- Email: [email protected]
Subpopulation of dogs with severe brain parenchymal β amyloidosis distinguished with cluster analysis
BRAIN RESEARCH
ELSEVIER
Brain Research 728 (1996) 20-26
Research report
Subpopulation of dogs with severe brain parenchymal 13 amyloidosis distinguished with cluster analysis Jerzy W~giel *, Henryk M. Wisniewski, Jerzy Dziewigtkowski 1, Michat Tarnawski, Anna Dziewigtkowska 2, Janusz Mory~1, Zenon Sottysiak 3, Kwang Soo Kim Department of Pathological Neurobiology, New York State Institute for Basic Research in Developmental Disabilities, 1050 Forest Hill Road, Staten Island, NY 10314, USA
Accepted 19 March 1996
Abstract
A study of the brains of 30 dogs, mongrels from 6.5 to 26.5 years of age, revealed that all dogs older than 13 years of age develop amyloid-13-positive plaques. Cluster analysis based on the age of the dogs and the numerical density of amyloid-positive plaques stained with monoclonal antibody 4G8 (17-24aa) revealed that the population of old dogs consists of two subpopulations: one with a very low (0.8/mm 2 on average) and other with a high (19.2/mm 2 on average) numerical density of plaques. These two groups (19.5 and 19.1 years of age, respectively) appear to emerge from the younger group (12.2 years of age on average), with moderate (2.2/mm 2 on average) numerical density of 4G8-positive plaques. These data may indicate that only a portion of the mongrel population (43%) is susceptible to amyloidosis 13 or that only this severely affected subpopulation was exposed to a factor or factors inducing this pathology and developed severe cortical amyloidosis that correlates with age. Dog plaques are only of the diffuse type, with nonfibrillar, thioflavin S-, and Congo red-negative amyloid in all groups distinguished by cluster analysis. Only from 10% of 4G8-positive plaques in the mildly affected group to 29% in the severely and 37% in the moderately affected group are Bielschowsky positive. In the younger, moderately affected group, 6El0 (1-17aa)-positive plaques prevail. In the two old groups with severe and weak changes, almost all 4G8-positive plaques are also 6Et0-positive. Carboxy-terminal region immunocytochemistry reveals that BC42-positive plaques are numerous, whereas BC40-positive plaques are few or absent. The differences in the silver-positivity of plaques and their immunoreactivity in both the aminoand carboxy-terminal regions may reflect differences in amyloid-13 deposition and resolution. Dog parenchymal amyloidosis 13 appears to be a model for the study of diffuse plaques Keywords: Dog; Aging; Diffuse plaque; Morphometry;Cluster analysis
1. Introduction Extracellular deposition of amyloid-13 (A[3) in plaques and vessels and neurofibrillary pathology in neurons are the hallmark of Alzheimer disease (AD). Amyloid-13 pathology in brains of old dogs appears to be a natural analogue of Alzheimer-type amyloidosis 13. Amyloidosis 13 has been studied in more than 300 dogs [2-4,6,16,18,2022,24,26,27,29,31,33]. The majority of studies show a high
* Corresponding author. Fax: + 1 (718) 698-3803. 1 Visiting scientist from the Department of Anatomy,Medical Academy of Gdansk, Poland. 2 Visiting scientist from the Department of Child Psychiatry, Medical Academy of Gdansk, Poland. 3 Visiting scientist from Veterinary Faculty of Higher Agricultural School, Wroclaw, Poland. 0006-8993/96/$15.00 Published by Elsevier Science B.V. PII S0006-8993(96)00373-3
frequency of brain amyloidosis, reaching 100% when very old dogs are examined [29,31,33]. They indicate that only three types of deposits are commonly seen in brains of aged dogs: diffuse nonfibfillar plaques in the cerebral cortex and subcortical gray matter, diffuse nonfibfillar amyloid deposits in the molecular layer of the hippocampal formation, and amyloid angiopathy with fibfillar amyloid deposits. Thioflavin-positive, classical-like plaques were noted in only a very few dogs and in very small numbers [ 18,20,21 ]. Microglia-associated fibfillized classical and primitive plaque formation [28,32,35] and pefivascular cell/pefivascular microglia-associated fibrillar amyloid deposition in the wall of human capillaries [34,35] appear to be extremely rare or absent in dog brain [29]. The analogy between human and dog amyloidosis 13 appears to be close in only two types of lesions: amyloid angiopathy of arteries and veins, and diffuse plaques. The smooth muscle cell-associated course of fibfillar amyloid
J. Wcgiel et al. / Brain Research 728 (1996) 20-26
deposition, cell degeneration, and necrosis and the degradation of both smooth muscle cells and fibrillar amyloid are morphologically identical in human and dog leptomeningeal and cortical azteries and veins [5,30,36]. The resultant remodeling of the vascular wall makes human and dog vessels vulnerable to disruption and hemorrhages [7,8,26]. Some reports show that diffuse plaques are the only type of plaques present in old dog brain [3,29]. Diffuse plaques seen in dog brain share morphological characteristics with the diffuse A[3 deposits present in tile brains of young subjects with Down syndrome (DS) [37], in the brains of adults after trauma [9], and in the molecular layer of the human cerebellum in AD and DS [12,14]. They also develop in the rat brain after intraventricular infusion with okadaic acid [1]. Significant interindividual differences in the amount of parenchymal amyloid deposits and the lack of correlation of these changes with age [20,29] can reduce the usefulness of the dog as a model of human pathology. The variable numerical density of plaques in the brains of aged dogs may indicate that the dog population is heterogeneous in terms of age-related parenchymal amyloidosis and consists of subpopulations with different rates of plaque formarion. We hypothesize that the correlation between age and progression of amyloid deposition is a feature of only a subpopulation of dogs with a higher rate of diffuse plaque formation. The aim of this study based on the morphometry of plaques sl:ained with several histological and immunocytochemical methods and cluster analysis is to test this hypothesis.
2. Materials and methods
The study was performed on 30 dogs, mongrels from 6.5 to 26.5 years of age,,. Older dogs showed several age-related features: fecal and urinary incontinence, decrease of motor activity, visual and hearing loss. The incontinence of stool or uriae, or double incontinence was noticed in 13 of 17 dogs older than 15.5 years of age; for four dogs in this age category, data were not available. All dogs older than 15 years of age had been less active, and two were nonambulatory. Progressing visual loss was one of the reasons for the spatial disorientation and reduced motor activity of old dogs. Unilateral or bilateral cataract was found in 96% of dogs 11 years of age or older. Fifteen of 18 dogs older than 15 years of age showed partial or total loss of hearing. For three animals, data about hearing were not available. The brains were removed rapidly from the skull and fixed. One brain hemisphere fixed in 10% formalin for four to six weeks was cut on the coronal plane in 5-mmthick slabs that were embedded in paraffin and cut into 6-p~m-thick serial section,;. They were stained with the Bielschowsky method to iidentify silver-positive plaques.
21
Thioflavin S staining and examination in fuorescence was used to identify amyloid deposits with B-pleated sheet conformation. Sections stained with Congo red were examined in polarized light to reveal dichroism of fibrillar amyloid-[3 deposits. Monoclonal antibodies (mAbs) 4G8 and 6El0 were used to characterize the immunoreactivity of the amino-terminal region, and antisera BC40 and BC42 to characterize the carboxy-terminal region of A[3 in plaques, mAb 4G8 was raised against a synthetic peptide corresponding to amino acids 17-24 of A[3 (IgG2b; 1:4000) [26]. Mab 6El0 recognizes amino acids 1-17 of the B-protein (1:4000). Anrisera BC40 and BC42 were raised against synthetic peptides A[333-40 and A[337-42; 1:1000 [40]. Paraffin sections were treated with concentrated formic acid before immunostaining to enhance immunoreactivity of amyloid-[3 deposits. In 23 brains with 4G8-positive plaques, the numerical density ( n / m m 2) of 4G8-, 6El0-, and Bielschowsky-positive plaques was examined in the frontal lobe in the gyms frontalis, proreus, rectus, and genualis; in the temporal lobe in the gyms lateralis, suprasylvius posterior, and ectosylvius posterior; in the parietal lobe in the gyms sigmoideus, suprasylvius and ectosylvius anterior, sylvii anterior and posterior, coronalis, parietalis medialis and orbitalis; in the occipital lobe in the gyms lateralis, ectolateralis, splenialis and suprasplenialis, and lingualis; in the limbic cortex in the gyms cinguli, piriformis, parasplenialis, subcallosus, and parahippocampalis. The numerical density of plaques was established at magnification 165 X using a projective microscope (Pictoval, C. Zeiss). About 300 test areas were examined in one staining in one hemisphere of each dog. The randomness of sample was examined with runs tests based on age and the numerical density of 4G8-positive plaques [41]. The distribution of data was assessed by the Kolmogorov-Smimov test. The association between parameters was examined with the correlation coefficient (Pearson's r). Cluster analysis based on two variables (the age and the numerical density of 4G8 plaques) was used to identify groups with different ages and amounts of plaques. The ANOVA test was applied for examination of the differences between groups distinguished by cluster analysis. The differences in the age of dogs and in the numerical density of plaques stained with mAbs 4G8 and 6El0 and the Bielschowsky method among animals in three groups were examined with the Mann-Whitney U-test (CSS Statistica).
