|Home | About | Journals | Submit | Contact Us | Français|
We evaluated the therapeutic efficacy of combined treatment of Aβ-immunization with simvastatin in an Alzheimer mouse model at age 22 months. DNA prime-adenovirus boost immunization induced modest anti-Aβ titers and simvastatin increased the seropositive rate. Aβ-KLH was additionally administered to boost the titers. Irrespective of simvastatin, the immunization did not decrease cerebral Aβ deposits but increased soluble Aβ and tended to exacerbate amyloid angiopathy in the hippocampus. The immunization increased cerebral invasion of leukocytes and simvastatin counteracted the increase. Thus, modest anti-Aβ titers can increase soluble Aβ and simvastatin may reduce inflammation associated with vaccination in aged Alzheimer mouse models.
Alzheimer’s disease (AD) is the most common cause of dementia in the elderly. Patients with AD develop deposits of abnormally aggregated amyloid β-peptide (Aβ) as forms of amyloid plaques and cerebral amyloid angiopathy (CAA). Aβ is produced from a larger transmembrane glycoprotein, amyloid β-protein precursor (APP), by proteolytic processing. Aβ found in the AD brain is heterogeneous at its carboxyl-terminus, resulting in peptides of 39–43 amino acids. Amyloid plaques, however, mainly contain 2 forms of Aβ consisting of 40 (Aβ40) and 42 (Aβ42) amino acids (Iwatsubo et al., 1994; Savage et al., 1995). Aβ42 is considered to be highly amyloidogenic, and oligomeric forms of Aβ are neurotoxic (Lambert et al., 2001; Walsh et al., 2002). The amyloid hypothesis states that accumulation of Aβ in the brain is the primary influence driving AD pathogenesis (Hardy and Selkoe, 2002). Thus, the amyloid hypothesis provides the prime target for AD therapy and preventing Aβ production and aggregation as well as removing Aβ from the brain are considered to be preventive and therapeutic (Citron, 2010).
Anti-Aβ immunotherapy is considered to be one of the most efficacious means of clearing Aβ deposits from the brain. Immunization of AD mouse models with synthetic Aβ prevented or reduced Aβ deposits and improved their memory and learning deficits (Janus et al., 2000; Morgan et al., 2000; Schenk et al., 1999). These impressive results prompted initiation of phase I/II clinical trials to evaluate the safety and tolerability of aggregated Aβ vaccination. Clinical trials of peptide Aβ vaccine (AN1792) brought to light the problems associated with the modality. Approximately, 6% of AD patients subjected to the Aβ vaccine developed aseptic T-lymphocyte meningoenchephalitis and/or Fc-mediated brain inflammation, which caused discontinuation of the trial (Check, 2002; Orgogozo et al., 2003). Only 20% of AD patients developed anti-Aβ antibodies in the vaccine trial (Gilman et al., 2005), indicating that most AD patients might not have an adequate immune response to Aβ vaccination due to aging and low immunogenicity of Aβ (Monsonego et al., 2001; Monsonego, 2005). Cerebral hemorrhages were found in patients subjected to the Aβ vaccine trial (Ferrer et al., 2004; Uro-Coste et al., 2010). Such hemorrhages appeared to be associated with increases in vascular amyloid deposits after immunization, which may be detrimental to endothelial cells and blood-brain barrier (BBB) (Ferrer et al., 2004; Lee et al., 2005; Pfeifer et al., 2002; Racke et al., 2005; Wilcock et al., 2004). Therefore, the next generation of vaccines will need to overcome these problems associated with Aβ (AN-1792) vaccination.
To avoid the T-cell mediated side effects, we previously constructed an adenovirus vector encoding 11 tandem repeats of Aβ1-6 fused to the receptor-binding domain (Ia) of Pseudomonas exotoxin A, AdPEDI-(Aβ1-6)11 (Kim et al., 2005). Aβ1-6 lacks T cell epitopes of Aβ (Agadjanyan et al., 2005; Cribbs et al., 2003; Ghochikyan et al., 2006; Seabrook et al., 2007). Nasal vaccination of 2-month-old AD model mice with AdPEDI-(Aβ1-6)11 for 10 months induced Th2-biased responses and reduced Aβ deposits in the brain. Thus, nasal vaccination of mice with AdPEDI-(Aβ1-6)11 is preventive against cerebral β-amyloidosis without inducing T cell responses. Additionally, we demonstrated that a DNA prime-adenovirus boost regimen enhanced Th2-biased responses by increasing the seropositive rate and titers in 2–3 month-old C57BL/6 mice using AdPEDI-(Aβ1-6)11 and a plasmid DNA encoding PEDI-(Aβ1-6)11. Therefore, in this study, we have nasally vaccinated an aged AD mouse model with the DNA prime-adenovirus boost regimen followed by boosters of Keyhole limpet hemocyanin (KLH)-conjugated Aβ to evaluate its therapeutic efficacy in reducing cerebral Aβ deposits.
