Hulette, Christine M. MD; Ervin, John F. BA; Edmonds, Yvette BS; Antoine, Samantha BA; Stewart, Nicolas BA; Szymanski, Mari H. RN; Hayden, Kathleen M. PhD; Pieper, Carl F. DPH; Burke, James R. MD, PhD; Welsh-Bohmer, Kathleen A. PhD
Efforts to understand the pathophysiology of blood-brain barrier compromise in Alzheimer disease (AD) have received increased attention in recent years (1), and some authors have suggested that vascular disease is the primary pathological event in the development of Alzheimer-type dementia (2, 3). Our group previously demonstrated that there is significantly less smooth muscle actin (SMA) immunoreactivity in leptomeningeal and intracerebral arterioles of patients with AD than in nondemented, neuropathologically normal older subjects. Moreover, we found that AD patients with the apolipoprotein E (ApoE) ϵ4/ϵ4 genotype have less vascular SMA immunoreactivity than ApoE ϵ3/ϵ3 AD patients (4, 5). These observations suggested that vascular disease may be important in the pathogenesis of AD and that the ApoE ϵ4/ϵ4 genotype is associated with enhanced degeneration of smooth muscle cells within the cerebral vasculature. Possession of at least one copy of the ApoE ϵ4 allele is associated with increased β-amyloid deposition in cerebral vessels (6). In addition, we have shown that vascular β-amyloid deposition is maximum in cortical arterioles of AD subjects with the ApoE ϵ4/ϵ4 genotype (7). Thus, changes in cerebral arteries and arterioles in AD patients could result in impaired blood-brain barrier function or may contribute to inadequate drainage or removal of β-amyloid. This hypothesis is supported by observations of reduced cerebral blood flow in patients with AD and incipient AD (8). Vascular disease, including amyloid angiopathy, is very common in AD brains; β-amyloid, the major pathological protein associated with amyloid angiopathy, exhibits a high avidity binding to ApoE lipoprotein (9), the most common genetic risk factor for AD (10). Thus, our observations also imply that different pathological mechanisms may play varying roles in the different genetic forms of AD (5, 7).
This study was undertaken to extend and strengthen previous preliminary observations of vascular pathology in AD (11) by examining increased numbers of nondemented subjects and subjects with dementia due to a neurodegenerative disorder other than AD. Using methods identical to those in our previous report (5), we analyzed the severity of microvascular disease by immunostaining for SMA (the major structural protein expressed in arteries and arterioles) in 3 additional patient groups: subjects with frontotemporal dementia (FTD), nondemented subjects with no significant pathology at autopsy (Normal), and nondemented subjects with frequent plaques at autopsy. In accordance with our previous work and that of others, this latter group is termed "Preclinical AD" (12, 13).
MATERIALS AND METHODS
Subjects were enrolled in the autopsy and brain donation program of the Joseph and Kathleen Price Bryan Alzheimer Disease Research Center (ADRC), as described previously (14). The Duke University Medical Center Institutional Review Board approved the procedures for enrollment. After death, consent for autopsy was obtained from the next of kin, and the autopsy was performed according to institutional guidelines. Brains were evaluated neuropathologically and banked according to published protocols (15-19). Using NIA-Reagan Institute recommendations (20), the brains were examined and diagnosed according to The Consortium to Establish a Registry for Alzheimer's Disease (CERAD) criteria (21). Neurofibrillary change was staged according to Braak (22). Apolipoprotein E genotyping was performed using established methods (10).