3. Results
3.1. Properties of plaques in dog cerebral cortex All plaques detected in the cerebral cortex of the examined dogs are diffuse, however, their morphology and number depend on the method of staining. The plaques
J. Wcgielet al./ Brain Research 728 (1996) 20-26
22
Fig. 1. Plaques observed in the cerebral cortex of aged dogs are of the diffuse type. They are immunopositivewhen stained with mAbs 6El0 (1-17aa) (a) and 4G8 (17-24aa) (b)to the amino-terminal region of amyloid [3-peptide and with antiserum BC40 (c) and BC42 (d) to the carboxy-terminal region. Gyrus cinguli of the 19-year-01ddog from the severely affected group.
revealed with mAbs 4G8 and 6El0, antiserum BC42 (Fig. 1), and the Bielschowsky silver method are round or oval, with a fuzzy borderline: The majority of 4G8- and 6E10positive plaques are evenly stained and rather pale. In about 30% of these plaques, the staining is much more intensive in the center. All plaques revealed with antiserum BC42 are pale. In contrast, BC40-positive plaques are stained intensely, and they have a sharp border (Fig. 1). In the plaque perimeter, the bodies of neurons, astrocytes, microglia, and oligodendroglia without morphologically identifiable changes are seen. In Bielschowsky-positive plaques, the network of silver-impregnated fibers is denser and darker than in the surrounding neuropil; however, there is no deformation of the neuronal processes. The average diameter of 4G8- and 6E10-positive plaques is similar, 68 p,m and 641xm, respectively, whereas Bielschowsky-positive plaques are significantly smaller, 58 p,m in diameter ( P < 0.01). All cortical plaques are negative when stained with thioflavin S and examined in
fluorescence and stained with Congo red and examined in polarized light. 4G8-positive plaques were found in the cerebral cortex in 23 of 30 dogs 6.5 to 26.5 years of age (77%). The youngest plaque-positive dog was 9 years of age; all dogs older than 13 years of age had plaques. Bielschowsky-positive plaques were found in 21 dogs (70%). The numerical density of plaques varies interindividually in so broad a range that there is no correlation between the numerical density of 4G8-, 6El0- and Bielschowsky-positive plaques and age in the plaque-positive group of dogs.
3.2. Groups of dogs distinguished with cluster analysis Cluster analysis based on two variables (age and the numerical density of 4G8-positive plaques in the cerebral cortex) distinguished three groups of dogs: with weak (10 dogs), moderate (6 dogs), and severe (7 dogs) amyloidosis (Table 1, Fig. 2). The group of dogs with weak amyloido-
Table 1 Characteristics of three groups of dogs distinguished by cluster analysis based on the age of animals and the numerical density of 4G8-positive plaques in the cerebral cortex Parameter Number of dogs Age (years) Numerical density of 4G8-positive plaques (n/mm 2) Numerical density of 6E10-positive plaques (n/mm 2) Numerical density of Bielschowsky-positive plaques (n/mm 2) S.E.M. in oarentheses.
Group, severity of cortical amyloidosis
P<
A, weak 10 19.5 (1.1) 0.8 (0.2)
B, moderate
C, severe
A/B
A/C
6
12.2 (0.8) 2.7 (1.2)
0.8 (0.3) 0.08 (0.02)
B/C
7
-
-
-
19.1 (0.5) 19.2 (3.1)
0.01 -
0.001
0.01 0.001
9.2 (6.1)
18.4 (4.2)
-
0.001
-
1.0 (0.6)
5.6 (1.2)
-
0.001
0.01
J. W¢giel et aL / Brain Research 728 (1996) 20-26 40. 30
20.
o
10
14
i
i
18
22
26
30
Age Fig. 2. Graphical presentation of the three groups of dogs distinguished by cluster analysis based on the age of animals and the numerical density of 4G8-positive plaques. Group A, with moderate severity of changes as estimated by the numerical density of plaques, appears to be a heterogenous subpopulation. This group consists of three dogs with a low numerical density of plaques (circ!tes below dotted line) comparable with those seen in group B, and three dogs with a higher numerical density of plaques than in group B but a lower numerical density than in group C (circles above dotted line). During aging in a portion of the dog population (group B, weak changes), the numerical density of plaques is low, 0.8/mm 2, and appears to remain almost constant during the more than 10-year-long period of observation. A subpopulation of aged dogs develops severe parenchymal amyloidosis-[3 with about 19 plaques per mm 2, on average (group C). The numerical density in combined groups A and C correlates with age.
sis has only 0.8 plaques per mm 2 on average and is the largest (43% of the plaque-positive population) and, surprisingly, the oldest group: 19.5 years of age, on average. The group with a moderate number of plaques (2.7/mm 2 on average) comprises 26% of the population. These dogs, which are 12.2 years of age on average, are significantly younger than the dogs with the lowest numerical density of plaques ( P < 0.01). The third group distinguished by cluster analysis consists of seven dogs (30% of the plaquepositive population) that are severely affected by parenchymal amyloidosis: the numerical density of cortical plaques is 19.2/mm 2 on average. The average age of dogs in this group (19.1 years) is non-significantly lower than that of the group with the lowest number of plaques. They are significantly older than moderately affected animals. The numerical density of 4G8-positive plaques in the combination of two groups (the youngest, with moderate amyloidosis, and the old with severe amyloidosis) correlates with age ( r = 0.7; P < 0.001). 4G8-positive plaques occupy significantly more cortex in the severely affected group (3.7%) than in the groups with weak (0.4%; P < 0.001) and moderate (0.7%; P < 0.01) parenchymal amyloidosis. The numerical density of 4G8-positive plaques in the most affected group of dogs was significantly ( P < 0.001) higher in the frontal cortex ( 3 8 / m m 2, on average) than in the temporal, limbic (19/rnm2), parietal (18/mm2), and occipital cortex (15/mm2). In two other groups with moderate and mild parenchymal amyloidosis, the plaques were distributed so unevenly in the examined cortical gyri that interregional differences were undetectable.
23
The numerical density of 6E10-positive plaques correlates with the numerical density of 4G8-positive plaques ( r = 0.71, P < 0.001). In the youngest group distinguished by cluster analysis, the number of 6E10-positive plaques is three times higher than the number of 4G8-positive plaques. In the two oldest groups (with severe and weak changes) 96% and 100% of 4G8-positive plaques, respectively, are 6E10-positive. Bielschowsky-positive plaques were found in almost all cases with 4G8-positive plaques (91.4%). The numerical density of Bielschowsky-positive plaques correlates with the numerical density of 4G8-positive plaques ( r = 0.82, P < 0.001); however, in groups with weak, moderate, and severe changes, they constitute only 10%, 37%, and 29%, respectively, of 4G8-positive plaques. They occupy significantly less cortex (0.46% + 0.18%) than do 4G8-positive plaques (1.3% _ 0.3%, Wilcoxon test P < 0.001).
