Abstract
Accumulation of amyloid-β peptide (Aβ) in the brain is one of the most important features of Alzheimer’s dementia (AD). Cerebral amyloid angiopathy (CAA) is characterized by Aβ accumulation in the walls of cerebral arteries and capillaries, and is present in over 90% of patients with AD. Several novel agents for AD/CAA developed around the amyloid hypothesis have shown positive signs in animal studies but have failed in clinical trials due to adverse events and/or lack of efficiency. As CAA is presumably caused by a failure in Aβ clearance, drugs that promote Aβ clearance may hold promise in the treatment of CAA and possibly AD. With this in mind, cilostazol, an anti-platelet drug with vasodilating action, has been found to promote Aβ clearance along perivascular drainage pathway, reduce Aβ accumulation in the brain, and restore memory impairment in Tg-SwDI mice, an animal model of CAA. We therefore tested whether the most common anti-platelet agent, aspirin, also reduced Aβ and rescued cognitive impairment in Tg-SwDI mice, and also whether aspirin affected hemorrhagic complications that can occur in Tg-SwDI mice. Mice aged 4 months were assigned into vehicle-treated and low-dose aspirin-treated groups. Low-dose aspirin for 8 months did not increase hemorrhagic lesions, nor increase resting cerebral blood flow or cerebral vascular reserve in response to hypercapnia or acetylcholine. Subsequently, aspirin did not restore cognitive dysfunction. These results suggest that low-dose aspirin does not have a direct influence on cerebrovascular Aβ metabolism nor aggravate hemorrhagic complications in CAA.
INTRODUCTION
Alzheimer’s dementia (AD), the most common form of neurodegenerative disease, is characterized by the accumulation of amyloid-β (Aβ) peptide as senile plaques in the brain parenchyma and cerebral amyloid angiopathy (CAA) in the vessel walls. CAA has a prevalence of 90–96% in patients with AD and is present in 30% of non-demented individuals over the age of 60 years [1]. Cognitive impairment is also caused by cerebrovascular diseases such as chronic cerebral hypoperfusion and cerebral infarcts; such cognitive dysfunction has been termed ‘vascular cognitive impairment’ (VCI) [2, 3]. VCI was previously thought to be distinct from AD/CAA but accumulating evidence from several large population-based pathologic and radiologic studies has highlighted the important contribution of vascular risk factors (hypertension and diabetes as primary examples) to AD [4]. Indeed, the Nun study revealed that cerebrovascular diseases play an important role in determining the presence and severity of the clinical symptoms of AD [5]. VCI with concomitant AD/CAA pathology has emerged as an increasing cause of age-related cognitive impairment [2]. Therefore, therapeutic and prophylactic drugs that target vascular factors and cerebral vessels in AD/CAA patients have been recently explored [6].
Anti-platelet drugs are generally able to suppress the development of ischemic atherosclerotic diseases such as ischemic heart disease, ischemic stroke, and peripheral artery disease. One of anti-platelet drugs, cilostazol, is a selective inhibitor of cyclic nucleotide phosphodiesterase III, and is widely used for secondary prevention of cerebral infarction [7]. A recent study showed that cilostazol reduces Aβ deposition and restores cognitive function in an animal model of CAA [8], and clinically improves cognitive function in patients with mild cognitive impairment [9, 10], suggesting that an anti-platelet drug could hold promise as a treatment strategy in AD/CAA. Aspirin (acetylsalicylic acid) has been the most traditional and prevalent drug among anti-platelet agents, and low-dose aspirin is widely used for secondary prevention of cerebral infarction, myocardial infarction, and peripheral artery diseases. The main adverse effect of aspirin is an increasing risk of bleeding such as intracerebral hemorrhage. The Rotterdam scan study showed that cerebral microbleeds (CMBs) are more prevalent among elderly users of anti-platelet drugs, the most common of which is aspirin [11]. Therefore, it remains controversial whether low-dose aspirin can be prescribed safely to AD/CAA patients who may show cerebral hemorrhagic lesions, such as past lobar hemorrhage and lobar-type CMBs together with cortical microinfarcts (CMIs), though aspirin may prevent formation of CMIs with its anti-platelet activity. We therefore investigated whether low-dose aspirin affected cerebral hemodynamics, Aβ metabolism and cognitive function in an animal model of CAA.
