Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Ann Neurol. Author manuscript; available in PMC 2011 May 15.
Published in final edited form as:
PMCID: PMC3094694

Days-to-criterion as an indicator of toxicity associated with human Alzheimer amyloid-β oligomers

Sam Gandy, M.D., Ph.D.,1,2, Adam J. Simon, Ph.D.,3,* John W. Steele, B.A.,1,* Alex L. Lublin, Ph.D.,1,* James J. Lah, M.D., Ph.D.,4 Lary C. Walker, Ph.D.,4,5 Allan I. Levey, M.D., Ph.D.,4 Grant A. Krafft, Ph.D.,6 Efrat Levy, Ph.D.,7 Frédéric Checler, Ph.D.,8 Charles Glabe, Ph.D.,9 Warren Bilker, Ph.D.,10 Ted Abel, Ph.D.,11 James Schmeidler, Ph.D.,12 and Michelle E. Ehrlich, M.D.13



Recent evidence suggests that high molecular weight soluble oligomeric Aβ (oAβ) assemblies (also known as Aβ-derived diffusible ligands, or ADDLs) may represent a primary neurotoxic basis for cognitive failure in AD. To date, in vivo studies of oAβ/ADDLs have involved injection of assemblies purified from the cerebrospinal fluid (CSF) of human subjects with Alzheimer’s disease or from the conditioned media of Aβ-secreting cells into experimental animals. We sought to study the bioactivities of endogenously formed oAβ/ADDLs generated in situ from the physiological processing of human APP transgenes.


We produced and histologically characterized single transgenic mice overexpressing APPE693Q or APPE693Q X PS1ΔE9 bigenic mice. APPE693Q mice were studied in the Morris water maze (MWM) task at 6 and 12 months of age. Following the second MWM evaluation, mice were sacrificed, and brains were assayed for Aβtotal, Aβ40, Aβ42, and oAβ/ADDL by ELISA and were also histologically examined. Based on results from the oAβ/ADDL ELISA, we assigned individual APPE693Q mice to either an “undetectable oAβ/ADDLs group” or a “readily detectable oAβ/ADDLs group”. A days-to-criterion (DTC) analysis was used to determine delays in acquisition of the MWM task.


Both single transgenic and bigenic mice developed intraneuronal accumulation of APP/Aβ, though only Dutch APPE693Q X PS1Δ9 bigenic mice developed amyloid plaques. The APPE693Q mice did not develop amyloid plaques at any age studied, up to 30 months. APPE693Q mice were tested for spatial learning and memory, and only 12-month old APPE693Q mice with readily detectable oAβ/ADDLs displayed a significant delay in acquisition of the MWM task when compared to NTg littermates.


These data suggest that cerebral oAβ/ADDL assemblies generated in brain in situ from human APP transgenes may be associated with cognitive impairment. We propose that a DTC analysis may be a sensitive method for assessing the cognitive impact in mice of endogenously generated oligomeric human Aβ assemblies.

SEARCH TERMS: (1) Amyloid, (2) Alzheimer’s Disease, (3) Spatial Recognition, (4) Days-to-Criterion, (5) Amyloid Precursor Protein


Alzheimer’s disease (AD) is a progressive neurodegenerative disorder and is the most common cause of senile dementia. Rare familial forms of AD are caused by genes that modulate metabolism of the amyloid-β peptide (Aβ) (for review, see ref. 1), and progression of all forms of AD involves the accumulation in brain of insoluble spherical deposits of aggregated Aβ known as amyloid plaques2. A reformulation of the amyloid cascade hypothesis has shifted focus from the hallmark amyloid plaques to high molecular weight soluble assemblies of oligomeric Aβ (oAβ, also known as Aβ-derived diffusible ligands; ADDLs)37 as the proximate neurotoxins underlying AD.

Recent evidence has implicated oAβ/ADDLs βin cognitive decline8,9. Electrophysiological studies have shown that addition of oAβ/ADDLs to hippocampal slices results in an inhibition of long-term potentiation (LTP), a cellular model of learning and memory6. These results were corroborated in vivo via demonstration of deficits in learning and memory performance following injection of oAβ/ADDLs directly into the hippocampi of living rats4,10.

In this report, we utilized an in vivo model of AD that produce soluble oAβ/ADDLs βeither with (APPE693Q X PS1ΔE9) or without (APPE693Q) β-amyloid plaques in the brain11. We show that the levels of oAβ/ADDLs are associated with impaired acquisition of the Morris water maze (MWM) task by APPE693Q mice. We propose that days-to-criterion (DTC) analyses might be especially sensitive for assessing deficits associated with oAβ/ADDLs generated from the physiological processing of transgenic human APP.


