Systemic immune challenges trigger and drive Alzheimer-like neuropathology in mice
© Krstic et al.; licensee BioMed Central Ltd. 2012
Received: 21 February 2012
Accepted: 10 May 2012
Published: 2 July 2012
Alzheimer’s disease (AD) is the most prevalent form of age-related dementia, and its effect on society increases exponentially as the population ages. Accumulating evidence suggests that neuroinflammation, mediated by the brain’s innate immune system, contributes to AD neuropathology and exacerbates the course of the disease. However, there is no experimental evidence for a causal link between systemic inflammation or neuroinflammation and the onset of the disease.
The viral mimic, polyriboinosinic-polyribocytidilic acid (PolyI:C) was used to stimulate the immune system of experimental animals. Wild-type (WT) and transgenic mice were exposed to this cytokine inducer prenatally (gestation day (GD)17) and/or in adulthood. Behavioral, immunological, immunohistochemical, and biochemical analyses of AD-associated neuropathologic changes were performed during aging.
We found that a systemic immune challenge during late gestation predisposes WT mice to develop AD-like neuropathology during the course of aging. They display chronic elevation of inflammatory cytokines, an increase in the levels of hippocampal amyloid precursor protein (APP) and its proteolytic fragments, altered Tau phosphorylation, and mis-sorting to somatodendritic compartments, and significant impairments in working memory in old age. If this prenatal infection is followed by a second immune challenge in adulthood, the phenotype is strongly exacerbated, and mimics AD-like neuropathologic changes. These include deposition of APP and its proteolytic fragments, along with Tau aggregation, microglia activation and reactive gliosis. Whereas Aβ peptides were not significantly enriched in extracellular deposits of double immune-challenged WT mice at 15 months, they dramatically increased in age-matched immune-challenged transgenic AD mice, precisely around the inflammation-induced accumulations of APP and its proteolytic fragments, in striking similarity to the post-mortem findings in human patients with AD.
Chronic inflammatory conditions induce age-associated development of an AD-like phenotype in WT mice, including the induction of APP accumulations, which represent a seed for deposition of aggregation-prone peptides. The PolyI:C mouse model therefore provides a unique tool to investigate the molecular mechanisms underlying the earliest pathophysiological changes preceding fibrillary Aβ plaque deposition and neurofibrillary tangle formations in a physiological context of aging. Based on the similarity between the changes in immune-challenged mice and the development of AD in humans, we suggest that systemic infections represent a major risk factor for the development of AD.
KeywordsMouse model of sporadic Alzheimer`s disease Aging Immune challenge Systemic infection Neuroinflammation Cytokines Interleukins PolyI:C
Neuroinflammation was one of the prominent pathological features described by Alois Alzheimer in his first case report in 1907. Today we know that specific markers of neuroinflammation are selectively enriched in brain areas affected by Alzheimer’s disease (AD) neuropathology , and that individuals with high plaque burden without dementia show virtually no evidence of neuroinflammation . This is further supported by positron emission tomography (PET) imaging studies, which have shown that cognitive status is inversely correlated with microglial activation in patients with AD . In addition, recent genome-wide association studies have identified significant correlations between components of the innate immune system and the incidence of sporadic AD , supporting the link between the immune system and AD pathophysiology suggested by previous retrospective epidemiological studies in humans [5, 6].
However, the precise role of neuroinflammation in the disease etiology is still controversial, ranging from representing a possible cause to being a by-product of the disease  or even being beneficial . In 1996 it was proposed that interleukin (IL)-1, an inflammatory cytokine whose levels are increased in the brains of patients with AD , may represent a driving force in the pathogenesis of AD [10, 11]. Although this hypothesis was in accordance with a meta-analysis of 17 epidemiological studies indicating that non-steroidal anti-inflammatory drugs might decrease the risk of developing AD , subsequent randomized trials not only failed to show a beneficial effect of anti-inflammatory drugs on the etiology of AD [13, 14], but in fact found an increase in AD incidence in patients with mild cognitive impairment treated with these drugs as compared to placebo . The resolution of apparent inconsistency came only recently with the revision of the Alzheimer`s Disease Anti-inflammatory Prevention Trial (ADAPT) hypothesis that supports a beneficial role of anti-inflammatory drugs only in the early, asymptomatic, phases of the disease . However, in vivo experimental evidence to support an early and potentially causative role for systemic infections and neuroinflammation in the etiology of sporadic AD is still missing.
To elucidate the early role of inflammatory processes in the development of AD-like pathology in mice, we used the viral mimic polyriboinosinic-polyribocytidilic acid (PolyI:C), a synthetic analog of double-stranded RNA, to stimulate the immune system of our experimental animals [17, 18]. We have previously shown that a single exposure to PolyI:C during late gestation triggers the expression of several inflammatory cytokines in the fetal brain , evokes a reduction in adult neurogenesis, accompanied by memory impairments [19, 20], and accelerates protein depositions in the hippocampus of the adult offspring . In the current study, we tested the hypothesis that the prenatal immune challenge during late gestation results in pathological aging, and predisposes the offspring to aging-associated AD-like neuropathology and cognitive decline . In addition, we tested the effects of systemic immune challenge in adulthood on the progression of the AD-like phenotype either in prenatally challenged wild-type (WT) mice or in transgenic AD (3xTg-AD) mice .
All experimental procedures were approved by the local authorities of the Cantonal Veterinary Office in Zurich and carried out in agreement with the Principles of Laboratory Animal Care (National Institutes of Health publication number 86–23, revised 1985).
