- Email: [email protected]
presenilin-1 transgenic mice
Neuroscience 125 (2004) 1009 –1017
CENTRAL CHOLINERGIC FUNCTIONS IN HUMAN AMYLOID PRECURSOR PROTEIN KNOCK-IN/PRESENILIN-1 TRANSGENIC MICE J. HARTMANN,a,b C. ERB,c U. EBERT,c ¨ NIGc1 AND K. H. BAUMANN,c A. POPP,c G. KO a,b J. KLEIN *
Key words: acetylcholine, amyloid precursor protein, cholinergic system, microdialysis, presenilin-1, transgenic mice.
a Department of Pharmacology, Johannes Gutenberg University of Mainz, D-55101 Mainz, Germany
Alzheimer’s disease (AD), the most frequent type of dementia in humans, is characterized by the formation of plaque-forming amyloid peptides, neurofibrillary tangles, degeneration of neurons and cognitive loss (Cummings et al., 1998; Selkoe, 2001; Nussbaum and Ellis, 2003). Neuronal degeneration strongly affects cholinergic neurons in the basal forebrain which project to hippocampal and cortical areas and which are required for attention, spatial orientation, learning and memory, i.e. functions that are typically lost in AD (Collerton, 1986). The “cholinergic hypothesis of senile dementia” postulated that the symptoms of AD can be explained by central cholinergic dysfunction (Bartus et al., 1982), and drug development for AD focused on drugs which enhance cholinergic function such as inhibitors of acetylcholinesterase (AChE; Cummings et al., 1998; Gauthier, 2002; Palmer, 2002). Clinical studies have indeed shown that central cholinergic dysfunction correlates with the severity of AD (Bierer et al., 1995; Francis et al., 1999). However, findings that cholinergic dysfunction occurs early in the disease have been challenged (Davis et al., 1999; Gilmor et al., 1999). While many authors believe that a cholinergic intervention is required to successfully treat AD (Winkler et al., 1998), the cholinergic deficit is no longer assumed to be the primary factor of disease etiology. Instead, largely based on genetic and biochemical evidence, many workers in the field now support the “amyloid hypothesis” of AD. This hypothesis postulates that abnormal processing of the amyloid precursor protein (APP) by - and ␥-secretases causes buildup of amyloid peptides which may be toxic in monomeric or oligomeric form or after deposition as amyloid plaques (Selkoe, 2002; Hardy and Selkoe, 2002). While various toxic properties of amyloid peptides have been described (Mattson, 1997; Small et al., 2001), recent work suggested that amyloid peptides are most damaging to neurons when they form oligomeric assemblies (“protofibrils”; Klein et al., 2001; Hardy and Selkoe, 2002). As the amyloid hypothesis postulates that neuronal damage follows amyloid formation, it implicitly assumes that cholinergic neurons are particularly susceptible to amyloid toxicity. It should be noted that the amyloid hypothesis has also been challenged by claims that amyloid peptides are not the causal factor of AD-related neurodegeneration (Neve and Robakis, 1998; Joseph et al., 2001). Despite the popularity of the cholinergic and amyloid hypotheses, a limited number of studies were concerned with the relationship between amyloid deposition and
b Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Science Center, 1300 Coulter Drive, Amarillo, TX 79106, USA c
Alzheimer Research Group, Bayer Health Care AG, D-42096 Wuppertal, Germany
Abstract—Alzheimer’s disease is characterized by amyloid peptide formation and deposition, neurofibrillary tangles, central cholinergic dysfunction, and dementia; however, the relationship between these parameters is not well understood. We studied the effect of amyloid peptide formation and deposition on central cholinergic function in knock-in mice carrying the human amyloid precursor protein (APP) gene with the Swedish/London double mutation (APP-SL mice) which were crossbred with transgenic mice overexpressing normal (PS1wt) or mutated (M146L; PS1mut) human presenilin-1. APP-SLⴛPS1mut mice had increased levels of A peptides at 10 months of age and amyloid plaques at 14 months of age while APP-SLⴛPS1wt mice did not have increased peptide levels and did not develop amyloid plaques. We used microdialysis in 15–27 months old mice to compare hippocampal acetylcholine (ACh) levels in the two mouse lines and found that extracellular ACh levels were slightly but significantly reduced in the APP-SLⴛPS1mut mice (ⴚ26%; Pⴝ0.044). Exploratory activity in the open field increased hippocampal ACh release by two-fold in both mouse lines; total and relative increases were not significantly different for the two strains under study. Similarly, infusion of scopolamine (1 M) increased hippocampal ACh release to a similar extent (3–5-fold) in both groups. High-affinity choline uptake, a measure of the ACh turnover rate, was identical in both mouse lines. Neurons expressing choline acetyltransferase were increased in the septum of APP-SLⴛPS1mut mice (ⴙ26%; Pⴝ0.046). We conclude that amyloid peptide production causes a small decrease of extracellular ACh levels. The deposition of amyloid plaques, however, does not impair stimulated ACh release and proceeds without major changes of central cholinergic function. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. 1
Present address: Fidelity Investments, Watertown, MA, USA. *Correspondence to: Tel: ⫹1-806-356-4015x252; fax: ⫹1-806-3564034. E-mail address: [email protected] (J. Klein). Abbreviations: ACh, acetylcholine; AChE, acetylcholinesterase; aCSF, artificial cerebrospinal fluid; AD, Alzheimer’s disease; APP, amyloid precursor protein; APP-SL, mice carrying the human APP gene with the Swedish/London double mutation; AUC, area under the curve; ChAT, choline acetyltransferase; HACU, high-affinity choline uptake; PS-1, presenilin-1; PS1wt, mice overexpressing normal human presenilin-1; PS1mut, mice overexpressing mutated (M146L) human presenilin-1.
0306-4522/04$30.00⫹0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.02.038
1009
1010
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
cholinergic dysfunction (Auld et al., 2002). In vitro, soluble amyloid peptides were found to reduce synaptosomal choline uptake and release of acetylcholine (ACh) in brain slices; these effects were observed at low (nanomolar) concentrations which likely occur in vivo (Kar et al., 1996, 1998). Prolonged treatment with higher doses of amyloid peptides reduced ACh synthesis and impaired muscarinic receptor signaling in neuronal cultures (Kelly et al., 1996; Pedersen et al., 1996). Amyloid effects were also tested in in vivo-models of AD. Mouse models of AD express or overexpress wild-type or mutant forms of APP and of presenilin-1, a subunit of the ␥-secretase complex (Kimberly et al., 2003) which is mutated in most familial types of AD. Some evidence for amyloid-induced cholinergic degeneration in these mice was reported. For instance, the size of septal cholinergic fibers was reduced and fibers were rearranged in one mouse strain (Bronfman et al., 2000) while modest decreases of cholinergic enzymes and more clear-cut decreases of cholinergic fiber length, but no loss of basal forebrain neurons, were observed in the APP23 mouse (Boncristiano et al., 2002). In the Tg2576 mouse line, markers of cholinergic function (choline acetyltransferase, vesicular ACh transporter, high-affinity choline uptake) were found to be unchanged despite increased amyloid deposition during aging (Gau et al., 2002). In contrast, other investigators reported a reduction of high-affinity choline uptake (HACU) and of muscarinic and nicotinic receptors in aged (21 month old) Tg2576 mice which was accompanied by an increase of the vesicular ACh transporter (Apelt et al., 2002; Klingner et al., 2003). Finally, doubly transgenic mice overexpressing human APP and presenilin mutations displayed normal forebrain cholinergic cell densities but a diminution of cholinergic synapse density and signs of aberrant sprouting in cortex and hippocampus (Wong et al., 1999; Hernandez et al., 2001; Hu et al., 2003). Taken together, these findings provided anatomical and neurochemical evidence for some cholinergic dysfunction secondary to amyloid formation but the functional consequences of the observed changes for cholinergic neurotransmission remained unclear. The present study attempts to improve our understanding of amyloid– cholinergic interactions by characterizing presynaptic parameters of cholinergic function using in vivo-microdialysis in mouse brain. For this study, we used knock-in mice carrying the human APP gene with the Swedish and London mutation (APP-SL) which were cross-bred with transgenic mice overexpressing human wild-type or mutated (M146L) presenilin1. Only the mice expressing the mutant presenilin developed amyloid plaques in the brain. We characterized the activity of the septohippocampal pathway of these mouse lines and report that central cholinergic functions in these mice, while showing minor changes, are essentially preserved despite age-associated amyloid peptide formation and deposition.
EXPERIMENTAL PROCEDURES Animals We used two types of mice for the present experiments. APP-SL mice were produced at Bayer Health Care AG (Wuppertal, Germany) and are knock-in mice in which the murine gene for APP has been replaced with the human gene carrying the LondonSwedish mutations (K670N/M671L; V717F). APP-SL⫻transgenic mice overexpressing the human wild-type presenilin-1 gene (PS1wt) are mice produced by breeding APP-SL mice with PS1wt mice (Duff et al., 1996). APP-SL⫻transgenic mice overexpressing the mutated (M146L) human presenilin-1 gene (PS1mut) are mice produced by breeding APP-SL mice (Duff et al., 1996). Genotypes were confirmed by PCR analysis of tail DNA in all cases. The experimental procedures used in this study met the guidelines of and were approved by the responsible government agency (Bezirksregierung Rheinland-Pfalz). Every attempt was made to reduce the number of animals used and their suffering.
Determination of amyloid peptides and amyloid plaques Mice of different age (3–29 months) were decapitated, and the brains removed. The cortices of one hemisphere were dissected, and a one-step formic acid extraction was performed as described (Kawarabayashi et al., 2001). The concentrations of amyloid peptides (total A and A40) were determined by means of specific antibodies using a liquid phase electrochemo-luminescence assay (Khorkova et al., 1998). The other hemisphere was fixated in 4% phosphate-buffered paraformaldehyde solution, cut into 40 m transverse sections on a cryomicrotome and stained for solid plaques with thioflavine S and for amyloid plaques using the 4G8 antibody (Wisniewski et al., 1989). The size, density and area covered by amyloid plaques were determined by quantitative analysis under the microscope.
Choline acetyltransferase (ChAT) immunohistochemistry Mice were perfused transcardially with 4% phosphate-buffered paraformaldehyde solution (pH⫽7.4), brains were removed and cut into 40 m sections on a cryomicrotome. Sections were incubated with hydrogen peroxide (3%) for 30 min followed by blocking solution containing donkey serum and bovine serum albumin. Sections were incubated overnight with the primary ChAT goat– anti-human antibody (1:50; Chemicon). Visualization was carried out using biotinylated donkey–anti-goat antibody (1:1000; Dianova) and streptavidin-coupled horseradish peroxidase with consecutive diaminobenzidine incubation. ChAT-positive neurons were counted in three sections per animal in the septum and diagonal band of Broca (coordinates from bregma: ⫹1.1 mm; ⫹0.74 mm and ⫹0.5 mm). The number of ChAT-positive neurons was determined in 10 randomly positioned squares (0.01 m2) by an analysis system (KS400; Zeiss, Germany). The data were double-checked visually by a technician. Densities of ChATpositive neurons were calculated as averages of three sections per mm3.
Microdialysis Mice were anesthetized with halothane (1–3%) and placed in a stereotactic frame. I-shaped, miniature, concentric dialysis probes with an exchange length of 1 mm were constructed as previously described (Erb et al., 2001; Kopf et al., 2001) and implanted in the right ventral mouse hippocampus using the following coordinates (from bregma): AP ⫺2.0 mm; L ⫺1.8 mm; DV ⫺2.3 mm. The animals were allowed to recover overnight, and experiments were carried out on 2 consecutive days after probe implantation in freely moving animals.