4. Discussion
4.1. Subpopulations of dogs with severe and mild parenchymal amyloidosis The extensive study of the plaques in the cerebral cortex of 30 dogs, including 26 cortical regions, revealed plaques in the brains of all dogs older than 13 years of age. The incipient stage of cerebral brain amyloidosis comprises the age between the 9th and 13th years. The development of parenchymal brain amyloidosis in dogs appears to be an unavoidable, species-specific feature associated with age. However, the numerical density of 4G8-, 6El0-, or Bielschowsky-positive plaques in the cerebral cortex of dogs does not correlate with age. A similar lack of correlation was also observed in other dog populations [20]. Interindividual differences in the number of plaques in the brains of dogs of the same age suggest that the progress of parenchymal cortical amyloidosis might be upregulated only in a portion of the population of old dogs. Cluster analysis based on the age and the numerical density of 4G8-positive plaques confirmed this assumption. Three groups of dogs were distinguished: a severely affected group with numerical density of plaques of 19.2/mm 2 on average, a moderately affected group with 2.7 plaques per mm 2, and a less affected group with 0.8 plaques per mm 2. Almost the same average ages in the severely affected group (t9.1 years) and in the less affected group (19.5 years) suggests that these two groups are recruited from the youngest group: 12.2 years of age. The numerical density of plaques in the brains of three dogs in this group is in the range of that of the old and less affected group of animals, whereas the numerical density of plaques in the brain of the three other dogs is greater than the numerical density in old and less affected dogs but less than the numerical density seen in the oldest and most affected group (Fig. 2). The correlation between age and numerical
24
J. W¢gielet al./ Brain Research 728 (1996) 20-26
density of plaques in the combined younger group with moderate number of plaques and in the older group with severe amyloidosis indicates that the severe form of amyloidosis is an age-dependent process in a dog subpopulation. The results of cluster analysis suggest that the dog population consists of only two subpopulations, one that develops severe (43% of animals) and a second that develops weak cortical amyloidosis [3 (57%). The cause of the acceleration of amyloid deposition in the first years of the second decade of the life of some dogs is not clear. It might be the effect of higher breedspecific susceptibility to parenchymal amyloidosis [3. However, study of diffuse plaques in pure-bred beagles indicates that even in one breed there are remarkable differences in cerebral amyloidosis [3. Inter-familial differences suggest that a genetic factor might be involved in determining dogs' plaque susceptibility [20]. The impact of exogenous factors such as environmental conditions [24] or infections cannot be excluded. 4.2. Diffuse plaques properties It is believed that human diffuse plaque amyloid is nontoxic because they are free of neurofibrillary changes, and ultrastructurally, neuronal and synapfic pathology is almost undetectable [15,25,37-39]. The impact of dog diffuse plaques on neurons and the function of affected brain structures is not known. In light microscopy studies, dog diffuse plaques recall diffuse plaques in the cerebral cortex of young subjects with Down syndrome. The lack of dementia in young patients with DS suggests that even very numerous diffuse plaques ( 6 8 / m m 2) do not cause remarkable pathology and functional changes [37]. Dog diffuse plaques are free of clusters of microglial cells, which are the source of fibrillizing amyloid in classical and primitive plaques in human brain [28,32]. Diffuse plaques are nonfibrillar or contain only a few amyloid fibrils, and they are thioflavin S- and Congo red-negative [3,6,12,37]. The diffuse plaque amyloid might be derived either from A[31-42 or amyloid-[3 precursor protein A[3PP 695 [23] or from both sequences, with accumulation of soluble nonamyloidogenic peptides 1-16 and 17-42aa organized into putatively nontoxic aggregates. Diffuse plaques of the examined dogs are immunopositive when stained with mAbs to the N-terminal region 6El0 (1-17) and 4G8 (17-24). Similar onset and a strong correlation between the numerical density of 4G8- and 6E10-positive plaques ( r = 0.71, P < 0.001), and between the numerical density of 4G8- and Bielschowsky-positive plaques ( r = 0.82, P < 0.001) suggests common factors in their formation. Human diffuse plaques are positive when stained for the carboxy region 42, similarly to classical and primitive plaques and vascular amyloid, but they are negative when immunostained with antibodies to terminus 40 [10]. Our study indicates that a subset of diffuse dog plaques con-
tains both peptides with terminal regions 42 and 40. The presence of A[340-positive plaques in only some dogs suggests that processing of amyloid [3 might differ individually among dogs. Great variations among cases in the number of A[340-positive diffuse plaques and the lower number of A[340- than A[342-positive plaques were also found in human brain in AD and DS [11,40]. Labeling with antiserum BC40 gave only a fraction of BC42-positive plaques which may indicate local differences in [3peptide processing. Iwatsubo et al. [10] suggest that in the course of AD some A[342(43)-positive plaques acquire A[340-positivity, presumably by processing. The differences in the proportion of A[342/40-positive plaques in different brain structures may indicate that processing may also be brain-region-specific [13]. Amyloid [3-peptide 1-42(43) is assumed to be the factor that initiates fibrillar plaque formation [10]. The presence of amyloid [3-peptide with carboxy-terminal region 42 in dog vessels and fibrillar amyloid accumulation in the wall of dog arteries and veins confirm this suggestion; however, the presence of carboxy-terminal region 42 in diffuse plaques and the absence of fibrillar material during more than 10 years of parenchymal amyloidosis in dog cerebral cortex indicate that fibrillization does not depend only on the presence of peptide with C-terminal region 42. 4.3. Parenchymal and vascular amyloidosis in dog brain Morphometric study of amyloid angiopathy [30] and diffuse plaques (present study) in the same population of dogs has revealed that vascular and parenchymal amyloidosis appears at the same age. In contrast to parenchymal amyloidosis, which correlates with age in only a subpopulation of more affected dogs, amyloid angiopathy correlates with age in the whole amyloid-positive population. Amyloid deposition in the vascular wall in human [36] and dog brain [30] is associated with the smooth muscle cells of arteries and veins. Diffuse plaque formation appears to be associated with neurons [14,17,19,29,37]. The abovementioned differences in correlation between vascular and parenchymal amyloidosis and age might be related to the type of cell engaged in amyloid deposition in these two brain compartments. The similar onset of vascular and parenchymal amyloidoses suggests that the factor or factors initiating amyloid deposition by different types of cells is common. The different courses of vascular and parenchymal amyloidoses may be related more to a cellspecific response. 4.4. Dog as the model for the study of diffuse plaque formation Studies of the pathogenesis of amyloid suggest that Alzheimer-type pathology consists of several cell-specific pathological processes that coexist or develop for years
J. Wcgiel et aL/ Brain Research 728 (1996) 20-26
independently of each others. The key factor to classification of amyloid deposits appears to be the type o f cell associated causatively with amyloid deposition: neurons in diffuse plaque [14,17,19,29,37], microglial cells in classical and primitive plaque [28,39], perivascular cells and perivascular microglial cells in the wall of capillaries [34], and smooth muscle cells in the wall of parenchymal and meningeal arteries and veiins [5,30,36]. The morphology and the tinctorial and inmmnocytochemical properties of amyloid and the reaction o f the neuropil to amyloid appear to be secondary to the cell-specific factor initiating deposit formation. Marked differences in the spectrum of changes in dog brain in comparison to that in the brain of subjects with A D show that aged dog-brain amyloidosis [3 is a partial model of human Alzheimer-type pathology. In dog brain, there is no neurofibrillary pathology, no amyloid deposition in the wall of capillaries by perivascular cells of microglial lineage or with microglial cells in classical and primitive plaques, and no fibrillized plaques with degenerated neurons. However, the presence in dog brain o f two processes that are almost identical morphologically with human pathological proces,~es: fibrillar amyloid deposition in the wall of arteries and veins associated with smooth muscle cells and resulting in degeneration, necrosis, and degradation of cells of the vascular wall, and diffuse plaques with putatively nontoxic amyloid accumulation associated probably with neurons indicates that pathology of aged dogs can be a unique animal model for the study of these two forms of amyloidotic changes. In dog brain, the impact of diffuse plaques on neuropil and on the function of brain structures can be examined in isolation from other Alzheimer type changes. This study supports the hypothesis that diffuse plaque formation is a specific form of amyloidosis, probably associated with neurons, that is separate from the classical and primitive plaque formation associated with rnicroglial cells.
