Neurotoxicity Research

, Volume 15, Issue 1, pp 3–14

Consumption of Grape Seed Extract Prevents Amyloid-β Deposition and Attenuates Inflammation in Brain of an Alzheimer’s Disease Mouse

Authors

  • Yan-Jiang Wang
    • Department of Human Physiology and Centre for NeuroscienceFlinders University
    • Department of Neurology, Chongqing Daping HospitalThird Military Medical University
  • Philip Thomas
    • Australian Commonwealth Scientific and Research Organization (CSIRO) Human Nutrition
  • Jin-Hua Zhong
    • Department of Human Physiology and Centre for NeuroscienceFlinders University
  • Fang-Fang Bi
    • Department of Human Physiology and Centre for NeuroscienceFlinders University
  • Shantha Kosaraju
    • CSIRO Food Science Australia
  • Anthony Pollard
    • Department of Human Physiology and Centre for NeuroscienceFlinders University
    • Australian Commonwealth Scientific and Research Organization (CSIRO) Human Nutrition
    • Department of Human Physiology and Centre for NeuroscienceFlinders University
Article

DOI: 10.1007/s12640-009-9000-x

Cite this article as:
Wang, Y., Thomas, P., Zhong, J. et al. Neurotox Res (2009) 15: 3. doi:10.1007/s12640-009-9000-x

Abstract

Polyphenols extracted from grape seeds are able to inhibit amyloid-beta (Aβ) aggregation, reduce Aβ production and protect against Aβ neurotoxicity in vitro. We aimed to investigate the therapeutic effects of a polyphenol-rich grape seed extract (GSE) in Alzheimer’s disease (AD) mice. APPSwe/PS1dE9 transgenic mice were fed with normal AIN-93G diet (control diet), AIN-93G diet with 0.07% curcumin or diet with 2% GSE beginning at 3 months of age for 9 months. Total phenolic content of GSE was 592.5 mg/g dry weight, including gallic acid (49 mg/g), catechin (41 mg/g), epicatechin (66 mg/g) and proanthocyanidins (436.6 mg catechin equivalents/g). Long-term feeding of GSE diet was well tolerated without fatality, behavioural abnormality, changes in food consumption, body weight or liver function. The Aβ levels in the brain and serum of the mice fed with GSE were reduced by 33% and 44%, respectively, compared with the Alzheimer’s mice fed with the control diet. Amyloid plaques and microgliosis in the brain of Alzheimer’s mice fed with GSE were also reduced by 49% and 70%, respectively. Curcumin also significantly reduced brain Aβ burden and microglia activation. Conclusively, polyphenol-rich GSE prevents the Aβ deposition and attenuates the inflammation in the brain of a transgenic mouse model, and this thus is promising in delaying development of AD.

Keywords

Alzheimer’s diseasePolyphenolsGrape seed extractCurcuminAmyloid-betaMicrogliaInflammation

Introduction

Alzheimer’s disease (AD) is the most common form of senile dementia occurring in later life and is a major cause of disability and death in the elderly (World Health Organization 2003). With the world population ageing, it is estimated that the number of people affected with AD will double every 20 years from today’s estimate of 26.6 million to 106.8 million by 2050 (Brookmeyer et al.2007). However, no strong disease-modifying treatment or preventative measure is currently available (Citron 2004).

AD is characterized neuropathologically by deposits of amyloid-beta peptides (Aβ), neurofibrillary tangles, reactive microgliosis and astrogliosis, cerebral amyloid angiopathy and neuronal loss that result in the progressive deterioration of cognition and memory (see Thomas and Fenech 2007, review). According to the amyloid hypothesis, the accumulation of Aβ in the brain is the primary factor driving AD pathogenesis (Hardy and Selkoe 2002). It has been suggested that neuroinflammation may significantly contribute to disease progression and chronicity of AD (Heneka and O’Banion 2007). Therefore, clearance of Aβ from the brain and anti-inflammation represent potential important strategies that may be used to prevent and treat the disease (Citron 2004; Wang et al. 2006b).

Epidemiological studies have shown that consumption of diets rich in anti-inflammatory agents, such as those found in fruits and vegetables, or anti-inflammation drugs, may lower the risk of developing age-related neurodegenerative diseases such as Parkinson’s disease and AD (Lau et al. 2005; Barberger-Gateau et al. 2007; McGeer and McGeer 2007). Polyphenols from grape seeds extract (GSE) have been suggested to be able to inhibit Aβ aggregation, reduce Aβ production, protect against Aβ neurotoxicity and attenuate oxidative stress in vitro (Bastianetto et al. 2000; Jang and Surh 2003; Ono et al. 2003, 2005; Savaskan et al. 2003; Li et al. 2004; Marambaud et al. 2005; Mancuso et al. 2007; Riviere et al. 2007). GSE has been widely used as food additives in order to benefit health and chronic illness due to its anti-oxidation effects (for review, see Shi et al.2003). However, whether polyphenols from grape seeds can prevent or slow down the pathogenesis of the disease or reduce brain pathology in AD patients or animal models is not clear. In the present study, we investigated the effects of polyphenol-rich GSE on Aβ deposition and inflammation in a transgenic AD mouse model.