Overview of the experimental animals used in this study
Time of tissue collection (use)
4 NaCI, 4 PolyI:C
15 months (IHC)
Systemic infection in TgAD mice
5 NaCI, 7 PolyI:C
15 months (IHC)
5 NaCI, 5 PolyI:C
3 weeks (E)
Prenatal infection in non-Tg mice: longitudinal changes in cytokine and brain APP/pTau levels and cognitive performance
4 NaCI, 4 PolyI:C
3 months (IB/E)
5 NaCI, 5 PolyI:C
5 months (E)
16 NaCI, 15 PolyI:C
5 months (BT)
4 NaCI, 4 PolyI:C
6 months (IB/E)
6 NaCI, 6 PolyI:C
12 months (IB)
4 NaCI, 5 PolyI:C
15 months (IB/E)
6 NaCI, 6 PolyI:C
15 months (IHC)
18 NaCI, 13 PolyI:C
20 months (BT)
6 NN*, 6 NP, 6N*, 7 PP
12 months (IB)
Prenatal and adult infection in non-Tg mice
10 NP, 11 PP
15 months (IHC)
6 NN, 5 NP, 4 PN, 5 PP
18 months (IHC)
Polyriboinosinic-polyribocytidilic acid injections
Pregnant mouse dams of the C57Bl/6 J strain were given a single intravenous injection of 5 mg/kg PolyI:C potassium salt (P9582, 50 mg; Sigma-Aldrich Chemie GmbH, Buchs, Switzerland) dissolved in 0.9% saline (NaCl) with an injection volume of 5 ml/kg body weight or an equivalent volume of saline at gestation day (GD) 17. The animals were mildly restrained during the injection procedure using an acrylic mouse restrainer, and after the injection, were immediately placed back in their home cage and left undisturbed until the first cage change at 1 week after delivery.
To assess a putative cognitive decline in aged mice exposed to a prenatal immune challenge, adult (5 months) and aged (20 months) mice were tested in the elevated Y-maze spontaneous alternation task (Table 1). This task is used to measure spatial recognition memory that depends on proper hippocampal functions, as confirmed by the severe memory impairment after excitotoxic lesion of this brain region . It is based on the innate tendency of rodents to explore novel environments, that is, their preference to investigate a new arm of the maze rather than returning to one that was previously visited. This paradigm avoids the effects of other methods such as punishment (such as electric shock) or reward (such as food following phases of deprivation) that are commonly used in other paradigms and may have non-specific effects on the results. In addition, it does not require learning of a rule, thus it is useful for studying performance in rodents, particularly in aged animals.
The Y-maze apparatus was made of transparent Plexiglas and consisted of three identical arms (500 mm long × 90 mm wide) surrounded by transparent Plexiglas walls 100 mm in height. The three arms radiated from a central triangle (80 mm on each side) and were spaced 120 degrees from each other. The floor of the maze was covered with sawdust bedding, which was changed between each test run. The maze was elevated 900 mm above the floor, and was positioned in a well-lit room enriched with distal spatial cues. A digital camera was mounted above the Y-maze apparatus. Images were captured at a rate of 5 Hz, and transmitted to a PC running the EthoVision tracking system (Noldus Information Technology), which calculated the total distance moved and the number of entries into the three arms and the center zone of the Y-maze.
Mice were placed in the center of the Y maze. An observer in an adjacent room viewed the mice through a video camera, and recorded the number and sequence of arm entries (defined as entry of the whole body into an arm) during a period of 5 minutes. Alternation was defined as entry into the three arms in any non-repeating order (for example, ABC, BAC, CBA). The percentage alternation was calculated as the total number of alternations divided by the possible alternations given the number of arm entries (total number of arm entries, 2). In addition to the analysis of percentage alternation, the total distance moved and the total number of arm entries were recorded and analyzed to assess general activity during the 5-minute test period.
Mouse brain-tissue preparation
The tissue was fixed by perfusion and processed as described previously . Briefly, animals were deeply anesthetized by intraperitoneal injection of pentobarbital (4 ml/kg body weight; Nembutal™; Lundbeck Inc., Deerfield, IL, USA) and perfused transcardially with PBS (pH 7.4), followed by 4% paraformaldehyde in PBS (Sigma-Aldrich). The brains were post-fixed for 4 hours at 4°C followed by cryoprotection for 24 hours in 30% sucrose. Randomly sampled serial sections (cut at 40 μm) were collected throughout the hippocampal formation starting at bregma level −0.82 to −3.64 , and stored at −20°C in cryoprotectant solution (50 mmol/l sodium phosphate buffer, pH 7.4, containing 15% glucose and 30% ethylene glycol; Sigma-Aldrich) until immunohistochemical or histological evaluation.
Summary of antibodies used in this study
Millipore, Billerica, MA, USA
Mouse monoclonal, clone 4, MAB1522
Covance, Princeton, NJ, USA
Mouse monoclonal, clone 6E10, SIG-39320
Rabbit polyclonal, AB5076
APP (A4), N-term
Mouse monoclonal, clone 22 C11, MAB348
Sigma-Aldrich Chemie GmbH, Buchs, CH
Rabbit polyclonal, A8717
Cell Signaling Technology, USA
Rabbit polyclonal, Nr. 9751
AbD Serotec Ltd, Oxford, UK
Rat monoclonal, MCA341R
Dako Schweiz AG, Baar, CH
Rabbit polyclonal, Z334
Abcam, Cambridge, UK
Rabbit polyclonal, ab4841
Thermo Fisher Scientific, Fremont, CA, USA
Mouse monoclonal, MN1060
Thermo Fisher Scientific
Mouse monoclonal, total Tau, clone Tau-5
After three washes in PBS, tissue sections processed for immunoperoxidase labeling were incubated for 30 minutes at room temperature in biotinylated secondary antibodies (diluted 1:500; Jackson ImmunoResearch Laboratories Inc. West Grove, PA, USA), followed by three rinses in PBS. A commercial kit (Vectastain Kit; Vector Laboratories; Burlingame, CA, USA) was then used with 3,3-diaminobenzidine (DAB; Sigma–Aldrich Inc.), and sections were stained for 5 to 10 minutes . After three washes in PBS, sections were mounted onto gelatinized glass slides and air-dried overnight. The sections were then dehydrated through ethanol, cleared in xylene, mounted with resinous mounting medium (EukittTM™; Sigma-Alrich), and coverslipped.
For immunofluorescence staining, sections were incubated for 30 minutes at room temperature with secondary antibodies coupled to Alexa488 (diluted 1:1000; Molecular Probes, Eugene, OR, USA), Cy3 (diluted 1:500; Molecular Probes) or Cy5 (diluted 1:200; Jackson ImmunoResearch Laboratories).
For the visualization of β-sheet enriched proteinous aggregates, sections were counterstained with thioflavinS (Sigma-Aldrich), involving incubation for 10 minutes in filtered 0.1% aqueous thioflavinS solution at room temperature, followed by two washes for 5 minutes each in 80% EtOH, one wash for 5 minutes in 95% ethanol, and three washes for 5 minutes each in distilled water. Brain sections were air-dried in the dark and mounted with aqueous permanent mounting medium (Dako) containing 1.5 μg/ml DAPI (Thermo Scientific (Schweiz) AG, Reinach, Switzerland).