Determination of Cognitive Status
All nondemented subjects were living independently or with some assistance from family for physical disability due to age or illness at least 6 months before death. Within 1 month before or after death, the next of kin was interviewed by a nurse coordinator (M.H.S.). The next of kin thought that the subject was cognitively normal. Twenty-nine of the 40 nondemented subjects had undergone clinical evaluation in the Bryan ADRC and had at a minimum Mini Mental Status Examination scores (23). All 29 subjects were reviewed through a clinical diagnostic procedure that consisted of a review of self (or proxy)-report of memory and functional change, medical history, family history, and medication information by a nurse clinician specialist (M.H.S.). Consensus diagnoses were assigned after a review by a team of clinician specialists. Cases were assigned to diagnostic groups of clinically normal, mild cognitive impairment (MCI), or cognitively impaired not demented, based on all available clinical information including the provisional diagnoses assigned by the nurse, medical history, clinical notes, and neuropsychological test data. The MCI diagnosis was based on the original proposed criteria (24) that noted a rather isolated but pronounced impairment in recent memory function (deficits exceeding 1.5 SD below), which appeared in the setting of otherwise normal cognition (Mini Mental Status Examination score >23), included no impairments or only very minimal impairments in higher order Instrumental Activities of Daily Living functions (corresponding to a clinical dementia rating score of 0.5), and for which there was no known medical explanation. The cognitively impaired not demented designation was reserved for MCI not reaching the level of dementia and ascribed to a known medical illness such as stroke.
All subjects in the FTD group had neurological impairment. Four had a clinical diagnosis of dementia of unclear and therefore questionable etiology. We have previously used the designation of dementia of unclear and therefore questionable etiology for dementia of all cause, including AD. It is based on Diagnostic and Statistical Manual of Mental Disorders, Third Edition criteria and defined as evidence of a change in functional abilities accompanied by cognitive impairment encompassing more than 1 cognitive domain. Three patients in the FTD group had a clinical diagnosis of frontal lobe dementia, 1 had a clinical diagnosis of Parkinsonism with dementia, and 1 had a diagnosis of dementia with sleep disorder. No clinical information was available on 1 FTD subject.
The cases selected from the Kathleen Price Bryan Brain Bank were matched for age and sex with those in the previous report (5). There were 11 male and 9 female normal subjects with an average age of 78.5 years at death (range, 60 to >90 years). These 20 subjects had histologically normal brains, that is, no neuritic plaques or only a few diffuse plaques in the cortex and no histological evidence of vascular amyloid as determined using an immunostain for β-amyloid peptide. The minor changes corresponded to a CERAD score of 1A or 1B (21). Neurofibrillary changes were documented by AT8 tau immunostain. Mild neurofibrillary change was only seen in the entorhinal cortex and hippocampus corresponding to Braak stage I or II (22). Eight of the subjects had mild to moderate vascular disease, 6 had focal mild cerebrovascular atherosclerosis in the arteries at the base of the brain, and 2 had moderate cerebrovascular atherosclerosis. None of these subjects had intracerebral vascular amyloid in either the middle frontal (MF) cortex or the inferior parietal (IP) lobule (Table 1).
Preclinical AD Subjects
There were 9 male and 11 female nondemented subjects with frequent plaques at autopsy according to CERAD criteria (21). In accordance with previous published work from our center (12) and from other investigators (13), we designate this group Preclinical AD. These subjects had an average age of 81 years at death (range, 71 to >90 years). Cognitive evaluation was performed as described above and in previous publications (12). All were determined to be nondemented 1 year before death by neurocognitive testing or informant interview. When they were examined at autopsy, all 20 subjects had frequent amyloid plaques in the cortex as demonstrated by immunostain for β-amyloid peptide. Mild neurofibrillary change was seen in the hippocampus and entorhinal cortex in all of these subjects. Some of the plaques in a few of these subjects had associated neurites, as demonstrated by AT8 immunostain. The severity of AT8 deposition in plaques was variable, and they were most frequent in the hippocampal formation. A few tangles were observed in the inferior temporal neocortex in 12 of these subjects; therefore, the Braak stages ranged from I to III. Eleven subjects who overlapped with but were not congruent with these had significant vascular disease; 3 had mild, 1 had moderate, and 1 had severe atherosclerosis in the vessels at the base of the brain. Microinfarcts or lacunes were seen in 3 subjects; 4 had 2 or 3 vascular abnormalities. Although 6 had mild and 1 had moderate amyloid angiopathy of the intracerebral arteries that was restricted to the meningeal vessels, none of the Preclinical AD subjects had severe vascular amyloid deposition and none had vascular amyloid in the cerebral cortical arterioles or white matter in the sections used for the SMA analysis. Vascular amyloid was restricted to the occipital lobe and cerebellum in 3 of the 7 cases with amyloid angiopathy. Thus, only 4 had mild amyloid angiopathy in the frontal and/or parietal lobes that were used for SMA analysis (Table 1).