Experimental animals

Generation of C57BL/6J-TgN(Thy1-APPE693Q, APP751 numbering) transgenic mice was performed as described by Gandy et al (2007)12. Briefly, pTSC21, the mouse Thy1.2 expression cassette13 was digested and blunt-ended at the unique XhoI site, and the APP751(E693Q) cDNA (provided by Dr. Efrat Levy, New York University)14 was inserted into the Thy1.2 cassette. The 5′ end of the cDNA was modified to introduce a Kozak sequence, with primers


CCGCGGTACGACGGGCCAAAC-3′, using the Quik Change Site-Directed Mutagenesis kit (Stratagene). The DNA for injection was released with PvuI, purified from an agarose gel, dialyzed and injected following routine protocol. Generation of transgenic PS1ΔE9 mice was previously described12. Experimentally naïve male and female non-transgenic littermates (NTg n=8), APPE693Q single transgenic mice (n=17), or APPE693Q X PS1ΔE9 bigenic mice (n=12) were maintained and bred under standard conditions consistent with National Institutes of Health guidelines for animal care and approved by the Institutional Animal Care and Use Committee of the James J. Peters Veterans Affairs Medical Center. Mice were handled for 2 minutes per day for 3 days prior to pre-training on MWM.

Western blot analysis of APP expression

Brains were homogenized on ice and extracts were denatured in SDS loading buffer. Samples were separated by standard SDS-polyacrylamide gel electrophoresis and electrophoretically transferred to polyvinylidene fluoride membrane (Millipore, Bedford, MA). The following primary antibodies were used: pan-species anti-APP C-terminus specific pAb369 (previously described15) or anti-human Aβ1–16 human APP/Aβ specific mAb6E10 (Covance).

Immunohistochemistry and Electron Microscopy

Mice were anesthetized by CO2 exposure and transcardially perfused with cold saline, followed by fixation in 4% phosphate-buffered paraformaldehyde. Coronal sections (40μm) were cut with a Leica vibratome 2000 (Nussloch, Germany), cryoprotected, and stored at −20°C. Cresyl violet, hematoxylin and eosin (H&E) staining are done according to standard protocols. For light microscopy, tissue blocks were frozen on dry ice and sectioned at 40 μm on a freezing microtome. For electron microscopy, blocks of forebrain, motor cortex, hippocampus, and cerebellum were sectioned at 40 μm on a vibratome (Technical Products International, St. Louis, MO).

Immunohistochemical processing was performed with free-floating sections and immunoperoxidase using previously described methods16. The following antibodies were used: mAb 4G8 and mAb 6E10 (Covance), polyclonal antibodies to the Aβ carboxy terminus [polyclonal antibody FCA3340 and FCA3542 specific for Aβ40 or Aβ42, respectively, generous gift from F. Checler] and anti-GLUT4 (Chemicon). Biotinylated goat anti-rat IgG (1:200 dilution; Vector Labs, Burlingame, CA) and avidin-biotinylated horseradish peroxidase complex (Vectastain Elite, Vector Labs) were used to localize the primary antibody. Immunoreactivity was visualized with 0.05% 3,3-diaminobenzidine tetrahydrochloride (DAB) and 0.01% H2O2 in 50 mM Tris, pH 7.6. Sections for light microscopy were slide-mounted, air-dried, dehydrated through a graded alcohol series and xylenes, and, finally, coverslipped for microscopic examination. Sections for electron microscopic immunohistochemistry were post-fixed in 1% osmium tetroxide, stained with 2% uranyl acetate, dehydrated through graded alcohol and propylene oxide, and embedded in Eponate 12 resin. Ultrathin sections were cut on a Reichert Ultracut S ultramicrotome (Leica, Deerfield, IL) and collected on mesh copper grids for examination on a JEOL JEM-100C transmission electron microscope. Control sections were processed in parallel, in which the primary antibody was either omitted or pre-absorbed with the corresponding antigen. Sections for electron microscopic immunohistochemistry were post-fixed in 1% osmium tetroxide, stained with 2% uranyl acetate, dehydrated through graded alcohol and propylene oxide, and embedded in Eponate 12 resin.

Thioflavin-S staining

Floating sections were washed in PBS and mounted on Superfrost Plus slides coated with Vectabond (Vector Laboratories, Burlingame, CA), before being processed for Thioflavin-S (Thio-S). Briefly, sections were post-fixed in 10% formalin for 10 minutes, then washed in PBS. After incubation for 10 minutes in 0.25% potassium permanganate, sections were washed in PBS and incubated in 2% potassium metabisulfite and 1% oxalic acid until they appeared white. Sections were then washed in water and stained for 10 minutes with a solution of 0.015% Thio-S in 50% ethanol. Finally, sections were washed in 50% ethanol and in water, then dried, and dipped into Histo-Clear before being coverslipped with Permount. All chemicals were from Sigma (St. Louis, MO). Finally, the sections were coverslipped with Vectashield (Vector Laboratories), and sealed with nail polish.