Statins inhibit 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase and reduce biosynthesis of cholesterol. Statins are extensively used as lipid-lowering drugs in clinical practice and have been proven to be effective in reducing cardiovascular-related morbidity and mortality. Increasing evidence suggests a link between AD, vascular risk factors, and atherosclerosis (Altman and Rutledge, 2010; Bhat, 2010; Luzzi et al., 2010). Retrospective epidemiological studies indicated that subjects taking statins had a lower prevalence and incidence of AD (Jick et al., 2000; Rockwood et al., 2002; Wolozin et al., 2000). Data, however, from clinical trials of statins for AD are inconsistent. The conflicting results of the clinical trials can be explained by the differences in doses between the different trials (Miida et al., 2007; Wolozin et al., 2006). The beneficial effects of statins for AD may be exerted by the pleiotropic mechanisms of statin action at high doses (Menge et al., 2005; Miida et al., 2007; Wolozin et al., 2006). Such pleiotropic effects of statins include reducing Aβ production, suppressing inflammatory responses, stabilizing the BBB integrity, protecting neurons from excitotoxins, apoptosis and oxidative stresses, and promoting synaptogenesis (Fonseca et al., 2009; Willey and Elkind, 2010). Previously, we showed that simvastatin was effective in reversing learning and memory deficits of an AD mouse model at 11–15 months of age without altering cerebral Aβ load (Li et al., 2006) and that simvastatin boosted the seropositive rate and titers in 2–5 month-old C57BL/6 mice vaccinated with AdPEDI-(Aβ1-6)11 (Kou et al., 2010). Here, we have evaluated the therapeutic efficacy of combined treatment of DNA prime-adenovirus boost regimen followed by Aβ-KLH boosters with simvastatin in reducing cerebral Aβ load in an AD mouse model at 11–22 months of age.
AdPEDI-(Aβ1-6)11 and pCA-PEDI-(Aβ1-6)11 were prepared as previously described (Kim et al., 2005).
B6.Cg-Tg(APPswe, PSEN1dE9) 85Dbo/J mice (TgAPPswe/PS1dE9 mice) mice (Jankowsky et al., 2004) were purchased from Jackson Laboratory (Bar Harbor, ME) and used as an AD mouse model. Eleven-month-old TgAPPswe/PS1dE9 mice were divided into 4 groups: PBS group, mice fed with regular food and no vaccination (treated with PBS) (n = 14, 7 males and 7 females); statin-only group, mice fed with food containing simvastatin with no vaccination (n = 14, 7 males and 7 females); vaccine-only group, mice fed with regular food with vaccination (n = 11, 5 males and 6 females); vaccine-plus-statin group, mice fed with food containing simvastatin with vaccination (n = 14, 7 males and 7 females). The simvastatin treatment groups were fed with a diet admixture containing 0.04% of simvastatin so that each mouse consumed a daily dose of approximately 50 mg simvastatin/kg body weight. One month after simvastatin treatment, mice in vaccination groups were nasally immunized with plasmid DNA, pCA-PEDI-(Aβ1-6)11 (100 µg/mouse, twice, weekly), followed by an adenovirus vector, AdPEDI-(Aβ1-6)11 (1 × 108 pfu/mouse, twice weekly). Mice in the PBS group were subjected to only nasal treatment with PBS. Due to the decline of antibody titers 2 months after the initial vaccination, 50 µg of synthetic Aβ1-8 conjugated with KLH was given to the mice subcutaneously for further boosting at 3 to 5 week intervals for 5 months. The treatment schedule is shown in Fig 1. Two months after the final vaccination, the experimental mice were euthanized to determine the therapeutic effects of the treatment. All animal protocols used for this study were prospectively reviewed and approved by the Institutional Animal Care and Use Committee of the University of Illinois College of Medicine at Peoria.
Blood samples were collected using heparinized capillary tubes at 1 and 2 months after simvastatin treatment. Plasma was prepared by centrifuging blood samples at 1500 × g for 10 min at room temperature. The total cholesterol levels in plasma were determined colorimetrically with commercial reagents (InfinityTM cholesterol reagent, Thermo Electron Corporation, Melbourne, Australia).
Blood samples were collected at 2 or 3 weeks after each vaccination booster by cutting the tail tips or puncturing the saphenous veins. The blood samples were incubated at room temperature for 1 h then transferred to 4°C. After overnight incubation, sera were separated by centrifugation at 10,000 × g for 10 min. Sera were stored at −80°C and thawed at the time of assay. Enzyme-linked immunosorbent assay (ELISA) was carried out to determine the titers of anti-Aβ antibodies and the immunoglobulin isotypes in the sera as previously described (Kim et al., 2005). To be brief, a 96-well plate was coated with 500 ng synthetic Aβ1-42 peptide per well at 4°C overnight, followed by incubation with blocking buffer (1× PBS containing 0.5% BSA, 0.05% Tween-20 and 5% goat serum) at room temperature for 1 h. Then, diluted serum samples were added to microtiter wells and incubated at 4°C overnight. The next day, microtiter wells were washed 5 times using washing buffer (1× PBS containing 0.05% Tween-20), and then incubated with an appropriate horseradish peroxidase (HRP)-conjugated secondary antibody at room temperature for 1 h. The microtiter wells were washed again with the washing buffer 5 times and then incubated with 100 µl of 3,3’,5,5’-tetramethylbenzidine (TMB) for 15 min. The reaction was stopped by adding 100 µl of 1 N H2SO4. The optic densities were determined at 450 nm using Microplate Reader. Serial dilutions of 6E10 (monoclonal anti-Aβ antibody) were used as the standard to determine the titer of anti-Aβ antibodies in the sera. Therefore, the concentrations (µg/ml) of the serum titers presented here reflect the concentrations of 6E10 antibody, which produce the same ELISA readings, and may not accurately represent the absolute amounts. Comparison of treatment groups was performed by one-way analysis of variance (ANOVA) and two-tailed Student’s t-test. P < 0.05 was considered statistically significant.