Materials and Methods

Transgenic Mouse Model

Approval for this study was obtained from CSIRO Human Nutrition and Flinders University Animal Ethics committees. APPSwe/PS1dE9 transgenic mice were provided by Jackson Laboratory, USA. These mice were constructed on a C57BL/6 background and bear a chimeric mouse/human (Mo/Hu) APP695 with mutations linked to familial AD (KM 593/594 NL) and human PS1 carrying the exon-9-deleted variant associated with familial AD (PS1dE9) in one locus under control of a brain- and neuron-specific murine Thy-1 promoter element (Jankowsky et al.2001). Genotyping of mice was performed by PCR following the supplier’s instructions. Mice were maintained on ad libitum food and water with a 12-h light/dark cycle.

The 3-month-old APPSwe/PS1dE9 transgenic mice were randomly assigned to a polyphenol group (N = 12, fed with polyphenol diet), curcumin group (N = 12, fed with curcumin diet) and normal diet group (N = 12, fed with normal diet). The age- and sex-matched wild-type littermates were used as a normal control (N = 20, fed with normal diet). All animal husbandry procedures performed were approved by the Flinders University and CSIRO Human Nutrition Animal Welfare Committees in accordance with NHMRC Australian guidelines in Australia.

HPLC Analysis of Polyphenols in GSE

GSE (Vinlife N05010) was purchased from Tarac Technologies P.L. and was characterized by high performance liquid chromatography (HPLC) analysis without any further extraction process. GSE, dissolved in 80% methanol acidified with 0.1% HCl to obtain final concentration of 2.0 mg/ml, was injected (10 μl) into the HPLC column for the analysis of polyphenolic compounds. HPLC analysis was carried out according to methods described previously (Kammerer et al.2004). Analytical HPLC was run at 25°C and monitored at 280 nm (hydroxybenzoic acids and flavanols), 320 nm (hydroxycinnamic acids, stilbenes) and 370 nm (flavonols).

Diets

Before experiments, all animals were fed with commercial standard diet pellets (Gordon’s Specialty Stock Feeds Pty Ltd., NSW, Australia). All experimental diets were prepared by Specialty Feeds, Glen Forrest, Western Australia. The Control diet was the standard AIN-93G rodent diet (Reeves 1997) consisting of 39.7% corn starch, 20% casein (vitamin free), 13.2% dextrin, 10% sucrose, 7% soybean oil, 5% powdered cellulose, 3.5% AIN-93G mineral mix, 1% AIN-93G vitamin mix, 0.3% l-cysteine, 0.25% choline bitartrate, 0.001% t-butylhydroquinone and 5% maize starch. Curcumin diet consisted of the AIN-93G diet with the exception that it contained 4.93% maize starch and 0.07% curcumin (Sigma, Cat No.: C1386, USA). GSE diet consisted of the AIN-93G diet with the exception that it contained 3% maize starch and 2% GSE. The dosage of curcumin and GSE was based on previous reports indicating lack of toxicity and in the case of curcumin preventive effects against AD in a mouse model (Yamakoshi et al. 2002; Deshane et al. 2004; Yang et al. 2005). Sufficient diet was prepared for the duration of the study and housed at the Flinders Medical Centre animal holding facility. The animals were fed with the above diets for 9 months starting from when they were 3 months old, at which time no Aβ deposition was formed in the brain of the animal (Garcia-Alloza et al. 2006). Food consumption and animal body weight were monitored every 3 months throughout the study.

Tissue Sampling

Animals were sacrificed by overdosing with chloral hydrate (1.5 g/kg). Blood was sampled from the right atrium of the heart, followed by intracardial perfusion with 100 ml of 0.1% NaNO2 in phosphate buffer. Brains were sampled and weighed on a digital electronic balance with a readability of 1 mg (BX-420H, Shimadzu Scientific Instruments, USA). Left brain hemisphere for histological analysis was fixed in 4% paraformaldehyde (pH 7.4) for 24 h and incubated for 24 h in 30% sucrose for subsequent cryoprotection. Coronal sections of the brain at 35-μm thickness were collected with a cryosectioning microtome and stored at 4°C in PBS containing 0.1% sodium azide until use. Right brain hemisphere was snap frozen in liquid nitrogen and stored at −80°C for future biochemical analysis.

