992) Septal cholinergic afferents regulate expression of brain-derived neurotrophic factor and beta-nerve growth factor mRNA in rat hippocampus. Exp Brain Res 88, 78–90.

32.

Lapchak

, Araujo

, Hefti

(1993) Cholinergic regulation of hippocampal brain-derived neurotrophic factor mRNA expression: Evidence from lesion and chronic cholinergic drug treatment studies. Neuroscience 52, 575–585.

33.

Ferencz

, Kokaia

, Keep

, Elmer

, Metsis

, Kokaia

, Lindvall

(1997) Effects of cholinergic denervation on seizure development and neurotrophin messenger RNA regulation in rapid hippocampal kindling. Neuroscience 80, 389–399.

34.

Kokaia

, Ferencz

, Leanza

, Elmer

, Metsis

, Kokaia

, Wiley

, Lindvall

(1996) Immunolesioning of basal forebrain cholinergic neurons facilitates hippocampal kindling and perturbs neurotrophin messenger RNA regulation. Neuroscience 70, 313–327.

35.

Baxter

, Bucci

, Gorman

, Wiley

, Gallagher

(2013) Selective immunotoxic lesions of basal forebrain cholinergic cells: Effects on learning and memory in rats. Behav Neurosci 127, 619–627.

36.

Montkowski

, Holsboer

(1997) Intact spatial learning and memory in transgenic mice with reduced BDNF. Neuroreport 8, 779–782.

37.

Dahlgren

, Manelli

, Stine

Jr , Baker

, Krafft

, LaDu

(2002) Oligomeric and fibrillar species of amyloid-β peptides differentially affect neuronal viability. J Biol Chem 277, 32046–32053.

38.

Benilova

, Karran

, De Strooper

(2012) The toxic Aβ oligomer and Alzheimer’s disease: An emperor in need of clothes. Nat Neurosci 15, 349–357.

39.

Zempel

, Thies

, Mandelkow

(2010) Abeta oligomers cause localized Ca²⁺ elevation, missorting of endogenous Tau into dendrites, Tau phosphorylation, and destruction of microtubules and spines. J Neurosci 30, 11938–11950.

40.

Laursen

, Mork

, Plath

, Kristiansen

, Bastlund

(2014) Impaired hippocampal acetylcholine release parallels spatial memory deficits in Tg2576 mice subjected to basal forebrain cholinergic degeneration. Brain Res 1543, 253–262.

41.

Grothe

, Zaborszky

, Atienza

, Gil-Neciga

, Rodriguez-Romero

, Teipel

, Amunts

, Suarez-Gonzalez

, Cantero

(2010) Reduction of basal forebrain cholinergic system parallels cognitive impairment in patients at high risk of developing Alzheimer’s disease. Cereb Cortex 20, 1685–1695.