Abstract
INTRODUCTION
Amyloid-β protein precursor (AβPP) is a transmembrane protein widely expressed in neurons. It has three major isoforms (AβPP695, AβPP751, and AβPP770) generated by alternative mRNA splicing [1]. AβPP metabolism is largely represented by its sequential cleavage by two AβPP secretases, i.e., α- or β- followed by γ-secretase. These two mutually exclusive pathways yield either soluble (sAβPPα) or amyloidogenic (Aβ) peptides. α- and β-secretases pertain respectively to the families of ADAM (A-disintegrin and metalloprotease) and BACE (beta-site AβPP-cleaving enzyme) [2–4], whereas the γ-secretase complex comprises four core proteins including presenilin-1 (PSEN1) [5].
Platelets are the main source of AβPP in peripheral tissues, accounting for 95% of circulating AβPP [6–8]. In addition, platelets contain the enzymatic machinery required for AβPP metabolism, including its secretases [8]. Studies indicate that the amyloidogenic pathway of AβPP metabolism is upregulated in platelets of AD patients, paralleling intracerebral abnormalities [9]. In platelets, AβPP metabolism generates soluble fragments of 120–130 kDa and 110 kDa, and the proportion of these two AβPP fragments (120–130 kDa:110 kDa) is commonly referred to as platelet AβPP ratio [10]. Studies from our group and others [10–12] demonstrated that platelet AβPP ratio is decreased in AD patients, including those at pre-dementia stages [13], and this abnormality correlates with the degree of cognitive impairment [14].
Both protein expression and activity of the non-amyloidogenic AβPP secretase ADAM10 have been reported to be decreased in AD platelets [9, 15]. Experimental [16–19] and clinical studies [20, 21] further indicated that these changes may be modified by treatment with cholinesterase inhibitors. Data on platelet BACE1 and PSEN1 are less abundant and more controversial. While BACE1 has been shown to be overactive and overexpressed in the postmortem brain [22] and in the cerebrospinal fluid [23] of AD patients, increased β-secretase activity was paradoxically associated with decreased protein expression of BACE1 in AD platelets [9].
The aim of the present study is to determine AβPP ratio and the protein expression of the AβPP secretases (ADAM10, BACE1, and PSEN1) in platelets of AD patients, and to determine whether the protein expression of these markers can be modified by the treatment with a cholinesterase inhibitor.
METHODS
Participants and study protocol
The present study was conducted at the Institute of Psychiatry, Faculty of Medicine, University of Sao Paulo, Brazil. Participants are community-dwelling outpatients or volunteers attending a memory clinic program, seeking treatment and/or prevention. The protocol was approved by the local Ethics Committee (CAPPesq-HCFMUSP) and the study was conducted according to the tenets of the Helsinki declaration. All participants (or respective caregivers in case of dementia) signed a written informed consent prior to enrollment.
Patients and controls were assessed at baseline with a comprehensive clinical examination followed by laboratory tests and brain imaging for diagnostic purposes. Cognitive state was ascertained by neuropsychological testing. The diagnosis of dementia was established according to the DSM-IV criteria [24], and the specification of probable AD was made according to the NINCDS-ADRDA diagnostic criteria [25]. To be allocated in the comparison group (controls), subjects were expected to have a normal performance on neuropsychological testing (according to age- and education-corrected norms), in addition to being physically healthy. AD patients were then prescribed treatment with a cholinesterase inhibitor. Eligibility to the study subsumed that, in the previous 6 months, participants: (i) had not received any treatment with anti-dementia drugs; and (ii) had not been prescribed any other drugs with known effect on the central nervous system (CNS), e.g., neuroleptics, lithium, antidepressants, benzodiazepines, anticonvulsants, etc.