AD-type Pathology and Quantitative Image Analysis

The staining for brain total Aβ, microgliosis, astrogliosis and microhaemorrhage was processed as described previously (Wang et al. 2009). Briefly, three series of six equally spaced tissue sections (~200 μm apart) spanning the hippocampus were randomly selected and stained using free-floating immunohistochemistry for total Aβ (Biotin-conjugated mouse anti-Aβ antibody 6E10, Serotec, USA; 1:1000 dilution), activated microglia (rat monoclonal anti-CD45, Chemicon, USA; 1:2000 dilution) and astrocyte (rabbit polyclonal anti-glial fibrillary acidic protein, Dako, Denmark; 1:1000 dilution), respectively. Sections were incubated overnight with primary antibodies at 4°C, further developed with biotinylated secondary antibodies and the ABC kit (Vector Lab, Burlingame, CA) using diaminobenzidine and glucose oxidase as substrates. Quantification of total Aβ deposit, microgliosis and astrogliosis were performed on images acquired with a digital camera and analysed with NIH Image J.

Images were collected at 4× magnification using constant bulb temperature and exposure, with all images acquired in the same session. The area of neocortex and hippocampus was selected for automatic quantification of Aβ, microglia and astrocyte immunostaining, yielding the area fraction of the total positive staining against the area of tissue analysed. The average of the individual measurements was used to calculate group means and standard errors.

A series of six equally spaced tissue sections (~1 mm apart) spanning the entire brain was mounted and stained for haemosiderin using 2% potassium ferrocyanide in 2% HCl for 15 min, followed by a counterstain in a 1% Neutral Red solution for 10 min at room temperature. Microhaemorrhage events in the form of the number of Prussian blue-positive profiles were counted in the brains of each mouse on all sections, and the average number of haemosiderin deposits was calculated per each brain hemisphere. All image analyses were processed in a blind manner.

Quantification of Aβ Peptide Levels in the Mouse Brain and Plasma by ELISA

ELISA analysis of the brain Aβ was processed as described previously (Wang et al. 2009). Briefly, frozen brain was homogenized and sonicated in water containing 2% sodium dodecyl sulphate (SDS) and protease inhibitors (Boehringer Mannheim, Germany). Homogenates were centrifuged at 100,000g for 1 h at 4°C, and the resultant supernatant was collected, representing the SDS-soluble fraction (Aβ-SDS). The resultant pellet was then extracted in 70% formic acid, centrifuged, and the resultant supernatant was collected, representing the SDS-insoluble fraction (Aβ-FA). Before ELISA assay, SDS extracts were diluted 1:50 and formic acid extracts were neutralized by 1:20 dilution into 1 M Tris phosphate buffer, pH 11, and then diluted 1:20 in sample buffer. Concentration of Aβ40 and of Aβ42 in brain extract and plasma were quantitatively measured by ELISA (catalogue nos. 8940 and 8942; Signet Laboratories, Dedham, MA) according to manufacturer’s instructions. Using the wet weight of brain tissue in the original homogenate, the final values of brain Aβ were expressed as picomoles per gram wet weight of brain.

Quantification of TNFα, IL-1β and IFN-γ in the Mouse Plasma by ELISA

TNFα, IL-1β and IFN-γ in the plasma of mice were measured using ELISA kits (Cat No. 88-7342, 88-7913, 88-7914, eBioscience, USA) as per manufacturer’s instructions.

Assessment of Toxicity of Polyphenol from Grape Seeds

Total bilirubin, alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were analyzed using commercial enzyme assays according to the manufacturer’s instructions (Roche Diagnostics, GmbH, d-68298 Mannheim).

Statistical Analysis

The data in the text and figures are expressed as mean ± SEM, unless otherwise stated. Inter-group comparisons were assayed using one-way ANOVA and post hoc for testing the significance of values. The Spearman correlation coefficient was used to analyze the correlation of brain weight with Aβ level, microglia activation and astrocytosis in the brain and the correlation between brain Aβ level and serum Aβ level. P values less than 0.05 were considered as statistically different. All the analyses were performed using SPSS for Windows version 13.0 (SPSS Inc.).

Results

Chemical Analysis of Polyphenols in GSE

HPLC analysis detected compounds in GSE only at 280 nm (Fig. 1). At this wavelength phenolic acids (gallic acid) and flavanols (catechins) were detected. The level of proanthocyanidins was calculated as a difference between the total peak area at 280 nm and the area of individual peaks that represent the monomers.
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig1_HTML.gif
Fig. 1

HPLC profile of phenolic compounds in GSE. Analytical HPLC was run at 25°C and monitored at 280 nm (hydroxybenzoic acids and flavanols), 320 nm (hydroxycinnamic acids, stilbenes) and 370 nm (flavonols). Compounds were detected at 280 nm including phenolic acids (gallic acid) and flavanols (catechins). The level of proanthocyanidins was calculated as a difference between the total peak area at 280 nm and the area of individual peaks that represent the monomers. The mobile phase consisted of 2% acetic acid in water (solvent A) and 1.0% acetic acid in water and acetonitrile (50:50 v/v, solvent B). The flow rate was 1 ml min−1. The following gradient programme was used: from 10% to 24% solvent B (20 min), from 24% to 30% B (20 min), from 30% to 55% B (20 min), from 55% to 100% B (15 min), 100% B isocratic (8 min), from 100% to 10% B (2 min). The total run time was 85 min

Through spectral characteristics and comparison with standards we detected three main compounds, which were gallic acid (49 mg/g DW), catechin (41 mg/g DW) and epicatechin (66 mg/g DW). The concentration of proanthocyanidins was 436.6 mg catechin Eq./g DW. The total polyphenolic content was 592.5 mg/g DW. A similar total phenolic content was obtained when analysed by the Folin-Ciocalteu method (data not shown).