Analysis of cerebral microhemorrhage

Cerebral hemorrhage, when present, is typically accompanied by a delayed appearance of hemosiderin-positive microglia17. Perls’s Berlin blue-stained clusters of hemosiderin staining were qualitatively evaluated (presence/absense) from sections throughout the neocortex, hippocampus, and thalamus. An additional set of every 10th section was stained for H&E and screened for acute intraparenchymal hemorrhage.

Aβ ELISAs (Aβ40, Aβ42, oligomeric Aβ/ADDLs)

Aβ was detected by incubating horseradish peroxidase-conjugated JRF/Atot/17 (human Aβ) or JRF/rA1-15/2 (murine Aβ) as detection antibody. For ELISA determination of oAβ/ADDLs, the identical monoclonal antibody (6E10) was used for both capture and detection18. Therefore, only species with at least two mAb6E10 epitopes were detected. ELISA plates were developed using a color reaction (TMB Microwell Peroxidase Substrate System, Kirkegaard & Perry, Gaithersburg, MD), and the A450 was read and quantified by comparison to covalently cross-linked Aβ dimer standards.

Morris water maze behavioral analysis

Experimentally naïve mice were trained on the MWM task at 6 months of age, extinguished, then trained and tested again at 12 months of age, all according to a standard protocol19. The water maze was a circular pool (120 cm in diameter). White nontoxic tempera paint was mixed with water to make the water opaque. Hidden 0.5 cm beneath the surface of the water was a circular platform (11.2 cm in diameter). The path of the mouse was recorded with a video tracking system (HVS Image, Buckingham, UK). The water maze was located in a 5.2 m × 2.1-m room. There were different cues on each wall of the room: along one wall is a 90 cm × 60-cm poster; along another wall is a coat rack and a 30 cm × 30-cm black triangle; along the third wall is a deflated multicolored inner tube, measuring 45 cm in diameter; and hung in the center of the curtain is a 30-cm diameter inflated yellow ball. A video camera was mounted above the center of the pool. During pre-training, mice were trained to sit on the platform. This training occurred in a 5-gallon (19-L) bucket in a room that was different from the experimental room. Pre-training consisted of three trials. In the first trial, mice were placed into the bucket and allowed to swim to a visible platform located in the center of the bucket, where they sat for 30 s. The second trial was identical to the first, except that the platform is submerged 0.5 cm beneath the surface of the water. The third trial was identical to the second, except that mice sat on the platform for 60s.

During the training/acquisition phase, on 12 consecutive days, mice received four trials per day during which the platform was hidden 0.5 cm beneath the surface of the water in a constant location (one of two locations were used, balanced across subjects). A different starting location was used on each trial, which consists of a swim followed by a 20-s platform sit. Any mouse that does not find the platform within 60s was guided to it by the experimenter. The inter-trial interval (ITI) is 4–6 min (this ITI is used in all subsequent experiments). During the ITI, mice sat in their home cages, which are kept near the computer and out of sight of the water maze throughout each session. The 4-6-min ITI was long enough for the mice to dry themselves before the next trial.

The MWM task was extinguished by standard extinction protocol at 6 months. Briefly, four 60-s trials in which they swam in the pool in the absence of the platform, but shower curtains were hung around the pool to block the distal cues in the room from view. Care is taken to ensure that mice are removed from different locations on each trial. Preference during extinction and probe trials was assessed by analyzing time spent searching in the target quadrant compared with time spent searching in the other three quadrants. Mice were sacrificed and analyzed at 13 months of age, following MWM testing at 12 months and two probe trials. ANOVAs at each time point were used to determine whether there exists a link between MWM and levels of Aβ species in each brain region. Mice were group housed with ad libitum access to food and water and maintained on a 12:12 light:dark cycle with lights on at 7AM. All experiments were conducted during the light period between the hours of 9AM and 5PM.

Statistical analysis of MWM behavior

An overhead video camera was used to capture swim pattern and time in each quadrant for all mice during training/acquisition and probe trials. Average escape latency was calculated for each animal on each day of the training/acquisition period. For probe trials, time spent in each quadrant was calculated. A days-to-criterion (DTC) analysis has been previously used to analyze acquisition of the MWM task20. Mice met criterion if they had escape latencies of less than 25s on two consecutive trials, indicating reliable performance of the task. The 25s criterion was based on the third quartile escape latencies for day 11 of training, indicating that at least 75% of animals tested found the escape platform within approximately 25s (quartile ranks not shown). Each animal received a DTC score of up to 12 reflecting the day on which they met criterion for acquisition of the task.