Double‐or triple‐labeling was visualized by confocal microscopy (LSM-710; Zeiss, Jena, Germany) using 40× (numerical aperture (NA) 1.3) and 63× (NA 1.4) objectives and sequential acquisition of separate channels. Z-stacks of consecutive optical sections (6 to 12; 1024 × 1024 pixels, spaced 0.5 to 1 μm in z) were summed and projected in the z dimension (maximal intensity), and merged using the image analysis software Imaris (Bitplane, Zurich, Switzerland) for visual display. Cropping of images, adjustments of brightness and contrast were performed using Adobe Photoshop and were identical for each staining.
Fluoro-Jade and silver staining
For the detection of dystrophic neurites and proteinous aggregates in immune-challenged 3xTg-AD mice, we used two commercial kits (Fluoro-Jade® Kit (Millipore, Billerica, MA, USA, catalog number AG325) and FD NeuroSilverTM™; Kit II (FD NeuroTechnologies Inc, Ellicott City, MD, USA).
For the Fluoro-Jade staining, tissue sections processed for anti-Aβ immunofluorescence staining were mounted on gelatin-coated slides and air-dried for 2 to 3 hours, then stained in accordance with the manufacturer’s protocol. In brief, slides were immersed in a solution of 1% sodium hydroxide in 80% ethanol for 5 minutes. Next, they were washed in 70% ethanol and distilled water for 2 minutes each. This was followed by 10 minutes incubation in a solution of 0.06% potassium permanganate. After rinsing the slides in distilled water, they were incubated in the staining solution for 20 minutes (dye concentration: 0.0004%). The slides were washed three times in distilled water before air-drying for 1 hour. After immersion in xylene, slides were coverslipped with aqueous mounting medium containing DAPI (Dako, Glostrup, Denmark.
For the silver staining, free-floating perfusion-fixed brain slices (40 μm thick) prepared for normal immunohistochemistry, were processed in accordance with the manufacturer’s instructions. For both procedures, control sections obtained from adult WT mice that had received unilateral intra-hippocampal injections of kainic acid (a manipulation that results in rapid degeneration of hippocampal interneurons and CA1 pyramidal cells ) served as controls.
Quantification of immunohistochemical staining
All immunoperoxidase-stained slides were scanned with an automated upright slide-scanning microscope (Mirax Midi Slide Scanner; Zeiss) in bright-field mode. Images were acquired with a digital camera (1288 × 1040 pixels, with a pixel size of 0.23 μm; AxioCam monochrome charge-coupled display; Zeiss) with a 20× objective (NA 0.8) using the software Pannoramic Viewer (version 1.8.3; 3D Histech Ltd, Budapest, Hungary). A densitometric segmentation analysis was performed for the anti-Aβ (15 months, prenatal infection time course), anti-phosphorylated (p)Tau (hilar mossy cells, 15 months, prenatal infection), anti-glial fibrillary acidic protein (GFAP) (astrocytes, double immune-challenge experiment), and anti-CD68 (activated microglia, double immune-challenge experiment) immunoperoxidase stains to measure the relative percentage and the mean size of the labeled cells covered by the immunoreactive signals in the hippocampal formation. Here, the immunoreactivity (IR) in the neuropil served as the reference for non-specific background staining.
where ∑P is the sum of plaques. Densitometric and stereological values were averaged per animals and used in the statistical analyses.
Processing of human post-mortem tissue
Paraffin wax-embedded brain sections (4 μm thick) of the temporal lobe obtained from a 88-year-old patient diagnosed with AD (generously provided by Professor Dr Manuela Neumann, Institute of Neuropathology, University of Zurich, Zurich, Switzerland) were dewaxed in xylene (3 × 3 minutes) and rehydrated in decreasing concentrations of ethanol (2 × 100%, 96%, 70% ethanol, and 2 × H2O for 3 minutes each) followed by washing for 5 minutes in 50 mmol/l Tris-buffered saline (pH 7.4). Optimal antigen retrieval was achieved by treating sections for 5 minutes with 95% formic acid (involving the anti-amyloid-β1–40/42 antibody, 1:200, AB5076; Millipore) or a citrate buffer (0.1 mol/l citric acid and 0.1 mol/l tri-sodium citrate-2-hydrate) and microwaving at 84°C followed by treatment with pepsin (involving the anti-N-APP antibody, 1:500, MAB348, clone 22C11; Millipore) as described. Before the primary antibody incubation, sections were treated for 1 hour with blocking solution (PBS (pH 7.4) containing 5% normal horse serum, 5% normal goat serum and 4% BSA). Primary antibodies were diluted in PBS (pH 7.4, with 2.5% normal horse serum, 2.5% normal goat serum and 2% BSA added), and sections were then incubated with the primary antibody overnight in a wet histochamber at 4°C under constant agitation. Sections were washed three times for 5 minutes each in PBS(pH 7.4). Secondary antibodies coupled to either Alexa Fluor 488 or Cy3 (diluted 1:1000 and 1:500, respectively, in the same buffer as for primary antibodies; Molecular Probes and Invitrogen, respectively) were added to the sections and incubated for 45 minutes. To reduce auto-fluorescence of intracellular lipofuscin in aged cells, we adapted the protocol described previously . Sections were treated with 0.1% Sudan Black (Carl Roth GmbH, Karlsruhe, Germany) in 70% methanol for 2 minutes in a wet histochamber and briefly washed with PBS before coverslipping. Sections were mounted with aqueous mounting medium containing DAPI (Dako) to visualize cell nuclei. Laser scanning confocal microscopy and digital imaging was performed as described above for the murine immunofluorescence staining.