Acknowledgements The authors thank Dr. Haruyasu Yamaguchi (Gunma University, Gunma, Japan) for antisera BC40 and BC42, and Ms. Maureen Marlow for copy editing. Supported in part by funds from the New York State Office of Mental Retardation and Developmental Disabilities and a grant from the National Institutes of Helath, National Institute of Aging No. PO1-AGO-4220.
References [1] Arendt, T., Holzer, M., Fruth, R., Bruckner, M.K. and Gartner, U., Paired helical filament-like phosphorylation of tau, deposition of [3/A4-amyloid and memory impairment in rat induced by chronic inhibition of phosphatase 1 asad 2A, Neuroscience, 69 (1995) 691698.
25
[2] Ball, M.J., MacGregor, J., Fyfe, I.M., Rapoport, S.I. and London, E.D., Paucity of morphological changes in the brains of ageing beagle dogs: further evidence that Alzheimer lesions are unique for primate central nervous system, Neurobiol. Aging, 4 (1983) 127131. [3] Cummings, B., Su, J., Cotman, C., White, R. and Russell, M., [3-amyloid accumulation in aged canine brain: A model of early plaque formation in Alzheimer's disease, Neurobiol. Aging, 14 (1993) 547-560. [4] Dayan, A.D., Comparative neuropathology of aging, Brain, 94 (1971) 31-42. [5] Frackowiak, J., Zoltowska, A. and Wisniewski, H.M., Non-fibrillar 13-amyloidprotein is associated with smooth muscle cells of vessel walls in Alzheimer's disease, J. Neuropathol. Exp. Neurol., 53 (1994) 637-645. [6] Giaccone, G., Verga, L., Finazzi, M., Pollo, B., Tagliavini, F., Frangione, B. and Bugiani, O., Cerebral preamyloid deposits and congophilic angiopathy in aged dogs, Neurosci. Lett., 114 (1990) 178-183. [7] Gilbert, J.J. and Vinters, H.V., Cerebral amyloid angiopathy: Incidence and complications in the aging brain. I. Cerebral hemorrhage, Stroke, 14 (1983) 915-923. [8] Gilles, C., Bruchner, J.M., Khoubesserian, P. and Vanderhaeghen, J.J., Cerebral amyloid angiopathy as a cause of multiple intracerebral hemorrhages, Neurology, 34 (1984) 730-735. [9] Graham, D.I., Gentlemen, S.M., Lynch, A. and Roberts, G.W., Distribution of 13-amyloidprotein in the brain following severe head injury, Neuropathol. Appl. Neurobiol., 21 (1995) 27-34. [10] Iwatsubo, T., Odaka, A., Suzuki, N., Mizusawa, H., Nukina, N. and Ihara, Y., Visualization of AB42(43) and A1340 in senile plaques with end-specific A[3 monoclonals: evidence that an initially deposited species in AI342(43), Neuron, 13 (1994) 43-53. [11] Iwatsubo, T., Mann, D.M.A., Odaka, A., Suzuki, N. and lhara, Y., Amyloid [3 protein (AI3) deposition: A1342(43) precedes A1340 in Down's syndrome, Ann. Neurol., 37 (1995) 294-299. [12] Joachim, C.L., Morris, J.H. and Selkoe, D.J., Diffuse senile plaques occur commonly in the cerebellum in Alzheimer's disease, Am. Z Pathol., 135 (1989) 309-319. [13] Kida, E., Wisniewski, K.E. and Wisniewski, H.M., Early amyloid-[3 deposits show different immunoreactivity to the amino- and carboxy-terminal regions of [3-peptide in Alzheimer's disease and Down syndrome brain, Neurosci. Lett., 193 (1995) 1-4. [14] Li, Y.T., Woodruff-Pak, D.S. and Trojanowski, J.Q., Amyloid plaques in cerebellar cortex and the integrity of Purkinje cell dendrites, Neurobiol. Aging, 15 (1994) 1-9. [15] Masliah, E., Terry, R.D., Mallory, M., Alford, M., Hansen, L.A., Diffuse plaques do not attenuate synapse loss in Alzheimer's disease, Am. J. Pathol., 137 (1990) 1293-1297. [16] Osetowska, E., Morphologic changes in the brains of old dogs, Neuropathol. Pol., 4 (1966) 97-110. [17] Pappolla, M.A., Omar, R.A. and Vinters, H.V., Image analysis microscopy shows that neurons participate in the genesis of subset of early primitive (diffuse) senile plaques, Am. J. Pathol., 359 (1991) 599-607. [18] Pauli, B. and Luginbuhl, H., Fluorescenzmikroskopische Untersuchungen der cerebralen Amyloidose bei alten Hunden und senilen Menschen, Acta Neuropathol. (Berl.), 19 (1971) 121-128. [19] Probst, A., Langui, D., Ipsen, S., Robakis, N. and Ulrich, J., Deposition of beta/A4 protein along neuronal plasma membranes in diffuse senile plaques, Acta Neuropathol. (Berl.), 83 (1991) 21-29. [20] Russell, M., White, R., Patel, E., Markesbery, W., Watson, C. and Geddes, J., Familial influence on plaque formation in the beagle brain, NeuroReport, 3 (1992) 1093-1096. [21] Selkoe, D.J., Bell, D.S., Podlisny, M.B., Price, D.L., Cork, L.C., Conservation of brain amyloid proteins in aged mammals and humans with Alzheimer's disease, Science, 235 (1987) 873-877. [22] Shimada, A., Kuwamura, M., Umemura, T., Takada, K., Ohama, E.
26
J. W¢giel et a l . / Brain Research 728 (1996) 20-26
and Itakura, C., Modified Bielschowsky and immunohistochemical studies on senile plaques in aged dogs, Neurosci. Lett., 129 (1991) 25-28. [23] Sisodia, S.S., Koo, E.H., Beyreuther, K., Unterbeck, A. and Price, D.L., Evidence that 13-amyloid protein in Alzheimer's disease is not derived by normal processing, Science, 248 (1990) 492-496. [24] Soltysiak, Z., Comparative pathomorphology of senile changes in central nervous system (CNS) of dogs from a country, town and industry environment, Zesz. Nauk. Akad. Roln. Wroclawiu, Weter., 42 (1985) 249-271. [25] Tagliavini, F., Giaccone, G., Frangione, B. and Bugiani, O., Preamyloid deposits in the cerebral cortex of patients with Alzheimer's disease and nondemented individuals, Neurosci. Lett., 93 (1988) 191-196. [26] Uchida, K., Miyauchi, Y., Nakayama, H. and Goto, N., Amyloid angiopathy with cerebral hemorrhage and senile plaque in aged dogs, Jpn. J. Vet. Sci., 52 (1990) 605-611. [27] von Braunmtihl A., Kongophile Angiopathie und 'senile Plaques' bei greisen Hunden, Arch. Psychol. Z. Neurol., 194 (1956) 396-414. [28] Wegiel, J. and Wisniewski, H.M., The complex of microglia cells and amyloid star in three-dimensional reconstruction, Acta Neuropathol. (Berl.), 81 (1990) 116-124. [29] Wegiel, J., Wisniewski, H.M., Dziewiatkowski, J., Tarnawski, M., Dziewiatkowski, A., Soltysiak, Z. and Kim, K.S., Fibrillar and non-fibrillar amyloid in the brain of aged dogs. In K. Iqbal, J.A. Mortimer, A. Winblad and H.M. Wisniewski (Eds.), Research Advances in Alzheimer's Disease and Related Disorders, John Wiley and Sons, Chichester, 1995, pp. 703-707. [30] Wegiel, J., Wisniewski, H.M., Dziewiatkowski, J., Tarnawski, M., Nowakowski, J., Dziewiatkowska, A. and Soltysiak, Z., The origin of amyloid in cerebral vessels of aged dogs, Brain Res., 705 (1995) 225-234. [31] Wisniewski, H.M., Johnson, A.B., Raine, C.S., Kay, W.J. and Terry, R.D., Senile plaques and cerebral amyloidosis in aged dogs. A histochemical and ultrastructural study, Lab. Invest., 23 (1970) 287-296.