Sixty-one physically healthy older adults were enrolled to the present study, with 23 patients with mild or moderate AD and 38 cognitively unimpaired elders (controls). No statistically significant differences were observed in gender distribution and mean age of patients (74% of females; mean age 73.1±6.9 years) and controls (68% of females; mean age 72.3±6.6 years); however, these groups displayed statistically significant differences regarding the degree of education (years of schooling: AD, 5.9±4.4; controls, 13.8±5.1; p < 0.001).
After the completion of the baseline assessment protocol (ahead), AD patients were started on donepezil hydrochloride (5 mg/day) administered orally. The target dose of 10 mg/day was to be reached after four weeks; otherwise (in case of bad tolerability) the low dose was maintained, and the patient was excluded from the trial. AD patients were clinically re-examined throughout the clinical treatment at appointments with the prescribing doctor held on a monthly basis. Cognitive and biological variables were measured according to the study protocol at baseline (i.e., prior to treatment with donepezil) and after 3 and 6 months of continuous treatment. Cognitive assessments were made with the Mini-Mental State Examination (MMSE) [26] and the Cambridge Cognitive Test (CAMCOG) [27]. At baseline, AD patients as expected had a significantly worse performance on the MMSE (mean score 18±5.1) and the CAMCOG (56.1±18.5) compared to controls (28.8±1.5 and 96.6±5.7 respectively; p < 0.0001 in both comparisons).
Preparation of platelets
Samples of peripheral blood (10 ml) were collected by venipuncture of the median cubital vein into 0.1 M sodium citrate-coated tubes (S-Monovett, Sarsted). Blood samples were homogenized with 200 μl of 0.09M ethylenediaminetetraacetic acid (EDTA) and centrifuged at 200×g for 10 min at room temperature. Platelet-rich plasma was separated from blood cells, and platelets were collected by centrifugation at 1,159×g for 15 min at room temperature. Pellets were washed with 5 ml of 10 mM Tris, pH = 7.4, and re-suspended in lysis buffer containing 10 mM Tris-HCl, pH = 7.4, 1 mM ethylene glycol tetraacetic acid (EGTA), 100 mM phenylmethanesulphonyl fluoride (PMSF), and protease inhibitors. Platelet homogenates were stored at –80°C. Protein concentrations were determined in each sample by a modified Lowry method (Bio-Rad DC Protein Assay) [28] before western blot assays.
Determination of AβPP ratio and AβPP secretases
Aliquots holding 25 μg of total protein obtained from each sample of platelet homogenates were submitted to electrophoresis in 8% polyacrylamide gels and transferred to nitrocellulose membranes by Western blotting. After blocking unspecific bindings, membranes were incubated with the primary antibodies as follows: (i) for 2 h with anti-APP A4 clone m22C11 (Millipore); (ii) overnight with anti-ADAM10 (Abcam); (iii) for 1 h withanti-BACE1 (Abcam); and (iv) for 1 h with anti-PSEN1 (Abcam). The first membranes were then incubated for 1 h with goat anti-mouse secondary antibodies labeled with horseradish peroxidase (Sigma), and the remaining membranes were incubated for 1 h with anti-rabbit IgG, peroxidase-conjugated, secondary antibody (GE Healthcare Life Sciences). Next, an enhanced chemiluminescence reagent (ECL, GE) was poured. Imaging was performed in Chemiimager TM4000 equipment (Alpha Innotech), which captures the chemiluminescent light emission from the reaction of the peroxidase-conjugated antibody with the ECL reagent. Light emission was captured for 5 min and the densitometry of distinct bands was performed with specific software tools. The proportion of 130 kDa to 110 kDa band densities was used to determine AβPP ratio, in addition to the densitometry of 6 5 kDa, 56 kDa, and 53 kDa bands to estimate the protein expression of ADAM10, BACE1, and PSEN1, respectively (Fig. 1). An internal standard (IS) consisting of a standard suspension of platelets (in-house prepared from a pool of platelets from young healthy volunteers) was added to each gel to control for analytical differences between blots performed in different days (inter-assay variation). The IS readings in proportion to the 130 kDa to 110 kDa ratio, 65 kDa, 56 kDa,and 53 kDa, was then calculated, yielding the normalized estimate of the AβPP ratio, ADAM10, BACE1, and PSEN1, respectively. Each sample was analyzed in duplicates, and additional blots of the same samples were performed whenever the reading between replicates displayed a variation > 15%.