The demographic characteristics of FTD cases were matched with the AD cases in our previous report (2). All of the FTD cases were ApoE ϵ3/ϵ3. The subjects ranged in age from 55 to >90 years at death with an average of 70.8 years. There were 7 men and 3 women. Brains were examined routinely as previously described using methods identical to those used for the nondemented subjects (15-18). Two FTD cases had severe atherosclerosis; 1 had mild atherosclerosis in the vessels at the base of the brain. Microscopically, 1 of the FTD cases had mild amyloid angiopathy and 1 had moderate amyloid angiopathy. Rare or only a few primitive plaques were present in these brains. Neurofibrillary change was assessed by AT8 immunostain as previously described (25); all of the FTD cases had marked neurofibrillary change in the hippocampal formation and variable numbers of neurofibrillary tangles in the substantia nigra and basal ganglia. Neither family history nor genetic data about tau mutation or polymorphism status were available for these subjects (Table 2).
Histology and Immunohistochemistry
Paraffin-embedded sections from the MF cortex and the IP cortex of each subject were immunostained for SMA. Sections were cut at 8 μm, deparaffinized in xylene, and hydrated through graded ethanols. Endogenous peroxidase activity was blocked with 2% H2O2 in methanol. Boiling the sections in citrate buffer (pH 6.0) for 30 minutes was used for antigen retrieval. A monoclonal antibody to SMA (clone 1A4; DAKO Corp, Carpinteria, CA) was used. The antibody was diluted 1:100 and incubated for 1 hour at 37°C. A biotinylated horse anti-mouse immunoglobulin G secondary antibody (Vector Labs, Burlingame, CA) 1:300 was incubated for 20 minutes at 37°C. Horseradish peroxidase-labeled streptavidin (DAKO) diluted 1:500 was incubated for 20 minutes at 37°C. The sections were developed with a 0.5 mg/mL diaminobenzidine (Sigma, St Louis, MO) solution for 3 minutes.
Phase contrast microscopy was used to identify vessels for image analysis. This enables the operator to visualize cellular details better than counterstaining with hematoxylin because hematoxylin may obscure the SMA signal, unless there is large amount of positive staining. Early trials demonstrated that SMA staining may be missed or underestimated when hematoxylin counterstain is used. Within each cortical section, 5 small arteries or arterioles in each of 3 anatomical regions (i.e. arachnoid, gray matter, and white matter) were selected for detailed image analysis. The vessels had average external diameters of 39.7, 19.6, and 21.8 μm in the arachnoid, cortex, and white matter, respectively. Images of the vessels were captured using a Pixera Penguin 600CL camera attached to a Nikon Optiphot-2 microscope. The images were saved using ViewFinder 3.0 software. They were analyzed using Metamorph 4.12 image analysis software from Universal Imaging. Images were captured and analyzed randomly by an observer who was blinded to the diagnosis. Two images of each vessel were taken; the first image allowed visualization of the entire structure, and the second image was captured after adjusting the light intensity to allow visualization of the positively stained portion of the vessels.
The amount of positive staining in the artery wall was quantified using Metamorph 4.12 by determining the pixel area from the first set of captured images. The vessel was divided into 2 regions using a drawing tool. Region 1 represented the entire vessel by defining the outer vessel wall including the lumen; region 2 represented only the lumen only. By subtracting the pixel area of the lumen from the total pixel area, the pixel area of the vessel wall was determined. The area of the positively stained portion of vessel wall was then measured in the second captured image by drawing a region around the image and thresholding for dark objects. The percentage of staining was then determined by dividing the pixel area of the positively stained portion by the pixel area of the total vessel wall. This analysis was performed on MF and IP samples in each case.