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017 On the experimental days, the microdialysis probes were perfused with artificial cerebrospinal fluid (aCSF; 147 mM NaCl, 4 mM KCl, 1.2 mM CaCl2 and 1.2 mM MgCl2) containing the AChE inhibitor neostigmine (1 M) at a perfusion rate of 1 l/min. On day 1, efflux from the microdialysis probe was sampled in 15 min intervals to obtain baseline values of ACh. After 90 min, the mice were placed into a novel environment (“open field”), and microdialysis was continued. The testing chamber was a gray plastic container (45⫻32⫻20 cm), and each mouse was allowed to explore the chamber during the time of the experiment (90 min). Subsequently, the mice were returned to their home cage, and ACh efflux was monitored for another 90 min. On day 2, baseline values of ACh were again sampled for 90 min. Subsequently, the dialysis fluid (aCSF) was switched to aCSF containing 1 M of scopolamine bromide while the mice remained in the home cage. After 90 min, the perfusion fluid was switched back to aCSF, and the mice were monitored for a further 90 min. On day 3, the mice were killed, and the hemispheres (cortex and hippocampus) were used to measure HACU. The correct location of the microdialysis probe was also verified during this procedure.
Determination of ACh ACh in dialysates was determined by microbore HPLC using a metal-free system which consisted of a low-speed pump (BAS PM80), separation column (SepStik; 530⫻1 mm), enzyme reactor (50⫻1 mm) carrying immobilized AChE and choline oxidase and electrochemical detector (BAS LC-4C) with a platinum electrode operating at 0.5 V. The flow rate was 120 l/min. The retention time for ACh was 10.3 min. At an injection volume of 5 l, the detection limit of this system was 10 fmol/injection.
Microdialysis data analysis ACh efflux data were recorded for each individual mouse as fmol per 15 min time interval during rest, behavioral stimulation in the open field and scopolamine infusion. Areas under the curves (AUC) were calculated by subtracting baseline ACh levels from levels measured during the 90 min of stimulation and were used to correlate the effects of the open field and scopolamine with animal age and with plaque density (GraphPad Prism).
HACU HACU was determined in synaptosomal (P2) fractions obtained from cortex and hippocampus as described (Erb et al., 2001). Briefly, brain tissue was homogenized in isotonic sucrose solution containing 10 mM HEPES buffer and centrifuged at 1000⫻g at 4 °C for 10 min. The supernatant was again centrifuged at 17,000⫻g for 10 min, and the resulting P2 pellet was washed and used for HACU determinations. For this purpose, aliquots were incubated in Krebs–Ringer buffer at 37 °C with 50 nM [3H]-choline (DuPont-NEN, Bad Nauheim, Germany; specific activity 50 Ci/ mmol) at 30 °C for exactly 5 min, in the presence or absence of hemicholinium-3 (HC-3, 1 M). Choline uptake was stopped by addition of ice-cold buffer. After centrifugation, the pellet was washed three times with HC-3 containing buffer, and radioactivity associated with the pellet was determined by liquid scintillation counting. HACU is equivalent to the HC-3-inhibitable part of choline uptake and was expressed as [3H]-uptake (d.p.m./g protein). Protein was determined by the Lowry assay.
RESULTS Animals, amyloid peptides and amyloid plaques APP-SL knock-in mice were phenotypically normal and fertile but did not develop amyloid plaques during aging as
1011
Fig. 1. (A) APP-SL⫻PS1wt and (B) APP-SL⫻PS1mut mouse forebrain stained immunohistochemically with antibody 4G8 directed against -amyloid peptide. The transverse sections were taken from 26 month old mice.
analyzed by immunohistology and Thioflavin-S staining (not illustrated). Similarly, the APP-SL knock-in mice which were crossbred with mice overexpressing human presenilin1 (APP-SL⫻PS1wt mice) did not develop measurable A peptide levels or amyloid plaques up to the age of 26 months (Fig. 1A). In contrast, APP-SL knock-in mice carrying the mutated human PS-1 gene (APP-SL⫻PS1mut mice) had increased levels of A peptides at 10 months of age (Fig. 2A), and all of the mice developed amyloid plaques (Fig. 1B). In this mouse line, total A peptides increased strongly with age reaching more than 10 nmol/g wet weight at 20 months. Only a small percentage (about 10%) of these peptides were A1– 40 (Fig. 2A) whereas a preliminary analysis at 16 months of age showed three-fold higher A1– 42 levels compared with A1– 40 (data not shown). Amyloid plaque deposition in APP-SL⫻PS1mut mice was measurable at 14 months of age, and plaque load increased steadily with age to 400 plaques per cortical section at 26 months (Fig. 2B). The distribution of amyloid plaques in these animals resembled that seen in Alzheimer’s disease, with high densities in the parietal cortex and in the hippocampal CA1 region (approximately 3 plaques per mm2) and moderate plaque accumulation in the dentate gyrus and in piriform cortex (approximately one plaque per mm2; Fig. 1B). Microdialysis: basal concentrations of ACh For in vivo-microdialysis, we used 11 aged mice (eight male, three female; weight 36.2⫾6.0 g) of the APPSL⫻PS1wt line (656 –919 days of age; average age 797 days) and 11 mice (nine male, two female; weight 40.2⫾10.1 g) of the APP-SL⫻PS1mut strain (441– 826 days of age; average days 601 days). As illustrated in Fig. 3A, basal efflux of ACh, reflecting the extracellular concentration of ACh in the hippocampus of awake animals, was 65.4⫾6.2 fmol/5 min in APP-SL⫻PS1wt mice. The corresponding value for APP-SL⫻PS1mut mice (46.3⫾6.4 fmol/5 min) was slightly lower (by 26%; P⫽0.044).
1012
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
Fig. 2. (A) Levels of soluble A peptides and A1– 40 in APPSL⫻PS1mut mice of various age. APP-SL⫻PS1wt mice did not have measurable levels of A peptides. (B) Number of amyloid plaques as detected by the 4G8 antibody in corresponding transverse sections of cortex and hippocampus of APP-SL⫻PS1mut mice of various age. No plaques could be detected in the brains of APP-SL⫻PS1wt mice. Data are means⫾S.D. from seven to 10 mice at each age group.
Microdialysis: effect of behavioral stimulation Exploration of a new environment is known to stimulate hippocampal ACh release (Thiel et al., 1998; Giovannini et al., 2001). When the doubly transgenic mice were exposed to a novel environment (“open field”) hippocampal ACh release was increased in all animals by approximately two-fold (Fig. 4A). The absolute increases of ACh efflux within 15 min of exploration were similar in APP-SL⫻PS1wt mice (⫹58.8 fmol) and APPSL⫻PS1mut mice (⫹61.9 fmol). In relative terms, maximum increases were identical for APP-SL⫻PS1wt mice (⫹134.6% of baseline ACh efflux) and APP-SL⫻PS1mut mice (⫹132.7%); however, the increase of ACh efflux was longer-lasting in APP-SL⫻PS1wt mice, an observation which is reflected in the calculation of AUC over 180 min: the average AUC value for APP-SL⫻PS1wt was 421⫾136 fmol⫻h, for APP-SL⫻PS1mut 216⫾63 fmol⫻h (P⫽0.18). If the relative increases were calculated in percentages, and AUC values taken from percentage values, then AUCs of APP-SL⫻PS1mut mice (526⫾85 AU) were only slightly lower than AUCs of APP-SL⫻PS1wt mice (699⫾211 AU).
Fig. 3. Cholinergic parameters of APP-SL⫻PS1wt and APPSL⫻PS1mut mice. (A) Basal release of ACh in mouse hippocampus (* P⫽0.044; N⫽11 each). (B) HACU measured in cortical and hippocampal homogenates (P2 fraction; N⫽8 each). (C) Density of septal neurons stained positive for ChAT. (* P⫽0.046; N⫽9 each). Data are means⫾S.E.M. of eight to 11 experiments in mice of 15–27 months of age (see Results).
Microdialysis: effect of pharmacological stimulation Scopolamine blocks presynaptic muscarinic receptors on cholinergic nerve endings and thereby disrupts the negative feedback that ACh exerts on its own release (Erb et al., 2001). When scopolamine was infused into the hippocampus, a several-fold increase of ACh release was observed in both groups of mice over a period of 90 min (Fig. 4B). Maximum increases of ACh efflux were observed after 60 –75 min and were ⫹227.5 fmol in APP-SL⫻PS1wt mice and ⫹180.1 fmol in APPSL⫻PS1mut mice. In relative terms, maximum increases from baseline efflux were ⫹289% for APPSL⫻PS1wt mice and ⫹375% for APP-SL⫻PS1mut mice. Calculations of the AUC did not reveal significant differences between the groups; the data for APPSL⫻PS1wt mice (1195⫾314 fmol⫻h) and APPSL⫻PS1mut mice (1146⫾261 fmol⫻h) were, for all practical purposes, identical.
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
Fig. 4. Stimulated release of ACh in mouse hippocampi measured by microdialysis. (A) Changes of ACh efflux upon behavioral activation (exposure to a novel environment) for 90 min, followed by return to the home cage. (B) Changes of ACh efflux upon infusion of scopolamine (1 M) for 90 min, followed again by artificial CSF. PSwt, APPSL⫻PS1wt mice. PSmut, APP-SL⫻PS1mut mice. Data are means⫾S.E.M. of 11 experiments in mice of 15–27 months of age (see Results).
Correlations To investigate possible relationships between age, solid (Thioflavin S positive) amyloid plaque formation, and ACh release, we correlated basal ACh release data of plaquebearing APP-SL⫻PS1mut mice with age and amyloid plaque density of individual animals. As shown in Fig. 5, basal ACh efflux showed a tendency to increase with age and plaque burden, but the correlation was not significant. In addition, no correlation was found between age and stimulated ACh release in APP-SL⫻PS1mut mice, and of age with basal or stimulated ACh release in APPSL⫻PS1wt mice (data not shown). HACU HACU, measured ex vivo, reflects the turnover of ACh in vivo (Simon et al., 1976). In the present study, the hemicholinium-sensitive uptake of [3H]-choline was 40.2⫾3.6 d.p.m./g protein in APP-SL⫻PS1wt mice (uptake within 5 min; equivalent to 4.3 pmol/h/mg protein). The corresponding value for APP-SL⫻PS1mut mice was 37.0⫾5.3
1013
Fig. 5. Correlations of data obtained for APP-SL⫻PS1mut mice. (A) Correlation of mouse age with basal release of ACh (N⫽11). (B) Correlation of solid plaque density with basal release of ACh. Note that the data on plaque density were not available for all animals in the study (N⫽9). For (A) and (B), the slopes of the correlation lines (calculated by linear regression) were not significantly different from zero (GraphPad Prism software).
d.p.m./g protein. Thus, the data did not reveal a significant difference of ACh turnover between the two mouse lines (P⬎0.5; Fig. 3B). Cholinergic cell number We stained the septal area, location of the cholinergic cell bodies which project to the hippocampus, with an antibody directed against ChAT, a specific marker of cholinergic neurons. Quantitative image analysis demonstrated that the plaque-bearing APP-SL⫻PS1mut mouse line had a higher number of septal ChAT-positive neurons (by 26%) than the plaque-free APP-SL⫻PS1wt mice (P⫽0.046; Fig. 3C). Nissl stains did not reveal visible changes of neuronal morphology, density or size of nuclei in transgenic lines (not illustrated).