[32] Wisniewski, H.M., Wegiel, J., Wang, K.C., Kujawa, M. and Lach, B., Ultrastructural studies of the cells forming amyloid fibers in classical plaques, Can. J. Neurol. Sci., 16 (1989) 535-542. [33] Wisniewski, H.M., Wegiel, J., Morys, J., Bancher, C., Soltysiak, Z. and Kim, K.S., Aged dogs: an animal model to study beta-protein amyloidogenesis. In: K. Maurer, P. Riederer and H. Beckmann (Eds.), Alzheimer's Disease. Epidemiology, Neuropathology, Neurochemistry, and Clinics, Springer-Verlag, Vienna, 1990, pp. 151-168. [34] Wisniewski, H.M., Wegiel, J., Wang, K.C. and Lach, B., Ultrastructural studies of the cells forming amyloid in the cortical vessel wall in Alzheimer's disease, Acta Neuropathol. (Berl.), 84 (1992) 117127. [35] Wisniewski, H.M. and Wegiel, J., Migration of perivascular cells into the neuropil and their involvement in 13-amyloid plaque formation, Acta Neuropathol. (Berl.), 85 (1993) 586-595. [36] Wisniewski, HM. and Wegiel, J., 13-amyloidformation by myocytes of leptomeningeal vessels, Acta Neuropathol. (Berl.), 87 (1994) 233-241. [37] Wisniewski, H.M., Wegiel, J. and Popovitch, E., Age associated changes in 13-amyloid plaque fibrillization in Down syndrome, Dev. Brain. Dysfunct., 7 (1994) 330-339. [38] Yamaguchi, H., Hirai, S., Morimatsu, M., Shoji, M. and Ihara, Y.A., Variety of cerebral amyloid deposits in the brains of the Alzheimertype dementia demonstrated by 13 protein immunostaining, Acta Neuropathol. (BerL), 76 (1988) 541-549. [39] Yamaguchi, H., Nakazato, Y., Hirai, S., Shoji, M. and Harigaya, Y., Electron micrograph of diffuse plaques. Initial stage of senile plaque formation in the Alzheimer brain, Am. J. Pathol., 135 (1989) 593-597. [40] Yamaguchi, H., Sugihara, S., Ishiguro, K., Takashima, A. and Hirai, S., Immunohistochemical analysis of COO-termini of amyloid beta protein (A[3) using end-specific antisera for AI340 and AI342 in Alzheimer's disease and normal aging, Amyloid: Int. J. Exp. Clin. Invest., 2 (1995) 7-16. [41] Zar, J.H., Biostatistical Analysis, Prentice Hall, Englewood Cliffs, 1984.
ELSEVIER
Brain Research 728 (1996) 20-26
Research report
Subpopulation of dogs with severe brain parenchymal 13 amyloidosis distinguished with cluster analysis Jerzy W~giel *, Henryk M. Wisniewski, Jerzy Dziewigtkowski 1, Michat Tarnawski, Anna Dziewigtkowska 2, Janusz Mory~1, Zenon Sottysiak 3, Kwang Soo Kim Department of Pathological Neurobiology, New York State Institute for Basic Research in Developmental Disabilities, 1050 Forest Hill Road, Staten Island, NY 10314, USA
Accepted 19 March 1996
Abstract
A study of the brains of 30 dogs, mongrels from 6.5 to 26.5 years of age, revealed that all dogs older than 13 years of age develop amyloid-13-positive plaques. Cluster analysis based on the age of the dogs and the numerical density of amyloid-positive plaques stained with monoclonal antibody 4G8 (17-24aa) revealed that the population of old dogs consists of two subpopulations: one with a very low (0.8/mm 2 on average) and other with a high (19.2/mm 2 on average) numerical density of plaques. These two groups (19.5 and 19.1 years of age, respectively) appear to emerge from the younger group (12.2 years of age on average), with moderate (2.2/mm 2 on average) numerical density of 4G8-positive plaques. These data may indicate that only a portion of the mongrel population (43%) is susceptible to amyloidosis 13 or that only this severely affected subpopulation was exposed to a factor or factors inducing this pathology and developed severe cortical amyloidosis that correlates with age. Dog plaques are only of the diffuse type, with nonfibrillar, thioflavin S-, and Congo red-negative amyloid in all groups distinguished by cluster analysis. Only from 10% of 4G8-positive plaques in the mildly affected group to 29% in the severely and 37% in the moderately affected group are Bielschowsky positive. In the younger, moderately affected group, 6El0 (1-17aa)-positive plaques prevail. In the two old groups with severe and weak changes, almost all 4G8-positive plaques are also 6Et0-positive. Carboxy-terminal region immunocytochemistry reveals that BC42-positive plaques are numerous, whereas BC40-positive plaques are few or absent. The differences in the silver-positivity of plaques and their immunoreactivity in both the aminoand carboxy-terminal regions may reflect differences in amyloid-13 deposition and resolution. Dog parenchymal amyloidosis 13 appears to be a model for the study of diffuse plaques Keywords: Dog; Aging; Diffuse plaque; Morphometry;Cluster analysis
1. Introduction Extracellular deposition of amyloid-13 (A[3) in plaques and vessels and neurofibrillary pathology in neurons are the hallmark of Alzheimer disease (AD). Amyloid-13 pathology in brains of old dogs appears to be a natural analogue of Alzheimer-type amyloidosis 13. Amyloidosis 13 has been studied in more than 300 dogs [2-4,6,16,18,2022,24,26,27,29,31,33]. The majority of studies show a high
* Corresponding author. Fax: + 1 (718) 698-3803. 1 Visiting scientist from the Department of Anatomy,Medical Academy of Gdansk, Poland. 2 Visiting scientist from the Department of Child Psychiatry, Medical Academy of Gdansk, Poland. 3 Visiting scientist from Veterinary Faculty of Higher Agricultural School, Wroclaw, Poland. 0006-8993/96/$15.00 Published by Elsevier Science B.V. PII S0006-8993(96)00373-3
frequency of brain amyloidosis, reaching 100% when very old dogs are examined [29,31,33]. They indicate that only three types of deposits are commonly seen in brains of aged dogs: diffuse nonfibfillar plaques in the cerebral cortex and subcortical gray matter, diffuse nonfibfillar amyloid deposits in the molecular layer of the hippocampal formation, and amyloid angiopathy with fibfillar amyloid deposits. Thioflavin-positive, classical-like plaques were noted in only a very few dogs and in very small numbers [ 18,20,21 ]. Microglia-associated fibfillized classical and primitive plaque formation [28,32,35] and pefivascular cell/pefivascular microglia-associated fibrillar amyloid deposition in the wall of human capillaries [34,35] appear to be extremely rare or absent in dog brain [29]. The analogy between human and dog amyloidosis 13 appears to be close in only two types of lesions: amyloid angiopathy of arteries and veins, and diffuse plaques. The smooth muscle cell-associated course of fibfillar amyloid
J. Wcgiel et al. / Brain Research 728 (1996) 20-26
deposition, cell degeneration, and necrosis and the degradation of both smooth muscle cells and fibrillar amyloid are morphologically identical in human and dog leptomeningeal and cortical azteries and veins [5,30,36]. The resultant remodeling of the vascular wall makes human and dog vessels vulnerable to disruption and hemorrhages [7,8,26]. Some reports show that diffuse plaques are the only type of plaques present in old dog brain [3,29]. Diffuse plaques seen in dog brain share morphological characteristics with the diffuse A[3 deposits present in tile brains of young subjects with Down syndrome (DS) [37], in the brains of adults after trauma [9], and in the molecular layer of the human cerebellum in AD and DS [12,14]. They also develop in the rat brain after intraventricular infusion with okadaic acid [1]. Significant interindividual differences in the amount of parenchymal amyloid deposits and the lack of correlation of these changes with age [20,29] can reduce the usefulness of the dog as a model of human pathology. The variable numerical density of plaques in the brains of aged dogs may indicate that the dog population is heterogeneous in terms of age-related parenchymal amyloidosis and consists of subpopulations with different rates of plaque formarion. We hypothesize that the correlation between age and progression of amyloid deposition is a feature of only a subpopulation of dogs with a higher rate of diffuse plaque formation. The aim of this study based on the morphometry of plaques sl:ained with several histological and immunocytochemical methods and cluster analysis is to test this hypothesis.