MATERIALS AND METHODS
Animals
Tg-SwDI mice on a pure C57BL/6J background were obtained from Jackson Laboratories. These mice express low levels of human Swedish/Dutch/Iowa mutant AβPP in neurons under the control of the mouse Thy1.2 promoter, and develop robust accumulation of both vascular and parenchymal Aβ with predominance of perivascular/vascular Aβ deposits starting at 3 to 5 months old [12]. Male heterozygous Tg-SwDI C57BL/6 mice aged 12 months were used in the present study. The mice were housed in a room with a 12-h light/dark cycle (lights on at 7 : 00 a.m.) with access to food and water ad libitum. All animal experimental protocols were approved by the Institutional Animal Care and Use Committee at the National Cerebral and Cardiovascular Center, and were performed in accordance with the Guidelines for Proper Conduct of Animal Experiments established by Science Council of Japan.
Study design
The animals aged 4 months were assigned into two groups (n = 16 each) and randomized to receive (1) vehicle or (2) aspirin (30 mg/L) in their drinking water, which would be equal to 90 to 120 μg ‘low-dose’ aspirin per day [13], for 8 months; on a body scale-adjusted scale, this amount would be equal to 180 to 240 mg/day if the animals weighed 60 kg, which would significantly inhibit platelet aggregation and production of thromboxane A2 but not alter the amount of prostacyclin compared to placebo control [13]. At 12 months old, resting cerebral blood flow (CBF), vascular reactivity, histologic changes, and cognitive function evaluations were conducted.
CBF measurement assessed with laser speckle flowmetry
Relative CBF was recorded by laser speckle flowmetry (Omegazone, Omegawave; Tokyo), which obtains high-resolution, two-dimensional imaging and has a linear relationship with absolute CBF values [14]. In the day prior to the first CBF measurement, anesthesia was induced with 2% isoflurane and maintained with 1.5% isoflurane in 80% nitrous oxide and 20% oxygen, and the scalp removed by a midline incision, exposing the skull throughout the experiment. During the measurement of CBF, the skull surface was illuminated by 780 nm of laser light. The scattered light was filtered and detected by a CCD camera positioned over the head. The filter detected only scattered light that had a perpendicular polarization to the incident laser light. The raw speckle images were used to compute speckle contrast, which corresponds to the measured velocity of moving red blood cells, approximating CBF. Signal processing was performed by the algorithm developed byForrester et al. [15]. Color-coded blood flow images were obtained in high-resolution mode (639 × 480 pixels; 1 image/sec) and the sample frequency was 60 Hz. One blood flow image was generated by averaging numbers obtained from 20 consecutive raw speckle images. Circular regions of interest with a diameter of 1 mm were located bilaterally at 1 mm posterior and 2 mm lateral to the bregma in each mouse. The recordings were initiated after the examiner confirmed that CBF did not change over 1 min, and the five recordings of CBF image in both hemispheres were averaged. In order to prevent the fluctuation of CBF and blood pressure during the measurement of CBF, anesthesia was induced, as stated above. During the measurement of CBF, mice were held in a small plastic holder on a warming pad and rectal temperature was thermostatically controlled at 36.5°C to 37.5°C. Blood pressure was measured by the tail cuff method and confirmed to be kept constant.