Observers (Y.E., S.A., and N.S.) were trained by a neuropathologist (C.M.H.) to detect and examine only small arteries or arterioles; veins were not examined. A single individual who is skilled with image analysis (J.F.E.) supervised all experiments and trained the observers to use the drawing tube so that only images of vessel walls were captured for the analysis. Duplicate determinations for 1 or 2 vessels from each subject were performed to ensure that image analysis was uniform between experiments and to allow comparison of data between observers. After the initial training period, there was less than 5% variability between observers.
The design of this study is a repeated-measures or split plot design. The outcome of interest (SMA expression) was assumed to be continuous, and there was a single between-group factor with 3 levels (i.e. Normal, Preclinical AD, and FTD). There were 2 within-person factors: region (MF and IP), with 3 levels, and type (arachnoid, gray matter, and white matter), also with 3 levels. With this balanced data set and no missing values, a repeated-measure analysis of variance was used (26). Initially, a fully interactive repeated-measures model testing the group by region by type interaction was assessed. When the omnibus F-test for this 3-way interaction was not declared significant, a follow-up test of the 3 two-way interactions (i.e. group by type, group by region, and region by type) was assessed. The result of this test of this full 2-way interaction model is reported. Follow-up tests of pairwise comparisons and tests of the assumptions of the model are also reported.
By light microscopy, SMA immunostaining in the arteries examined was seen only in the tunica media. The positively stained arteries in the arachnoid were of large caliber; stained arteries in the gray matter were of intermediate caliber; and stained arteries in the white matter were the smallest in diameter (data not shown). Comparison of SMA densities between the different groups highlighted a pattern in vascular SMA density that was similar in the MF and IP lobes when each area was examined separately (data not shown). Smooth muscle actin densities in both male and female subjects were similar when they were examined separately (data not shown); therefore, data from male and female subjects were also combined for analytical purposes.
In all groups, there was more vascular SMA in the arachnoid than in the gray matter or white matter arteries; gray matter arteries had more SMA immunoreactivity than white matter arteries (Fig. 1). The Metamorph image analysis also showed patterns of SMA staining to be similar for the MF and IP lobes in each region when they were examined separately, and this was confirmed by statistical analysis.
Arachnoid vessel staining and white matter vessel staining were different in both the Normal and Preclinical AD subjects; this was also the case in comparisons between gray and white matter arteries. Only the Normal group showed a significant difference between arachnoid and gray matter arteries.
Comparison of vascular staining between the Normal and FTD groups showed that there was more SMA staining in the arteries from all regions in the Normal group (p = 0.0014). The FTD group also had less vascular SMA immunoreactivity in all regions than the Preclinical AD group (p < 0.0001). Comparison of vascular staining between the Normal and Preclinical AD groups showed that there was more SMA staining in the arteries from all regions in the latter (p < 0.0001). Excluding subjects with mild or moderate vascular amyloid from the analysis yielded a slightly less significant result (p = 0.0003). Data of SMA immunostaining in the 3 anatomic regions in each of the 3 subject groups are summarized in Figure 1.
We also examined the possible effect of age on the microvasculature. The overall difference between the 3 groups was p = 0.015 by analysis of variance (df = 2); however, p = 0.29 for the difference between the Normal and Preclinical AD groups. In accordance with the Health Insurance Portability and Accountability Act regulations, true age was not recorded for the subjects older than 90 years. Therefore, these findings were replicated (with minor differences in the p values) when the ages were analyzed nonparametrically.
Because the presence of vascular amyloid might have had an effect on SMA expression, we also compared data from nondemented subjects with and without amyloid angiopathy. Because amyloid was present only in the arachnoid vessels of 4 cases in the Normal group, results of vascular SMA expression in the arachnoid vessels of the 4 nondemented subjects with vascular amyloid were compared to those of the 16 nondemented subjects without vascular amyloid. The differences were not significant (p = 0.13). We also compared SMA expression in the FTD cases with and without vascular amyloid; this result was also not significant (arachnoid: p = 0.33; gray matter: p = 0.77; white matter: p = 0.1).