Repeated measures analysis of variance (ANOVA) was utilized to analyze escape latency for the 12 days of training/acquisition. Multivariate ANOVA (MANOVA) was used to compare escape latency from each day of training/acquisition with animals grouped by genotype or ADDL level. For DTC analysis, a fourth root (x^0.25) transformation of DTC score was used to account for long right tail distribution, allowing for assumption of normal distribution in subsequent parametric tests. One-way ANOVAs were used to compare mean group differences on probe trials and also for DTC analysis. Levene’s statistic was computed to determine homogeneity of variance between groups. For all ANOVAs, a Bonferroni’s (homogeneity of variance assumed) or a Dunnett’s T3 (homogeneity of variance not assumed) multiple comparisons was used to determine between-groups differences for ADDL level and independent samples t-tests were used to compare APPE693Q versus NTg mice for DTC and at probe trials. A Kaplan-Meier survival analysis and Mantel-Cox log rank multiple comparisons were employed to further analyze between-groups differences for DTC by ADDL level (see also Supplementary Table 2b–d). These tests, which used untransformed data, were primarily utilized to validate the fourth-root-transformed mean DTC data for use in parametric tests. Significance is reported for all tests with a p≤0.05 using two-tailed α=0.05; p-values for all tests are reported in Supplementary Table 2.


Biochemical characterization of APPE693Q mice

Several missense mutations within the Aβ domain of the amyloid precursor protein (APP) have been associated with an increase in the propensity of the peptide to form oAβ/ADDL assemblies21. All these mutations are located near the middle of the Aβ domain, where they have been proposed to disrupt salt bridges that, when present, stabilize parallel β-sheets and promote fibrillogenesis. The model suggests that, because the salt bridges cannot form, fibrillogenesis is destabilized, and the formation of oAβ/ADDL assemblies is favored. Based on these observations, we sought to determine whether the APPE693Q mutation generates Aβ with a high propensity to form soluble oligomers, without plaque pathology.

Brains of six individual lines of APPE693Q transgenic mice (all F1 generation) were analyzed for levels of huAPP expression by western blot analysis (Figure 1a). Using rabbit anti-pan-APP cytoplasmic tail pAb369 and human-APP (Aβ1–16)-specific mouse mAb6E10, we were able to confirm transgene protein expression in the brains of transgenic animals. Since APP-CTFs are the immediate precursors for metabolism and generation of the Aβ peptide, their detectability is an important measure in order to account for all the catabolic fragments of APP along the pathway to Aβ generation. In comparison to several other transgenic AD models, which utilize the Swedish APPK670N, M671L mutation (i.e. Tg2576 and TgCRND8) to increase total production of Aβ via increasing BACE cleavage of APP22, the APPE693Q mutation is not preferentially cleaved by BACE. Moreover, APPE693Q C99 (β-CTF) levels are not obviously increased as in the Tg2576 or TgCRND8 lines (Figure 1b).

Figure 1
Immunoblot characterization of APPE693Q single transgenic mice

Immunohistological analysis of APPE693Q and APPE693Q X PS1ΔE9 mice

The hereditary APPE693Q mutation has been described as an autosomal dominant form of cerebral amyloid angiopathy (CAA) with cerebral hemorrage11. APPE693Q single transgenic mice were analyzed for vascular pathology using immunohistochemistry, revealing initial appearance of amyloid-laden cerebral vessels in APPE693Q mice at 12 months or older, as compared to their non-transgenic littermates (Figure 2a). Moreover, Perls’ blue stain of APPE693Q brain tissue revealed occasional vessels outlined by hemosiderin, representing extravasation of blood, which is likely due to a combination of aging and CAA (Figure 2b). There was no evidence of gross hemorrhage.

Figure 2
Pathological analysis of CAA and amyloid deposition in APPE693Q and APPE693Q X PS1ΔE9 mice

Human APP-overexpressing mouse models of AD have been reported to display learning deficits prior to plaque deposition, though all of these murine models do eventually develop plaques23,24. Our APPE693Q mice, however, never develop senile plaques, up to at least 30 months of age, in comparison to the APPE693Q mouse model developed by the Jucker laboratory, which develop some diffuse plaques11. To accelerate the progression of plaque-like Alzheimer’s related pathology and, importantly, to validate the integrity of the APPE693Q transgene, we crossed APPE693Q mice with a mice overexpressing the familial AD-associated exon 9-deleted PS1 mutant (PS1ΔE9). Both APPE693Q single transgenic and APPE693Q X PS1ΔE9 were further studied using immunohistochemistry with mAb6E10 and a rabbit pAb anti-Aβ42-C-terminus-specific antibody. In both lines of mice, mAb6E10 and pAb anti-Aβ42-immunopositive staining of intraneuronal vesicles was detectable as early as 2 months of age, and mAb6E10 showed typical amyloid plaques in the brains of APPE693Q X PS1ΔE9 bigenic mice as early as 11 months of age (Figure 2c), but again not in APPE693 single transgenic littermates (see Supplementary Figure 1). We compared intraneuronal mAb6E10-positive and mAb4G8-positive (pan-species anti-Aβ17–21) vesicular staining patterns in APPE693Q single transgenic mice versus the well-characterized Tg2576 mouse model of AD (Figure 2d)24. The intensity and granularity of staining in APPE693Q mice appears to be more robust compared to that observed in the Tg2576 mouse. Moreover, immunoelectron microscopy revealed APP/Aβ-immunopositivity associated with the multivesicular bodies (MVBs/lipofuscin) of the late endosomal/lysosomal system (Figure 3) of both APPE693Q single transgenic and APPE693Q X PS1ΔE9 bigenic mice. In comparison to APPE693Q alone, APPE693Q X PS1ΔE9 mice have an increased Aβ42/Aβ40 ratio and develop plaques (Figure 2c and Supplementary Figure 1).