Evaluation of vascular responses to hypercapnia and acetylcholine (ACh)
To induce hypercapnia, non-specific vasodilator, mice were ventilated with 5% carbon dioxide for 5 min, followed by ventilation with 20% oxygen containing air. At first, the CBF changes in response to hypercapnia were evaluated by laser speckle flowmetry. The CBF values were obtained every 30 s for 5 min. The first five images were taken at baseline. Peak CBF increase during the 5 min after hypercapnia was taken as response amplitude using laser speckle flowmetry. The rate of CBF increase after hypercapnia was calculated as the peak CBF increase (%) divided by the baseline CBF. In addition, to evaluate vascular responses to vasodilatory stimuli, a cranial window preparation was performed as previously reported, with modification [16, 17]. In brief, a 3 mm × 3 mm diameter craniotomy was performed in the left parietal bone with a dental drill and the dura mater removed. For real-time in vivo imaging of the cerebral vessels, fibered fluorescence microscopy (MVX10; Olympus) was used. After intravenous tail vein injection of fluorescein isothiocyanate dextran (2 × 106 molecular weight, 200 μL of 20 mg/mL; Sigma-Aldrich), the leptomeningeal vessels were visualized. The images were obtained every 1 min for 5 min after infusion. Averaged vessel diameters across a 25 μm longitudinal segment (5 consecutive segments per mouse) of the dorsal middle cerebral arteries were analyzed, as previously described [16]. Peak vessel diameter increase during the 5 min was taken as response amplitude. Data were calculated as % vasodilation versus baseline vessel diameter. We distinguished penetrating arteries from bridging (collecting) veins by identifying the location and also by following the direction of flow from the pial surface. The endothelium-dependent vasodilator ACh (100 μM; Sigma-Aldrich) was then infused into the cranial window at a rate of 100 μL/min for 5 min using other mice. Data were obtained and calculated by the same methods when inducing hypercapnia.
Histologic investigation
Mice were deeply anesthetized with an intraperitoneal injection of sodium pentobarbital (40 mg/kg) and transcardially perfusion-fixed with 0.01 mol/L phosphate buffered saline, then 4% paraformaldehyde. Mouse brains were removed and sectioned sagittally at 0.2, 1.2, 2.2 and 3.2 mm lateral from the midline. The sectioned brains were embedded in paraffin and sliced into 6 μm-thick sections, which were then subjected to hematoxylin and eosin, and Perls-Stieda’s iron staining. Immunohistochemical staining for mouse Aβ was performed using 1–16 (6E10) monoclonal antibody (1 : 500; COVANCE) on the sagittal section 0.2 mm lateral from the midline. The number of hemorrhagic lesions, such as CMBs and superficial siderosis, defined by Perls-Stieda’s iron staining, were counted in every section. The densitometric analysis of Aβ was performed blindly to animal groups by setting regions of interest in three areas of the cerebral cortex and the hippocampus (600 × 400 μm each), and the percentage area of immunostained Aβ in the cerebral cortex and hippocampus was calculated using the Image J software package (National Institutes of Health).
Y-maze test
Spatial working memory and spontaneous activity was assessed by the Y-maze test, as described earlier [18]. The Y-maze test was conducted during the dark period (7–11 pm) at 12 months old. The maze consists of three identical arms (40 cm long, 9.5 cm high, and 4 cm wide), labeled A, B, or C, diverging at 120° angles from a central point. The experiments were performed in a dimly illuminated room. After each mouse was tested, the floor of the maze was cleaned using super hypochlorous water-soaked paper for the elimination of smell to avoid olfactory cues. Each mouse was placed at the end of the start arm and allowed to move freely through the maze during an 8-min session without reinforcers such as food, water, or electric foot shock. The sequence of arm entries was manually recorded. A mouse was considered to have entered an arm when all four paws were positioned in the arm runway. An alternation was defined as entry into all three arms on consecutive occasions (e.g., the sequence, ABCBCBCA was counted as two alternations, with the first consecutive ABC and the last consecutive BCA out of six consecutive occasions; 33% alternations). The maximum alternation was calculated as the total number of arm entries minus two, and the percentage of alternation was calculated as (actual alternation/maximum alternation) × 100. The total number of arms entered during the sessions, which reflects spontaneous activity, was also recorded. Mice that entered arms less than eight times during the test were eliminated because their data were not considered to reflect precisealternation.
Statistical analysis
All values are expressed as means±standard error in the figures. Individual comparisons were analyzed by unpaired t test. Differences with a probability value of p < 0.05 were considered to be statistically significant.