GSE is Well Tolerated in APPSwe/PS1dE9 Transgenic Mice

The overall goal of the present study was to test the hypothesis that polyphenols from grape seeds may prevent AD-type Aβ-associated pathology. The feed of the polyphenols and curcumin diet was from 3 months of age, when Aβ deposition had not yet formed in the brain, to 12 months of age. In APPSwe/PS1dE9 transgenic mice, Aβ deposition begins at 4 months of age and becomes obvious at 9 months of age (Garcia-Alloza et al. 2006; Wang et al. 2006c).

The mean daily food consumption of the mice was 0.11–0.14 g per gram body weight, the corresponding daily polyphenol consumption was 1.2–1.7 mg per gram body weight and daily curcumin consumption was 77–98 μg per gram body weight. The equivalent consumption in a 60 kg human is about 5.9 g per day for polyphenol and 0.35 g per day for curcumin, as derived using FDA criteria for converting drug equivalent dosages across species, based on body surface area [human equivalent dose in mg/kg = animal dose in mg/kg × (animal weight in kg/human weight in kg)0.33] (Food and Drug Administration 2003).

During the period of the study, no animal death occurred and no behavioural abnormality was observed. We found that the long-term daily consumption for 9 months in APPSwe/PS1dE9 transgenic mice, delivered in the food, did not significantly influence animal body weight (Fig. 2a) and daily food consumption (Fig. 2b).
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig2_HTML.gif
Fig. 2

Polyphenols from grape seeds are well tolerated in APPSwe/PS1dE9 transgenic mice. APPSwe/PS1dE9 transgenic mice were fed Control, Curcumin or GSE diets for 9 months from 3 months of age. In parallel control studies, gender- and age-matched wild-type littermates were fed with Control diet. a Body weight was measured at 3, 6, 9 and 12 months of age. b Food intake was monitored at 3, 6, 9 and 12 months of age, and was calculated as food intake (gram) per gram body weight per day. c Serum indices of liver functional status such as AST and ALT. Points and bar graphs represent group mean (±SEM)

It is important to note that the chronic consumption of polyphenols and curcumin did not cause liver function damage, as reflected by the normal serum levels of bilirubin, aminotransferase (AST) and alanine aminotransferase aspartate (ALT) (Fig. 2c), and bilirubin which was detected at a very low level (<1 μmol/l) in serum of all animals (data not shown).

GSE Reduces Brain and Serum Aβ Levels and Prevents Aβ Deposition in APPSwe/PS1dE9 Transgenic Mice

After 9 months feeding on different diets, Aβ in SDS fraction (Aβ-SDS) and in formic acid fraction (Aβ-FA) were quantified utilizing a sandwich ELISA. Aβ-SDS represents the soluble forms of Aβ, while Aβ-FA represents the insoluble forms of Aβ. The total Aβ level from individual animals was calculated by the sum of total SDS-soluble Aβ (SDS-soluble Aβ42 and Aβ40) and total FA soluble Aβ (FA soluble Aβ42 and Aβ40).

Compared with transgenic mice consuming Control diet, there was a significant reduction in the total brain Aβ burden in mice consuming a GSE diet (P < 0.001) or Curcumin diet (P = 0.01) (ANOVA, F = 10.762, P < 0.001, Fig. 3a). GSE diet consumption led to a 33% reduction in brain Aβ burden, while Curcumin diet consumption resulted in a 22% reduction. Consistently, inter-group comparisons of SDS-Aβ and FA-Aβ, Aβ40 and Aβ42 are essentially the same as those of total Aβ (Fig. 3b, c).
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig3_HTML.gif
Fig. 3

Effects of polyphenol consumption on Aβ levels in the brain and serum of APPSwe/PS1dE9 transgenic mice. Aβ peptide concentration in the brain and serum of each animal was measured using ELISA. a Comparison of total Aβ, Aβ in SDS fraction (Aβ-SDS) and Aβ in formic acid fraction (Aβ-FA) among groups. b Comparison of total Aβ40, Aβ40-SDS and Aβ40-FA. c Comparison of total Aβ42, Aβ42-SDS and Aβ42-FA. d Comparison of total Aβ, Aβ40 and Aβ42 in serum. * and ** denote P < 0.05 or P < 0.01 versus APPSwe/PS1dE9 transgenic mice fed with Control diet

GSE or Curcumin diet consumption also reduced the Aβ concentrations in the serum (ANOVA, F = 10.004, P < 0.001, Fig. 3d). The total Aβ concentration in serum correlated significantly and positively with brain total Aβ burden (Pearson = 0.348, P = 0.021), suggesting the potential value of serum Aβ concentration monitoring to reflect brain Aβ burden during the course of Alzheimer’s dietary prevention studies with polyphenols and curcumin.