Female mice (n = 5 for the PBS, statin-only, and vaccine-only group; and n = 7 for the vaccine-plus-statin group) were terminated at 22 months of age by intraperitoneal injection of pentobarbital (100 mg/kg). The neocortex and hippocampus were dissected and lysed using the Bio-Plex cell lysis kit (Bio-Rad Laboratories, Hercules, CA). The brain tissues were dounce-homogenized in carbonate buffer (100 mM Na2CO3, 50mM NaCl, pH 11.5) containing protease inhibitors [10 µg/ml aprotinin and 1 mM 4-(2-aminoethyl) benzenesulphonyl fluoride hydrochloride (AEBSF)] and centrifuged at 16,000 g for 30 min at 4°C. The supernatants containing buffer soluble Aβ were collected and the protein concentrations in the supernatants were determined by Bio-Rad Protein Assay (Bio-Rad). The pellets containing insoluble Aβ were further dounce-homogenized in guanidine hydrochloride (final concentration, 5 M) and then rock-shaken for 3–4 h at room temperature. Levels of buffer-soluble and insoluble Aβ in the neocortex and hippocampus were quantified by the Aβ42 and Aβ40 ELISA kits (Invitogen, Carlsbad, CA) according to the manufacturer’s protocol.
Five male mice for each group were deeply anesthetized with pentobarbital and cardinally perfused with cold PBS followed by 4% paraformaldehyde. The brains were quickly removed and fixed in 4% paraformaldehyde for 16 h. The brains were then stored overnight in 30% sucrose in 0.1M PBS and frozen in Tissue-Teck optimal cutting temperature compound. Frozen sections (35 µm thickness) were prepared and subjected to immunohistochemical staining to determine the brain Aβ load and the activation of microglial cells and astrocytes using the avidin-biotin-peroxidase method (Vectastatin ABC kit, Vector, Burlingame, CA). Endogenous peroxidase was eliminated by treatment with 3% H2O2/10% methanol Tris-buffered saline (TBS) for 60 min at room temperature. After washing with water and 0.1 M TBS (pH 7.4), slides were blocked with blocking buffer for 60 min at room temperature to prevent non-specific protein binding. The slides were then incubated with primary antibody 6E10 (1: 2000; Signet Laboratories, Dedham, MA), CD45 (1:100; Serotec, MCA1388, Raleigh, NC), GFAP (1:500; Dako, Carpinteria, CA) or CD11b (1:200; Serotec, MCA711, Raleigh, NC) overnight at 4°C. The sections were rinsed in 0.1 M TBST and incubated with appropriate biotinylated secondary antibody for 1 h at room temperature. Finally, the avidin biotin peroxidase method using 3,3’-diaminobenzidine as a substrate (Vector, Burlingame, CA) was performed according to manufacturer’s protocol. Histomorphometry for quantification of amyloid deposition, astrocyte and microglial activation was performed using an Olympus BX61 automated microscope, Olympus Fluoview system and the Image Pro Plus v4 image analysis software (Media Cybernetics, Silver Spring, MD) capable of color segmentation and automation via programmable macros. For each mouse, 4–5 coronal brain sections starting at 1.2 mm posterior to the bregma to caudal, each separated by approximately 300 µm intervals, were analyzed. Areas stained with specific antibodies were expressed as a percentage of total hippocampus or neocortex examined.
For detection of CAA, 3 coronal brain sections for each male mouse (n = 5 for each group) starting at 2.7 mm posterior to the bregma to caudal, each separated by 300 µm intervals, were stained with 1% thioflavin S followed by destaining in 70% ethanol. Blood vessels positive for thioflavin fluorescence in the hippocampus were enumerated as described previously (Van Dorpe et al., 2000).
For detection of possible cerebral hemorrhages that might be associated with vaccination treatment, Prussian blue reaction was carried out on brain sections using an Iron Stain kit (Sigma-Aldrich, St. Louis, MO) according to the manufacturer’s protocol. For each male mouse (n = 5 for each group), 8 coronal brain sections starting at 1.5 mm posterior to the bregma to caudal, each separated by approximately 150 µm intervals, were analyzed.
Hippocampal homogenates (30 µg, n = 5 females for each group) were mixed with 5 × sample buffer (final concentration: 60mM Tris-HCl, 2% SDS, 10% glycerol, 0.001% bromphenol blue, pH: 6.8) and heated at 100°C for 5 min. Samples were subjected to 10–20% Tris-HCl gradient SDS-PAGE and electrotransferred to polyvinylidine difluoride (PVDF) membranes (Millipore, Bedford, MA). The membrane was blocked by PBS containing 5% non-fat dried milk (w/v), 0.02% sodium azide, and 0.02% Tween 20 for 1 h at room temperature, incubated at 4°C overnight with anti-CD45 antibody (Serotec, MCA1388, Raleigh, NC). The membrane was then incubated with appropriate biotinylated secondary antibody for 1 h at room temperature and visualized by the western lighting chemiluminescence reagent plus (Perkin Elmer, Boston, MA) according to the manufacturer’s protocol. The membranes were reprobed with monoclonal antibody against glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Chemicon, Temecula, CA). The optical densities of protein bands from the membranes were determined by densitometric scanning using a HP Scanjet G3010 Photo Scanner and the Image J V1.40 (National Institutes of Health, USA). The optical density of each protein band was divided by that of the GAPDH band on the same lane from the same membrane for normalization.
The neocortex and hippocampus tissue lysates were prepared and their protein concentrations were determined as described above. Levels of cytokines and chemokines in the lysates were determined by the bead-based suspension microarray technology (AssayGate, Ijamsville, MD) (Opalka et al., 2003). The threshold of detection was determined by adding two standard deviations to the mean fluorescence intensity of twenty zero standard (background) replicates. The minimum detectable concentrations were 3.6 pg/ml for IL-1α, 3.8 pg/ml for IL-1β, 2.3 pg/ml for IL-2, 4.3 pg/ml for IL-4, 1.3 pg/ml for IL-6, 8.2 pg/ml for IL-10, 2.3 pg/ml for IFN-γ, 16 pg/ml for TGF-β1, 2.5 pg/ml for monocyte chemoattractant protein (MCP)-1 (CCL2), and 1.3 pg/ml for macrophage inflammatory protein (MIP)-1α (CCL3). The results were expressed in picograms per milligram of brain protein and analyzed by ANOVA and two-tailed Student’s t-test using SigmaStat for Windows Version 3.00. P < 0.05 was considered statistically significant.