Statistical analysis
Baseline comparison of sociodemographic characteristics of patients and controls were assessed with Fisher’s exact test for categorical variables and with Student’s t test for numerical ones. Longitudinal analysis of patients was carried out with a linear mixed-effects model. All statistical analyses were performed with the SPSS (Statistical Package for Social Sciences, for Windows, v. 22, Chicago, IL) and significance level was p≤0.05.
RESULTS
Platelet AβPP ratio in AD patients was 40.8% lower than in controls at baseline (p < 0.0001). This difference was due both to a reduction in the protein expression of 120–130 kDa AβPP secreted peptides along with an increase in the protein expression of 110 kDa forms (18.4%, p = 0.039 and 42.3%, p = 0.0012, respectively) (Fig. 2). No statistically significant effects were on observed in the protein expression of AβPP secreted peptides or in platelet AβPP ratio as a consequence of donepezil treatment (Table 1).
The semi-quantitative analysis of platelet protein expression of AβPP secretases by immunoblotting indicated statistically significant decrements in ADAM10 (58.4%, p < 0.0001) and PSEN1 (31.6%, p = 0.004) in AD compared to controls at baseline, whereas the protein expression of BACE1 was statistically similar in these two groups (Fig. 2). After six months of treatment with donepezil, we observed a 24.3% reduction in platelet protein expression of BACE1 in AD patients, as compared to baseline values (p = 0.023) (Fig. 3). No statistically significant effects were observed as a consequence of donepezil treatment on the protein expression ADAM10 and PSEN1 (Table 1).
DISCUSSION
This study had two central aims: (i) to evaluate the protein expression of AβPP secreted forms (120–130 kDA, 110 kDA, and AβPP ratio) and of AβPP secretases (ADAM10, BACE1, and PSEN1) in platelets of unmedicated patients with mild or moderate AD, as compared to healthy controls; and (ii) to determine whether the protein expression of these biological markers might be modified by chronic treatment with a cholinesterase inhibitor.
First, we found as expected that platelet AβPP ratio is decreased in this sample of AD patients; this abnormality was attributed to reduction in the protein expression of sAβPP 120–130 kDa along with increased protein expression of sAβPP 110 kDa, corroborating previous studies addressing platelet AβPP metabolism in AD [10, 13]. We further found that unmedicated AD patients also have abnormal protein expression of AβPP secretases, compatible with imbalanced activation of α- and β-secretase pathways. In the present sample, these abnormalities were represented by decreased protein expression of platelet ADAM10 and PSEN1. The former finding is in agreement with previous publicationsindicating reduction in the protein expression of ADAM10 in platelets of AD patients [15, 29], whereas the reduction in the protein expression of platelet PSEN1 is so far a new finding in a clinical sample of AD patients. We did not find statistically significant differences in BACE1 protein expression comparing patients and controls at baseline. There is limited information on platelet BACE1 and, in fact, two studies suggested that the protein expression of this enzyme may actually be reduced in AD [9, 30]. Nonetheless, in the light of the available literature, our findings support the notion that the protein expression of AβPP-cleaving enzymes and AβPP-metabolites in platelets illustrates peripherally some of the core abnormalities observed in the AD brain, hence reinforcing the use of platelets as a source of AD-related biomarkers.