Protein extracts and immunoblotting
Western blotting parameters
Run time, hours/voltage
Transfer time hours
Blocking reagent/time, hours
Secondary antibody incubation, hours
1.5 to 2/125
2/125 or 2.5/1003
ELISA microtiter plates (catalog number EZBRAIN40/42; Millipore) were used for the quantitative analysis of Aβ1–40 and Aβ1–40/42. In addition, mouse-specific Aβ and IL-1β ELISA microtiter plates (catalog numbers KMB3481 and KMC0011, respectively; Invitrogen, Camarillo, CA, USA) and a fluorokine multi analyte profiling kit for the detection of IL-1α, IL-6, and tumor necrosis factor α (TNFα; catalog number LUM410; R & D Systems, Minneapolis, MN, USA) were employed to confirm the murine origin of the selected proteins. All assays were performed in accordance with the manufacturer’s protocol. Samples (50 μl) of the brain extracts (supernatant and pellet) or terminally collected blood plasma samples were used and run in duplicate. After 30 minutes incubation in the substrate solution (3,3′,5,5′-tetramethylbenzidine/peroxide mixture), the reaction was stopped and the absorption quantified using a microplate reader (Synergy HT Multi-Mode; BioTek Instruments Inc, Winooski, VT, USA) measuring the difference at 450 nm and 650 nm.
All analyses were performed with SPSS for Windows (version 16; SPSS Schweiz AG, Zug, Switzerland). ANOVA with ‘treatment’ as the main between-subjects factor was performed for the behavioral test, the immunohistochemical and immunoblotting analyses (anti-Aβ, anti-pTau, anti-GFAP, anti-CD68 IR) involving double immune-challenged mice (PP; PolyI:C at GD17 and in adulthood), and their controls for NN (NaCl at GD17 and in adulthood), NP (NaCl at GD17, PolyI:C in adulthood), and PN (PolyI:C at GD17, NaCl in adulthood). Significant main effects were explored further and the corresponding mean values between groups were compared using Fisher's least significant difference (LSD) test. Planned comparisons (NaCl versus PolyI:C, NP versus PP) with Mann–Whitney U-tests were used for the single and double immune-challenge experiments involving immunohistochemical, immunoblotting, and ELISA experiments assessing APP/Aβ and pTau levels. Statistical significance was set at P<0.05.
Taken together, the results presented here demonstrate that systemic immune challenges not only represent a crucial trigger but also a potent driver of AD-like neuropathology in both environmentally (PolyI:C exposure at GD17) and genetically (3xTg-AD) predisposed animals, in agreement with the notion that peripheral infections and inflammation in patients with AD are associated with more rapid cognitive decline and exacerbation of their symptoms .
In this study, we provide the first evidence that a systemic immune challenge during a late-gestational time window (GD17) predisposes WT mouse offspring to pathological brain aging and cognitive decline. A second immune challenge aggravates this phenotype and is sufficient to precipitate significant and widespread amyloid-associated and pTau-associated neuropathology in the aged offspring. By administering a single systemic infection to genetically predisposed transgenic mice (3xTg-AD) overexpressing the human variants of AD-relevant genes , we confirmed that the AD-like neuropathological hallmarks seen in double immune-challenged WT mice represent an early AD precursor stage, which is also seen in human patients with AD. Thus, the PolyI:C model offers a unique opportunity to identify molecular targets that are strongly affected by chronic inflammatory conditions and involved in the initiation of the typical neuropathology characteristic of sporadic AD.
Our experimental design builds on our previous studies, in which we showed that middle (GD9) and late (GD17) gestational periods correspond to two windows of enhanced vulnerability to inflammatory cytokines . Whereas mid-gestational immune challenges resulted in neurochemical and behavioral impairments that were releveant for schizophrenia , late-gestational PolyI:C exposure has a detrimental and long-term effect on cognitive performance during adulthood and aging , suggesting a causal link between disturbances of late embryonic development and risk of AD-like neuropathology .
In this study, we extended our previous findings that PolyI:C injection at GD17 induced a significant elevation of maternal and fetal cytokines at both protein and mRNA levels , by showing that this insult also has long-term consequences on basal cytokine levels in adulthood. We detected significantly increased levels of both circulating and brain IL-1β in PolyI:C-treated versus saline-treated mice throughout aging. It is conceivable that these changes are linked to the detrimental effect of the prenatal immune challenge on the developing microglia, the macrophage-derived immune cells of the brain that start to colonize the brain during late-gestational stages [43, 44], potentially affecting their proliferative and also phagocytotic capacity during adulthood and aging. In addition, the immune challenge may prime or activate microglia, thereby creating an innate immune memory, allowing a faster and exaggerated response upon further immune challenges and exposures to adverse stimuli . Indeed, in double immune-challenged mice we found a pronounced increase in activated microglia in the hippocampus before the accumulation of extracellular amyloid depositions. Furthermore, chronic elevation of several cytokines after a late-gestational immune challenge may also have a detrimental effect not only on neurons but also on microglia, by promoting cytokine-mediated release of reactive oxygen species . In combination with additional immune challenges in adulthood, these pathophysiological changes may detoriate, resulting in loss of the neuroprotective functions of microglia and even microglial degeneration, which in turn may contribute to the spread of neuropathology in AD. Experimental evidence for this hypothesis is provided by in vitro findings showing that repetitive stimulation of microglia by mitogens induces telomere shortening and replicative senescence , and that interferon-γ exposure is sufficient to trigger activation-induced microglia cell death . Hence, further morphological and biochemical investigations will be required to determine whether microglia in the aged immune-challenged WT mice may also share characteristics of fragmentation and degeneration, a distinct neuropathologic feature preceding the spread of tau and neuritic plaque pathology in patients with AD .
In this study, we also found that chronic rises in inflammatory cytokines are accompanied by significant rises in APP and its proteolytic fragments, and an increase in Tau phosphorylation and somatodendritic mislocalization in aged animals. These observations are in agreement with data showing that ectopic IL-1 application increases APP production and its processing [10, 49, 50], and increases Tau hyperphosphorylation . In addition, we found that an increase in C-terminal APP fragments, including AICD, which potentially contributes to the neuroinflammation response , was one of the first significant changes discovered in the course of aging of prenatally challenged animals. It is also plausible, however, that an early life insult may result in long-lasting epigenetic changes in the promoter region of AD risk genes, as previously proposed , providing an alternative explanation for the chronic inflammatory state occurring after prenatal immune challenge.
Based on recent findings showing that systemic inflammation exacerbates the course of the disease both in patients with AD  and in rodent models of the disease , we tested the effect of the second immune challenge in adulthood on AD-relevant neuropathology. Analysis of these double immune-challenged mice showed a strong aggravation of the phenotype that was present after a single prenatal infection. Moreover, we detected the appearance of extracellular APP/Aβ-positive deposits in the entorhinal and piriform cortex and the formation of intracellular Tau aggregates, selectively in the hippocampal formation. Hence, dysfunction of the brain’s immune system, triggered through systemic infections, may play a key role in initiation of AD pathogenesis, in line with previously proposed hypotheses [37, 54].