To evaluate potential liver damage caused by long-term administration of simvastatin, aspartate transaminase (AST) and alanine transaminse (ALT) levels in the plasma were determined using AST and ALT enzymatic assay kit (Bio Scientific Corporation, Austin, TX) according to the manufacturer’s protocol.
Skeletal muscle injury possibly associated with simvastatin treatment was assessed by examining thigh muscle sections stained with hematoxylin and eosin and by determining creatine kinase (CK) levels (Charles River research animal diagnostic service).
Data were expressed as mean ± SEM. ANOVA and two-tailed Student’s t-test were used to determine the intergroup differences except for seropositive rates where chi-square analysis was used. Correlations between the antibody titers and the areas (%) stained with specific antibodies (6E10, CD11b, and CD45) were estimated using Prism 3.02 program (GraphPad Software, Inc., San Diego, CA). P ≤ 0.05 was considered statistically significant.
Two groups of mice were treated daily with 50 mg/kg of simvastatin for 11 months. This daily dose was chosen for this study because this dose lowered the total plasma cholesterol levels in mice and produced learning and memory improvement without detectable toxicity after 3–4 months of simvastatin treatment in our previous studies (Kou et al., 2010; Li et al., 2006). The total cholesterol levels in plasma were measured 1 and 2 months after the initiation of simvastatin treatment. Simvastatin significantly reduced total plasma cholesterol levels by 10% only after 2 months of treatment compared to mice consuming regular diet without simvastatin (Table 1). We previously showed that plasma total cholesterol levels in young C57Bl/6 mice were reduced after 1-month treatment with simvastatin (Kou et al., 2010). However, 2-month treatment was required for aged mice to achieve the same cholesterol-lowering effect of simvastatin. There was no difference in the total cholesterol levels between the statin-only and vaccine-plus-statin groups. No difference in the total cholesterol levels was found between the PBS and vaccine-only groups, also. During the treatment period, there were no differences in physical appearance, body weight, food consumption, or mortality among the groups (data not shown). Long-term treatment with 50 mg/kg of simvastatin did not cause any generalized toxicity, muscle and liver damage assessed by histology and plasma levels of AST, ALT and CK (Table 1).
Sera were collected from all groups at 2 to 7 weeks after each vaccine booster (Fig. 1). The titers of anti-Aβ antibodies and the immunoglobulin isotypes in the sera were determined by ELISA. Anti-Aβ antibody titers in the mouse sera were calculated using serial dilutions of 6E10 (anti-Aβ antibody) as the standard. The data on immune responses, anti-Aβ antibody titers, isotypes and IgG1/IgG2a ratios are summarized in Table 2. The DNA prime-adenovirus boost regimen elicited modest anti-Aβ titers in both vaccine groups. One month after the initial DNA vaccination, only 2 mice in each vaccine group developed anti-Aβ antibody titers (seropositive) with the average of 1.53 ± 0.05 and 1.28 ± 0.05 µg/ml for the vaccine-only and vaccine-plus-statin groups, respectively. The numbers of seropositive mice increased to 7 and 8 out of 14 mice at 6 and 8 weeks after the initial vaccination, respectively, in the vaccine-plus-statin group but remained the same in the vaccine-only group. Thus, simvastatin increased the seropositive rates in an aged AD mouse model (P = 0.09 and 0.048, respectively, chi-squared test). IgG isotype profiles of anti-Aβ antibodies showed that the DNA prime-adenovirus boost regimen predominantly elicited IgG1 and IgG2b isotypes in both groups regardless of simvastatin treatment. Anti-Aβ IgG2a titers in both groups were below detectable levels by ELISA. Anti-Aβ titers continuously decreased 4 to 8 weeks after the initial vaccination in the vaccine-only group. The vaccine-plus-statin group had peak anti-Aβ titers at 6 weeks and, then, a decline at 8 weeks after the initial vaccination. Although the vaccine-plus-statin group had approximately 2- to 3-fold higher antibody titers than the vaccine-only group at 6 and 8 weeks after the initial vaccination, the differences were not significant. Due to the modest immune responses and the declines in the anti-Aβ antibody titers after cessation of the DNA prime-adenovirus boost regimen, Aβ1-8 conjugated with KLH (Aβ1-8-KLH) was subcutaneously given to both vaccination groups at intervals of 3 to 5 weeks for 5 months in order to increase the antibody titers. The peptide boosters increased the antibody titers by 8- to 17-fold in both groups and no differences were found in the anti-Aβ titers between the 2 vaccination groups. Almost all mice in both groups became seropositive after Aβ1-8-KLH administration (Table 2). IgG isotyping revealed that Aβ1-8-KLH increased anti-Aβ IgG1 titers but induced anti-Aβ IgG2a titers in both groups. The IgG1/IgG2a ratios in both groups were consistently more than 3, indicating Th2-polarized immune responses. Simvastatin shifted the immune responses further toward Th2-type only 1.5 weeks after the 4th Aβ1-8-KLH booster; the IgG1/IgG2a ratios were 3.2 ± 0.69 and 7.5 ± 1.81 for the vaccine-only and vaccine-plus-statin groups, respectively (P = 0.05). Similar to the way the anti-Aβ titers declined after cessation of adenovirus boosters, termination of Aβ1-8-KLH boosters resulted in reductions in anti-Aβ antibody titers 7 weeks after the 2nd Aβ1-8-KLH boosters, as well as 6 weeks after the 5th Aβ1-8-KLH boosters in both groups (Table 2).