Aβ plaques were observed primarily in the neocortical and hippocampal areas of the brain. Quantitative histology analysis using ANOVA also generated similar results (F = 6.873, P = 0.004). Compared with Control diet consumption (Fig. 4a), consumption of GSE (P = 0.002) or Curcumin (P = 0.015) diets reduced the Aβ deposition in the neocortex and hippocampus (Fig. 4b, c). Similar to the brain Aβ levels determined by Aβ ELISA, the total Aβ plaque burden determined by immunohistochemistry was reduced by 45% in the GSE diet group and 33% in the Curcumin diet group. These data suggest polyphenols from grape seeds are effective in reducing Aβ burden and preventing Aβ deposition in the brain.
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig4_HTML.gif
Fig. 4

Effects of polyphenol consumption on Aβ plaque burden in the brain of APPSwe/PS1dE9 transgenic mice. A series of six equally spaced tissue sections (~200 μm apart) spanning the hippocampus were stained using free-floating immunohistochemistry for Aβ plaque (anti-Aβ antibody 6E10, Serotec) and developed with DAB. The area of neocortex and hippocampus was selected for automatic quantification of Aβ plaque immunostaining with ImageJ, yielding the area fraction of the total positive staining against the area of tissue analysed. The average of the individual measurements was used to calculate group means and standard errors. ac Aβ plaques in hippocampus and neocortex of APPSwe/PS1dE9 transgenic mice fed with Control, Curcumin or GSE diets. d Comparison of Aβ plaque area fraction in neocortex and hippocampus among groups. * and ** denote P < 0.05 or P < 0.01 versus APPSwe/PS1dE9 transgenic mice fed with Control diet. Scale bar = 0.5 mm. Original magnification, 4×

GSE Prevents AD-type Neuropathology in APPSwe/PS1dE9 Transgenic Mice

Microgliosis and astrogliosis were observed primarily in the neocortical and hippocampal areas of the brain. We examined the area fraction of CD45+ microglia and GFAP+ astrocytes in neocortical and hippocampal regions.

No obvious microgliosis was observed in brains of wild-type littermates (Fig. 5a), while APPSwe/PS1dE9 transgenic mice showed obvious microgliosis (Fig. 5b). Statistical analysis using ANOVA (F = 10.260, P < 0.001)) showed that, compared with transgenic mice fed with Control diet, the mice fed with GSE had a significant lower level of microgliosis (0.51 ± 0.08 vs. 1.72 ± 0.34, P < 0.001), similarly so were the mice fed with curcumin (1.11 ± 0.16 vs. 1.72 ± 0.34, P = 0.032) (Fig. 5c–e).
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig5_HTML.gif
Fig. 5

Effects of polyphenol consumption on microgliosis in the brain. A series of six equally spaced tissue sections (~200 μm apart) spanning the hippocampus were stained using free-floating immunohistochemistry for activated microglia (rat monoclonal anti-CD45, Millipore) and developed with DAB. The area of neocortex and hippocampus was selected for automatic quantification of activated microglia immunostaining with ImageJ, yielding the area fraction of the total positive staining against the area of tissue analysed. The average of the individual measurements was used to calculate group means and standard errors. a No obvious microgliosis was observed in the brain of wild-type littermates fed with Control diet. bd Microgliosis in hippocampus and neocortex of APPSwe/PS1dE9 transgenic mice fed with Control, Curcumin or GSE diets. e Comparison of CD45 area fraction in neocortex and hippocampus among groups. * and ** denote P < 0.05 or P < 0.01 versus wild-type littermate fed with normal diet, # and ## denote P < 0.05 or P < 0.01 versus APPSwe/PS1dE9 transgenic mice fed with Control diet. Scale bar = 0.5 mm. Original magnification, 4×

Wild-type littermates fed with control diet had obvious astrogliosis (Fig. 6a), which, however, was significantly lower than that in APPSwe/PS1dE9 transgenic mice (Fig. 6b). There was no significant difference in astrogliosis among the groups of transgenic mice fed with Control, Curcumin and GSE diets (ANOVA, F = 2.540, P = 0.097, Fig. 6b–e).
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig6_HTML.gif
Fig. 6