2. Materials and methods
The study was performed on 30 dogs, mongrels from 6.5 to 26.5 years of age,,. Older dogs showed several age-related features: fecal and urinary incontinence, decrease of motor activity, visual and hearing loss. The incontinence of stool or uriae, or double incontinence was noticed in 13 of 17 dogs older than 15.5 years of age; for four dogs in this age category, data were not available. All dogs older than 15 years of age had been less active, and two were nonambulatory. Progressing visual loss was one of the reasons for the spatial disorientation and reduced motor activity of old dogs. Unilateral or bilateral cataract was found in 96% of dogs 11 years of age or older. Fifteen of 18 dogs older than 15 years of age showed partial or total loss of hearing. For three animals, data about hearing were not available. The brains were removed rapidly from the skull and fixed. One brain hemisphere fixed in 10% formalin for four to six weeks was cut on the coronal plane in 5-mmthick slabs that were embedded in paraffin and cut into 6-p~m-thick serial section,;. They were stained with the Bielschowsky method to iidentify silver-positive plaques.
21
Thioflavin S staining and examination in fuorescence was used to identify amyloid deposits with B-pleated sheet conformation. Sections stained with Congo red were examined in polarized light to reveal dichroism of fibrillar amyloid-[3 deposits. Monoclonal antibodies (mAbs) 4G8 and 6El0 were used to characterize the immunoreactivity of the amino-terminal region, and antisera BC40 and BC42 to characterize the carboxy-terminal region of A[3 in plaques, mAb 4G8 was raised against a synthetic peptide corresponding to amino acids 17-24 of A[3 (IgG2b; 1:4000) [26]. Mab 6El0 recognizes amino acids 1-17 of the B-protein (1:4000). Anrisera BC40 and BC42 were raised against synthetic peptides A[333-40 and A[337-42; 1:1000 [40]. Paraffin sections were treated with concentrated formic acid before immunostaining to enhance immunoreactivity of amyloid-[3 deposits. In 23 brains with 4G8-positive plaques, the numerical density ( n / m m 2) of 4G8-, 6El0-, and Bielschowsky-positive plaques was examined in the frontal lobe in the gyms frontalis, proreus, rectus, and genualis; in the temporal lobe in the gyms lateralis, suprasylvius posterior, and ectosylvius posterior; in the parietal lobe in the gyms sigmoideus, suprasylvius and ectosylvius anterior, sylvii anterior and posterior, coronalis, parietalis medialis and orbitalis; in the occipital lobe in the gyms lateralis, ectolateralis, splenialis and suprasplenialis, and lingualis; in the limbic cortex in the gyms cinguli, piriformis, parasplenialis, subcallosus, and parahippocampalis. The numerical density of plaques was established at magnification 165 X using a projective microscope (Pictoval, C. Zeiss). About 300 test areas were examined in one staining in one hemisphere of each dog. The randomness of sample was examined with runs tests based on age and the numerical density of 4G8-positive plaques [41]. The distribution of data was assessed by the Kolmogorov-Smimov test. The association between parameters was examined with the correlation coefficient (Pearson's r). Cluster analysis based on two variables (the age and the numerical density of 4G8 plaques) was used to identify groups with different ages and amounts of plaques. The ANOVA test was applied for examination of the differences between groups distinguished by cluster analysis. The differences in the age of dogs and in the numerical density of plaques stained with mAbs 4G8 and 6El0 and the Bielschowsky method among animals in three groups were examined with the Mann-Whitney U-test (CSS Statistica).
3. Results
3.1. Properties of plaques in dog cerebral cortex All plaques detected in the cerebral cortex of the examined dogs are diffuse, however, their morphology and number depend on the method of staining. The plaques
J. Wcgielet al./ Brain Research 728 (1996) 20-26
22
Fig. 1. Plaques observed in the cerebral cortex of aged dogs are of the diffuse type. They are immunopositivewhen stained with mAbs 6El0 (1-17aa) (a) and 4G8 (17-24aa) (b)to the amino-terminal region of amyloid [3-peptide and with antiserum BC40 (c) and BC42 (d) to the carboxy-terminal region. Gyrus cinguli of the 19-year-01ddog from the severely affected group.
revealed with mAbs 4G8 and 6El0, antiserum BC42 (Fig. 1), and the Bielschowsky silver method are round or oval, with a fuzzy borderline: The majority of 4G8- and 6E10positive plaques are evenly stained and rather pale. In about 30% of these plaques, the staining is much more intensive in the center. All plaques revealed with antiserum BC42 are pale. In contrast, BC40-positive plaques are stained intensely, and they have a sharp border (Fig. 1). In the plaque perimeter, the bodies of neurons, astrocytes, microglia, and oligodendroglia without morphologically identifiable changes are seen. In Bielschowsky-positive plaques, the network of silver-impregnated fibers is denser and darker than in the surrounding neuropil; however, there is no deformation of the neuronal processes. The average diameter of 4G8- and 6E10-positive plaques is similar, 68 p,m and 641xm, respectively, whereas Bielschowsky-positive plaques are significantly smaller, 58 p,m in diameter ( P < 0.01). All cortical plaques are negative when stained with thioflavin S and examined in
fluorescence and stained with Congo red and examined in polarized light. 4G8-positive plaques were found in the cerebral cortex in 23 of 30 dogs 6.5 to 26.5 years of age (77%). The youngest plaque-positive dog was 9 years of age; all dogs older than 13 years of age had plaques. Bielschowsky-positive plaques were found in 21 dogs (70%). The numerical density of plaques varies interindividually in so broad a range that there is no correlation between the numerical density of 4G8-, 6El0- and Bielschowsky-positive plaques and age in the plaque-positive group of dogs.
3.2. Groups of dogs distinguished with cluster analysis Cluster analysis based on two variables (age and the numerical density of 4G8-positive plaques in the cerebral cortex) distinguished three groups of dogs: with weak (10 dogs), moderate (6 dogs), and severe (7 dogs) amyloidosis (Table 1, Fig. 2). The group of dogs with weak amyloido-
Table 1 Characteristics of three groups of dogs distinguished by cluster analysis based on the age of animals and the numerical density of 4G8-positive plaques in the cerebral cortex Parameter Number of dogs Age (years) Numerical density of 4G8-positive plaques (n/mm 2) Numerical density of 6E10-positive plaques (n/mm 2) Numerical density of Bielschowsky-positive plaques (n/mm 2) S.E.M. in oarentheses.
Group, severity of cortical amyloidosis
P<
A, weak 10 19.5 (1.1) 0.8 (0.2)
B, moderate
C, severe
A/B
A/C
6
12.2 (0.8) 2.7 (1.2)
0.8 (0.3) 0.08 (0.02)
B/C
7
-
-
-
19.1 (0.5) 19.2 (3.1)
0.01 -
0.001
0.01 0.001
9.2 (6.1)
18.4 (4.2)
-
0.001
-
1.0 (0.6)
5.6 (1.2)
-
0.001
0.01
J. W¢giel et aL / Brain Research 728 (1996) 20-26 40. 30
20.
o
10
14
i
i
18
22
26
30
Age Fig. 2. Graphical presentation of the three groups of dogs distinguished by cluster analysis based on the age of animals and the numerical density of 4G8-positive plaques. Group A, with moderate severity of changes as estimated by the numerical density of plaques, appears to be a heterogenous subpopulation. This group consists of three dogs with a low numerical density of plaques (circ!tes below dotted line) comparable with those seen in group B, and three dogs with a higher numerical density of plaques than in group B but a lower numerical density than in group C (circles above dotted line). During aging in a portion of the dog population (group B, weak changes), the numerical density of plaques is low, 0.8/mm 2, and appears to remain almost constant during the more than 10-year-long period of observation. A subpopulation of aged dogs develops severe parenchymal amyloidosis-[3 with about 19 plaques per mm 2, on average (group C). The numerical density in combined groups A and C correlates with age.
sis has only 0.8 plaques per mm 2 on average and is the largest (43% of the plaque-positive population) and, surprisingly, the oldest group: 19.5 years of age, on average. The group with a moderate number of plaques (2.7/mm 2 on average) comprises 26% of the population. These dogs, which are 12.2 years of age on average, are significantly younger than the dogs with the lowest numerical density of plaques ( P < 0.01). The third group distinguished by cluster analysis consists of seven dogs (30% of the plaquepositive population) that are severely affected by parenchymal amyloidosis: the numerical density of cortical plaques is 19.2/mm 2 on average. The average age of dogs in this group (19.1 years) is non-significantly lower than that of the group with the lowest number of plaques. They are significantly older than moderately affected animals. The numerical density of 4G8-positive plaques in the combination of two groups (the youngest, with moderate amyloidosis, and the old with severe amyloidosis) correlates with age ( r = 0.7; P < 0.001). 4G8-positive plaques occupy significantly more cortex in the severely affected group (3.7%) than in the groups with weak (0.4%; P < 0.001) and moderate (0.7%; P < 0.01) parenchymal amyloidosis. The numerical density of 4G8-positive plaques in the most affected group of dogs was significantly ( P < 0.001) higher in the frontal cortex ( 3 8 / m m 2, on average) than in the temporal, limbic (19/rnm2), parietal (18/mm2), and occipital cortex (15/mm2). In two other groups with moderate and mild parenchymal amyloidosis, the plaques were distributed so unevenly in the examined cortical gyri that interregional differences were undetectable.