Ten months after the initial immunization, mice were euthanized at 22 months of age. Levels of buffer-soluble and insoluble Aβ in the neocortex and hippocampus were quantified by Aβ42- and Aβ40-specific ELISA and were expressed as pg of Aβ per mg of total protein. The amounts of hippocampal buffer-soluble Aβ42 were 86.5 ± 13.0 and 82.9 ± 12.9 pg/mg in the vaccine-only and vaccine-plus-statin groups, respectively, and 2 times greater than those in the PBS group (41.0 ± 5.9 pg/mg, P < 0.05 for both, Fig. 2A). No difference was found in the amounts of hippocampal buffer-soluble Aβ42 between the PBS and statin-only (65.4 ± 13.8 pg/mg) groups (P > 0.05). Similarly, hippocampal buffer-soluble Aβ40 in the vaccine-only and vaccine-plus-statin groups (88.0 ± 14.8 and 95.3 ± 16.8 pg/mg, respectively) increased 2-fold compared to the PBS group (47.2 ± 9.2 pg/mg, P < 0.05 for both) (Fig. 2B). There was no difference between the statin-only (79.8 ± 21.0 pg/mg) and PBS groups in the levels of hippocampal soluble Aβ40. Levels of neocortex buffer-soluble Aβ42 were 42.1 ± 6.5, 33.0 ± 3.6, 43.5 ± 6.7 and 29.2 ± 4.2 pg/mg for the PBS, statin-only, vaccine-only and vaccine-plus-statin groups, respectively. Levels of neocortex buffer-soluble Aβ40 were 34.3 ± 3.0, 44.2 ± 7.8, 30.3 ± 5.0 and 53.1 ± 11.6 pg/mg for the PBS, statin-only, vaccination-only and vaccine-plus-statin groups, respectively. There were no significant differences in the levels of buffer-soluble Aβ42 and Aβ40 in the neocortex between groups. Levels of buffer-insoluble Aβ42 and Aβ40 were determined by ELISA, also. No significant differences were identified in either buffer-insoluble Aβ42 or Aβ40 in both hippocampus and neocortex between groups (Table 3).
The Aβ load in the hippocampus and neocortex was determined by immunohistochemistry using anti-Aβ 6E10 antibody followed by morphometric analysis (Fig. 3). Average percentages of areas showing Aβ immunoreactivity measured by morphometry in the hippocampus and neocortex are shown in Figure 3B. The Aβ loads in the hippocampus were 0.69 ± 0.08, 0.75 ± 0.15, 0.82 ± 0.14 and 0.74 ± 0.16% for the PBS, statin-only, vaccine-only and vaccine-plus-statin groups, respectively. The Aβ loads in the neocortex were 1.00 ± 0.11, 1.31 ± 0.21, 1.06 ± 0.22 and 1.02 ± 0.20% for the PBS, statin-only, vaccine-only and vaccine-plus-statin groups, respectively. No significant differences were found in the Aβ loads in the hippocampus and neocortex between groups (P > 0.05).
To assess potential alterations in inflammatory responses in the brain after vaccination and/or statin treatment, brain sections were subjected to immunohistochemistry using anti-CD45 antibody for blood-derived leukocytes and CD11b antibody for activated microglia/monocytes, and anti-GFAP antibody for reactive astrocytes followed by morphometric analysis. The CD45-immunoreactive areas in both hippocampus (1.00 ± 0.06%) and neocortex (1.11 ± 0.15%) in the vaccine-only group were greater than those in the PBS group (0.64 ± 0.05 and 0.83 ± 0.03% in the hippocampus and neocortex, respectively, P < 0.05 for both) (Fig. 4). No differences were found in CD45-immureactive areas between the vaccine-plus-statin group (0.65 ± 0.10 and 0.86 ± 0.14% in the hippocampus and neocortex, respectively) and the PBS group (Fig. 4). In the hippocampus, the CD45-immunoreactive areas in the vaccine-plus-statin group were smaller than those in the vaccine-only group (P < 0.05) but this difference was not observed in the neocortex (P = 0.13). The statin-only group (0.82 ± 0.11 and 1.05 ± 0.22% in the hippocampus and neocortex, respectively) was not different from the PBS group in CD45 immunoreactivity (P > 0.05).
To confirm an increase in CD45-immunoractivity in the vaccine-only group and counteraction of simvastatin against CD45-immunoreactive cells, immunoblot analysis was performed using hippocampal homogenates and anti-CD45 antibody (Fig. 4C and D). Expression levels of CD45 in the vaccine-only group increased by 34% compared with those in the PBS group (P = 0.01). CD45 expression levels in the vaccine-only group were greater than those in the vaccine-plus-statin group in the hippocampus (P = 0.05). No difference was found in CD45 expression levels between the PBS and vaccine-plus-statin groups (P > 0.05). Thus, simvastatin counteracted an increase in CD45 expression, which was caused by vaccination.
In the hippocampus, CD11b-immunoreactive areas were 0.44 ± 0.05, 0.34 ± 0.04, 0.46 ± 0.07 and 0.44 ± 0.04% in the PBS, statin-only, vaccine-only and vaccine-plus-statin groups, respectively. CD11b-immunoreactive areas in the neocortex were 0.34 ± 0.07, 0.34 ± 0.01, 0.41 ± 0.05 and 0.27 ± 0.09% in the PBS, statin-only, vaccine-only and vaccine-plus-statin groups, respectively. There were no differences in CD11b immunoreactivity between groups. (Fig. 5A).