Effects of polyphenol consumption on astrogliosis in the brain. A series of six equally spaced tissue sections (~200 μm apart) spanning the hippocampus were stained using free-floating immunohistochemistry for astrocyte (rabbit polyclonal anti-glial fibrillary acidic protein, Dako) and developed with DAB. The area of neocortex and hippocampus was selected for automatic quantification of astrogliosis immunostaining with ImageJ, yielding the area fraction of the total positive staining against the area of tissue analysed. The average of the individual measurements was used to calculate group means and standard errors. a Astrogliosis in hippocampus and neocortex of wild-type littermates fed with Control diet. bd Astrogliosis in hippocampus and neocortex of APPSwe/PS1dE9 transgenic mice fed with Control, Curcumin or GSE diets.e Comparison of GFAP area fraction in neocortex and hippocampus among groups. * and ** denote P < 0.05 or P < 0.01 versus wild-type littermate fed with Control diet. Scale bar = 0.5 mm. Original magnification, 4×

Characteristic blue haemosiderin-positive profiles were observed primarily in the neocortical, leptomeningeal, hippocampal and thalamic areas of the brain (Fig. 7a). The microhaemorrhage was detected at a rate of 44.0 ± 14.6 per hemibrain in transgenic mice fed with normal diet, which was higher than in wild-type littermates (7.76 ± 1.07 per hemibrain, P = 0.006). Non-significant lower rates of microhaemorrhage were observed in mice fed with GSE (29.0 ± 6.2 per hemibrain, P = 0.270) or Curcumin (32.1 ± 8.4 per hemibrain, P = 0.383) diets when compared with transgenic mice fed with Control diet (ANOVA, F = 0.599, P = 0.557, Fig. 7b).
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig7_HTML.gif
Fig. 7

Effects of GSE and curcumin consumption on microhaemorrhage profiles in the brain. A series of six equally spaced tissue sections (~1 mm apart) spanning the entire brain was mounted and stained for haemosiderin using 2% potassium ferrocyanide in 2% hydrochloric acid, followed by a counterstain in a 1% Neutral Red solution. Microhaemorrhage events in the form of the number of Prussian blue-positive profiles were counted, and the average number and standard error of haemosiderin deposits was calculated per each brain hemisphere. a An example of microhaemorrhage profile (solid arrow) observed in hippocampus. b Comparison of microhaemorrhage profiles per each brain hemisphere among groups. * and ** denote P < 0.05 or P < 0.01 versus wild-type littermate fed with Control diet. Scale bar = 50 μm. Original magnification, 40×

Plasma Levels of Inflammatory Cytokines After GSE Consumption

Activated microglia and microphages secrete cytokines such as IL-1β, TNF-α and IFN-γ. This is a major pathologic event in the progression of inflammatory cascades within the AD brain. We measured the levels of IL-1β, TNF-α and IFN-γ in the plasma. In general, cytokines were detected but at very low levels which were at the lower range of the sensitivity of the ELISA kit used (8–15 pg/ml). Levels of IL-1β and TNF-α tended to be higher in transgenic mice than in wild-type littermates, but the difference between groups did not reach statistical significance (ANOVA, IL-1β, F = 1.881, P = 0.141; TNF-α, F = 1.178, P = 0.325, Fig. 8). The level of IFN-γ was higher in transgenic AD mice fed with Control diet (28.7 ± 10.3 pg/ml) than in their wild-type littermates (14.9 ± 1.2 pg/ml, P = 0.008) on the same diet. Compared with Control diet consumption, GSE (19.0 ± 2.0 pg/ml vs. 28.7 ± 10.3 pg/ml, P = 0.336) and Curcumin (15.1 ± 2.0 pg/ml vs. 28.7 ± 10.3 pg/ml, P = 0.021) diet consumption decreased the IFN-γ level in the plasma of APPSwe/PS1dE9 transgenic mice (ANOVA, F = 2.755, P = 0.049, Fig. 8).
https://static-content.springer.com/image/art%3A10.1007%2Fs12640-009-9000-x/MediaObjects/12640_2009_9000_Fig8_HTML.gif
Fig. 8

Effects of curcumin and GSE consumption on plasma levels of TNFα, IL-1β and IFN-γ. Plasma levels of TNFα, IL-1β and IFN-γ were measured using ELISA kits (eBioscience). ** denotes P < 0.01 versus wild-type littermate fed with normal diet, # denotes P < 0.05 versus APPSwe/PS1dE9 transgenic mice fed with Control diet

Discussion

Both genetics and environment determine the development of many chronic diseases including AD. The genetic mutations of APP and presenilin genes only attribute to a relatively small number (<5%) of the total number of AD patients (familial cases), whereas the majority of AD patients are likely due to environmental factors and other genetic factors affecting Aβ clearance (Hardy and Selkoe 2002). Major environmental factors are likely to include excess or deficiency of dietary constituents consumed on a regular basis which have bioactivity in relevant pathways. Our understanding of how food and drink can potentially influence the development of AD will help to develop and implement treatments that may aid in combating this devastating disease.