The existence of APP/Aβ-positive deposits in double immune-challenged non-transgenic mice, in brain areas which are among the first affected in humans  indicated that better characterization of these amyloid-like plaques is needed. Using N-terminal specific anti-APP and anti-Aβ1–40/42 antibodies, we showed that, whereas the deposits consisted of APP and N-terminus-containing APP fragments, Aβ/CTFs were enriched only in cells closely associated with plaques. The absence of significant accumulations of mouse Aβ peptides in these APP deposits might be due to the different aggregation properties of the rodent compared with the human form , or might be linked to the premature age and disease stage in these animals (15 months). It will therefore be important to determine the precise identity of the different APP fragments produced in aged single and double immune-challenged mice.
Importantly, other laboratories have also reported rodent APP in the core of Aβ deposits in variuos transgenic AD mouse strains [55, 57]. Hence, we conclude that accumulation of APP and its N-terminus-containing fragments precedes the formation of senile plaques, representing a crucial and conserved precursor stage of this typical neuropathologic hallmark.
In agreement with recent experimental evidence showing that hyperphosphorylated, non-fibrillary Tau may have a key role in eliciting behavioral impairments in vivo[34, 58], we were able to show in this study that the Tau hyperphosphorylation and mislocalization into somatodendritic compartments is accompanied by severe deficit in prenatally challenged mice compared with saline-treated controls. However, it would be interesting to confirm if a second immune challenge during aging results in progressive decline of cognitive functions, as indicated by the recent findings in patients with AD, providing evidence that acute and chronic systemic inflammation is associated with an increase in cognitive decline . Although we did not observe the formation of NFTs in immune-challenged mice, a second immune challenge in our model elicited the intraneuronal aggregation of Tau, probably involving the formation of PHFs. Further studies will be required to characterize the nature of these and potentially other phosphorylated epitopes affected by the chronic inflammatory state. We have recently shown that NFT-like structures can form in the brain of AD mice lacking tau transgene expression, by genetic reduction of the reelin gene . Interestingly, formation of these NFT-like structures was associated with a high Aβ plaque burden, and prominent neuroinflammation and neurodegeneration . Hence, it is reasonable to suggest that the formation of PHFs we saw in double immune-challenged WT mice would eventually develop into NFTs at older age and later stages of the disease.
In summary, and based on the results presented and discussed here, we propose a model (Figure 9G) in which the mendelian mutations underlying familial AD cause profound changes in APP metabolism, inducing a neuroinflammatory response capable of driving and aggravating AD neuropathology during aging. Repeated systemic immune challenges, by contrast, induce chronic neuroinflammation and may accelerate senescence of microglia cells. Reduction in their neuroprotective function is expected to impair both APP and pTau homeostasis. This in turn could trigger a vicious cycle that will lead, over the course of aging, to the formation of senile Aβ plaques and NFTs, induce degeneration, and ultimately result in clinical dementia. We suggest, therefore, that systemic infections and persistent neuroinflammatory conditions represent major risk factors for the development of sporadic AD in older persons. Importantly, our model complements the long-standing amyloid hypothesis of AD [59, 60], and is in accordance with the recent proposal to revise the current view on the etiology of sporadic AD .
We provide experimental evidence for a causative and exacerbating role of systemic immune challenges on the development of AD-like neuropathology in vivo, thereby supporting recent genome-wide association studies , retrospective epidemiological human studies [5, 6, 12], and the long-standing infection/neuroinflammation-based AD hypotheses [11, 22, 33, 61]. Furthermore, our data support the use of anti-inflammatory drugs  in the treatment of patients during the initial asymptomatic stages of AD , and confirm the importance of identifying early molecular markers of the disease. Finally, with the novel findings of focal accumulations of APP and its proteolytic fragments preceding senile plaque deposition, we offer a very suitable mouse model to elucidate the molecular mechanisms underlying the earliest stages of the typical neuropathologic changes characteristic of sporadic AD.
Bovine serum albumin
This study was supported by the Swiss National Science Foundation grant number 310030–132629 to IK), National Center for Competence in Research (NCCR project 2, Alzheimer’s Disease to IK) Hartmann-Müller Foundation (IK), Gottfried und Julia Bangerter-Rhyner Stiftung (DK, IK), Alzheimer- und Depressions-Fonds (SAMW, DK, IK), Novartis Foundation for Biomedical Research (DK, IK), and European Union’s Seventh Framework Programme (FP7/2007-2011, number 259679, UM). We thank Professor Manuela Neumann for providing the post-mortem human tissue, Professor Frank LaFerla for the 3xTg-AD mice, Professor Manfred Schedlowski for the brain cytokine measurements, Ms Corinne Sidler for her experimental support, and Professors Jean-Marc Fritschy, Hanns-Ulrich Zeilhofer, and Hanns Möhler for their advice, support and input into this project. We are also grateful to Professor Hugh V. Perry for the stimulating discussions and critical reading of the manuscript, as well as to Prof. Sanjay Pimplikar for his support regarding the AICD biochemistry. We also thank all the Animal Services staff of the Institute of Pharmacology and Toxicology for animal husbandry and care.