GFAP immunoreactivity was not significantly different across groups (P > 0.05). CD11b-immunoreactive area percentages of the PBS, statin-only, vaccine-only, vaccine-plus-statin groups were 11.7 ± 4.6, 10.1 ± 2.9, 11.0 ± 3.3 and 10.5 ± 2.1% in the hippocampus and 12.4 ± 4.8, 12.2 ± 2.8, 11.5 ± 3.6 and 11.2 ± 2.1% in the neocortex (Fig. 5B).
To further assess potential alterations in inflammatory responses in the brain after vaccination and/or statin treatment, levels of cytokines and chemokines in the hippocampus and neocortex were determined by the bead-based suspension microarray technology. Cytokines included IL-1α, IL-1β, IL-2, IL-4, IL-6, IL-10, IFN-α, IFN-γ, and TGF-β1. Chemokines determined were MCP-1 (CCL2) and MIP-1α (CCL3). IL-4, IL-6, and IFN-γ were the minimum detectable levels and omitted from analysis. The levels of some cytokines are shown in Fig. 6. Levels of IL-2 decreased in the vaccine-only group compared to the PBS group (P < 0.01 for the hippocampus and P < 0.05 for the neocortex). In the hippocampus and neocortex, levels of IL-10 consistently increased in the statin-only group compared to the PBS group (P < 0.05). There were no significant differences in any of the other cytokines and chemokines between any groups.
The relationships between the anti-Aβ titers and Aβ load (6E10-stained area) in the hippocampus and neocortex were analyzed in the vaccine-only and vaccine-plus-statin groups using the titers 1.5 weeks (Fig. 7) and 2 weeks (data not shown) after the 4th Aβ1-8-KLH booster. The anti-Aβ titers were not correlated with Aβ load in the hippocampus and neocortex in these mouse groups. Additionally, there was no correlation between the anti-Aβ titers and immunoreactive areas (%) with antibodies against activated microglial markers (CD11b and CD45) (data not shown).
Because anti-Aβ immunotherapy was reported to cause cerebral microhemorrhages in aged AD mouse models and patients with AD, possible cerebral hemorrhages were investigated by iron staining. Eight brain sections were examined from each mouse. Hemorrhages associated with blood vessels were rare in any group (0–1/section) and no differences were found in the numbers of hemorrhages between groups.
Hemorrhages associated with Aβ immunotherapy are often accompanied by an increase in CAA. Three brain sections were analyzed from each mouse. The numbers of hippocampal blood vessels positive for thioflavin fluorescence per brain section are 3.2 ± 0.98, 3.8 ± 1.28, 5.5 ± 1.00, 3.3 ± 0.80 for the PBS, statin-only, vaccine-only, and vaccine-plus-statin groups, respectively (Fig. 8). Thus, the vaccine-only group had more thioflavin-positive blood vessels in the hippocampus than the 3 other groups but the differences were not significant (P = 0.14, 0.33, and 0.12 for the PBS, statin-only, and vaccine-plus-statin groups, respectively).
Here, we have investigated the therapeutic efficacy of anti-Aβ immunotherapy with and without statin in a DNA prime-adenovirus boost regimen followed by Aβ peptide boosters using TgAPPswe/PS1dE9 mice at 22 months of age. The DNA prime-adenovirus boost regimen induced very modest anti-Aβ titers and simvastatin increased the seropositive rate. Because anti-Aβ titers declined after cessation of the adenovirus inoculations, we administered Aβ1-8-KLH as additional boosters. Aβ1-8-KLH boosters increased anti-Aβ titers. Regardless of simvastatin treatment, the DNA prime-adenovirus boost regimen followed by Aβ1-8-KLH boosters increased soluble Aβ and revealed a trend toward increased CAA in the hippocampus without decreasing Aβ deposits in the brain. The immunization increased CD45-positive leukocytes in the brain and simvastatin treatment counteracted the increase. Our results suggest that modest anti-Aβ titers in advanced AD patients can increase soluble Aβ and simvastatin may reduce inflammation associated with vaccination by inhibiting monocyte invasion.
Previously, we demonstrated that the DNA prime-adenovirus boost regimen increased the seropositive rate (100%) and anti-Aβ titers (19.5 µg/ml) compared with its homologous prime-boost regimens (0 and 57% seropositive rate, and 0 and 2.8 µg/ml anti-Aβ titers for DNA-only and adenovirus-only, respectively) in C57BL/6 mice at age 2–3 months (Kim et al., 2007a). In contrast to young C57BL/6 mice, only 18% of TgAPPswe/PS1dE9 mice older than 11 months developed anti-Aβ titers by the DNA prime-adenovirus boost regimen in this study. These poor responses in aged TgAPPswe/PS1dE9 mice could be explained by immunological tolerance and immunosenescence. Tg2576 mice, another AD mouse model overexpressing APP, showed a decreased antibody response when immunized with synthetic Aβ compared with non-transgenic controls, which is attributed to hyporesponsiveness of Aβ-reactive T or B cells caused by the high levels of Aβ in the blood of Tg2576 mice (Monsonego et al., 2001). We, however, showed that a homologous prime-boost regimen with AdPEDI-(Aβ1-6)11 in young TgAPPswe/PS1dE9 mice induced immune responses comparable to young C57BL/6 mice (Kim et al., 2007b). Therefore, our results suggest that the poor responses to the DNA prime-adenovirus boost regimen in the aged TgAPPswe/PS1dE9 mice are mainly due to immunosenescence and that the DNA prime-adenovirus boost regimen may not be effective in overcoming poor responsiveness to Aβ immunization in the elderly. Poor immune responses to Aβ in aged mice (20-month-old) compared to young mice (3-month-old) were reported (Pifer et al., 2002). However, the DNA prime-adenovirus boost regimen with AdPEDI-(Aβ1-6)11 followed by Aβ1-8-KLH boosters increased the seropositive rate and titers in aged mice. Thus, it is possible to induce low to moderate levels of anti-Aβ antibodies by active immunization modalities in aged mice (older than 18 months), which include intraperitoneal injection of Aβ42 in Freund’s adjuvant (Das et al., 2001b), subcutaneous injection of mannan conjugated to Aβ1-28 (Petrushina et al., 2008), and subcutaneous injection of Aβ1-6 conjugated to the virus-like particle (Wiessner et al., 2011).