DISCUSSION Most of the pathological features of AD can now be mimicked by individual animal models which reproduce the formation of amyloid plaques and neurofibrillary tangles, central cholinergic dysfunction, and neurodegeneration, respectively (Auld et al., 2002; Hock and Lamb, 2001; Chapman et al., 2001; Hernandez et al., 2002); however,
1014
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
the interrelationships between these pathologies remain poorly understood. Age-dependent formation of amyloid plaques has been observed in transgenic mouse lines overexpressing wild-type or mutated forms of APP coupled to strong promoters (Games et al., 1995; Hsiao et al., 1996; Sturchler-Pierrat et al., 1997). Amyloid peptide formation and plaque deposition can be enhanced by crossbreeding APP-transgenic mice with mice overexpressing presenilin-1 (Borchelt et al., 1996, 1997; Citron et al., 1997; Holcomb et al., 1998), an observation which is likely explained by presenilin’s role in APP processing (Kimberly et al., 2003). However, amyloid-bearing mice show no or only moderate signs of neuronal (especially cholinergic) dysfunction. While several of the transgenic AD models displayed altered long-term potentiation and cognitive dysfunction with aging, the relationship of these observations to a possible cholinergic degeneration remained unknown (Hock and Lamb, 2001; Chapman et al., 2001). Loss of neurons or synapses, if present, was moderate in all mouse lines, and severe and selective neurodegeneration as observed in human AD was not present. As for cholinergic systems, some work indicated that altered cholinergic activity may influence processing of APP (Roberson and Harrell, 1997; Hellstro¨m-Lindahl, 2000). The amyloid hypothesis, however, postulates that amyloid peptide formation and/or deposition precedes cholinergic dysfunction; thus, ongoing formation of amyloid peptides should cause overt cholinergic dysfunction during aging in murine Alzheimer models. In the present study, we developed a novel mouse model of AD which appeared to be particularly well suited to test this hypothesis. We first created a knock-in APP-SL mouse which did not develop amyloid plaques. We then crossbred these mice with PS1wt or PS1mut (Duff et al., 1996). We found that APP-SL⫻PS1wt mice did not form amyloid peptides or develop amyloid plaques with age. In contrast, the APP-SL⫻PS1mut line had elevated amyloid peptide levels in the brain at an age of 10 months, followed by an increasing deposition of amyloid plaques at 14 months of age. Both amyloid peptide and plaque formation increased with age and reached high levels at ⬎18 months of age (Figs. 1 and 2). Plaque depositions were observed in every individual animal and were limited to the brain areas (cortex, hippocampus) which are most severely affected in human AD. The availability of these mice offered the opportunity to study the influence of amyloid deposition on cholinergic parameters in two groups of mice which are genetically identical except for the mutation in presenilin-1. With respect to the extent of amyloid formation, our mice resemble the well-characterized Tg2576 line although our mice develop amyloid plaques somewhat later (at 15 vs. 10 months of age) and at a slightly lower density than Tg2576 mice (Hsiao et al., 1996; Kawarabayashi et al., 2001). Both mouse lines fail to develop neurofibrillary tangles and do not show overt neuronal degeneration. The main goal of the present study was to obtain in vivo-data of cholinergic activity in aged transgenic mice with amyloid depositions. The method of choice for the investigation of cholinergic activity in awake animals is
microdialysis. We have adapted the microdialysis procedure for the present study to follow neuronal activity (ACh release) in the septohippocampal pathway, a cholinergic pathway which is typically affected in AD. Our first finding was that the extracellular ACh levels in APP-SL⫻PS1mut mice were 26% lower than in APP-SL⫻PS1wt mice (P⫽0.044; Fig. 3A). This finding is compatible with a lower neuronal activity of the cholinergic fibers, or a loss of synapses induced by formation of amyloid peptides and plaques. However, this interpretation was challenged by the following experiments. We tested to what extent the septohippocampal pathway was capable of responding to behavioral and pharmacological manipulations. Exposure of rodents to a novel environment (“open field”) causes exploratory behavior which was previously shown to be associated with an increase of hippocampal ACh release in rats (Thiel et al., 1998; Giovannini et al., 2001) and mice (Kopf et al., 2001). The present results demonstrate that both groups of mice were able to activate the septohippocampal pathway during exploration to a similar extent. The maximum increase of ACh release during exploration was identical in both lines, whether expressed in absolute (fmol/min) or relative terms (% increase of baseline). The duration of the effect, however, was somewhat shorter in APP-SL⫻PS1mut mice (Fig. 4A) causing a tendency to reduced total ACh release during exploration. This latter finding may indicate reduced exploratory behavior of the APP-SL⫻PS1mut line, or a smaller capacity of these mice to sustain prolonged ACh release during exploration. To clarify the latter point, we evoked maximum ACh release by infusion of scopolamine which causes blockade of presynaptic muscarinic (M2/M4) autoreceptors and, thereby, uncouples presynaptic ACh release from feedback inhibition. In our hands, scopolamine caused a severalfold increase of ACh release in both groups of mice (Fig. 4B). The response to scopolamine was almost identical in plaque-free and plaque-bearing mice and actually was somewhat higher in APP-SL⫻PS1mut mice if the results were expressed as relative increases (% of baseline values). The effect of scopolamine clearly demonstrates that the maximum output of ACh from cholinergic synapses under stimulatory conditions is not reduced in APPSL⫻PS1mut mice. The results obtained with scopolamine seemed to exclude synaptic losses in septohippocampal cholinergic fibers of APP-SL⫻PS1mut mice. However, at least two further ideas had to be addressed: first, it has been suggested that amyloid peptides may directly interfere with cholinergic transmission in the absence of neurotoxic effects, e.g. by inhibiting HACU and ACh synthesis (Kar et al., 1996, 1998; Auld et al., 2002). Second, partial lesions of the septohippocampal pathway have previously been shown to cause compensatory increases of ACh release in rats which were demonstrated by an increase of HACU activity. It has to be noted that HACU activity measured ex vivo is directly coupled to cholinergic activity in vivo (Simon et al., 1976). Therefore, our result that HACU activity in the two mouse lines under scrutiny is almost identical (Fig. 3B)
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
demonstrates that the turnover rate of ACh is apparently unchanged despite of the presence of amyloid peptides and plaques in APP-SL⫻PS1mut mice. This finding seems to contradict earlier in vitro-findings showing inhibition of HACU in brain slices and synaptosomes by low amyloid concentrations (Kar et al., 1996, 1998). We cannot exclude that amyloid peptides exert an inhibitory influence on ACh release under basal conditions which is lost after brain homogenization due to dilution of the extracellular medium; however, it would then be difficult to explain why a reduction of ACh release is not seen after stimulation with scopolamine. It is possible that a different physical state of the amyloid peptides or the different methods of stimulation (intact fibers in vivo vs. chemically induced ACh release in isolated nerve endings) explain the discrepancies. To further investigate possible neuronal loss during chronic presence of amyloid peptides and plaques, we measured the density of neurons which express ChAT, a specific indicator of cholinergic neurons, in the septum, the location of cholinergic somata projecting to the hippocampus. The results again negated neuronal loss; we actually found a slightly increased number of ChAT-positive neurons in the septal area. While our method of cell counting does not strictly exclude an increased packing density of septal neurons as a possible explanation for this finding, the observed 26% increase of septal cholinergic neurons is more likely due to a neurotrophic response in APPSL⫻PS1mut mice (Stein and Johnson, 2002). This surprising result clearly needs further investigation but may indicate compensatory changes in the cholinergic system of transgenic mice as indicated by earlier reports (Stein and Johnson, 2002; Klingner et al., 2003). Lastly, we performed correlation analysis between the age of the individual animals, their plaque load, and the ACh release rates. For technical reasons, we had to use APP-SL⫻PS1mut mice which were, on average, somewhat younger than the APP-SL⫻PS1wt mice (601 vs. 797 days; P⬍0.05; see Results); thus, a positive correlation of age with ACh release would indicate a potential error in our interpretation. We correlated basal and stimulated (AUC during open field and after scopolamine infusion) ACh release in individual mice with their age and density of amyloid plaques. In no case, however, did we obtain a correlation line which was significantly significant from zero; in other words, the null hypothesis (slope of the correlation line⫽0) was not rejected in any case (P⬎0.4). One example, the correlation of age and Thioflavin Spositive amyloid plaques with basal ACh release in APPSL⫻PS1mut mice, is illustrated in Fig. 5. We conclude that increasing age (from 15 to 27 months) did not change basal or stimulated ACh release in our mouse lines. Similarly, increasing amyloid peptide formation or amyloid plaque load were not associated with a dysfunction of hippocampal ACh release (Fig. 5). Summarizing, the present work for the first time addressed changes in cholinergic neurotransmission in amyloid-bearing mice. Our in vivo-results support previous studies which found no or limited changes of presynaptic cholinergic parameters (Boncristiano et al., 2002; Gau et
1015
al., 2002; Apelt et al., 2002). Many of these earlier studies focused on cholinergic parameters which reflect cholinergic cell density but do not indicate functional changes of cholinergic transmission. Thus, the presence of ChAT reflects cholinergic innervation, but ChAT activity is not ratelimiting for ACh release, and the same holds true for the expression of the vesicular ACh transporter or AChE activity. The present study investigates ACh release with a functional in vivo-approach as recently requested by Selkoe (2002). In two mouse lines which differ only by the introduction of a single presenilin-1 mutation, we demonstrate an age-dependent formation of amyloid peptides and plaques which clearly occurs in the absence of a major effect on stimulated ACh release. Moreover, we could not detect any correlation between increasing amyloid plaque formation and cholinergic dysfunction, an observation which supports the recent hypothesis that insoluble amyloid plaques are not the toxic species of false APP processing pathways (Klein et al., 2001; Hardy and Selkoe, 2002). In clinical terms, our present results suggest that either patients must be exposed to elevated levels of A and to plaques for a much longer time period than transgenic mice to develop cholinergic hypofunction, or that the “amyloid hypothesis of neurodegeneration” needs to be modified to include as yet unknown factors causing selective cholinergic degeneration. Our study does not exclude amyloid peptide-induced changes of postsynaptic neurotransmission, e.g. changes of cholinergic receptor signaling (Kelly et al., 1996; Wang et al., 2000; Auld et al., 2002), nor does it contradict anatomical rearrangements of cholinergic fibers as suggested by some reports (Wong et al., 1999; Boncristiano et al., 2002; Hu et al., 2003). However, a presynaptic cholinergic dysfunction which is rapidly initiated by increasing amounts of amyloid peptides or plaques could not be demonstrated in our mouse model of AD. Taken together, our findings support the concept that APPand presenilin-transgenic mice are valuable models of amyloid deposition but fail to express cholinergic dysfunction, a typical symptom of clinical AD. Acknowledgements—We thank C. Dvorak (U. Mainz) for help with HACU determinations, E. Hartmann (Bayer AG) for help with histology and I. Hagelschuer (Bayer AG) for breeding and providing the transgenic mice.