In the present study, we report that polyphenols-rich GSE fed for 9 months as a food additive dramatically could prevent the AD development in a genetic mouse model. We found that the chronic consumption of polyphenols extracted from grape seeds was well tolerated, effectively reduced the Aβ burden in the brain and blood, prevented the Aβ deposition and attenuated the microgliosis.

Aβ accumulation and deposition in the brain is one of the histopathological hallmarks of AD. Our study clearly demonstrated that the polyphenols in GSE, when fed to a transgenic mouse model for 9 months, reduced the total brain amyloid burden by 33–45%, depending on the analysis methods used. Aβ40 and Aβ42 are the major forms of amyloid-beta peptides in the brain. Aβ42 is much more prone to aggregation and more toxic to neurons than Aβ40 (Jarrett et al. 1993; El-Agnaf et al. 2000). In the present study, it appears that GSE reduced Aβ40 slightly more than Aβ42, which is consistent with a recent study (Wang et al. 2008). However, the mechanism of this differential effect is not known. We used curcumin as a positive control at a dietary concentration known to reduce the brain amyloid burden (Yang et al. 2005). Our data indicate that the polyphenols-rich GSE are comparable to curcumin in the reduction of brain and plasma amyloid burden. As there are a number of polyphenols in GSE, it is not known which polyphenol plays a major role in these events. However, we assume that the combination of all polyphenols could be important in achieving these desirable effects. The role these polyphenols play in the reduction of the amyloid burden is not known. Recently, other known polyphenols have been shown to be able to influence Aβ metabolism and protect against Aβ neurotoxicity. Curcumin, a phenolic yellow pigment found in turmeric and a spice used extensively in Asian Indian food, can directly bind small Aβ species to block the formation of oligomer and fibril as well as to disaggregate Aβ aggregates in vitro, and can reduce amyloid levels and plaque burden in aged transgenic AD mice when administered peripherally (Lim et al. 2001; Yang et al. 2005). The consumption of the green tea polyphenol epigallocatechin-3-gallate has also been shown to reduce the overproduction of Aβ in vitro (Levites et al. 2003) and in the transgenic AD mouse brain (Rezai-Zadeh et al. 2005) by promoting the non-amyloidogenic alpha-secretase proteolytic pathway, and reduce the generation of holo-APP and Aβ presumably via iron chelating effects of the polyphenol in vitro (Reznichenko et al. 2006). In the present study, holo-APP expression did not change after chronic consumption of GSE or curcumin (data not shown). Wine is rich in polyphenols, and wine consumption is related to a lower risk for Alzheimer’s disease in epidemiological studies (Orgogozo et al. 1997; Lindsay et al.2002; Letenneur 2004). This has been confirmed in an animal study showing that wine consumption is effective in preventing Alzheimer’s disease, also by increasing the non-amyloidogenic alpha-secretase activity (Wang et al.2006a). It is known that grape-derived polyphenols inhibit Aβ aggregation, reduce Aβ production, degrade intracellular Aβ, protect against Aβ neurotoxicity and attenuate oxidative stress in vitro (Bastianetto et al. 2000; Jang and Surh 2003; Ono et al. 2003, 2005; Savaskan et al. 2003; Li et al. 2004; Marambaud et al. 2005; Mancuso et al. 2007; Riviere et al. 2007). It is likely that the polyphenols from GSE act on similar multiple pathways to reduce the brain amyloid burden. A reducing effect was observed preferentially on the insoluble fraction of both Aβ40 and Aβ42 in our study, suggesting that GSE and curcumin probably affect the assembly or disruption of preformed fibrils and Aβ aggregation (Wang et al. 2008). However, chronic consumption of GSE-derived polyphenols did not change the activities of Aβ generating and degradating enzymes (Wang et al. 2008). Thus, mechanisms of GSE in modulating Aβ metabolism need further investigation and novel pathways need to be explored.