Changes in the intracerebral vasculature and the possibility that these changes may affect the course of AD are of long-term interest to this laboratory (4-7). The pathobiology of blood vessel components and their role in human disease have been extensively examined and reviewed by others. These include, but are not limited to, studies of atherosclerosis and hypertension (27, 28), cancer (29), and mechanical vascular injury (30). The contribution of vascular disease to the pathogenesis of AD and the idea that cerebral hypoperfusion may be important to the development of AD dementia have recently enjoyed renewed focus (31). It is not clear whether the observed vascular changes reported by us (4-7) and others (1-3, 8) assert an effect on the clinical manifestations of the disease or if the changes themselves are merely a reflection of the progressive degeneration and development of AD type neuropathology.
Most of the cases in the present study came to autopsy and were banked by the ADRC Brain Bank at Duke University many years before the National Alzheimer Coordinating Center was established and Uniform Data Set implementation; therefore, the available clinical data are somewhat variable. All 40 of the nondemented subjects used for this analysis were considered by the next of kin to be "normal." Another, potentially confounding factor was systemic hypertension, which would be expected to induce changes in the microvasculature. Systematic clinical data about the presence or absence of hypertension were not obtained. Such information will be available in future research studies. Nevertheless, differences in the intracerebral vasculature persist in these neuropathologically defined groups.
The FTD cases reported here showed significantly more SMA staining in all 3 tissue anatomic regions than the previous reported AD cases (5). Comparing FTD cases to the combined AD group showed a small but significant difference in smooth muscle pathology between the 2 disease groups (Fig. 2). This suggests that in addition to differences in the clinical and neuropathological features of FTD and AD (24), there were also differences in their microvascular pathology. Vascular pathology seemed to play a much more important role in the pathobiology of AD.
The average ages of the 3 groups parallel the SMA findings and may represent a partial explanation for the differences in SMA immunoreactivity between the groups. The Normal and FTD subjects were on average younger (average, 78.5 and 70.8 years, respectively) than the Preclinical AD subjects (average, 81 years). The youngest FTD group showed the least SMA staining among the groups. However, the average age of the previously reported definite AD group was 75.5 years, with a range of 60 to 81 years at death (5). This is similar to the ages of the Normal subjects in the present study. Thus, it seems that increased SMA immunoreactivity in nondemented subjects with frequent plaques is unlikely to be due simply to age. When the relationship between SMA immunoreactivity and age is examined statistically, there is an age difference between FTD and the other 2 groups but not between the Normal and Preclinical AD groups
Age may be associated with progressive degenerative changes of the cortex that are ultimately the direct cause for the substantial loss of cognitive functioning seen in end-stage AD patients. The devastating loss of cognition parallels the loss of neurons, synaptic density, and brain mass that can be easily visualized in the severe cases but is more difficult to detect in milder cases (32). Less obvious are the factors that induce neuronal atrophy and cell death, which are the substrate of this loss of brain mass and cognitive function. Changes in the cerebral vasculature are common with age and may also be important in AD pathogenesis.