In the phase II clinical trial of aggregated Aβ (AN-1792) vaccination, only 19.7% of AD patients developed positive Aβ titers (Gilman et al., 2005). We previously found that simvastatin enhanced both seropositive rates and anti-Aβ titers in young C57BL/6 mice subjected to a homologous prime-boost regimen with AdPEDI-(Aβ1-6)11 (Kou et al., 2010). Here, we showed that simvastatin increased the seropositive rates of the DNA prime-adenovirus boost regimen in aged AD model mice without boosting the titers. Thus, simvastatin may be effective in improving immune responses to Aβ immunization in the elderly.
Because meningoencephalitis associated with the AN1792 trial is thought to be T-cell-mediated auto-immune responses, immunization modalities that elicit predominantly T helper (Th) type 2 (humoral) immune responses are considered to be safer for AD prevention and treatment (Gelinas et al., 2004; Goni and Sigurdsson, 2005; Monsonego and Weiner, 2003; Schenk et al., 2004; Wisniewski and Frangione, 2005). In mice, the production of IgG1 is primarily induced by Th2-type anti-inflammatory cytokines (IL-4 and IL-10), while IgG2a is produced through Th1-type pro-inflammatory cytokines (IL-2 and IFN-γ). IgG1 was predominant in anti-Aβ antibodies elicited by the DNA prime-adenovirus boost regimen in aged TgAPPswe/PS1dE9 mice and IgG2a was indiscernible by ELISA, indicating induction of Th2 immune responses. These results are consistent with those in young C57BL/6 mice in which Th2-biased immune responses were induced by the DNA prime-adenovirus boost regimen (Kim et al., 2007a) and with a decrease in cerebral IL-2 levels in the vaccine-only group (Fig. 6). Additional administration of Aβ1-8-KLH increased anti-Aβ titers in aged TgAPPswe/PS1dE9 mice but induced IgG2a anti-Aβ antibodies, suggesting that a minor Th1 component might be instrumental for boosting the titers. Simvastatin temporarily shifted the immune responses further toward Th2-type without reducing the titers 1.5 weeks after administration of the 4th Aβ1-8-KLH booster (Table 2). Statins promote Th2 differentiation by inducing expression of GATA-binding protein 3 (GATA3) and inhibit Th1 differentiation by downregulating activation of nuclear factor-γB (NF-γB) and T-bet (a T-box transcription factor) in CD4+ T cells (Arora et al., 2006). We and others showed that statins suppress expression of Th1 type cytokines including interferon (IFN)-γ (Arora et al., 2006; Hakamada-Taguchi et al., 2003; Kou et al., 2010). Naïve CD4 T-cells activated in the presence of IL-2 and IFN-γ tend to develop into Th1-cells and IFN-γ inhibits the proliferation of Th2-cells. Statins promote Th2 biased immune responses through the induction of STAT6 phosphorylation and Th2 cytokines including IL-10 and inhibition of STAT4 phosphorylation, also (Shimada et al., 2006). Furthermore, statins prevent or reverse paralysis in experimental autoimmune encephalomyelitis (EAE) (an animal model of multiple sclerosis) which is mediated by myelin-specific CD4+ T cells (Youssef et al., 2002). In line with this view, cerebral levels of IL-10 increased or tended to increase in the statin-only or vaccine-plus-statin group, respectively (Fig. 6). Thus, statins suppress Th1 and augment Th2 responses resulting in increased production of antigen-specific IgG1 in mice (Arora et al., 2006) and young healthy adults (Lee et al., 2006). Because induction of meningoencephalitis (infiltration of Aβ-specific T cells) by Aβ vaccination requires expression of IFN-γ in AD mouse models (Monsonego et al., 2006), statins may mitigate meningoencephalitis associated with Aβ vaccination by inhibiting IFN-γ signaling.
The DNA prime-adenovirus boost regimen followed by Aβ-KLH injections increased CD45-positive cells (migratory leukocytes and microglia) in the brain and simvastatin treatment counteracted the increase. Aβ immunotherapy activates microglia and recruits them to amyloid plaques (Morgan, 2006). Microglial phagocytosis of Aβ is one of the postulated mechanisms by which Aβ deposits are cleared from the brain (Bard et al., 2000; Nicoll et al., 2006). However, the role of microglia in clearing Aβ deposits is debatable (Rezai-Zadeh et al., 2011). In our study, the vaccination increased CD45-positive cells but did not reduce Aβ load in the brain, suggesting that CD45-positive cells have a limited ability to clear Aβ deposits from the brain. Such a limited role of microglia in antibody-mediated amyloid plaque clearance was previously demonstrated by enhancing and eliminating activation of microglia by IFN-γ and minocycline, respectively (Garcia-Alloza et al., 2007).