REFERENCES Apelt J, Kumar A, Schliebs R (2002) Impairment of cholinergic neurotransmission in adult and aged transgenic Tg2576 mouse brain expressing the Swedish mutation of human beta-amyloid precursor protein. Brain Res 953:17–30. Auld DS, Kornecook TJ, Bastianetto S, Quirion R (2002) Alzheimer’s disease and the basal forebrain cholinergic system: relations to beta-amyloid peptides, cognition, and treatment strategies. Prog Neurobiol 68:209 –245. Bartus RT, Dean RL, Beer B, Lippa AS (1982) The cholinergic hypothesis of geriatric memory dysfunction. Science 217:408 –417. Bierer LM, Haroutunian V, Gabriel S, Knott PJ, Carlin LS, Purohit DP, Perl DP, Schmeidler J, Kanof P, Davis KL (1995) Neurochemical correlates of dementia severity in Alzheimer’s disease: relative importance of the cholinergic deficits. J Neurochem 64:749 –760. Boncristiano S, Calhoun ME, Kelly PH, Pfeifer M, Bondolfi L, Stalder
1016
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
M, Phinney AL, Abramowski D, Sturchler-Pierrat C, Enz A, Sommer B, Staufenbiel M, Jucker MJ (2002) Cholinergic changes in the APP23 transgenic mouse model of cerebral amyloidosis. J Neurosci 22:3234 –3243. Borchelt DR, Ratovitski T, van Lare J, Lee MK, Gonzales V, Jenkins NA, Copeland NG, Price DL, Sisodia SS (1997) Accelerated amyloid deposition in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precursor proteins. Neuron 19:939 – 945. Borchelt DR, Thinakaran G, Eckman CB, Lee MK, Davenport F, Ratovitsky T, Prada CM, Kim G, Seekins S, Yager D, Slunt HH, Wang R, Seeger M, Levey AI, Gandy SE, Copeland NG, Jenkins NA, Price DL, Younkin SG, Sisodia SS (1996) Familial Alzheimer’s disease-linked presenilin 1 variants elevate Abeta1– 42/1-40 ratio in vitro and in vivo. Neuron 17:1005–1013. Bronfman FC, Moechars D, van Leuven F (2000) Acetylcholinesterasepositive fiber deafferentation and cell shrinkage in the septohippocampal pathway of aged amyloid precursor protein London mutant transgenic mice. Neurobiol Dis 7:152–168. Chapman GF, Falinska AM, Knevett SG, Ramsay MF (2001) Genes, models and Alzheimer’s disease. Trends Genet 17:254 –261. Citron M, Westaway D, Xia W, Carlson G, Diehl T, Levesque G, Johnson-Wood K, Lee M, Seubert P, Davis A, Kholodenko D, Motter R, Sherrington R, Perry B, Yao H, Strome R, Lieberburg I, Rommens J, Kim S, Schenk D, Fraser P, St George HP, Selkoe DJ (1997) Mutant presenilins of Alzheimer’s disease increase production of 42-residue amyloid beta-protein in both transfected cells and transgenic mice. Nat Med 3:67–72. Collerton D (1986) Cholinergic function and intellectual decline in Alzheimer’s disease. Neuroscience 19:1–28. Cummings JL, Vinters HV, Cole GM, Khachauturian ZS (1998) Alzheimer’s disease: etiologies, pathophysiology, cognitive reserve, and treatment opportunities. Neurology 51 (Suppl 1):S2–17. Davis KL, Mohs RC, Marin D, Purohit DP, Perl DP, Lantz M, Austin G, Haroutunian V (1999) Cholinergic markers in elderly patients with early signs of Alzheimer disease. JAMA 281:1401–1406. Duff K, Eckman C, Zehr C, Yu X, Prada CM, Perez-Tur J, Hutton M, Buee L, Harigaya Y, Yager D, Morgan D, Gordon MN, Holcomb L, Refolo L, Zenk B, Hardy J, Younkin S (1996) Increased amyloidbeta42(43) in brains of mice expressing mutant presenilin 1. Nature 383:710 –713. Erb C, Troost J, Kopf S, Schmitt U, Lo¨ffelholz K, Soreq H, Klein J (2001) Compensatory mechanisms facilitate hippocampal acetylcholine release in transgenic mice expressing human acetylcholinesterase. J Neurochem 77:638 –646. Francis PT, Palmer AM, Snape M, Wilcock GK (1999) The cholinergic hypothesis of Alzheimer’s disease: a review of progress. J Neurol Neurosurg Psychiatry 66:137–147. Games D, Adams D, Alessandrini R, Barbour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, et al (1995) Alzheimer-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 373:523–527. Gau JT, Steinhilb ML, Kao TC, Da´mato CJ, Gaut JR, Frey KA, Turner RS (2002) Stable beta-secretase activity and presynaptic cholinergic markers during progressive central nervous system amyloidogenesis in Tg2576 mice. Am J Pathol 160:731–738. Gauthier S (2002) Advances in the pharmacotherapy of Alzheimer’s disease. CMAJ 166:616 –623. Gilmor ML, Erickson JD, Varoqui H, Hersh LB, Bennett DA, Cochran EJ, Mufson EJ, Levey AI (1999) Preservation of nucleus basalis neurons containing choline acetyltransferase and the vesicular acetylcholine transporter in the elderly with mild cognitive impairment and early Alzheimer’s disease. J Comp Neurol 411:693–704. Giovannini MG, Rakovska A, Benton RS, Pazzagli M, Bianchi L, Pepeu G (2001) Effects of novelty and habituation on acetylcholine, GABA, and glutamate release from the frontal cortex and hippocampus of freely moving rats. Neuroscience 106:43–53. Hardy J, Selkoe DJ (2002) The amyloid hypothesis of Alzheimer’s
disease: progress and problems on the road to therapeutics. Science 297:353–356. Hellstro¨m-Lindahl E (2000) Modulation of beta-amyloid precursor protein processing and tau phosphorylation by acetylcholine receptors. Eur J Pharmacol 393:255–263. Hernandez F, Lim F, Lucas JJ, Perez-Martin C, Moreno F, Avila J (2002) Transgenic mouse models with tau pathology to test therapeutic agents for Alzheimer’s disease. Mini Rev Med Chem 2:51– 58. Hernandez D, Sugaya K, Qu T, McGowan E, Duff K, McKinney M (2001) Survival and plasticity of basal forebrain cholinergic systems in mice transgenic for presenilin-1 and amyloid precursor protein mutant genes. Neuroreport 12:1377–1384. Hock BJ, Lamb BT (2001) Transgenic mouse models of Alzheimer’s disease. Trends Genet 17:S7–S12. Holcomb L, Gordon MN, McGowan E, Yu X, Benkovic S, Jantzen P, Wright K, Saad I, Mueller R, Morgan D, Sanders S, Zehr C, O’Campo K, Hardy Prada CM, Eckman C, Younkin S, Hsiao K, Duff K (1998) Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4:97–100. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G (1996) Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 274: 99 –102. Hu L, Wong TP, Cote SL, Bell KF, Cuello AC (2003) The impact of Abeta-plaques on cortical cholinergic and non-cholinergic presynaptic boutons in Alzheimer’s disease-like transgenic mice. Neuroscience 121:421–432. Joseph J, Shukitt-Hale B, Denisova NA, Martin A, Perry G, Smith MA (2001) Copernicus revisited: amyloid beta in Alzheimer’s disease. Neurobiol Aging 22:131–146. Kar S, Issa AM, Seto D, Auld DS, Collier B, Quirion R (1998) Amyloid beta-peptide inhibits high-affinity choline uptake and acetylcholine release in rat hippocampal slices. J Neurochem 70:2179 –2187. Kar S, Seto D, Gaudreau P, Quirion R (1996) Beta-amyloid-related peptides inhibit potassium-evoked acetylcholine release from rat hippocampal slices. J Neurosci 16:1034 –1040. Kawarabayashi T, Younkin LH, Saido TC, Shoji M, Ashe KH, Younkin SG (2001) Age-dependent changes in brain, CSF, and plasma amyloid  protein in the Tg2576 transgenic mouse model of Alzheimer’s disease. J Neurosci 21:372–381. Kelly JF, Furukawa K, Barger SW, Rengen MR, Mark RJ, Blanc EM, Roth GS, Mattson MP (1996) Amyloid beta-peptide disrupts carbacholinduced muscarinic cholinergic signal transduction in cortical neurons. Proc Natl Acad Sci USA 93:6753–6758. Khorkova OE, Patel K, Heroux J, Sahasrabudhe S (1998) Modulation of amyloid precursor protein processing by compounds with various mechanisms of action: detection by liquid phase electrochemiluminescent system. J Neurosci Methods 82:159 –166. Kimberly WT, LaVoie MJ, Ostaszewski BL, Ye W, Wolfe MS, Selkoe DJ (2003) Gamma-secretase is a membrane protein complex comprised of presenilin, nicastrin, Aph-1, and Pen-2. Proc Natl Acad Sci USA 100:6382–6387. Klein WL, Krafft GA, Finch CE (2001) Targeting small Abeta oligomers: the solution to an Alzheimer’s disease conundrum? Trends Neurosci 24:219 –224. Klingner M, Apelt J, Kumar A, Sorger D, Sabri O, Steinbach J, Scheunemann M, Schliebs R (2003) Alterations in cholinergic and noncholinergic neurotransmitter receptor densities in transgenic Tg2576 mouse brain with -amyloid plaque pathology. Int J Dev Neurosci 21:357–369. Kopf SR, Buchholzer ML, Hilgert M, Lo¨ffelholz K, Klein J (2001) Glucose plus choline improve passive avoidance behaviour and increase hippocampal acetylcholine release in mice. Neuroscience 103:365–371. Neve RL, Robakis NK (1998) Alzheimer’s disease: a re-examination of the amyloid hypothesis. Trends Neurosci 21:15–19.
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017 Mattson MP (1997) Cellular actions of beta-amyloid precursor protein and its soluble and fibrillogenic derivatives. Physiol Rev 77:1081– 1132. Nussbaum RL, Ellis CE (2003) Alzheimer’s disease and Parkinson’s disease. N Engl J Med 348:1356 –1364. Palmer AM (2002) Pharmacotherapy for Alzheimer’s disease: progress and prospects. Trends Pharmacol Sci 23:426 –433. Pedersen WA, Kloczewiak MA, Blusztajn JK (1996) Amyloid betaprotein reduces acetylcholine synthesis in a cell line derived from cholinergic neurons of the basal forebrain. Proc Natl Acad Sci USA 93:8068 –8071. Roberson MR, Harrell LE (1997) Cholinergic activity and amyloid precursor protein metabolism. Brain Res Rev 25:50 –69. Selkoe DJ (2001) Alzheimer’s disease: genes, proteins, and therapy. Physiol Rev 81:741–766. Selkoe DJ (2002) Alzheimer’s disease is a synaptic failure. Science 298:789 –791. Simon JR, Atweh S, Kuhar MJ (1976) Sodium-dependent high affinity choline uptake: a regulatory step in the synthesis of acetylcholine. J Neurochem 26:909 –922. Small DH, Mok SS, Bornstein JC (2001) Alzheimer’s disease and Abeta toxicity: from top to bottom. Nat Neurosci 2:595–598. Stein TD, Johnson JA (2002) Lack of neurodegeneration in transgenic mice overexpressing mutant amyloid precursor protein is associ-
1017
ated with increased levels of transthyretin and the activation of cell survival pathways. J Neurosci 22:7380 –7388. Sturchler-Pierrat C, Abramowski D, Duke M, Wiederhold KH, Mistl C, Rothacher S, Ledermann B, Burki K, Frey P, Paganetti PA, Waridel C, Calhoun ME, Jucker M, Probst A, Staufenbiel M, Sommer B (1997) Two amyloid precursor protein transgenic mouse models with Alzheimer disease-like pathology. Proc Natl Acad Sci USA 94:13287–13289. Thiel CM, Huston JP, Schwarting RK (1998) Hippocampal acetylcholine and habituation learning. Neuroscience 85:1253–1262. Wang HY, Lee DH, D’Andrea MR, Peterson PA, Shank RP, Reitz AB (2000) beta-Amyloid(1– 42) binds to alpha7 nicotinic acetylcholine receptor with high affinity: implications for Alzheimer’s disease pathology. J Biol Chem 275:5626 –5632. Winkler J, Thal LJ, Gage FH, Fisher LJ (1998) Cholinergic strategies for Alzheimer’s disease. J Mol Med 76:555–567. Wisniewski HM, Wen GY, Kim KS (1989) Comparison of four staining methods on the detection of neuritic plaques. Acta Neuropathol 78:22–27. Wong TP, Debeir T, Duff K, Cuello AC (1999) Reorganization of cholinergic terminals in the cerebral cortex and hippocampus in transgenic mice carrying mutated presenilin-1 and amyloid precursor protein transgenes. J Neurosci 19:2706 –2716.