- McGeer PL, McGeer EG: Local neuroinflammation and the progression of Alzheimer’s disease. J Neurovirol. 2002, 8: 529-538. 10.1080/13550280290100969.View ArticlePubMedGoogle Scholar
- Lue LF, Brachova L, Civin WH, Rogers J: Inflammation, Abeta deposition, and neurofibrillary tangle formation as correlates of Alzheimer's disease neurodegeneration. J Neuropathol Exp Neurol. 1996, 55: 1083-1088.View ArticlePubMedGoogle Scholar
- Edison P, Archer HA, Gerhard A, Hinz R, Pavese N, Turkheimer FE, Hammers A, Tai YF, Fox N, Kennedy A, Rossor M, Brooks DJ: Microglia, amyloid, and cognition in Alzheimer's disease: An [11C](R)PK11195-PET and [11C]PIB-PET study. Neurobiol Dis. 2008, 32: 412-419. 10.1016/j.nbd.2008.08.001.View ArticlePubMedGoogle Scholar
- Lambert JC, Heath S, Even G, Campion D, Sleegers K, Hiltunen M, Combarros O, Zelenika D, Bullido MJ, Tavernier B, Letenneur L, Bettens K, Berr C, Pasquier F, Fiévet N, Barberger-Gateau P, Engelborghs S, De Deyn P, Mateo I, Franck A, Helisalmi S, Porcellini E, Hanon O, Investigators European Alzheimer's Disease Initiative, de Pancorbo MM, Lendon C, Dufouil C, Jaillard C, Leveillard T, Alvarez V, Bosco P, Mancuso M, Panza F, Nacmias B, Bossù P, Piccardi P, Annoni G, Seripa D, Galimberti D, Hannequin D, Licastro F, Soininen H, Ritchie K, Blanché H, Dartigues JF, Tzourio C, Gut I, Van Broeckhoven C, Alpérovitch A, Lathrop M, Amouyel P: Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer’s disease. Nat Genet. 2009, 41: 1094-1099. 10.1038/ng.439.View ArticlePubMedGoogle Scholar
- Schmidt R, Schmidt H, Curb JD, Masaki K, White LR, Launer LJ: Early inflammation and dementia: a 25-year follow-up of the Honolulu-Asia Aging Study. Ann Neurol. 2002, 52: 168-174. 10.1002/ana.10265.View ArticlePubMedGoogle Scholar
- Engelhart MJ, Geerlings MI, Meijer J, Kiliaan A, Ruitenberg A, van Swieten JC, Stijnen T, Hofman A, Witteman JC, Breteler MM: Inflammatory proteins in plasma and the risk of dementia: the Rotterdam study. Arch Neurol. 2004, 61: 668-672. 10.1001/archneur.61.5.668.View ArticlePubMedGoogle Scholar
- Wyss-Coray T: Inflammation in Alzheimer disease: driving force, bystander or beneficial response?. Nat Med. 2006, 12: 1005-1015.PubMedGoogle Scholar
- Rogers J, Strohmeyer R, Kovelowski CJ, Li R: Microglia and inflammatory mechanisms in the clearance of amyloid beta peptide. Glia. 2002, 40: 260-269. 10.1002/glia.10153.View ArticlePubMedGoogle Scholar
- Griffin WS, Stanley LC, Ling C, White L, MacLeod V, Perrot LJ, White CL, Araoz C: Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc Natl Acad Sci USA. 1989, 86: 7611-7615. 10.1073/pnas.86.19.7611.PubMed CentralView ArticlePubMedGoogle Scholar
- Sheng JG, Ito K, Skinner RD, Mrak RE, Rovnaghi CR, Van Eldik LJ, Griffin WS: In vivo and in vitro evidence supporting a role for the inflammatory cytokine interleukin-1 as a driving force in Alzheimer pathogenesis. Neurobiol Aging. 1996, 17: 761-766. 10.1016/0197-4580(96)00104-2.PubMed CentralView ArticlePubMedGoogle Scholar
- Mrak RE, Griffin WS: Glia and their cytokines in progression of neurodegeneration. Neurobiol Aging. 2005, 26: 349-354. 10.1016/j.neurobiolaging.2004.05.010.View ArticlePubMedGoogle Scholar
- McGeer PL, Schulzer M, McGeer EG: Arthritis and anti-inflammatory agents as possible protective factors for Alzheimer's disease: a review of 17 epidemiologic studies. Neurology. 1996, 47: 425-432. 10.1212/WNL.47.2.425.View ArticlePubMedGoogle Scholar
- Aisen PS, Schafer KA, Grundman M, Pfeiffer E, Sano M, Davis KL, Farlow MR, Jin S, Thomas RG, Thal LJ, Alzheimer's Disease Cooperative Study: Effects of rofecoxib or naproxen versus placebo on Alzheimer disease progression: a randomized controlled trial. JAMA. 2003, 289: 2819-2826. 10.1001/jama.289.21.2819.View ArticlePubMedGoogle Scholar
- Soininen H, West C, Robbins J, Niculescu L: Long-term efficacy and safety of celecoxib in Alzheimer's disease. Dement Geriatr Cogn Disord. 2007, 23: 8-21. 10.1159/000096588.View ArticlePubMedGoogle Scholar
- Thal LJ, Ferris SH, Kirby L, Block GA, Lines CR, Yuen E, Assaid C, Nessly ML, Norman BA, Baranak CC, Reines SA, Rofecoxib Protocol 078 study group: A randomized, double-blind, study of rofecoxib in patients with mild cognitive impairment. Neuropsychopharmacology. 2005, 30: 1204-1215. 10.1038/sj.npp.1300690.View ArticlePubMedGoogle Scholar
- Breitner JC, Baker LD, Montine TJ, Meinert CL, Lyketsos CG, Ashe KH, Brandt J, Craft S, Evans DE, Green RC, Ismail MS, Martin BK, Mullan MJ, Sabbagh M, Tariot PN, ADAPT Research Group: Extended results of the Alzheimer's disease anti-inflammatory prevention trial. Alzheimers Dement. 2011, 7: 402-411. 10.1016/j.jalz.2010.12.014.PubMed CentralView ArticlePubMedGoogle Scholar
- Alexopoulou L, Holt AC, Medzhitov R, Flavell RA: Recognition of double-stranded RNA and activation of NF-kappaB by Toll-like receptor 3. Nature. 2001, 413: 732-738. 10.1038/35099560.View ArticlePubMedGoogle Scholar
- Cunningham C, Campion S, Teeling J, Felton L, Perry VH: The sickness behaviour and CNS inflammatory mediator profile induced by systemic challenge of mice with synthetic double-stranded RNA (PolyI:C). Brain Behav Immun. 2007, 21: 490-502. 10.1016/j.bbi.2006.12.007.View ArticlePubMedGoogle Scholar