CD45 positive cells in the brain are thought to be migratory leukocytes from the circulation. Statins downregulate expression of chemokines and their receptors such as CCL2, CX3CL1, CCR2, and CX3CR1 via inhibition of the geranylgeranylpyrophosphate pathway in endothelial cells and leukocytes (Damas et al., 2005; Han et al., 2005; Veillard et al., 2006), leading to reduced BBB permeability and restrict leukocyte migration (Ghittoni et al., 2007; Ifergan et al., 2006; Zeng et al., 2005). In line with these observations, simvastatin inhibited invasion of CD45-positive cells, which was induced by vaccination. Thus, statins may be effective in ameliorating meningoencephalitis associated with Aβ vaccination by inhibiting leukocyte migration, also.
Parenchymal amyloid plaques were focally depleted in the brains of AD patients after Aβ42 vaccine (AN1792) treatment (Nicoll et al., 2003) but the severity of CAA increased (Boche et al., 2008). Levels of soluble Aβ in the cerebral cortex remarkably increased in AD patients subjected to AN1792 vaccination (Patton et al., 2006). These observations after AN1792 vaccination are explained by relocation of plaque Aβ to blood vessels based on one of the postulated mechanisms for amyloid plaque clearance by Aβ immunotherapy, which states that anti-Aβ antibodies directly interact with amyloid plaques, resulting in disaggregation and solubilization of Aβ fibrils (Bacskai et al., 2002; Boche et al., 2010). In this study, levels of buffer-soluble Aβ40 and Aβ42 increased in the hippocampus in mice subjected to the vaccination regardless of simvastatin treatment but no reduction was found in cerebral Aβ deposits and insoluble Aβ levels. There was no correlation between the anti-Aβ titers and cerebral Aβ load in the vaccine-only and vaccine-plus-statin groups (Fig. 7). The mice in the vaccine-only group showed an increasing trend in the severity of CAA and simvastatin treatment appeared to counteract the trend. The observed discrepancies in Aβ immunization outcomes between this study and the AN1792 clinical trial may be explained by relatively low levels of anti-Aβ titers in this study. A positive correlation between levels of anti-Aβ titers and the reduction of amyloid plaques has been reported in an AD mouse model (Petrushina et al., 2007) as well as AD patients subjected to AN1792 vaccination (Gilman et al., 2005; Nicoll et al., 2003). Prophylactic treatment of TgAPPswe/PS1dE9 mice with AdPEDI-(Aβ1-6)11 in our previous study (Kim et al., 2007b) elicited several-fold higher anti-Aβ titers than those in the current study and decreased cerebral amyloid deposits. Movsesyan et al. (2010) reported that prophylactic vaccination of an AD mouse model with DNA epitope vaccine induced low anti-Aβ titers and failed to reduce cerebral amyloid deposits. Uro-Coste et al. (2010) reported a clinical case in which an AD patient subjected to AN1792 vaccination had increases in the severity of CAA and microhemorrhages without apparent reduction in amyloid plaques. Therefore, anti-Aβ antibodies produced in aged TgAPPswe/PS1dE9 mice in this study were able to augment hippocampal levels of soluble Aβ, resulting in an increasing trend of CAA but might not be high enough to reduce cerebral Aβ deposits. Another possible explanation is that outcomes of Aβ immunotherapy may depend on the disease stage and ages. Several studies reported that a correlation between anti-Aβ antibody titers and the effect on Aβ load could not be established in aged AD mouse models partly due to the high intrinsic variations of Aβ load as well as titers over time (Das et al., 2001a; Minami et al., 2010; Okura et al., 2006; Wiessner et al., 2011).
Minami et al. (2010) reported that passive immunization with anti-Aβ antibody in a 22-month-old AD mouse model did not reduce cerebral Aβ load but increased levels of CCL2 and pro-inflammatory cytokine, tumor necrosis factor (TNF)-α, suggesting induction of neuroinflammation in advanced stages of cerebral amyloidosis by Aβ immunotherapy without beneficial effects. We, however, did not observe increased levels of CCL2 and pro-inflammatory cytokines by active immunization (Fig. 6). In this regard, it is tempting to speculate that statins may be effective in preventing both neuroinflammation and exacerbation of CAA, which are induced by Aβ immunotherapy, because statins inhibit the induction of TNF-α and CCL2 in microglia and macrophages (Damas et al., 2005; Han et al., 2005; Pahan et al., 1997; Veillard et al., 2006) and increase expression of low density lipoprotein-related protein 1 (LRP1) in endothelial cells (Shinohara et al., 2010), which is thought to mediate Aβ clearance at the BBB (Zlokovic, 2008).
In conclusion, this study revealed that it was difficult to achieve high anti-Aβ titers and high seropositivity by a DNA prime-adenovirus boost vaccination regimen in a very old AD mouse model and that modest levels of anti-Aβ antibody did not decrease amyloid plaques but increased soluble Aβ in the hippocampus. AD mouse models have been criticized because they failed to predict side effects. Our study emphasizes the importance of testing Aβ immunotherapy in very old AD mouse models with advanced amyloid pathology. Our study also demonstrated that simvastatin counteracted an increase in CD45-positive leukocytes and tended to mitigate an increase in CAA, both of which were induced by Aβ vaccination. These results warrant further investigation of statins in possibly preventing the side effects associated with Aβ immunotherapy.
We thank Linda Walter for assistance in preparation of this manuscript. This research was supported in part by grants from the National Institutes of Health (AG031846, AG031979, AG037814, AG030399, AG029818 and EY018478) and the Alzheimer’s Association (IIRG-07-59494).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.