Figure 3
Immunoelectron microscopy reveals APP accumulation in intraneuronal organelles

APPE693Q mice exhibit an oAβ/ADDL-dependent deficit in acquisition of the Morris water maze

To investigate the role of soluble oAβ/ADDLs in AD-related deficits in learning and memory, we employed the Morris water maze (MWM) to analyze deficits in spatial learning and memory at 6 and 12 months of age. Ultimately, APPE693Q X PS1ΔE9 mice were excluded from oAβ/ADDL-related MWM behavioral statistical analysis since only 3 of 12 total mice formed detectable levels of oAβ/ADDLs.

The MWM is a widely used measure of both short- and long-term spatial memory, in which the animal uses spatial cues within the test room to find a hidden escape platform. Visuospatial function has been correlated with functional status in AD patients25 and hippocampal dysfunction associated with AD typically results in poor performance on visuospatial and spatial orientation-related tasks26,27. APPE693Q single transgenic mice and their non-transgenic littermates (NTg) were trained and tested on the MWM at 6 months of age, extinguished, and then trained and tested again at 12 months of age. For 11 consecutive days, mice were trained to swim to an escape platform within 60s, and escape latency was recorded at each trial. At 12 and 21 days post-training, mice were placed in the tank without an escape platform, and time spent swimming in each quadrant during this “probe” trial was recorded. No significant differences were observed between NTg and APPE693Q mice at 6 months of age during training or probe trials, and swim speed did not vary by genotype (data not shown). Further, at 12 months of age no difference was observed between NTg and APPE693Q mice at either probe trial. However, a large amount of intra-genotype variability was observed for APPE693Q mice during training, notably in the later days of training (Figure 4).

Figure 4
Behavioral characterization of APPE693Q on the MWM task

Based on the hypothesis that oAβ/ADDL levels might explain behavioral differences between individual APPE693Q mice, we used a duplicate-epitope sandwich ELISA18 to measure oAβ/ADDL levels in all tested mice at 13 months of age. A duplicate epitope sandwich ELISA utilizes the same antibody for both capture and detection, resulting in detection of a substrate with two or more of the identical antigen site i.e. dimers or larger. However, we cannot exclude the possibility that APP fragments other than Aβ might also aggregate and contribute to the ELISA signal. This limitation is inherent in the method.

Based on oAβ/ADDL levels (Supplementary Table 1), NTg or APPE693Q were grouped as follows: NTg mice (NTg mice are unable to form oligomers in the absence of the human APP transgene, n=8); undetectable (ud)Aβ/ADDL mice (mice with oAβ/ADDL levels below the lower limit of reliable quantitation; LLRQ; 39pg/g, n=12); or readily detectable (d)oAβ/ADDL mice (mice with oAβ/ADDL levels above the LLRQ, n=5). No difference was observed at either probe trial (Figure 5a) at 12 months of age when mice were grouped by oAβ/ADDL level. Analysis of escape latency during the training period revealed significant between-subjects differences for (d)oAβ/ADDL mice (p=0.027), but not (ud)oAβ/ADDL mice (p=0.227), in comparison to NTg mice (Figure 5). A Bonferroni’s post-hoc analysis revealed significant differences for escape latency only between (d)oAβ/ADDL and NTg mice on day 6 (p=0.021) and also on day 5 and day 9 between (d)oAβ/ADDL and (ud)oAβ/ADDL (p=0.010, p=0.002, respectively), (d)oAβ/ADDL and NTg (p=0.010, p<0.001, respectively), but not (ud)oAβ/ADDL versus NTg (p=0.579). Notably, no significant differences in escape latency were observed between NTg, (ud)oAβ/ADDL, or (d)oAβ/ADDL groups on the final day of training/acquisition (Figure 5b), indicating that APPE693Q mice did eventually learn the MWM task.