23
The numerical density of 6E10-positive plaques correlates with the numerical density of 4G8-positive plaques ( r = 0.71, P < 0.001). In the youngest group distinguished by cluster analysis, the number of 6E10-positive plaques is three times higher than the number of 4G8-positive plaques. In the two oldest groups (with severe and weak changes) 96% and 100% of 4G8-positive plaques, respectively, are 6E10-positive. Bielschowsky-positive plaques were found in almost all cases with 4G8-positive plaques (91.4%). The numerical density of Bielschowsky-positive plaques correlates with the numerical density of 4G8-positive plaques ( r = 0.82, P < 0.001); however, in groups with weak, moderate, and severe changes, they constitute only 10%, 37%, and 29%, respectively, of 4G8-positive plaques. They occupy significantly less cortex (0.46% + 0.18%) than do 4G8-positive plaques (1.3% _ 0.3%, Wilcoxon test P < 0.001).
4. Discussion
4.1. Subpopulations of dogs with severe and mild parenchymal amyloidosis The extensive study of the plaques in the cerebral cortex of 30 dogs, including 26 cortical regions, revealed plaques in the brains of all dogs older than 13 years of age. The incipient stage of cerebral brain amyloidosis comprises the age between the 9th and 13th years. The development of parenchymal brain amyloidosis in dogs appears to be an unavoidable, species-specific feature associated with age. However, the numerical density of 4G8-, 6El0-, or Bielschowsky-positive plaques in the cerebral cortex of dogs does not correlate with age. A similar lack of correlation was also observed in other dog populations [20]. Interindividual differences in the number of plaques in the brains of dogs of the same age suggest that the progress of parenchymal cortical amyloidosis might be upregulated only in a portion of the population of old dogs. Cluster analysis based on the age and the numerical density of 4G8-positive plaques confirmed this assumption. Three groups of dogs were distinguished: a severely affected group with numerical density of plaques of 19.2/mm 2 on average, a moderately affected group with 2.7 plaques per mm 2, and a less affected group with 0.8 plaques per mm 2. Almost the same average ages in the severely affected group (t9.1 years) and in the less affected group (19.5 years) suggests that these two groups are recruited from the youngest group: 12.2 years of age. The numerical density of plaques in the brains of three dogs in this group is in the range of that of the old and less affected group of animals, whereas the numerical density of plaques in the brain of the three other dogs is greater than the numerical density in old and less affected dogs but less than the numerical density seen in the oldest and most affected group (Fig. 2). The correlation between age and numerical
24
J. W¢gielet al./ Brain Research 728 (1996) 20-26
density of plaques in the combined younger group with moderate number of plaques and in the older group with severe amyloidosis indicates that the severe form of amyloidosis is an age-dependent process in a dog subpopulation. The results of cluster analysis suggest that the dog population consists of only two subpopulations, one that develops severe (43% of animals) and a second that develops weak cortical amyloidosis [3 (57%). The cause of the acceleration of amyloid deposition in the first years of the second decade of the life of some dogs is not clear. It might be the effect of higher breedspecific susceptibility to parenchymal amyloidosis [3. However, study of diffuse plaques in pure-bred beagles indicates that even in one breed there are remarkable differences in cerebral amyloidosis [3. Inter-familial differences suggest that a genetic factor might be involved in determining dogs' plaque susceptibility [20]. The impact of exogenous factors such as environmental conditions [24] or infections cannot be excluded. 4.2. Diffuse plaques properties It is believed that human diffuse plaque amyloid is nontoxic because they are free of neurofibrillary changes, and ultrastructurally, neuronal and synapfic pathology is almost undetectable [15,25,37-39]. The impact of dog diffuse plaques on neurons and the function of affected brain structures is not known. In light microscopy studies, dog diffuse plaques recall diffuse plaques in the cerebral cortex of young subjects with Down syndrome. The lack of dementia in young patients with DS suggests that even very numerous diffuse plaques ( 6 8 / m m 2) do not cause remarkable pathology and functional changes [37]. Dog diffuse plaques are free of clusters of microglial cells, which are the source of fibrillizing amyloid in classical and primitive plaques in human brain [28,32]. Diffuse plaques are nonfibrillar or contain only a few amyloid fibrils, and they are thioflavin S- and Congo red-negative [3,6,12,37]. The diffuse plaque amyloid might be derived either from A[31-42 or amyloid-[3 precursor protein A[3PP 695 [23] or from both sequences, with accumulation of soluble nonamyloidogenic peptides 1-16 and 17-42aa organized into putatively nontoxic aggregates. Diffuse plaques of the examined dogs are immunopositive when stained with mAbs to the N-terminal region 6El0 (1-17) and 4G8 (17-24). Similar onset and a strong correlation between the numerical density of 4G8- and 6E10-positive plaques ( r = 0.71, P < 0.001), and between the numerical density of 4G8- and Bielschowsky-positive plaques ( r = 0.82, P < 0.001) suggests common factors in their formation. Human diffuse plaques are positive when stained for the carboxy region 42, similarly to classical and primitive plaques and vascular amyloid, but they are negative when immunostained with antibodies to terminus 40 [10]. Our study indicates that a subset of diffuse dog plaques con-
tains both peptides with terminal regions 42 and 40. The presence of A[340-positive plaques in only some dogs suggests that processing of amyloid [3 might differ individually among dogs. Great variations among cases in the number of A[340-positive diffuse plaques and the lower number of A[340- than A[342-positive plaques were also found in human brain in AD and DS [11,40]. Labeling with antiserum BC40 gave only a fraction of BC42-positive plaques which may indicate local differences in [3peptide processing. Iwatsubo et al. [10] suggest that in the course of AD some A[342(43)-positive plaques acquire A[340-positivity, presumably by processing. The differences in the proportion of A[342/40-positive plaques in different brain structures may indicate that processing may also be brain-region-specific [13]. Amyloid [3-peptide 1-42(43) is assumed to be the factor that initiates fibrillar plaque formation [10]. The presence of amyloid [3-peptide with carboxy-terminal region 42 in dog vessels and fibrillar amyloid accumulation in the wall of dog arteries and veins confirm this suggestion; however, the presence of carboxy-terminal region 42 in diffuse plaques and the absence of fibrillar material during more than 10 years of parenchymal amyloidosis in dog cerebral cortex indicate that fibrillization does not depend only on the presence of peptide with C-terminal region 42. 4.3. Parenchymal and vascular amyloidosis in dog brain Morphometric study of amyloid angiopathy [30] and diffuse plaques (present study) in the same population of dogs has revealed that vascular and parenchymal amyloidosis appears at the same age. In contrast to parenchymal amyloidosis, which correlates with age in only a subpopulation of more affected dogs, amyloid angiopathy correlates with age in the whole amyloid-positive population. Amyloid deposition in the vascular wall in human [36] and dog brain [30] is associated with the smooth muscle cells of arteries and veins. Diffuse plaque formation appears to be associated with neurons [14,17,19,29,37]. The abovementioned differences in correlation between vascular and parenchymal amyloidosis and age might be related to the type of cell engaged in amyloid deposition in these two brain compartments. The similar onset of vascular and parenchymal amyloidoses suggests that the factor or factors initiating amyloid deposition by different types of cells is common. The different courses of vascular and parenchymal amyloidoses may be related more to a cellspecific response. 4.4. Dog as the model for the study of diffuse plaque formation Studies of the pathogenesis of amyloid suggest that Alzheimer-type pathology consists of several cell-specific pathological processes that coexist or develop for years
J. Wcgiel et aL/ Brain Research 728 (1996) 20-26
independently of each others. The key factor to classification of amyloid deposits appears to be the type o f cell associated causatively with amyloid deposition: neurons in diffuse plaque [14,17,19,29,37], microglial cells in classical and primitive plaque [28,39], perivascular cells and perivascular microglial cells in the wall of capillaries [34], and smooth muscle cells in the wall of parenchymal and meningeal arteries and veiins [5,30,36]. The morphology and the tinctorial and inmmnocytochemical properties of amyloid and the reaction o f the neuropil to amyloid appear to be secondary to the cell-specific factor initiating deposit formation. Marked differences in the spectrum of changes in dog brain in comparison to that in the brain of subjects with A D show that aged dog-brain amyloidosis [3 is a partial model of human Alzheimer-type pathology. In dog brain, there is no neurofibrillary pathology, no amyloid deposition in the wall of capillaries by perivascular cells of microglial lineage or with microglial cells in classical and primitive plaques, and no fibrillized plaques with degenerated neurons. However, the presence in dog brain o f two processes that are almost identical morphologically with human pathological proces,~es: fibrillar amyloid deposition in the wall of arteries and veins associated with smooth muscle cells and resulting in degeneration, necrosis, and degradation of cells of the vascular wall, and diffuse plaques with putatively nontoxic amyloid accumulation associated probably with neurons indicates that pathology of aged dogs can be a unique animal model for the study of these two forms of amyloidotic changes. In dog brain, the impact of diffuse plaques on neuropil and on the function of brain structures can be examined in isolation from other Alzheimer type changes. This study supports the hypothesis that diffuse plaque formation is a specific form of amyloidosis, probably associated with neurons, that is separate from the classical and primitive plaque formation associated with rnicroglial cells.