- Meyer U, Nyffeler M, Engler A, Urwyler A, Schedlowski M, Knuesel I, Yee BK, Feldon J: The time of prenatal immune challenge determines the specificity of inflammation-mediated brain and behavioral pathology. J Neurosci. 2006, 26: 4752-4762. 10.1523/JNEUROSCI.0099-06.2006.View ArticlePubMedGoogle Scholar
- Meyer U, Nyffeler M, Yee BK, Knuesel I, Feldon J: Adult brain and behavioral pathological markers of prenatal immune challenge during early/middle and late fetal development in mice. Brain Behav Immun. 2008, 22: 469-486. 10.1016/j.bbi.2007.09.012.View ArticlePubMedGoogle Scholar
- Knuesel I, Nyffeler M, Mormède C, Muhia M, Meyer U, Pietropaolo S, Yee BK, Pryce CR, LaFerla FM, Marighetto A, Feldon J: Age-related accumulation of Reelin in amyloid-like deposits. Neurobiol Aging. 2009, 30: 697-716. 10.1016/j.neurobiolaging.2007.08.011.View ArticlePubMedGoogle Scholar
- Knuesel I: Reelin-mediated signaling in neuropsychiatric and neurodegenerative diseases. Prog Neurobiol. 2010, 91: 257-274. 10.1016/j.pneurobio.2010.04.002.View ArticlePubMedGoogle Scholar
- Oddo S, Caccamo A, Shepherd JD, Murphy MP, Golde TE, Kayed R, Metherate R, Mattson MP, Akbari Y, LaFerla FM: Triple-transgenic model of Alzheimer's disease with plaques and tangles: intracellular Abeta and synaptic dysfunction. Neuron. 2003, 39: 409-421. 10.1016/S0896-6273(03)00434-3.View ArticlePubMedGoogle Scholar
- Deacon RM, Bannerman DM, Kirby BP, Croucher A, Rawlins JN: Effects of cytotoxic hippocampal lesions in mice on a cognitive test battery. Behav Brain Res. 2002, 133: 57-68. 10.1016/S0166-4328(01)00451-X.View ArticlePubMedGoogle Scholar
- Paxinos G, Franklin K: The mouse brain in stereotaxic coordinates. 2001, San Diego: Academic, 2Google Scholar
- Doehner J, Madhusudan A, Konietzko U, Fritschy JM, Knuesel I: Co-localization of Reelin and proteolytic AbetaPP fragments in hippocampal plaques in aged wild-type mice. J Alzheimers Dis. 2010, 19: 1339-1357.PubMedGoogle Scholar
- Kocherhans S, Madhusudan A, Doehner J, Breu KS, Nitsch RM, Fritschy JM, Knuesel I: Reduced Reelin expression accelerates amyloid-beta plaque formation and tau pathology in transgenic Alzheimer's disease mice. J Neurosci. 2010, 30: 9228-9240.View ArticlePubMedGoogle Scholar
- Knuesel I, Riban V, Zuellig RA, Schaub MC, Grady RM, Sanes JR, Fritschy JM: Increased vulnerability to kainate-induced seizures in utrophin-knockout mice. Eur J Neurosci. 2002, 15: 1474-1484. 10.1046/j.1460-9568.2002.01980.x.View ArticlePubMedGoogle Scholar
- Schnell SA, Staines WA, Wessendorf MW: Reduction of lipofuscin-like autofluorescence in fluorescently labeled tissue. J Histochem Cytochem. 1999, 47: 719-730. 10.1177/002215549904700601.View ArticlePubMedGoogle Scholar
- Pimplikar SW, Suryanarayana A: Detection of APP intracellular domain in brain tissue. Methods Mol Biol. 2011, 670: 85-91.View ArticlePubMedGoogle Scholar
- Anisman H, Gibb J, Hayley S: Influence of continuous infusion of interleukin1-beta on depression-related processes in mice: corticosterone, circulating cytokines, brain monoamines, and cytokine mRNA expression. Psychopharmacology. 2008, 199: 199-244. 10.1007/s00213-008-1157-0.View ArticleGoogle Scholar
- Hanisch UK, Kettenmann H: Microglia: active sensor and versatile effector cells in the normal and pathologic brain. Nat Neurosci. 2007, 10: 1387-1394. 10.1038/nn1997.View ArticlePubMedGoogle Scholar
- Ghosal K, Stathopoulos A, Pimplikar SW: APP intracellular domain impairs adult neurogenesis in transgenic mice by inducing neuroinflammation. PLoS One. 2010, 5: e11866-10.1371/journal.pone.0011866.PubMed CentralView ArticlePubMedGoogle Scholar
- Hoover BR, Reed MN, Su J, Penrod RD, Kotilinek LA, Grant MK, Pitstick R, Carlson GA, Lanier LM, Yuan LL, Ashe KH, Liao D: Tau mislocalization to dendritic spines mediates synaptic dysfunction independently of neurodegeneration. Neuron. 2010, 68: 1067-1081. 10.1016/j.neuron.2010.11.030.PubMed CentralView ArticlePubMedGoogle Scholar
- Perry VH, Nicoll JA, Holmes C: Microglia in neurodegenerative disease. Nat Rev Neurol. 2010, 6: 193-201. 10.1038/nrneurol.2010.17.View ArticlePubMedGoogle Scholar
- Rossi D, Volterra A: Astrocytic dysfunction: insights on the role in neurodegeneration. Brain Res Bull. 2009, 80: 224-232. 10.1016/j.brainresbull.2009.07.012.View ArticlePubMedGoogle Scholar
- Fuller S, Steele M, Münch G: Activated astroglia during chronic inflammation in Alzheimer's disease–do they neglect their neurosupportive roles?. Mutat Res. 2010, 690: 40-49. 10.1016/j.mrfmmm.2009.08.016.View ArticlePubMedGoogle Scholar
- Braak H, Braak E: Neuropathological staging of Alzheimer-related changes. Acta Neuropathol. 1991, 82: 239-259. 10.1007/BF00308809.View ArticlePubMedGoogle Scholar
- Grammas P, Yamada M, Zlokovic B: The cerebromicrovasculature: a key player in the pathogenesis of Alzheimer’s disease. J Alzheimers Dis. 2002, 4: 217-223.PubMedGoogle Scholar
- Weller RO, Preston SD, Subash M, Carare RO: Cerebral amyloid angiopathy in the aetiology and immunotherapy of Alzheimer disease. Alzheimers Res Ther. 2009, 1: 6-10.1186/alzrt6.PubMed CentralView ArticlePubMedGoogle Scholar