Figure 5
Aged APPE693Q mice exhibited an oAβ/ADDL-dependent delay in acquisition of the MWM task

Throughout the 12-day acquisition phase, typically mean escape latency decreases from day-to-day for NTg mice (Figure 4b). In order to more efficiently analyze the relationship between oAβ/ADDL level and acquisition of the MWM task, we used a days-to-criterion (DTC) analysis of escape latency20. Briefly, we established the criterion for reliable performance of the acquired MWM task as two consecutive trials with escape latencies of 25 seconds or less, where DTC represents the day on which criterion was met. There was a significant increase in DTC for (d)oAβ/ADDL (M=10.6) compared to NTg mice (M=5.5; p=0.01), but no significant difference was observed between (ud)oAβ/ADDL and NTg mice (Figure 5). Taken together, these results indicate an oAβ/ADDL level-dependent delay in acquisition of the MWM task in APPE693Q transgenic mice at 12 months of age.


We provide evidence that APPE693Q single transgenic mice develop a significant oAβ/ADDL-dependent delay in acquisition of the MWM task at 12 months of age that is not dependent on the development of AD-like plaque pathology or macrohemorrhage. APPE693Q single transgenic mice, as old as 30 months, did not develop senile plaques in contrast to APPE693Q X PS1ΔE9 bigenic mice, which developed plaques by 12 months of age.

Both APPE693Q single transgenic and APPE693Q X PS1ΔE9 exhibited robust accumulation of intraneuronal APP/Aβ-like immunoreactivity within MVBs/lipofuscin. Recent evidence supports a toxic role of intraneuronal accumulation of APP/Aβ28, and activity-induced reduction of intraneuronal Aβ has been shown to protect against Aβ-related synaptic alterations29. Based on the previous findings that oAβ/ADDL formation may be initiated intracellularly3033 and the work reported here, we suggest that the intraneuronal accumulation of APP/Aβ observed in APPE693Q mice may represent one site for the initiation of oAβ/ADDL formation (Figures 2 and and3).3). By studying the effects of oAβ/ADDLs generated in brain in situ, the current study is highly novel, since all studies to date have involved external application or intracerebral injection of partially purified oAβ/ADDLs preparations410.

Impairment of spatial navigation on the hidden goal task (a human analogue of the MWM) was recently associated with hippocampal dysfunction, wherein patients with hippocampal-related mild cognitive impairment (MCI) and AD patients displayed nearly identical delays in acquisition compared to both controls and non-hippocampal-related MCI patients26. Moreover, 21–22-week old (early plaque pathology) TgCRND8 mice also showed a delayed acquisition of the MWM task without long-term deficits at probe trials, whereas 38–42-week old (late plaque pathology) TgCRND8 mice displayed a delayed acquisition of the MWM with long-term deficits at the Day 12 probe trial in comparison to NTg littermates23. Jacobsen and colleagues (2006)34 also described early, pre-plaque deficits in acquisition of spatial orientation of 3 month-old Tg2576 mice on the contextual fear conditioning task. However, none of these studies investigated the association of oAβ/ADDLs with the observed deficits in spatial learning and memory. Here, we report that 12-month old APPE693Q mice displayed an oAβ/ADDL-dependent delay in acquisition of the MWM task compared to NTg littermates, suggesting that more discrete deficits of spatial orientation may be an early marker of AD-like cognitive decline. Importantly, we provide evidence that, in a mouse model in which oAβ/ADDLs are generated in situ from physiological processing of transgenic human APP, these deficits in spatial orientation are oAβ/ADDL-dependent. A recent publication implicated a correlation of A-11-postive oAβ levels with deficits related to acquistion of spatial memory on the MWM task of 2-year old APP23/Abca1 mice35. Taken together with these findings, our results suggest that 12-month old APPE693Q mice may represent a “pre-clinical” model of AD, although further work is required to determine whether APPE693Q mice acquire even more severe long-term spatial deficits at a later age (i.e., 18 and 24 months). Without development of plaque pathology or long-term spatial navigation deficits such as those described in 16-month old APPK670L, M671N X PS1ΔE9 bigenic mice27, 12-month old APPE693Q mice provide a model for studying specific oAβ/ADDL-related deficits in spatial learning and memory. We propose that DTC analysis may represent a particularly sensitive measure of pre-pathological oAβ/ADDL-related clinical deficits in the acquisition of tasks requiring spatial orientation.

Supplementary Material

suppl 1a

suppl 1b

suppl 2a

suppl 2b

suppl 2c


The authors would like to thank Tabansi Kawafumu and Emily Sluzas for technical assistance on the project.