The intracerebral arteries, such as arteries elsewhere in the body, have a muscular wall composed of SMA and other contractile proteins. Actin is the most abundant protein found in a typical eukaryotic cell; it is highly conserved and participates in a large variety of functions (27-29). Myofibroblasts demonstrate transient expression of SMA during wound healing (30), which could be of some significance if arterial damage is ongoing in AD (31). Schildmeyer et al (33) demonstrated that smooth muscle α-actin null mice expressed skeletal α-actin in vascular smooth muscle in the absence of smooth muscle α-actin. The mice showed highly compromised vascular contractility, tone, and blood pressure. Smooth muscle actin thus plays an essential role in vascular function. Smooth muscle actin may also represent a marker of vascular integrity. The percentage of positive SMA staining is a reflection of the number and function of smooth muscle cells in the tunica media layer of arterial walls. The image analysis used here provides a semiquantitative value to smooth muscle immunoreactivity and possibly smooth muscle cell density in the arterial vessel walls. The FTD cases showed higher SMA staining in the arachnoid vessels than gray or white matter vessels and higher SMA staining in the gray matter arteries than white matter in both the MF and IP regions of the brain. This pattern is a reflection of the normal caliber of vessels in these regions. Arteries in the arachnoid layer have larger diameter and relatively larger tunica media layer than vessels in the gray or white matter layer. This implies that there is a greater smooth muscle cell density and hence greater positive SMA staining. Gray matter layer vessels are smaller in diameter and have a relatively smaller tunica media layer than arachnoid vessels because they carry less blood, which is under lower pressure. Gray matter blood vessels are surrounded by neuronal cell bodies and neuropil; therefore, gray matter vessels do not expand as much as arachnoid vessels that are surrounded by very loose arachnoid tissue. White matter blood vessels carry a still smaller quantity of blood deep into the tissue and are under less pressure than the vessels in the other 2 tissue layers. Thus, the behaviors of smooth muscle cells in arachnoid, gray matter and white matter are different, and this is reflected in the density of SMA immunoreactivity found in the vessels at each tissue layer.
Within the group of nondemented subjects without plaques, there was more SMA immunoreactivity in the arteries of the arachnoid than in those of the gray and white matter. In turn, the arteries in the gray matter exhibit more immunoreactivity than the white matter. This may also be due to normal anatomical variation as a reflection of their relative sizes, which incidentally show decreasing diameters in the arachnoid, to gray, and to white matter vessels. In the Preclinical AD group, there was no longer a significant difference in the staining intensity between vessels of the arachnoid and gray matter regions. This may be due to absence of the normal increased staining density of the arachnoid vessels. Alternatively, the arachnoid vessels in Preclinical AD may not acquire the same physiological hypertrophy as the cortical vessels. The finding of increased expression of arterial SMA in the Preclinical AD group raises the question of how these vascular changes may affect the pathogenesis of AD. The concept of hypertrophy as physiological response to stress or increased function has many parallels in other organ systems. For example, patients who undergo nephrectomy develop enlargement of the remaining kidney, and cardiac hypertrophy occurs in the muscle of athletes and as a stress response in patients who have untreated hypertension.
Autopsy neuropathology studies of AD and MCI at the molecular and subcellular level also suggest that neurons may be less able to maintain function as they age. This may be expressed clinically as age-related memory loss or MCI. Physiological response to this impaired function may result in upregulation of vital cellular processes. This upregulation may protect against the subsequent development of dementia in some individuals with MCI. Molecular studies support this hypothesis. Cholinergic plasticity when examined by hippocampal choline acetyltransferase (ChAT) activity varies with AD, normal cognition, and MCI. Severe AD cases from the Religious Orders study showed markedly depleted hippocampal ChAT levels. Mild AD cases were similar to normal controls, but MCI cases showed maximal ChAT activity (32). Similarly, increased size of the Golgi apparatus and increased metabolic activity in the nucleus basalis of Meynert have been reported in MCI cases compared with normal controls and subjects with moderate AD (34). In addition, a recent report demonstrates neuronal hypertrophy in pyramidal neurons of the hippocampus and in the cingulate gyrus in individuals with "asymptomatic AD" (35). In this study, asymptomatic AD was defined as an individual with substantial numbers of plaques and tangles at autopsy but who exhibited normal cognition before death. Herein and elsewhere (12), we have termed this group "Preclinical AD." In vivo imaging studies of the central nervous system also support this idea. Functional neuroimaging studies have demonstrated increased hippocampal activation in MCI compared with normal aging and AD (36). Consequently, there is a growing body of evidence that some individuals with MCI may be able to mount a physiological response to the neurodegeneration that precedes the progressive dementia that is characteristic of AD. Enhancing these physiological responses could prevent or delay the onset of AD.