Inflammation is another hallmark of AD. There is mounting evidence that chronic inflammatory processes play a fundamental role in the progression of neuropathological changes in the AD brain (Akiyama et al. 2000; von Bernhardi 2007). The major players involved in the inflammatory process in AD are thought to be microglia. Cytokine production by activated microglia, such as IL-1β, TNF-α and IFN-γ, is a key pathologic event in the progression of inflammatory cascades (Jekabsone et al. 2006; Meme et al. 2006; Yamamoto et al. 2007). In the present study, chronic polyphenol consumption effectively alleviated microgliosis. The serum levels of IL-1β, TNF-α and IFN-γ in transgenic mice were higher than in wild-type littermates. This is consistent with the findings from AD patients that IL-1β, TNF-α and IFN-γ are up-regulated (Fillit et al. 1991; Blum-Degen et al. 1995; Solerte et al. 2000). Polyphenol and curcumin consumption tend to decrease plasma levels of IFN-γ. However, we did not observe a difference in serum levels of IL-1β and TNF-α among transgenic mice fed with Control, Curcumin or GSE. It should be noted that these cytokines exist at very low levels in the serum, which is within the lower range of sensitivity of the ELISA kits used (8–15 pg/ml) which might not be sensitive enough to reflect the difference between groups. Another possible explanation is that the cytokines in the serum might not have a good correlation to the cytokines in the brain. The IL-1β, TNF-α and IFN-γ levels in the transgenic mice have not been well characterized. So far limited studies have examined these cytokines in the APP transgenic mouse model and generated controversial results, primarily as a result of the low levels of these cytokines (Mehlhorn et al. 2000; Sly et al. 2001; Abbas et al. 2002; Yamamoto et al. 2007; see review in Heneka and O’Banion 2007). However, our quantitative data on the microglia activation by immunostaining demonstrated that the microglia activation was suppressed by over 70% in animals on the GSE diet. The inflammation suppression by the GSE diet indicates that the polyphenols from GSE are also powerful ingredients which inhibit inflammation in the brain of AD. It is well known that GSE has a property of anti-inflammation in other inflammation models such as chemically-induced dermatitis (Li et al. 2001; Bralley et al. 2007), ultraviolet B (UVB) induced oxidative stress models (Sharma et al. 2007), atherosclerosis model (Vinson et al. 2002) or systemic sclerosis in patients (Kalin et al. 2002). How the GSE polyphenols in the present study affect the inflammation is not clear. It is known that the green tea polyphenol EGCG or GSE suppresses NF-kappaB activation and phosphorylation of p38 MAPK and JNK in human astrocytoma U373MG cells (Kim et al. 2007) or in a UVB-induced oxidative stress mouse model. As NF-kappaB, p38 MAPK and JNK are the major signal pathways involving inflammation, it is likely that the polyphenols from GSE suppress inflammation by inhibiting these signal pathways. In a human umbilical vein endothelial cell culture model, GSE significantly inhibited the expression of adhesion molecule VCAM-1 and activated peroxisome proliferators-activated receptor gamma (PPAR gamma) and reduced the content of Von Willebrand factor, indicating that GSE may suppress inflammation by inhibiting the cell inflammatory factor expression and activating PPAR gamma (Ma et al. 2007). Other mechanisms such as oxygen free radical scavenging, anti-lipid peroxidation and inhibition of the formation of inflammatory cytokines may also be involved (Li et al. 2001). Meanwhile, decrease of Aβ accumulation in the brain may also contribute to the reduced microgliosis observed in the present study.

In our present study, consumption of GSE or curcumin did not significantly reduce the GFAP positive astrogliosis, although Aβ pathology and microgliosis in the brain were significantly attenuated. Significant astrogliosis was also observed in the brain of wild-type control mice. Our findings are consistent with a recent study showing that area reactivity of GFAP did not correlate with Aβ immunoreactivity in AD patients’ brains, which suggests other factors such as age-associated events may also contribute to the astrocyte pathology in the AD brain (Simpson et al. 2008).

Moderate wine consumption has been recommended for the prevention of AD (Hampl et al. 2002). Considering that the disease primarily affects the older generation most of whom are contraindicated to alcohol, polyphenol extracts from grape seeds might be a better alternative to wine for the prevention of AD. In a parallel study, we have shown that the consumption of polyphenols from GSE significantly suppressed the genomic instability events associated with DNA damage in those animals with a high polyphenol or curcumin diet (Thomas et al. in press), which is in agreement with other studies from our group showing that wine polyphenols protect against DNA damage induced by oxidative stress in vitro and ex vivo (Greenrod and Fenech 2003; Greenrod et al. 2005). In the present study, we have clearly demonstrated the benefit of GSE as a food additive in the prevention of AD. Our studies showed that 9 months of continuous consumption of GSE did not cause any damage to the liver, as the bilirubin level and amino acid transferase activities are normal. The food intake and body weight in the experimental animals are well maintained and comparable to those of animals fed with the Control diet. In addition, all animals fed with GSE were found to be normal with no tumour development or unexpected death. Our study indicates that GSE is a safe food additive in mice and therefore it is possible that it may be safely consumed in a long-term manner in humans to prevent the development of AD, although this has yet to be tested. As this food additive has a strong anti-inflammatory effect shown in the previous studies and our current studies, the consumption of GSE may be beneficial for chronic inflammatory diseases. Given the safety of GSE in long-term use and that no strong disease-modifying therapeutic and preventive measures are currently available in clinical settings, consuming GSE would be promising in developing practical preventive and therapeutic measures for AD.

Acknowledgements

The authors wish to thank Dr Qiao-Xin Li and Professor Colin Masters from the Department of Pathology, the University of Melbourne for their advice and input. This work was supported by grants from the Australian Centre of Excellence for Functional Foods (MF) and Australian NHMRC (No. 480422, X. F. Zhou & Y. J. Wang). Y. J. Wang is supported by EIPRS at Flinders University.

Copyright information

© Springer Science+Business Media, LLC 2009