Funding and Support:

S.G., A.L.L., J.W.S. and M.E.E. were supported by the Cure Alzheimer’s Fund, VA MERIT review grant 1I01BX000348-01, and National Institute on Aging grant P01AG10491. T.A. was supported by National Institute on Aging grant P50AG017623 (A.I. Pack, PI). J.J.L. and A.I.L. were supported by National Institute on Aging grant P50AG025688 and L.C.W. was supported by National Center for Research Resources grant RR-00165. J.S. was supported by National Institute on Aging grants P50AG05138 and P01AG02219. J.W.S. is a trainee in the Integrated Pharmacological Sciences Training Program supported by grant T32GM062754 from the National Institute of General Medical Sciences.

Financial Disclosures:

MEE: Medivation (Grantee for study of latrepirdine mechanisms of action in Huntington’s Disease); SG: J&J/Elan (Safety Monitoring Committee); Diagenic, Amicus (Consultant and grantee). A.S. is founder of Neuronostics. G.K. is the founder of Acumen Pharmaceuticals.


S.G., J.W.S., and A.L. prepared the manuscript. J.W.S., W.B., A.L.L. and J.S. performed statistical analysis of all data. L.C.W. and F.C. contributed to the histological analyses. A.J.S. and G.K. designed and performed ADDL ELISAs. C.G., T.A., E.L., G.K., L.C.W., A.I.L., and M.E.E., provided critical reading of the manuscript and were instrumental in the design and execution of histology, microscopy, and behavioral experiments. M.E.E. and S.G. procured funding for the project.