(Accepted 26 February 2004)
CENTRAL CHOLINERGIC FUNCTIONS IN HUMAN AMYLOID PRECURSOR PROTEIN KNOCK-IN/PRESENILIN-1 TRANSGENIC MICE J. HARTMANN,a,b C. ERB,c U. EBERT,c ¨ NIGc1 AND K. H. BAUMANN,c A. POPP,c G. KO a,b J. KLEIN *
Key words: acetylcholine, amyloid precursor protein, cholinergic system, microdialysis, presenilin-1, transgenic mice.
a Department of Pharmacology, Johannes Gutenberg University of Mainz, D-55101 Mainz, Germany
Alzheimer’s disease (AD), the most frequent type of dementia in humans, is characterized by the formation of plaque-forming amyloid peptides, neurofibrillary tangles, degeneration of neurons and cognitive loss (Cummings et al., 1998; Selkoe, 2001; Nussbaum and Ellis, 2003). Neuronal degeneration strongly affects cholinergic neurons in the basal forebrain which project to hippocampal and cortical areas and which are required for attention, spatial orientation, learning and memory, i.e. functions that are typically lost in AD (Collerton, 1986). The “cholinergic hypothesis of senile dementia” postulated that the symptoms of AD can be explained by central cholinergic dysfunction (Bartus et al., 1982), and drug development for AD focused on drugs which enhance cholinergic function such as inhibitors of acetylcholinesterase (AChE; Cummings et al., 1998; Gauthier, 2002; Palmer, 2002). Clinical studies have indeed shown that central cholinergic dysfunction correlates with the severity of AD (Bierer et al., 1995; Francis et al., 1999). However, findings that cholinergic dysfunction occurs early in the disease have been challenged (Davis et al., 1999; Gilmor et al., 1999). While many authors believe that a cholinergic intervention is required to successfully treat AD (Winkler et al., 1998), the cholinergic deficit is no longer assumed to be the primary factor of disease etiology. Instead, largely based on genetic and biochemical evidence, many workers in the field now support the “amyloid hypothesis” of AD. This hypothesis postulates that abnormal processing of the amyloid precursor protein (APP) by - and ␥-secretases causes buildup of amyloid peptides which may be toxic in monomeric or oligomeric form or after deposition as amyloid plaques (Selkoe, 2002; Hardy and Selkoe, 2002). While various toxic properties of amyloid peptides have been described (Mattson, 1997; Small et al., 2001), recent work suggested that amyloid peptides are most damaging to neurons when they form oligomeric assemblies (“protofibrils”; Klein et al., 2001; Hardy and Selkoe, 2002). As the amyloid hypothesis postulates that neuronal damage follows amyloid formation, it implicitly assumes that cholinergic neurons are particularly susceptible to amyloid toxicity. It should be noted that the amyloid hypothesis has also been challenged by claims that amyloid peptides are not the causal factor of AD-related neurodegeneration (Neve and Robakis, 1998; Joseph et al., 2001). Despite the popularity of the cholinergic and amyloid hypotheses, a limited number of studies were concerned with the relationship between amyloid deposition and
b Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Science Center, 1300 Coulter Drive, Amarillo, TX 79106, USA c
Alzheimer Research Group, Bayer Health Care AG, D-42096 Wuppertal, Germany
Abstract—Alzheimer’s disease is characterized by amyloid peptide formation and deposition, neurofibrillary tangles, central cholinergic dysfunction, and dementia; however, the relationship between these parameters is not well understood. We studied the effect of amyloid peptide formation and deposition on central cholinergic function in knock-in mice carrying the human amyloid precursor protein (APP) gene with the Swedish/London double mutation (APP-SL mice) which were crossbred with transgenic mice overexpressing normal (PS1wt) or mutated (M146L; PS1mut) human presenilin-1. APP-SLⴛPS1mut mice had increased levels of A peptides at 10 months of age and amyloid plaques at 14 months of age while APP-SLⴛPS1wt mice did not have increased peptide levels and did not develop amyloid plaques. We used microdialysis in 15–27 months old mice to compare hippocampal acetylcholine (ACh) levels in the two mouse lines and found that extracellular ACh levels were slightly but significantly reduced in the APP-SLⴛPS1mut mice (ⴚ26%; Pⴝ0.044). Exploratory activity in the open field increased hippocampal ACh release by two-fold in both mouse lines; total and relative increases were not significantly different for the two strains under study. Similarly, infusion of scopolamine (1 M) increased hippocampal ACh release to a similar extent (3–5-fold) in both groups. High-affinity choline uptake, a measure of the ACh turnover rate, was identical in both mouse lines. Neurons expressing choline acetyltransferase were increased in the septum of APP-SLⴛPS1mut mice (ⴙ26%; Pⴝ0.046). We conclude that amyloid peptide production causes a small decrease of extracellular ACh levels. The deposition of amyloid plaques, however, does not impair stimulated ACh release and proceeds without major changes of central cholinergic function. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. 1
Present address: Fidelity Investments, Watertown, MA, USA. *Correspondence to: Tel: ⫹1-806-356-4015x252; fax: ⫹1-806-3564034. E-mail address: [email protected] (J. Klein). Abbreviations: ACh, acetylcholine; AChE, acetylcholinesterase; aCSF, artificial cerebrospinal fluid; AD, Alzheimer’s disease; APP, amyloid precursor protein; APP-SL, mice carrying the human APP gene with the Swedish/London double mutation; AUC, area under the curve; ChAT, choline acetyltransferase; HACU, high-affinity choline uptake; PS-1, presenilin-1; PS1wt, mice overexpressing normal human presenilin-1; PS1mut, mice overexpressing mutated (M146L) human presenilin-1.
0306-4522/04$30.00⫹0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.02.038
1009
1010
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
cholinergic dysfunction (Auld et al., 2002). In vitro, soluble amyloid peptides were found to reduce synaptosomal choline uptake and release of acetylcholine (ACh) in brain slices; these effects were observed at low (nanomolar) concentrations which likely occur in vivo (Kar et al., 1996, 1998). Prolonged treatment with higher doses of amyloid peptides reduced ACh synthesis and impaired muscarinic receptor signaling in neuronal cultures (Kelly et al., 1996; Pedersen et al., 1996). Amyloid effects were also tested in in vivo-models of AD. Mouse models of AD express or overexpress wild-type or mutant forms of APP and of presenilin-1, a subunit of the ␥-secretase complex (Kimberly et al., 2003) which is mutated in most familial types of AD. Some evidence for amyloid-induced cholinergic degeneration in these mice was reported. For instance, the size of septal cholinergic fibers was reduced and fibers were rearranged in one mouse strain (Bronfman et al., 2000) while modest decreases of cholinergic enzymes and more clear-cut decreases of cholinergic fiber length, but no loss of basal forebrain neurons, were observed in the APP23 mouse (Boncristiano et al., 2002). In the Tg2576 mouse line, markers of cholinergic function (choline acetyltransferase, vesicular ACh transporter, high-affinity choline uptake) were found to be unchanged despite increased amyloid deposition during aging (Gau et al., 2002). In contrast, other investigators reported a reduction of high-affinity choline uptake (HACU) and of muscarinic and nicotinic receptors in aged (21 month old) Tg2576 mice which was accompanied by an increase of the vesicular ACh transporter (Apelt et al., 2002; Klingner et al., 2003). Finally, doubly transgenic mice overexpressing human APP and presenilin mutations displayed normal forebrain cholinergic cell densities but a diminution of cholinergic synapse density and signs of aberrant sprouting in cortex and hippocampus (Wong et al., 1999; Hernandez et al., 2001; Hu et al., 2003). Taken together, these findings provided anatomical and neurochemical evidence for some cholinergic dysfunction secondary to amyloid formation but the functional consequences of the observed changes for cholinergic neurotransmission remained unclear. The present study attempts to improve our understanding of amyloid– cholinergic interactions by characterizing presynaptic parameters of cholinergic function using in vivo-microdialysis in mouse brain. For this study, we used knock-in mice carrying the human APP gene with the Swedish and London mutation (APP-SL) which were cross-bred with transgenic mice overexpressing human wild-type or mutated (M146L) presenilin1. Only the mice expressing the mutant presenilin developed amyloid plaques in the brain. We characterized the activity of the septohippocampal pathway of these mouse lines and report that central cholinergic functions in these mice, while showing minor changes, are essentially preserved despite age-associated amyloid peptide formation and deposition.
EXPERIMENTAL PROCEDURES Animals We used two types of mice for the present experiments. APP-SL mice were produced at Bayer Health Care AG (Wuppertal, Germany) and are knock-in mice in which the murine gene for APP has been replaced with the human gene carrying the LondonSwedish mutations (K670N/M671L; V717F). APP-SL⫻transgenic mice overexpressing the human wild-type presenilin-1 gene (PS1wt) are mice produced by breeding APP-SL mice with PS1wt mice (Duff et al., 1996). APP-SL⫻transgenic mice overexpressing the mutated (M146L) human presenilin-1 gene (PS1mut) are mice produced by breeding APP-SL mice (Duff et al., 1996). Genotypes were confirmed by PCR analysis of tail DNA in all cases. The experimental procedures used in this study met the guidelines of and were approved by the responsible government agency (Bezirksregierung Rheinland-Pfalz). Every attempt was made to reduce the number of animals used and their suffering.
Determination of amyloid peptides and amyloid plaques Mice of different age (3–29 months) were decapitated, and the brains removed. The cortices of one hemisphere were dissected, and a one-step formic acid extraction was performed as described (Kawarabayashi et al., 2001). The concentrations of amyloid peptides (total A and A40) were determined by means of specific antibodies using a liquid phase electrochemo-luminescence assay (Khorkova et al., 1998). The other hemisphere was fixated in 4% phosphate-buffered paraformaldehyde solution, cut into 40 m transverse sections on a cryomicrotome and stained for solid plaques with thioflavine S and for amyloid plaques using the 4G8 antibody (Wisniewski et al., 1989). The size, density and area covered by amyloid plaques were determined by quantitative analysis under the microscope.
Choline acetyltransferase (ChAT) immunohistochemistry Mice were perfused transcardially with 4% phosphate-buffered paraformaldehyde solution (pH⫽7.4), brains were removed and cut into 40 m sections on a cryomicrotome. Sections were incubated with hydrogen peroxide (3%) for 30 min followed by blocking solution containing donkey serum and bovine serum albumin. Sections were incubated overnight with the primary ChAT goat– anti-human antibody (1:50; Chemicon). Visualization was carried out using biotinylated donkey–anti-goat antibody (1:1000; Dianova) and streptavidin-coupled horseradish peroxidase with consecutive diaminobenzidine incubation. ChAT-positive neurons were counted in three sections per animal in the septum and diagonal band of Broca (coordinates from bregma: ⫹1.1 mm; ⫹0.74 mm and ⫹0.5 mm). The number of ChAT-positive neurons was determined in 10 randomly positioned squares (0.01 m2) by an analysis system (KS400; Zeiss, Germany). The data were double-checked visually by a technician. Densities of ChATpositive neurons were calculated as averages of three sections per mm3.
Microdialysis Mice were anesthetized with halothane (1–3%) and placed in a stereotactic frame. I-shaped, miniature, concentric dialysis probes with an exchange length of 1 mm were constructed as previously described (Erb et al., 2001; Kopf et al., 2001) and implanted in the right ventral mouse hippocampus using the following coordinates (from bregma): AP ⫺2.0 mm; L ⫺1.8 mm; DV ⫺2.3 mm. The animals were allowed to recover overnight, and experiments were carried out on 2 consecutive days after probe implantation in freely moving animals.