- Arnaud L, Robakis NK, Figueiredo-Pereira ME: It may take inflammation, phosphorylation and ubiquitination to ‘tangle’ in Alzheimer’s disease. Neurodegener Dis. 2006, 3: 313-319. 10.1159/000095638.View ArticlePubMedGoogle Scholar
- Holmes C, Cunningham C, Zotova E, Culliford D, Perry VH: Proinflammatory cytokines, sickness behavior, and Alzheimer disease. Neurology. 2011, 77: 212-218. 10.1212/WNL.0b013e318225ae07.PubMed CentralView ArticlePubMedGoogle Scholar
- Herbomel P, Thisse B, Thisse C: Zebrafish early macrophages colonize cephalic mesenchyme and developing brain, retina, and epidermis through a M-CSF receptor-dependent invasive process. Dev Biol. 2001, 238: 274-288. 10.1006/dbio.2001.0393.View ArticlePubMedGoogle Scholar
- Ginhoux F, Greter M, Leboeuf M, Nandi S, See P, Gokhan S, Mehler MF, Conway SJ, Ng LG, Stanley ER, Samokhvalov IM, Merad M: Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science. 2010, 330: 841-845. 10.1126/science.1194637.PubMed CentralView ArticlePubMedGoogle Scholar
- Colton CA, Chernyshev ON, Gilbert DL, Vitek MP: Microglial contribution to oxidative stress in Alzheimer's disease. Ann N Y Acad Sci. 2000, 99: 292-307.Google Scholar
- Flanary BE, Streit WJ: Progressive telomere shortening occurs in cultured rat microglia, but not astrocytes. Glia. 2004, 45: 75-88. 10.1002/glia.10301.View ArticlePubMedGoogle Scholar
- Yun HJ, Yoon JH, Lee JK, Noh KT, Yoon KW, Oh SP, Oh HJ, Chae JS, Hwang SG, Kim EH, Maul GG, Lim DS, Choi EJ: Daxx mediates activation-induced cell death in microglia by triggering MST1 signalling. EMBO J. 2011, 30: 2465-2476. 10.1038/emboj.2011.152.PubMed CentralView ArticlePubMedGoogle Scholar
- Streit WJ, Braak H, Xue QS, Bechmann I: Dystrophic (senescent) rather than activated microglial cells are associated with tau pathology and likely precede neurodegeneration in Alzheimer’s disease. Acta Neuropathol. 2009, 118: 475-485. 10.1007/s00401-009-0556-6.PubMed CentralView ArticlePubMedGoogle Scholar
- Buxbaum JD, Oishi M, Chen HI, Pinkas-Kramarski R, Jaffe EA, Gandy SE, Greengard P: Cholinergic agonists and interleukin 1 regulate processing and secretion of the Alzheimer beta/A4 amyloid protein precursor. Proc Natl Acad Sci USA. 1992, 89: 10075-10078. 10.1073/pnas.89.21.10075.PubMed CentralView ArticlePubMedGoogle Scholar
- Liu L, Aboud O, Jones RA, Mrak RE, Griffin ST, Barger SW: Apolipoprotein E expression is elevated by interleukin 1 and other interleukin 1-induced factors. J Neuroinflammation. 2011, 8: 175-10.1186/1742-2094-8-175.PubMed CentralView ArticlePubMedGoogle Scholar
- Li Y, Liu L, Barger SW, Griffin WS: Interleukin-1 mediates pathological effects of microglia on tau phosphorylation and on synaptophysin synthesis in cortical neurons through a p38-MAPK pathway. J Neurosci. 2003, 23: 1605-1611.PubMed CentralPubMedGoogle Scholar
- Lahiri DK, Maloney B, Zawia NH: The LEARn model: an epigenetic explanation for idiopathic neurobiological diseases. Mol Psychiatry. 2009, 14: 992-1003. 10.1038/mp.2009.82.View ArticlePubMedGoogle Scholar
- Johnston H, Boutin H, Allan SM: Assessing the contribution of inflammation in models of Alzheimer's disease. Biochem Soc Trans. 2011, 39: 886-890. 10.1042/BST0390886.View ArticlePubMedGoogle Scholar
- Streit WJ: Microglia and Alzheimer's disease pathogenesis. J Neurosci Res. 2004, 77: 1-8. 10.1002/jnr.20093.View ArticlePubMedGoogle Scholar
- Jankowsky JL, Younkin LH, Gonzales V, Fadale DJ, Slunt HH, Lester HA, Younkin SG, Borchelt DR: Rodent A beta modulates the solubility and distribution of amyloid deposits in transgenic mice. J Biol Chem. 2007, 282: 22707-22720. 10.1074/jbc.M611050200.PubMed CentralView ArticlePubMedGoogle Scholar
- Kyrkanides S, Tallents RH, Miller JH, Olschowka ME, Johnson R, Yang M, Olschowka JA, Brouxhon SM, O’Banion MK: Osteoarthritis accelerates and exacerbates Alzheimer’s disease pathology in mice. J Neuroinflammation. 2011, 8: 112-10.1186/1742-2094-8-112.PubMed CentralView ArticlePubMedGoogle Scholar
- van Groen T, Kiliaan AJ, Kadish I: Deposition of mouse amyloid beta in human APP/PS1 double and single AD model transgenic mice. Neurobiol Dis. 2006, 23: 653-662. 10.1016/j.nbd.2006.05.010.View ArticlePubMedGoogle Scholar
- Brunden KR, Trojanowski JQ, Lee VM: Evidence that non-fibrillar tau causes pathology linked to neurodegeneration and behavioral impairments. J Alzheimers Dis. 2008, 14: 393-399.PubMed CentralPubMedGoogle Scholar
- Selkoe DJ: The molecular pathology of Alzheimer's disease. Neuron. 1991, 6: 487-498. 10.1016/0896-6273(91)90052-2.View ArticlePubMedGoogle Scholar
- Hardy JA, Higgins GA: Alzheimer's disease: the amyloid cascade hypothesis. Science. 1992, 256: 184-185. 10.1126/science.1566067.View ArticlePubMedGoogle Scholar
- Herrup K: Reimagining Alzheimer's disease–an age-based hypothesis. J Neurosci. 2010, 30: 16755-16762. 10.1523/JNEUROSCI.4521-10.2010.PubMed CentralView ArticlePubMedGoogle Scholar
- Spillantini MG, Goedert M: Tau protein pathology in neurodegenerative diseases. TINS. 1998, 21: 428-433.PubMedGoogle Scholar
- Porzig R, Singer D, Hoffman R: Epitope mapping of mAbs AT8 and Tau5 directed against hyperphosphorylated regions of the human tau protein. Biochem Biophys Res Commun. 2007, 358: 644-649. 10.1016/j.bbrc.2007.04.187.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.