1. Gandy S. The role of cerebral amyloid beta accumulation in common forms of Alzheimer disease. J Clin Invest. 2005;115:1121–1129. [PMC free article] [PubMed]
2. Giannakopoulos P, Hof PR. Frontiers of neurology and neuroscience. Dementia in clinical practice. Preface. Front Neurol Neurosci. 2009;24:IX–X. [PubMed]
3. Lublin AL, Gandy S. Amyloid-beta oligomers: possible roles as key neurotoxins in A lzheimer’s disease. Mt Sinai J Med. 2010;77:43–49. [PMC free article] [PubMed]
4. Shankar GM, Li S, Mehta TH, et al. Amyloid-beta protein dimers isolated directly from Alzheimer’s brains impair synaptic plasticity and memory. Nat Med. 2008;14:837–842. [PMC free article] [PubMed]
5. Cleary JP, Walsh DM, Hofmeister JJ, et al. Natural oligomers of the amyloid-beta protein specifically disrupt cognitive function. Nat Neurosci. 2005;8:79–84. [PubMed]
6. Shankar GM, Bloodgood BL, Townsend M, et al. Natural oligomers of the Alzheimer amyloid-beta protein induce reversible synapse loss by modulating an NMDA-type glutamate receptor-dependent signaling pathway. J Neurosci. 2007;27:2866–2875. [PubMed]
7. Li S, Hong S, Shepardson NE, et al. Soluble oligomers of amyloid Beta protein facilitate hippocampal long-term depression by disrupting neuronal glutamate uptake. Neuron. 2009;62:788–801. [PMC free article] [PubMed]
8. Selkoe DJ. Soluble oligomers of the amyloid beta-protein impair synaptic plasticity and behavior. Behav Brain Res. 2008;192:106–113. [PMC free article] [PubMed]
9. Koffie RM, Meyer-Luehmann M, Hashimoto T, et al. Oligomeric amyloid beta associates with postsynaptic densities and correlates with excitatory synapse loss near senile plaques. Proc Natl Acad Sci USA. 2009;106:4012–4017. [PubMed]
10. Lesné S, Koh MT, Kotilinek L, Kayed R, Glabe CG, Yang A, Gallagher M, Ashe KH. A specific amyloid-beta protein assembly in the brain impairs memory. Nature. 2006;440:352–357. [PubMed]
11. Herzig MC, Winkler DT, Burgermeister P, et al. Abeta is targeted to the vasculature in a mouse model of hereditary cerebral hemorrhage with amyloidosis. Nat Neurosci. 2004;7:954–960. [PubMed]
12. Andrä K, Abramowski D, Duke M, et al. Expression of APP in transgenic mice: a comparison of neuron-specific promoters. Neurobiol Aging. 1996;17:183–190. [PubMed]
13. Gandy S, Zhang YW, Ikin A, et al. Alzheimer’s presenilin 1 modulates sorting of APP and its carboxyl-terminal fragments in cerebral neurons in vivo. J Neurochem. 2007;102:619–626. [PubMed]
14. Vidal RG, Fernandez-Madrid I, Frangione B, Levy E. Sequencing of the Alzheimer’s APP gene Dutch variant (APP-D) Hum Mutat. 1993;2:496–497. [PubMed]
15. Gouras GK, Xu H, Jovanovic JN, et al. Generation and regulation of beta-amyloid peptide variants by neurons. J Neurochem. 1998;71:1920–1925. [PubMed]
16. Olichney JM, Hansen LA, Galasko D, et al. The apolipoprotein E epsilon 4 allele is associated with increased neuritic plaques and cerebral amyloid angiopathy in Alzheimer’s disease and Lewy body variant. Neurology. 1996;47:190–196. [PubMed]
17. Koeppen AH, Dickson AC, McEvoy JA. The cellular reactions to experimental intracerebral hemorrhage. J Neurol Sci. 1995;134 (Suppl):102–112. [PubMed]
18. Xia W, Yang T, Shankar G, et al. A specific enzyme-linked immunosorbent assay for measuring beta-amyloid protein oligomers in human plasma and brain tissue of patients with Alzheimer disease. Arch Neurol. 2009;66:190–199. [PubMed]
19. Wood MA, Kaplan MP, Park A, et al. Transgenic mice expressing a truncated form of CREB-binding protein (CBP) exhibit deficits in hippocampal synaptic plasticity and memory storage. Learn Mem. 2005;12:111–119. [PubMed]
20. Dawson R, Jr, Pelleymounter MA, Cullen MJ, Gollub M, Liu S. An age-related decline in striatal taurine is correlated with a loss of dopaminergic markers. Brain Res Bull. 1999;48:319–324. [PubMed]
21. Wang Z, Natte R, Berliner JA, van Duinen SG, Vinters HV. Toxicity of Dutch (E22Q) and Flemish (A21G) mutant amyloid beta proteins to human cerebral microvessel and aortic smooth muscle cells. Stroke. 2000;31:534–538. [PubMed]
22. Haass C. Take five--BACE and the gamma-secretase quartet conduct Alzheimer’s amyloid beta-peptide generation. EMBO J. 2004;23:483–488. [PubMed]
23. Hyde LA, Kazdoba TM, Grilli M, et al. Age-progressing cognitive impairments and neuropathology in transgenic CRND8 mice. Behav Brain Res. 2005;160:344–355. [PubMed]
24. Westerman MA, Cooper-Blacketer D, Mariash A, et al. The relationship between Abeta and memory in the Tg2576 mouse model of Alzheimer’s disease. J Neurosci. 2002;22:1858–1867. [PubMed]
25. Fukui T, Lee E. Visuospatial function is a significant contributor to functional status in patients with Alzheimer’s disease. Am J Alzheimers Dis Other Demen. 2009;24:313–321. [PubMed]
26. Laczo J, Vlcek K, Vyhnalek M, et al. Spatial navigation testing discriminates two types of amnestic mild cognitive impairment. Behav Brain Res. 2009;202:252–259. [PubMed]
27. O’Leary TP, Brown RE. Visuo-spatial learning and memory deficits on the Barnes maze in the 16-month-old APPswe/PS1dE9 mouse model of Alzheimer’s disease. Behav Brain Res. 2009;201:120–127. [PubMed]
28. Almeida CG, Tampellini D, Takahashi RH, et al. Beta-amyloid accumulation in APP mutant neurons reduces PSD-95 and GluR1 in synapses. Neurobiol Dis. 2005;20:187–198. [PubMed]
29. Tampellini D, Rahman N, Gallo EF, et al. Synaptic activity reduces intraneuronal Abeta, promotes APP transport to synapses, and protects against Abeta-related synaptic alterations. J Neurosci. 2009;29:9704–9713. [PMC free article] [PubMed]
30. Walsh DM, Tseng BP, Rydel RE, Podlisny MB, Selkoe DJ. The oligomerization of amyloid beta-protein begins intracellularly in cells derived from human brain. Biochemistry. 2000;39:10831–10839. [PubMed]
31. Yang AJ, Chandswangbhuvana D, Shu T, Henschen A, Glabe CG. Intracellular accumulation of insoluble, newly synthesized abetan-42 in amyloid precursor protein-transfected cells that have been treated with Abeta1–42. J Biol Chem. 1999;274:20650–20656. [PubMed]
32. Yang AJ, Knauer M, Burdick DA, Glabe C. Intracellular A beta 1–42 aggregates stimulate the accumulation of stable, insoluble amyloidogenic fragments of the amyloid precursor protein in transfected cells. J Biol Chem. 1995;270:14786–14792. [PubMed]
33. Hu X, Crick SL, Bu G, et al. Amyloid seeds formed by cellular uptake, concentration, and aggregation of the amyloid-beta peptide. Proc Natl Acad Sci USA. 2009;106:20324–20329. [PubMed]
34. Jacobsen JS, Wu CC, Redwine JM, et al. Early-onset behavioral and synaptic deficits in a mouse model of Alzheimer’s disease. Proc Natl Acad Sci USA. 2006;103:5161–5166. [PubMed]
35. Lefterov I, Fitz NF, Cronican A, et al. Memory deficits in APP23/Abca1+/− mice correlate with the level of Abeta oligomers. ASN Neuro. 2009:1. doi: 10.1042/AN20090015. [PMC free article] [PubMed] [Cross Ref]