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017 On the experimental days, the microdialysis probes were perfused with artificial cerebrospinal fluid (aCSF; 147 mM NaCl, 4 mM KCl, 1.2 mM CaCl2 and 1.2 mM MgCl2) containing the AChE inhibitor neostigmine (1 M) at a perfusion rate of 1 l/min. On day 1, efflux from the microdialysis probe was sampled in 15 min intervals to obtain baseline values of ACh. After 90 min, the mice were placed into a novel environment (“open field”), and microdialysis was continued. The testing chamber was a gray plastic container (45⫻32⫻20 cm), and each mouse was allowed to explore the chamber during the time of the experiment (90 min). Subsequently, the mice were returned to their home cage, and ACh efflux was monitored for another 90 min. On day 2, baseline values of ACh were again sampled for 90 min. Subsequently, the dialysis fluid (aCSF) was switched to aCSF containing 1 M of scopolamine bromide while the mice remained in the home cage. After 90 min, the perfusion fluid was switched back to aCSF, and the mice were monitored for a further 90 min. On day 3, the mice were killed, and the hemispheres (cortex and hippocampus) were used to measure HACU. The correct location of the microdialysis probe was also verified during this procedure.
Determination of ACh ACh in dialysates was determined by microbore HPLC using a metal-free system which consisted of a low-speed pump (BAS PM80), separation column (SepStik; 530⫻1 mm), enzyme reactor (50⫻1 mm) carrying immobilized AChE and choline oxidase and electrochemical detector (BAS LC-4C) with a platinum electrode operating at 0.5 V. The flow rate was 120 l/min. The retention time for ACh was 10.3 min. At an injection volume of 5 l, the detection limit of this system was 10 fmol/injection.
Microdialysis data analysis ACh efflux data were recorded for each individual mouse as fmol per 15 min time interval during rest, behavioral stimulation in the open field and scopolamine infusion. Areas under the curves (AUC) were calculated by subtracting baseline ACh levels from levels measured during the 90 min of stimulation and were used to correlate the effects of the open field and scopolamine with animal age and with plaque density (GraphPad Prism).
HACU HACU was determined in synaptosomal (P2) fractions obtained from cortex and hippocampus as described (Erb et al., 2001). Briefly, brain tissue was homogenized in isotonic sucrose solution containing 10 mM HEPES buffer and centrifuged at 1000⫻g at 4 °C for 10 min. The supernatant was again centrifuged at 17,000⫻g for 10 min, and the resulting P2 pellet was washed and used for HACU determinations. For this purpose, aliquots were incubated in Krebs–Ringer buffer at 37 °C with 50 nM [3H]-choline (DuPont-NEN, Bad Nauheim, Germany; specific activity 50 Ci/ mmol) at 30 °C for exactly 5 min, in the presence or absence of hemicholinium-3 (HC-3, 1 M). Choline uptake was stopped by addition of ice-cold buffer. After centrifugation, the pellet was washed three times with HC-3 containing buffer, and radioactivity associated with the pellet was determined by liquid scintillation counting. HACU is equivalent to the HC-3-inhibitable part of choline uptake and was expressed as [3H]-uptake (d.p.m./g protein). Protein was determined by the Lowry assay.
RESULTS Animals, amyloid peptides and amyloid plaques APP-SL knock-in mice were phenotypically normal and fertile but did not develop amyloid plaques during aging as
1011
Fig. 1. (A) APP-SL⫻PS1wt and (B) APP-SL⫻PS1mut mouse forebrain stained immunohistochemically with antibody 4G8 directed against -amyloid peptide. The transverse sections were taken from 26 month old mice.
analyzed by immunohistology and Thioflavin-S staining (not illustrated). Similarly, the APP-SL knock-in mice which were crossbred with mice overexpressing human presenilin1 (APP-SL⫻PS1wt mice) did not develop measurable A peptide levels or amyloid plaques up to the age of 26 months (Fig. 1A). In contrast, APP-SL knock-in mice carrying the mutated human PS-1 gene (APP-SL⫻PS1mut mice) had increased levels of A peptides at 10 months of age (Fig. 2A), and all of the mice developed amyloid plaques (Fig. 1B). In this mouse line, total A peptides increased strongly with age reaching more than 10 nmol/g wet weight at 20 months. Only a small percentage (about 10%) of these peptides were A1– 40 (Fig. 2A) whereas a preliminary analysis at 16 months of age showed three-fold higher A1– 42 levels compared with A1– 40 (data not shown). Amyloid plaque deposition in APP-SL⫻PS1mut mice was measurable at 14 months of age, and plaque load increased steadily with age to 400 plaques per cortical section at 26 months (Fig. 2B). The distribution of amyloid plaques in these animals resembled that seen in Alzheimer’s disease, with high densities in the parietal cortex and in the hippocampal CA1 region (approximately 3 plaques per mm2) and moderate plaque accumulation in the dentate gyrus and in piriform cortex (approximately one plaque per mm2; Fig. 1B). Microdialysis: basal concentrations of ACh For in vivo-microdialysis, we used 11 aged mice (eight male, three female; weight 36.2⫾6.0 g) of the APPSL⫻PS1wt line (656 –919 days of age; average age 797 days) and 11 mice (nine male, two female; weight 40.2⫾10.1 g) of the APP-SL⫻PS1mut strain (441– 826 days of age; average days 601 days). As illustrated in Fig. 3A, basal efflux of ACh, reflecting the extracellular concentration of ACh in the hippocampus of awake animals, was 65.4⫾6.2 fmol/5 min in APP-SL⫻PS1wt mice. The corresponding value for APP-SL⫻PS1mut mice (46.3⫾6.4 fmol/5 min) was slightly lower (by 26%; P⫽0.044).
1012
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
Fig. 2. (A) Levels of soluble A peptides and A1– 40 in APPSL⫻PS1mut mice of various age. APP-SL⫻PS1wt mice did not have measurable levels of A peptides. (B) Number of amyloid plaques as detected by the 4G8 antibody in corresponding transverse sections of cortex and hippocampus of APP-SL⫻PS1mut mice of various age. No plaques could be detected in the brains of APP-SL⫻PS1wt mice. Data are means⫾S.D. from seven to 10 mice at each age group.
Microdialysis: effect of behavioral stimulation Exploration of a new environment is known to stimulate hippocampal ACh release (Thiel et al., 1998; Giovannini et al., 2001). When the doubly transgenic mice were exposed to a novel environment (“open field”) hippocampal ACh release was increased in all animals by approximately two-fold (Fig. 4A). The absolute increases of ACh efflux within 15 min of exploration were similar in APP-SL⫻PS1wt mice (⫹58.8 fmol) and APPSL⫻PS1mut mice (⫹61.9 fmol). In relative terms, maximum increases were identical for APP-SL⫻PS1wt mice (⫹134.6% of baseline ACh efflux) and APP-SL⫻PS1mut mice (⫹132.7%); however, the increase of ACh efflux was longer-lasting in APP-SL⫻PS1wt mice, an observation which is reflected in the calculation of AUC over 180 min: the average AUC value for APP-SL⫻PS1wt was 421⫾136 fmol⫻h, for APP-SL⫻PS1mut 216⫾63 fmol⫻h (P⫽0.18). If the relative increases were calculated in percentages, and AUC values taken from percentage values, then AUCs of APP-SL⫻PS1mut mice (526⫾85 AU) were only slightly lower than AUCs of APP-SL⫻PS1wt mice (699⫾211 AU).
Fig. 3. Cholinergic parameters of APP-SL⫻PS1wt and APPSL⫻PS1mut mice. (A) Basal release of ACh in mouse hippocampus (* P⫽0.044; N⫽11 each). (B) HACU measured in cortical and hippocampal homogenates (P2 fraction; N⫽8 each). (C) Density of septal neurons stained positive for ChAT. (* P⫽0.046; N⫽9 each). Data are means⫾S.E.M. of eight to 11 experiments in mice of 15–27 months of age (see Results).
Microdialysis: effect of pharmacological stimulation Scopolamine blocks presynaptic muscarinic receptors on cholinergic nerve endings and thereby disrupts the negative feedback that ACh exerts on its own release (Erb et al., 2001). When scopolamine was infused into the hippocampus, a several-fold increase of ACh release was observed in both groups of mice over a period of 90 min (Fig. 4B). Maximum increases of ACh efflux were observed after 60 –75 min and were ⫹227.5 fmol in APP-SL⫻PS1wt mice and ⫹180.1 fmol in APPSL⫻PS1mut mice. In relative terms, maximum increases from baseline efflux were ⫹289% for APPSL⫻PS1wt mice and ⫹375% for APP-SL⫻PS1mut mice. Calculations of the AUC did not reveal significant differences between the groups; the data for APPSL⫻PS1wt mice (1195⫾314 fmol⫻h) and APPSL⫻PS1mut mice (1146⫾261 fmol⫻h) were, for all practical purposes, identical.
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
Fig. 4. Stimulated release of ACh in mouse hippocampi measured by microdialysis. (A) Changes of ACh efflux upon behavioral activation (exposure to a novel environment) for 90 min, followed by return to the home cage. (B) Changes of ACh efflux upon infusion of scopolamine (1 M) for 90 min, followed again by artificial CSF. PSwt, APPSL⫻PS1wt mice. PSmut, APP-SL⫻PS1mut mice. Data are means⫾S.E.M. of 11 experiments in mice of 15–27 months of age (see Results).
Correlations To investigate possible relationships between age, solid (Thioflavin S positive) amyloid plaque formation, and ACh release, we correlated basal ACh release data of plaquebearing APP-SL⫻PS1mut mice with age and amyloid plaque density of individual animals. As shown in Fig. 5, basal ACh efflux showed a tendency to increase with age and plaque burden, but the correlation was not significant. In addition, no correlation was found between age and stimulated ACh release in APP-SL⫻PS1mut mice, and of age with basal or stimulated ACh release in APPSL⫻PS1wt mice (data not shown). HACU HACU, measured ex vivo, reflects the turnover of ACh in vivo (Simon et al., 1976). In the present study, the hemicholinium-sensitive uptake of [3H]-choline was 40.2⫾3.6 d.p.m./g protein in APP-SL⫻PS1wt mice (uptake within 5 min; equivalent to 4.3 pmol/h/mg protein). The corresponding value for APP-SL⫻PS1mut mice was 37.0⫾5.3
1013
Fig. 5. Correlations of data obtained for APP-SL⫻PS1mut mice. (A) Correlation of mouse age with basal release of ACh (N⫽11). (B) Correlation of solid plaque density with basal release of ACh. Note that the data on plaque density were not available for all animals in the study (N⫽9). For (A) and (B), the slopes of the correlation lines (calculated by linear regression) were not significantly different from zero (GraphPad Prism software).
d.p.m./g protein. Thus, the data did not reveal a significant difference of ACh turnover between the two mouse lines (P⬎0.5; Fig. 3B). Cholinergic cell number We stained the septal area, location of the cholinergic cell bodies which project to the hippocampus, with an antibody directed against ChAT, a specific marker of cholinergic neurons. Quantitative image analysis demonstrated that the plaque-bearing APP-SL⫻PS1mut mouse line had a higher number of septal ChAT-positive neurons (by 26%) than the plaque-free APP-SL⫻PS1wt mice (P⫽0.046; Fig. 3C). Nissl stains did not reveal visible changes of neuronal morphology, density or size of nuclei in transgenic lines (not illustrated).