A novel and potentially relevant contribution of the present study is that long-term treatment with donepezil was associated with an inhibitory effect on the protein expression of BACE1 in platelets. As compared to pre-treatment values, the protein expression of this amyloidogenic enzyme was decreased by 24.3% in AD patients treated with donepezil at optimal therapeutic doses (10 mg) for six months. To the best of our knowledge, this is the first study to report such effect in human platelets, suggesting that chronic donepezil therapy may downregulate the β-secretase pathway, by decreasing the availability of BACE1. Accordingly, in a rat model of AD [18], treatment with donepezil hydrochloride led to a significant decrease in the mRNA expression of BACE1 in the hippocampus (along with similar effect on AβPP), and this effect was associated with benefits to acetylcholine metabolism and cognition. Other studies published a few years ago addressing the effect of donepezil on AβPP metabolism may provide additional support to this hypothesis. In an experimental model using cultured neuroblastoma cells [17], treatment with donepezil led to the upregulation of the secretory pathway, through an effect attributed to increased protein expression and activity of ADAM10. The authors interpreted this effect as a consequence of the pharmacological modulation of multiple homeostatic mechanisms, i.e., beyond the receptor-mediated pathway of cholinergic enhancement, involving the promotion of ADAM10 traffic to the cell membrane [16]. In a clinical model of AD, the same group showed that low-dose (5 mg/day) treatment with donepezil for one month led to increased protein expression of platelet ADAM10 [21]. In addition, the same group found that donepezil treatment for 30 days was associated with an increase in the platelet AβPP ratio in mild to moderately demented AD patients, and this effect was more intense among non-carriers of the apolipoprotein ɛ4 allele [31]. In an experimental study using both in vitro and in vivo models [19], the acute administration of donepezil (or tacrine) was associated with decreased protein expression of PSEN1 in cultures of SH-SY5Y neuroblastoma cells and in the brain of wild-type mice; however, this effect was not sustained upon long-term treatment.
Taken together, these pieces of evidence along with data from the present study substantiate the notion that treatment with donepezil may be associated with specific effects on AβPP metabolism. Distinct timeframes and dose-response curves may explain the dissociation of the observed effects on ADAM10, BACE1, and PSEN1 in response to treatment as described in the aforementioned studies. We hypothesize that the downregulation of BACE1 may require a long-term treatment with donepezil at higher doses, since this effect was only seen in patients treated with 10 mg/day for at least six months. As opposed to that, no effects were observed on platelet protein expression of ADAM10 and PSEN1; nonetheless, evidence of modification in the protein expression of these enzymes were only seen upon short-term (1 month-) or acute (3 day-) treatments, in studies respectively conducted in a clinical [21] and an experimental model [19].
The inhibition of BACE1 represents an important therapeutic approach targeting disease modification in AD, at least according to pre-clinical data [9, 33]. Studies suggest that the inhibition of BACE1 activity leads to a reduction in Aβ formation, yielding improvements in cholinergic function, reduction of neuronal loss, and prevention of memory deficits [33–35]. Two pharmacological compounds designed to inhibit β-secretase activity are currently being tested in Phase I (CTS21166) and Phase III (MK-8931) trials in prodromal and mild AD, whereas another trial with LY2886721 was prematurely interrupted due to hepatotoxicity [36]. According to preliminary data from these trials, lack of efficacy in promoting Aβ clearance, in addition to the risk of severe adverse events, represent major challenges to this pharmacological approach [37, 38]. Although a novel finding per se, the present indication that donepezil may actually lead to downregulation of the protein expression BACE1 in platelets must be interpreted with caution, given that the therapeutic effect of cholinesterase inhibitors have so far been attributed to a symptomatic, rather than disease-modifying effect.
Footnotes
ACKNOWLEDGMENTS
This work was supported by Fundação de Amparo á Pesquisa de São Paulo (FAPESP Grant n° 2013/20695-3 and 2009/52825-8), Conselho Nacional de Pesquisa Científica (CNPq, Project 554535/2005-0), Associação Beneficente Alzira Denise Hertzog da Silva (ABADHS), and JNK Empreendimentos e Incorporações.