RESULTS
Low-dose aspirin did not increase cerebral hemorrhagic lesions in the Tg-SwDI mice
The 12-month-old Tg-SwDI mice occasionally develop microhemorrhages [12] and CMIs when subjected to chronic cerebral hypoperfusion [19]. We investigated whether low-dose aspirin increased hemorrhagic lesions, such as superficial siderosis and CMBs, and ischemic lesions, such as CMIs, after application of low-dose aspirin in Tg-SwDI mice. The number of hemorrhagic lesions was not significantly different between the two groups (vehicle versus low-dose aspirin; 2.8 versus 2.3/per one brain; p > 0.05), and some mice showed CMBs in the cortex and superficial siderosis (Fig. 1). CMIs, however, were not observed in either group. These results indicate that low-dose aspirin does not induce cerebral hemorrhage in Tg-SwDI mice.
Low-dose aspirin did not restore cerebrovascular reserve in the Tg-SwDI mice
Resting CBF was not significantly different between the two groups, though the CBF level of low-dose aspirin group was slightly higher compared to that of vehicle-treated group (vehicle versus low-dose aspirin; 48.1 versus 49.2, as arbitrary unit; p > 0.05) (Fig. 2A). We then evaluated the cerebral hemodynamic reserve response to hypercapnia, which is a non-specific vasodilator via the relaxation of endothelial smooth muscle cells. In response to hypercapnia, CBF tended to be lower in the low-dose aspirin group (vehicle versus low-dose aspirin; 14.8% versus 12.1%; p > 0.05) (Fig. 2B), and low-dose aspirin did not significantly dilate leptomeningeal arteries, compared to vehicle (vehicle versus low-dose aspirin; 10.8% versus 11.9%; p > 0.05) (Fig. 2C). To measure cerebral arterial endothelial function in Tg-SwDI mice, we infused ACh, an endothelial-dependent vasodilator, onto the brain surface and evaluated the vasodilatory response to ACh. There was nodifference in vasodilatory response to ACh between the two groups. The leptomeningeal vascular diameter in low-dose aspirin-treated group tended to be lower than in the vehicle-treated group (vehicle versus low-dose aspirin; 12.2% versus 9.3%; p > 0.05) (Fig. 2D). These findings suggest that low-dose aspirin did not restore the hemodynamic reserve, represented as vascular reactivity to carbon dioxide and ACh, in Tg-SwDI mice.
Low-dose aspirin did not reduce Aβ accumulation in the Tg-SwDI mice brain
The low-dose aspirin group showed the similar amount of Aβ deposits compared to the vehicle-treated group in the frontal cortex (vehicle versus low-dose aspirin; 9.1% versus 7.9%; p > 0.05) (Fig. 3) and hippocampus (vehicle versus low-dose aspirin; 5.5% versus 7.2%; p > 0.05) (Fig. 3). The Aβ deposits were vasculotropic and mostly distributed in and around the cerebral microvessels (Fig. 3) which was compatible with the findings observed in previous reports [8, 12]. These results indicate that low-dose aspirin did not eliminate vascular Aβ accumulation in the brains of Tg-SwDI mice.
Low-dose aspirin did not rescue cognitive dysfunction in the Tg-SwDI mice
The percentage of alternation behaviors (used as an indicator of working memory) was not significantly different between the vehicle-treated and low-dose aspirin group at 12 months (vehicle versus low-dose aspirin; 54.5% versus 56.3%; p > 0.05) (Fig. 4A). The number of arm entries, which reflects spontaneous activity, was significantly decreased in the low-dose aspirin group compared to the vehicle-treated group (vehicle versus low-dose aspirin; 15.1 versus 10.0) (Fig. 4B). These results suggest that low-dose aspirin does not prevent cognitive decline, but may prevent hyperactivity in Tg-SwDI mice.
DISCUSSION
The number of patients with AD has risen worldwide, and several disease-modifying drugs, including immunotherapy agents such as Aβ vaccination, or agents that lower Aβ synthesis, such as β- or γ-secretase inhibitors, have been successfully established based on the amyloid hypothesis in animal studies. However, such interventions have generally failed to suppress cognitive decline and induced adverse events in clinical trials. For example,immunotherapy against Aβ42 was successful in experimental studies, but could not halt cognitive decline and caused cortical microbleeds and vasogenic brain edema, possibly due to accumulation of soluble Aβ along capillary and artery walls [20–22].