DISCUSSION Most of the pathological features of AD can now be mimicked by individual animal models which reproduce the formation of amyloid plaques and neurofibrillary tangles, central cholinergic dysfunction, and neurodegeneration, respectively (Auld et al., 2002; Hock and Lamb, 2001; Chapman et al., 2001; Hernandez et al., 2002); however,
1014
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
the interrelationships between these pathologies remain poorly understood. Age-dependent formation of amyloid plaques has been observed in transgenic mouse lines overexpressing wild-type or mutated forms of APP coupled to strong promoters (Games et al., 1995; Hsiao et al., 1996; Sturchler-Pierrat et al., 1997). Amyloid peptide formation and plaque deposition can be enhanced by crossbreeding APP-transgenic mice with mice overexpressing presenilin-1 (Borchelt et al., 1996, 1997; Citron et al., 1997; Holcomb et al., 1998), an observation which is likely explained by presenilin’s role in APP processing (Kimberly et al., 2003). However, amyloid-bearing mice show no or only moderate signs of neuronal (especially cholinergic) dysfunction. While several of the transgenic AD models displayed altered long-term potentiation and cognitive dysfunction with aging, the relationship of these observations to a possible cholinergic degeneration remained unknown (Hock and Lamb, 2001; Chapman et al., 2001). Loss of neurons or synapses, if present, was moderate in all mouse lines, and severe and selective neurodegeneration as observed in human AD was not present. As for cholinergic systems, some work indicated that altered cholinergic activity may influence processing of APP (Roberson and Harrell, 1997; Hellstro¨m-Lindahl, 2000). The amyloid hypothesis, however, postulates that amyloid peptide formation and/or deposition precedes cholinergic dysfunction; thus, ongoing formation of amyloid peptides should cause overt cholinergic dysfunction during aging in murine Alzheimer models. In the present study, we developed a novel mouse model of AD which appeared to be particularly well suited to test this hypothesis. We first created a knock-in APP-SL mouse which did not develop amyloid plaques. We then crossbred these mice with PS1wt or PS1mut (Duff et al., 1996). We found that APP-SL⫻PS1wt mice did not form amyloid peptides or develop amyloid plaques with age. In contrast, the APP-SL⫻PS1mut line had elevated amyloid peptide levels in the brain at an age of 10 months, followed by an increasing deposition of amyloid plaques at 14 months of age. Both amyloid peptide and plaque formation increased with age and reached high levels at ⬎18 months of age (Figs. 1 and 2). Plaque depositions were observed in every individual animal and were limited to the brain areas (cortex, hippocampus) which are most severely affected in human AD. The availability of these mice offered the opportunity to study the influence of amyloid deposition on cholinergic parameters in two groups of mice which are genetically identical except for the mutation in presenilin-1. With respect to the extent of amyloid formation, our mice resemble the well-characterized Tg2576 line although our mice develop amyloid plaques somewhat later (at 15 vs. 10 months of age) and at a slightly lower density than Tg2576 mice (Hsiao et al., 1996; Kawarabayashi et al., 2001). Both mouse lines fail to develop neurofibrillary tangles and do not show overt neuronal degeneration. The main goal of the present study was to obtain in vivo-data of cholinergic activity in aged transgenic mice with amyloid depositions. The method of choice for the investigation of cholinergic activity in awake animals is
microdialysis. We have adapted the microdialysis procedure for the present study to follow neuronal activity (ACh release) in the septohippocampal pathway, a cholinergic pathway which is typically affected in AD. Our first finding was that the extracellular ACh levels in APP-SL⫻PS1mut mice were 26% lower than in APP-SL⫻PS1wt mice (P⫽0.044; Fig. 3A). This finding is compatible with a lower neuronal activity of the cholinergic fibers, or a loss of synapses induced by formation of amyloid peptides and plaques. However, this interpretation was challenged by the following experiments. We tested to what extent the septohippocampal pathway was capable of responding to behavioral and pharmacological manipulations. Exposure of rodents to a novel environment (“open field”) causes exploratory behavior which was previously shown to be associated with an increase of hippocampal ACh release in rats (Thiel et al., 1998; Giovannini et al., 2001) and mice (Kopf et al., 2001). The present results demonstrate that both groups of mice were able to activate the septohippocampal pathway during exploration to a similar extent. The maximum increase of ACh release during exploration was identical in both lines, whether expressed in absolute (fmol/min) or relative terms (% increase of baseline). The duration of the effect, however, was somewhat shorter in APP-SL⫻PS1mut mice (Fig. 4A) causing a tendency to reduced total ACh release during exploration. This latter finding may indicate reduced exploratory behavior of the APP-SL⫻PS1mut line, or a smaller capacity of these mice to sustain prolonged ACh release during exploration. To clarify the latter point, we evoked maximum ACh release by infusion of scopolamine which causes blockade of presynaptic muscarinic (M2/M4) autoreceptors and, thereby, uncouples presynaptic ACh release from feedback inhibition. In our hands, scopolamine caused a severalfold increase of ACh release in both groups of mice (Fig. 4B). The response to scopolamine was almost identical in plaque-free and plaque-bearing mice and actually was somewhat higher in APP-SL⫻PS1mut mice if the results were expressed as relative increases (% of baseline values). The effect of scopolamine clearly demonstrates that the maximum output of ACh from cholinergic synapses under stimulatory conditions is not reduced in APPSL⫻PS1mut mice. The results obtained with scopolamine seemed to exclude synaptic losses in septohippocampal cholinergic fibers of APP-SL⫻PS1mut mice. However, at least two further ideas had to be addressed: first, it has been suggested that amyloid peptides may directly interfere with cholinergic transmission in the absence of neurotoxic effects, e.g. by inhibiting HACU and ACh synthesis (Kar et al., 1996, 1998; Auld et al., 2002). Second, partial lesions of the septohippocampal pathway have previously been shown to cause compensatory increases of ACh release in rats which were demonstrated by an increase of HACU activity. It has to be noted that HACU activity measured ex vivo is directly coupled to cholinergic activity in vivo (Simon et al., 1976). Therefore, our result that HACU activity in the two mouse lines under scrutiny is almost identical (Fig. 3B)
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
demonstrates that the turnover rate of ACh is apparently unchanged despite of the presence of amyloid peptides and plaques in APP-SL⫻PS1mut mice. This finding seems to contradict earlier in vitro-findings showing inhibition of HACU in brain slices and synaptosomes by low amyloid concentrations (Kar et al., 1996, 1998). We cannot exclude that amyloid peptides exert an inhibitory influence on ACh release under basal conditions which is lost after brain homogenization due to dilution of the extracellular medium; however, it would then be difficult to explain why a reduction of ACh release is not seen after stimulation with scopolamine. It is possible that a different physical state of the amyloid peptides or the different methods of stimulation (intact fibers in vivo vs. chemically induced ACh release in isolated nerve endings) explain the discrepancies. To further investigate possible neuronal loss during chronic presence of amyloid peptides and plaques, we measured the density of neurons which express ChAT, a specific indicator of cholinergic neurons, in the septum, the location of cholinergic somata projecting to the hippocampus. The results again negated neuronal loss; we actually found a slightly increased number of ChAT-positive neurons in the septal area. While our method of cell counting does not strictly exclude an increased packing density of septal neurons as a possible explanation for this finding, the observed 26% increase of septal cholinergic neurons is more likely due to a neurotrophic response in APPSL⫻PS1mut mice (Stein and Johnson, 2002). This surprising result clearly needs further investigation but may indicate compensatory changes in the cholinergic system of transgenic mice as indicated by earlier reports (Stein and Johnson, 2002; Klingner et al., 2003). Lastly, we performed correlation analysis between the age of the individual animals, their plaque load, and the ACh release rates. For technical reasons, we had to use APP-SL⫻PS1mut mice which were, on average, somewhat younger than the APP-SL⫻PS1wt mice (601 vs. 797 days; P⬍0.05; see Results); thus, a positive correlation of age with ACh release would indicate a potential error in our interpretation. We correlated basal and stimulated (AUC during open field and after scopolamine infusion) ACh release in individual mice with their age and density of amyloid plaques. In no case, however, did we obtain a correlation line which was significantly significant from zero; in other words, the null hypothesis (slope of the correlation line⫽0) was not rejected in any case (P⬎0.4). One example, the correlation of age and Thioflavin Spositive amyloid plaques with basal ACh release in APPSL⫻PS1mut mice, is illustrated in Fig. 5. We conclude that increasing age (from 15 to 27 months) did not change basal or stimulated ACh release in our mouse lines. Similarly, increasing amyloid peptide formation or amyloid plaque load were not associated with a dysfunction of hippocampal ACh release (Fig. 5). Summarizing, the present work for the first time addressed changes in cholinergic neurotransmission in amyloid-bearing mice. Our in vivo-results support previous studies which found no or limited changes of presynaptic cholinergic parameters (Boncristiano et al., 2002; Gau et
1015
al., 2002; Apelt et al., 2002). Many of these earlier studies focused on cholinergic parameters which reflect cholinergic cell density but do not indicate functional changes of cholinergic transmission. Thus, the presence of ChAT reflects cholinergic innervation, but ChAT activity is not ratelimiting for ACh release, and the same holds true for the expression of the vesicular ACh transporter or AChE activity. The present study investigates ACh release with a functional in vivo-approach as recently requested by Selkoe (2002). In two mouse lines which differ only by the introduction of a single presenilin-1 mutation, we demonstrate an age-dependent formation of amyloid peptides and plaques which clearly occurs in the absence of a major effect on stimulated ACh release. Moreover, we could not detect any correlation between increasing amyloid plaque formation and cholinergic dysfunction, an observation which supports the recent hypothesis that insoluble amyloid plaques are not the toxic species of false APP processing pathways (Klein et al., 2001; Hardy and Selkoe, 2002). In clinical terms, our present results suggest that either patients must be exposed to elevated levels of A and to plaques for a much longer time period than transgenic mice to develop cholinergic hypofunction, or that the “amyloid hypothesis of neurodegeneration” needs to be modified to include as yet unknown factors causing selective cholinergic degeneration. Our study does not exclude amyloid peptide-induced changes of postsynaptic neurotransmission, e.g. changes of cholinergic receptor signaling (Kelly et al., 1996; Wang et al., 2000; Auld et al., 2002), nor does it contradict anatomical rearrangements of cholinergic fibers as suggested by some reports (Wong et al., 1999; Boncristiano et al., 2002; Hu et al., 2003). However, a presynaptic cholinergic dysfunction which is rapidly initiated by increasing amounts of amyloid peptides or plaques could not be demonstrated in our mouse model of AD. Taken together, our findings support the concept that APPand presenilin-transgenic mice are valuable models of amyloid deposition but fail to express cholinergic dysfunction, a typical symptom of clinical AD. Acknowledgements—We thank C. Dvorak (U. Mainz) for help with HACU determinations, E. Hartmann (Bayer AG) for help with histology and I. Hagelschuer (Bayer AG) for breeding and providing the transgenic mice.