Although much is known about the gross and histological abnormalities present in the AD brain, there is ongoing debate as to which feature precedes all the rest and begins the inexorable chain reaction of pathological events. We report here the intriguing observation of enhanced SMA in the nondemented subjects with frequent plaques at autopsy. This suggests that increased SMA expression and other as yet unexplored molecular events that occur along with the development of neuritic plaques may protect against the overt clinical expression of dementia. The findings reported here emphasize the need for further molecular and clinical studies of the progression from normal cognition to MCI and subsequent, but not inevitable, development of AD.
Donors and families are gratefully acknowledged for their participation in the Kathleen Price Bryan Brain Bank Autopsy and Brain donation program. Pathologists throughout the United States are acknowledged for their help with tissue retrieval.
Preliminary findings were presented in abstract form at the American Association of Neuropathologists annual meeting in June 2005.
1. Donahue JE, Johnason CE. Apolipoprotein E, amyloid-β, and blood-brain barrier permeability in Alzheimer disease. J Neuropathol Exp Neurol 2008;67:262-70
2. Bueé L, Hof PR, Bouras C, et al. Pathological alterations of the cerebral microvasculature in Alzheimer's disease and relating dementing disorders. Acta Neuropathol 1994;87:469-80
3. Bueé L, Hof PR, Delacourte A. Brain microvascular changes in Alzheimer's disease and other dementias. Ann N Y Acad Sci 1997;826:7-24
4. Hulette CM, Van Eck M, Pannell C, et al. Vascular pathology in ApoE 3, 3 and ApoE 4, 4 dementia. In: Korczyn AD, ed. First International Congress of Vascular Dementia
. Bologna, Italy: Monduzzi Editore, 1999:45-9
5. Ervin JF, Pannell C, Szymanski M, et al. A quantitative analysis of vascular smooth muscle actin in Alzheimer's disease. J Neuropathol Exp Neurol 2004;63:735-41
6. Schmechel DE, Saunders AM, Strittmatter WJ, et al. Increased β-peptide deposition in cerebral cortex as a consequence of ApoE genotype in late-onset Alzheimer's disease. Proc Natl Acad Sci U S A 1993;90:9649-53
7. Trembath D, Ervin JF, Broom L, et al. Vascular beta-amyloid deposition in Alzheimer's disease varies with APOE genotype: A quantitative analysis. Acta Neuropathol 2006;113:23-31
8. Rutenberg A, den Heijer T, Bakker SLM, et al. Cerebral hypoperfusion and clinical onset of dementia: The Rotterdam Study. Ann Neurol 2005;57:789-94
9. Strittmatter WJ, Saunders AM, Schmechel DE, et al. Apolipoprotein E: High avidity binding to β-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer's disease. Proc Natl Acad Sci U S A 1993;90:1977-81
10. Saunders AM, Strittmatter WJ, Schmechel DE, et al. Association of apolipoprotein E allele ϵ4 with late-onset familial and sporadic Alzheimer's disease. Neurol 1993;43:1467-72
11. Hulette CM, Ervin JF, Edmonds Y, et al. Cerebrovascular smooth muscle actin in increased in possible Alzheimer disease brain tissue: A controlled autopsy study. J Neuropathol Exp Neurol 2005;64:467
12. Hulette CM, Welsh-Bohmer KA, Murray MG, et al. Neuropathological and neuropsychological changes in "normal" aging. Evidence for preclinical Alzheimer's disease in cognitively normal individuals. J Neuropathol Exp Neurol 1998;57:1168-74
13. Price JL, Morris JC. Tangles and plaques in nondemented aging and "preclinical" Alzheimer's disease. Ann Neurol 1999;45:358-68
14. Hulette CM, Welsh-Bohmer KA, Crain B, et al. Rapid brain autopsy: The Bryan Alzheimer's Disease Research Center experience. Arch Pathol Lab Med 1997;121:615-18