A recent study has reported that reduced Aβ clearance from the brain rather than increased Aβ production is responsible for sporadic AD [23]. Several mechanisms of eliminating Aβ proteins have been identified, consisting mainly of three categories: (1) enzymatic/glial degradation; (2) transcytotic delivery; and (3) perivascular drainage [6]. Among them, the drainage of extracellular Aβ along the basement membranes of capillaries and arteries appears to be the most important mechanism for removal of Aβ from the brain in diseased states [24, 25]. Besides impacting on the structural and functional components of the neurovascular unit, deposition of Aβ within the perivascular drainage pathway (namely, CAA) lowers the motive force of Aβ clearance, thereby contributing to increased parenchymal Aβ deposition [1, 26]. While experimentally cilostazol was found to promote perivascular drainage of Aβ [8], the current study showed that another anti-platelet agent, aspirin, had no influence on Aβ metabolism, suggesting no direct effect on the perivascular drainage pathway. One of the differences between the two antiplatelet agents is that, unlike aspirin, cilostazol is a vasoactive drug that has a direct effect on vascular smooth muscle cells. Thus, antiplatelet activity may not play a role in removal of Aβ from the brain.
Besides its platelet-inhibitory function by inhibiting cyclooxygenase-1, preceding thromboxane A2 [13], low-dose aspirin is known to increase nitric oxide production [13, 27]. In the present study, however, low-dose aspirin did not significantly increase vascular reactivity in response to hypercapnia or ACh, compared with vehicle, suggesting that it did not produce sufficient nitric oxide to dilate leptomeningeal arteries. The lack of direct action on vessels may explain why low-dose aspirin did not promote perivascular Aβ drainage or have any effect on cognitive dysfunction.
Of note is that low-dose aspirin did not increase hemorrhagic complications, such as CMBs or superficial siderosis, even in old Tg-SwDI mice, although several clinical reports have indicated an association between aspirin administration and intracerebral hemorrhagic lesions [11]. Aspirin is reported to increase recurrent intracerebral hemorrhage in CAA patients [28], although some other unknown factorsmay associate with recurrent hemorrhage. Intriguingly, blood-brain barrier leakage/injury rarely develop in normotensive Tg-SwDI mice but does develop in Tg-SwDI mice exposed to chronic hypertension chemically induced by Nω-Nitro-l-arginine methyl ester hydrochloride [29]. Use of hypertensive Tg-SwDI mice would be required to further investigate effects of aspirin on CAA because in humans lobar CMBs are related to diastolic blood pressure [30] and blood pressure lowering has been shown to reduce the risk of CAA-related intracerebral hemorrhage in the PROGRESS study [31].
One of the limitations of the study is that, while hypoperfusion is known to induce formation of CMIs, they did not develop in 1-year-old Tg-SwDI mice [19]. CMIs are frequently observed and preferentially distributed in arterial borderzone areas, and are thought to be a pathological substrate of dementia in AD/CAA patients. Therefore, aspirin may have rescued phenotypes of Tg-SwDI mice if such CMIs were successfully induced in older mice. Similarly, older mice may have suffered from greater hemorrhagic complications, which may have been accelerated with aspirin.
In conclusion, our study showed that low-dose aspirin did not have positive effects on cerebrovascular Aβ or cognitive impairment in Tg-SwDI mice. However, low-dose aspirin did not increase cerebral hemorrhagic lesions in normotensive Tg-SwDI mice, suggesting that aspirin may be used in AD/CAA patients, if required, under strict control of vascular risk factors, such as hypertension.
Footnotes
ACKNOWLEDGMENTS
We acknowledge Dr. Ahmad Khundakar for editing the manuscript and are indebted to Ms. Takako Kawada and the late Dr. Yoko Okamoto for their excellent technical assistances in staining and tissue sections. This work was supported by grant support from the Ministry of Health, Labour and Welfare (M.I.), the Ministry of Education, Culture, Sports, Science and Technology (M.I., Grant-in-Aid for Scientific Research (B); Y.H., Grant-in-Aid for Challenging Exploratory Research), and the Takeda Science Foundation (M.I.).