REFERENCES Apelt J, Kumar A, Schliebs R (2002) Impairment of cholinergic neurotransmission in adult and aged transgenic Tg2576 mouse brain expressing the Swedish mutation of human beta-amyloid precursor protein. Brain Res 953:17–30. Auld DS, Kornecook TJ, Bastianetto S, Quirion R (2002) Alzheimer’s disease and the basal forebrain cholinergic system: relations to beta-amyloid peptides, cognition, and treatment strategies. Prog Neurobiol 68:209 –245. Bartus RT, Dean RL, Beer B, Lippa AS (1982) The cholinergic hypothesis of geriatric memory dysfunction. Science 217:408 –417. Bierer LM, Haroutunian V, Gabriel S, Knott PJ, Carlin LS, Purohit DP, Perl DP, Schmeidler J, Kanof P, Davis KL (1995) Neurochemical correlates of dementia severity in Alzheimer’s disease: relative importance of the cholinergic deficits. J Neurochem 64:749 –760. Boncristiano S, Calhoun ME, Kelly PH, Pfeifer M, Bondolfi L, Stalder
1016
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017
M, Phinney AL, Abramowski D, Sturchler-Pierrat C, Enz A, Sommer B, Staufenbiel M, Jucker MJ (2002) Cholinergic changes in the APP23 transgenic mouse model of cerebral amyloidosis. J Neurosci 22:3234 –3243. Borchelt DR, Ratovitski T, van Lare J, Lee MK, Gonzales V, Jenkins NA, Copeland NG, Price DL, Sisodia SS (1997) Accelerated amyloid deposition in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precursor proteins. Neuron 19:939 – 945. Borchelt DR, Thinakaran G, Eckman CB, Lee MK, Davenport F, Ratovitsky T, Prada CM, Kim G, Seekins S, Yager D, Slunt HH, Wang R, Seeger M, Levey AI, Gandy SE, Copeland NG, Jenkins NA, Price DL, Younkin SG, Sisodia SS (1996) Familial Alzheimer’s disease-linked presenilin 1 variants elevate Abeta1– 42/1-40 ratio in vitro and in vivo. Neuron 17:1005–1013. Bronfman FC, Moechars D, van Leuven F (2000) Acetylcholinesterasepositive fiber deafferentation and cell shrinkage in the septohippocampal pathway of aged amyloid precursor protein London mutant transgenic mice. Neurobiol Dis 7:152–168. Chapman GF, Falinska AM, Knevett SG, Ramsay MF (2001) Genes, models and Alzheimer’s disease. Trends Genet 17:254 –261. Citron M, Westaway D, Xia W, Carlson G, Diehl T, Levesque G, Johnson-Wood K, Lee M, Seubert P, Davis A, Kholodenko D, Motter R, Sherrington R, Perry B, Yao H, Strome R, Lieberburg I, Rommens J, Kim S, Schenk D, Fraser P, St George HP, Selkoe DJ (1997) Mutant presenilins of Alzheimer’s disease increase production of 42-residue amyloid beta-protein in both transfected cells and transgenic mice. Nat Med 3:67–72. Collerton D (1986) Cholinergic function and intellectual decline in Alzheimer’s disease. Neuroscience 19:1–28. Cummings JL, Vinters HV, Cole GM, Khachauturian ZS (1998) Alzheimer’s disease: etiologies, pathophysiology, cognitive reserve, and treatment opportunities. Neurology 51 (Suppl 1):S2–17. Davis KL, Mohs RC, Marin D, Purohit DP, Perl DP, Lantz M, Austin G, Haroutunian V (1999) Cholinergic markers in elderly patients with early signs of Alzheimer disease. JAMA 281:1401–1406. Duff K, Eckman C, Zehr C, Yu X, Prada CM, Perez-Tur J, Hutton M, Buee L, Harigaya Y, Yager D, Morgan D, Gordon MN, Holcomb L, Refolo L, Zenk B, Hardy J, Younkin S (1996) Increased amyloidbeta42(43) in brains of mice expressing mutant presenilin 1. Nature 383:710 –713. Erb C, Troost J, Kopf S, Schmitt U, Lo¨ffelholz K, Soreq H, Klein J (2001) Compensatory mechanisms facilitate hippocampal acetylcholine release in transgenic mice expressing human acetylcholinesterase. J Neurochem 77:638 –646. Francis PT, Palmer AM, Snape M, Wilcock GK (1999) The cholinergic hypothesis of Alzheimer’s disease: a review of progress. J Neurol Neurosurg Psychiatry 66:137–147. Games D, Adams D, Alessandrini R, Barbour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, et al (1995) Alzheimer-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 373:523–527. Gau JT, Steinhilb ML, Kao TC, Da´mato CJ, Gaut JR, Frey KA, Turner RS (2002) Stable beta-secretase activity and presynaptic cholinergic markers during progressive central nervous system amyloidogenesis in Tg2576 mice. Am J Pathol 160:731–738. Gauthier S (2002) Advances in the pharmacotherapy of Alzheimer’s disease. CMAJ 166:616 –623. Gilmor ML, Erickson JD, Varoqui H, Hersh LB, Bennett DA, Cochran EJ, Mufson EJ, Levey AI (1999) Preservation of nucleus basalis neurons containing choline acetyltransferase and the vesicular acetylcholine transporter in the elderly with mild cognitive impairment and early Alzheimer’s disease. J Comp Neurol 411:693–704. Giovannini MG, Rakovska A, Benton RS, Pazzagli M, Bianchi L, Pepeu G (2001) Effects of novelty and habituation on acetylcholine, GABA, and glutamate release from the frontal cortex and hippocampus of freely moving rats. Neuroscience 106:43–53. Hardy J, Selkoe DJ (2002) The amyloid hypothesis of Alzheimer’s
disease: progress and problems on the road to therapeutics. Science 297:353–356. Hellstro¨m-Lindahl E (2000) Modulation of beta-amyloid precursor protein processing and tau phosphorylation by acetylcholine receptors. Eur J Pharmacol 393:255–263. Hernandez F, Lim F, Lucas JJ, Perez-Martin C, Moreno F, Avila J (2002) Transgenic mouse models with tau pathology to test therapeutic agents for Alzheimer’s disease. Mini Rev Med Chem 2:51– 58. Hernandez D, Sugaya K, Qu T, McGowan E, Duff K, McKinney M (2001) Survival and plasticity of basal forebrain cholinergic systems in mice transgenic for presenilin-1 and amyloid precursor protein mutant genes. Neuroreport 12:1377–1384. Hock BJ, Lamb BT (2001) Transgenic mouse models of Alzheimer’s disease. Trends Genet 17:S7–S12. Holcomb L, Gordon MN, McGowan E, Yu X, Benkovic S, Jantzen P, Wright K, Saad I, Mueller R, Morgan D, Sanders S, Zehr C, O’Campo K, Hardy Prada CM, Eckman C, Younkin S, Hsiao K, Duff K (1998) Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4:97–100. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G (1996) Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 274: 99 –102. Hu L, Wong TP, Cote SL, Bell KF, Cuello AC (2003) The impact of Abeta-plaques on cortical cholinergic and non-cholinergic presynaptic boutons in Alzheimer’s disease-like transgenic mice. Neuroscience 121:421–432. Joseph J, Shukitt-Hale B, Denisova NA, Martin A, Perry G, Smith MA (2001) Copernicus revisited: amyloid beta in Alzheimer’s disease. Neurobiol Aging 22:131–146. Kar S, Issa AM, Seto D, Auld DS, Collier B, Quirion R (1998) Amyloid beta-peptide inhibits high-affinity choline uptake and acetylcholine release in rat hippocampal slices. J Neurochem 70:2179 –2187. Kar S, Seto D, Gaudreau P, Quirion R (1996) Beta-amyloid-related peptides inhibit potassium-evoked acetylcholine release from rat hippocampal slices. J Neurosci 16:1034 –1040. Kawarabayashi T, Younkin LH, Saido TC, Shoji M, Ashe KH, Younkin SG (2001) Age-dependent changes in brain, CSF, and plasma amyloid  protein in the Tg2576 transgenic mouse model of Alzheimer’s disease. J Neurosci 21:372–381. Kelly JF, Furukawa K, Barger SW, Rengen MR, Mark RJ, Blanc EM, Roth GS, Mattson MP (1996) Amyloid beta-peptide disrupts carbacholinduced muscarinic cholinergic signal transduction in cortical neurons. Proc Natl Acad Sci USA 93:6753–6758. Khorkova OE, Patel K, Heroux J, Sahasrabudhe S (1998) Modulation of amyloid precursor protein processing by compounds with various mechanisms of action: detection by liquid phase electrochemiluminescent system. J Neurosci Methods 82:159 –166. Kimberly WT, LaVoie MJ, Ostaszewski BL, Ye W, Wolfe MS, Selkoe DJ (2003) Gamma-secretase is a membrane protein complex comprised of presenilin, nicastrin, Aph-1, and Pen-2. Proc Natl Acad Sci USA 100:6382–6387. Klein WL, Krafft GA, Finch CE (2001) Targeting small Abeta oligomers: the solution to an Alzheimer’s disease conundrum? Trends Neurosci 24:219 –224. Klingner M, Apelt J, Kumar A, Sorger D, Sabri O, Steinbach J, Scheunemann M, Schliebs R (2003) Alterations in cholinergic and noncholinergic neurotransmitter receptor densities in transgenic Tg2576 mouse brain with -amyloid plaque pathology. Int J Dev Neurosci 21:357–369. Kopf SR, Buchholzer ML, Hilgert M, Lo¨ffelholz K, Klein J (2001) Glucose plus choline improve passive avoidance behaviour and increase hippocampal acetylcholine release in mice. Neuroscience 103:365–371. Neve RL, Robakis NK (1998) Alzheimer’s disease: a re-examination of the amyloid hypothesis. Trends Neurosci 21:15–19.
J. Hartmann et al. / Neuroscience 125 (2004) 1009 –1017 Mattson MP (1997) Cellular actions of beta-amyloid precursor protein and its soluble and fibrillogenic derivatives. Physiol Rev 77:1081– 1132. Nussbaum RL, Ellis CE (2003) Alzheimer’s disease and Parkinson’s disease. N Engl J Med 348:1356 –1364. Palmer AM (2002) Pharmacotherapy for Alzheimer’s disease: progress and prospects. Trends Pharmacol Sci 23:426 –433. Pedersen WA, Kloczewiak MA, Blusztajn JK (1996) Amyloid betaprotein reduces acetylcholine synthesis in a cell line derived from cholinergic neurons of the basal forebrain. Proc Natl Acad Sci USA 93:8068 –8071. Roberson MR, Harrell LE (1997) Cholinergic activity and amyloid precursor protein metabolism. Brain Res Rev 25:50 –69. Selkoe DJ (2001) Alzheimer’s disease: genes, proteins, and therapy. Physiol Rev 81:741–766. Selkoe DJ (2002) Alzheimer’s disease is a synaptic failure. Science 298:789 –791. Simon JR, Atweh S, Kuhar MJ (1976) Sodium-dependent high affinity choline uptake: a regulatory step in the synthesis of acetylcholine. J Neurochem 26:909 –922. Small DH, Mok SS, Bornstein JC (2001) Alzheimer’s disease and Abeta toxicity: from top to bottom. Nat Neurosci 2:595–598. Stein TD, Johnson JA (2002) Lack of neurodegeneration in transgenic mice overexpressing mutant amyloid precursor protein is associ-
1017
ated with increased levels of transthyretin and the activation of cell survival pathways. J Neurosci 22:7380 –7388. Sturchler-Pierrat C, Abramowski D, Duke M, Wiederhold KH, Mistl C, Rothacher S, Ledermann B, Burki K, Frey P, Paganetti PA, Waridel C, Calhoun ME, Jucker M, Probst A, Staufenbiel M, Sommer B (1997) Two amyloid precursor protein transgenic mouse models with Alzheimer disease-like pathology. Proc Natl Acad Sci USA 94:13287–13289. Thiel CM, Huston JP, Schwarting RK (1998) Hippocampal acetylcholine and habituation learning. Neuroscience 85:1253–1262. Wang HY, Lee DH, D’Andrea MR, Peterson PA, Shank RP, Reitz AB (2000) beta-Amyloid(1– 42) binds to alpha7 nicotinic acetylcholine receptor with high affinity: implications for Alzheimer’s disease pathology. J Biol Chem 275:5626 –5632. Winkler J, Thal LJ, Gage FH, Fisher LJ (1998) Cholinergic strategies for Alzheimer’s disease. J Mol Med 76:555–567. Wisniewski HM, Wen GY, Kim KS (1989) Comparison of four staining methods on the detection of neuritic plaques. Acta Neuropathol 78:22–27. Wong TP, Debeir T, Duff K, Cuello AC (1999) Reorganization of cholinergic terminals in the cerebral cortex and hippocampus in transgenic mice carrying mutated presenilin-1 and amyloid precursor protein transgenes. J Neurosci 19:2706 –2716.
(Accepted 26 February 2004)