15. Hulette CM, Pericak-Vance MA, Roses AD, et al. Neuropathologic features of frontotemporal dementia and Parkinsonism linked to chromosome 17q12-22 (FTDP-17): Duke family 1684. J Neuropathol Exp Neurol 1999;58:859-66
16. Rosenberg CK, Cummings T, McIntyre LM, et al. Dementia with Lewy bodies and Alzheimer disease. Acta Neuropathol 2001;102:621-26
17. Trembath Y, Rosenberg C, Ervin JF, et al. Lewy body pathology is a frequent co-pathology in familial Alzheimer disease. Acta Neuropathol 2003;105:484-88
18. Hulette CM. Brain banking in the United States. J Neuropathol Exp Neurol 2003;62:717-22
19. Corder EH, Lockhart E, Ervin JF, et al. Cardiovascular damage in Alzheimer disease: Autopsy findings from the Bryan ADRC. J Biomed Biotechnol 2005:2;189-97
20. Hyman BT, Trojanowski JQ. Consensus recommendations for the postmortem diagnosis of Alzheimer disease from the National Institute on Aging and the Reagan Institute Working Group on diagnostic criteria for the neuropathological assessment of Alzheimer disease. J Neuropathol Exp Neurol 1997;56:1095-97
21. Mirra SS, Heyman A, McKeel D, et al. The Consortium to Establish a Registry for Alzheimer's Disease (CERAD) Part II. Standardization of the neuropathological assessment of Alzheimer's disease. Neurol 1991;41:479-86
22. Braak H, Braak E. Neuropathological staging of Alzheimer-related changes. Acta Neuropathol 1991;82:239-59
23. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state." A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189-98
24. Petersen RC, Smith GE, Waring SC, et al. Mild cognitive impairment: Clinical characterization and outcome. Arch Neurol 1999;56:303-8
25. Mott R, Ervin JE, Trojanowski JQ, et al. Molecular and neuropathological characterization of the frontotemporal dementias. J Neuropathol Exp Neurol 2005:64;420-28
26. Bock D. Multivariate Statistical Methods in Behavioral Research
. New York: McGraw-Hill, 1973
27. Kumar MS, Owens GK. Combinatorial control of smooth muscle-specific gene expression. Arterioscler Thromb Vasc Biol 2003;23:737-47
28. Owens GK, Kumar MS, Wamhoff BR. Molecular regulation of smooth muscle cell differentiation in development and disease. Physiol Rev 2004;84:767-801
29. Carmeliet P, Jain RK. Angiogenesis in cancer and other diseases. Nature 2000;407:249-57
30. Regan CP, Adam PJ, Madsen CS, et al. Molecular mechanisms of decreased smooth muscle differentiation marker expression after vascular injury. J Clin Inv 2000;106:1139-47
31. de la Torre JC. Is Alzheimer's disease preceded by neurodegeneration or cerebral hypoperfusion? Ann Neurol 2005;57:783-84
32. Ikonomovic MD, Mufson EJ, Wuu J, et al. Cholinergic plasticity in hippocampus of individuals with mild cognitive impairment: Correlation with Alzheimer's neuropathology. J Alzheimer Dis 2003;5:39-48
33. Schildmeyer LA, Braun R, Taffet G, et al. Impaired vascular contractility and blood pressure homeostasis in smooth muscle α-actin null mouse. FASEB J 2000;14:2213-20
34. Dubelaar EJ, Mufson EJ, ter Meulen WG, et al. Increased metabolic activity in nucleus basalis of Meynert neurons in elderly individuals with mild cognitive impairment as indicated by the size of the Golgi apparatus. J Neuropathol Exp Neurol 2006;65:257-66
35. Iacono D, O'Brien R, Resnick SM, et al. Neuronal hypertrophy in asymptomatic Alzheimer disease. J Neuropathol Exp Neurol 2008;67:578-89
36. Dickerson BC, Salat DH, Greve DN, et al. Increased hippocampal activation in mild cognitive impairment compared to normal aging and AD. Neurol 2005;65:404-11
37. Esiri MM, Hyman BT, Beyreuther K, et al. Ageing and dementia. In: Graham DI, Lantos PL, eds. Greenfield's Neuropathology. London, UK: Arnold, 1997:153-88