Acknowledgements The authors thank Dr. Haruyasu Yamaguchi (Gunma University, Gunma, Japan) for antisera BC40 and BC42, and Ms. Maureen Marlow for copy editing. Supported in part by funds from the New York State Office of Mental Retardation and Developmental Disabilities and a grant from the National Institutes of Helath, National Institute of Aging No. PO1-AGO-4220.
References [1] Arendt, T., Holzer, M., Fruth, R., Bruckner, M.K. and Gartner, U., Paired helical filament-like phosphorylation of tau, deposition of [3/A4-amyloid and memory impairment in rat induced by chronic inhibition of phosphatase 1 asad 2A, Neuroscience, 69 (1995) 691698.
25
[2] Ball, M.J., MacGregor, J., Fyfe, I.M., Rapoport, S.I. and London, E.D., Paucity of morphological changes in the brains of ageing beagle dogs: further evidence that Alzheimer lesions are unique for primate central nervous system, Neurobiol. Aging, 4 (1983) 127131. [3] Cummings, B., Su, J., Cotman, C., White, R. and Russell, M., [3-amyloid accumulation in aged canine brain: A model of early plaque formation in Alzheimer's disease, Neurobiol. Aging, 14 (1993) 547-560. [4] Dayan, A.D., Comparative neuropathology of aging, Brain, 94 (1971) 31-42. [5] Frackowiak, J., Zoltowska, A. and Wisniewski, H.M., Non-fibrillar 13-amyloidprotein is associated with smooth muscle cells of vessel walls in Alzheimer's disease, J. Neuropathol. Exp. Neurol., 53 (1994) 637-645. [6] Giaccone, G., Verga, L., Finazzi, M., Pollo, B., Tagliavini, F., Frangione, B. and Bugiani, O., Cerebral preamyloid deposits and congophilic angiopathy in aged dogs, Neurosci. Lett., 114 (1990) 178-183. [7] Gilbert, J.J. and Vinters, H.V., Cerebral amyloid angiopathy: Incidence and complications in the aging brain. I. Cerebral hemorrhage, Stroke, 14 (1983) 915-923. [8] Gilles, C., Bruchner, J.M., Khoubesserian, P. and Vanderhaeghen, J.J., Cerebral amyloid angiopathy as a cause of multiple intracerebral hemorrhages, Neurology, 34 (1984) 730-735. [9] Graham, D.I., Gentlemen, S.M., Lynch, A. and Roberts, G.W., Distribution of 13-amyloidprotein in the brain following severe head injury, Neuropathol. Appl. Neurobiol., 21 (1995) 27-34. [10] Iwatsubo, T., Odaka, A., Suzuki, N., Mizusawa, H., Nukina, N. and Ihara, Y., Visualization of AB42(43) and A1340 in senile plaques with end-specific A[3 monoclonals: evidence that an initially deposited species in AI342(43), Neuron, 13 (1994) 43-53. [11] Iwatsubo, T., Mann, D.M.A., Odaka, A., Suzuki, N. and lhara, Y., Amyloid [3 protein (AI3) deposition: A1342(43) precedes A1340 in Down's syndrome, Ann. Neurol., 37 (1995) 294-299. [12] Joachim, C.L., Morris, J.H. and Selkoe, D.J., Diffuse senile plaques occur commonly in the cerebellum in Alzheimer's disease, Am. Z Pathol., 135 (1989) 309-319. [13] Kida, E., Wisniewski, K.E. and Wisniewski, H.M., Early amyloid-[3 deposits show different immunoreactivity to the amino- and carboxy-terminal regions of [3-peptide in Alzheimer's disease and Down syndrome brain, Neurosci. Lett., 193 (1995) 1-4. [14] Li, Y.T., Woodruff-Pak, D.S. and Trojanowski, J.Q., Amyloid plaques in cerebellar cortex and the integrity of Purkinje cell dendrites, Neurobiol. Aging, 15 (1994) 1-9. [15] Masliah, E., Terry, R.D., Mallory, M., Alford, M., Hansen, L.A., Diffuse plaques do not attenuate synapse loss in Alzheimer's disease, Am. J. Pathol., 137 (1990) 1293-1297. [16] Osetowska, E., Morphologic changes in the brains of old dogs, Neuropathol. Pol., 4 (1966) 97-110. [17] Pappolla, M.A., Omar, R.A. and Vinters, H.V., Image analysis microscopy shows that neurons participate in the genesis of subset of early primitive (diffuse) senile plaques, Am. J. Pathol., 359 (1991) 599-607. [18] Pauli, B. and Luginbuhl, H., Fluorescenzmikroskopische Untersuchungen der cerebralen Amyloidose bei alten Hunden und senilen Menschen, Acta Neuropathol. (Berl.), 19 (1971) 121-128. [19] Probst, A., Langui, D., Ipsen, S., Robakis, N. and Ulrich, J., Deposition of beta/A4 protein along neuronal plasma membranes in diffuse senile plaques, Acta Neuropathol. (Berl.), 83 (1991) 21-29. [20] Russell, M., White, R., Patel, E., Markesbery, W., Watson, C. and Geddes, J., Familial influence on plaque formation in the beagle brain, NeuroReport, 3 (1992) 1093-1096. [21] Selkoe, D.J., Bell, D.S., Podlisny, M.B., Price, D.L., Cork, L.C., Conservation of brain amyloid proteins in aged mammals and humans with Alzheimer's disease, Science, 235 (1987) 873-877. [22] Shimada, A., Kuwamura, M., Umemura, T., Takada, K., Ohama, E.
26
J. W¢giel et a l . / Brain Research 728 (1996) 20-26
and Itakura, C., Modified Bielschowsky and immunohistochemical studies on senile plaques in aged dogs, Neurosci. Lett., 129 (1991) 25-28. [23] Sisodia, S.S., Koo, E.H., Beyreuther, K., Unterbeck, A. and Price, D.L., Evidence that 13-amyloid protein in Alzheimer's disease is not derived by normal processing, Science, 248 (1990) 492-496. [24] Soltysiak, Z., Comparative pathomorphology of senile changes in central nervous system (CNS) of dogs from a country, town and industry environment, Zesz. Nauk. Akad. Roln. Wroclawiu, Weter., 42 (1985) 249-271. [25] Tagliavini, F., Giaccone, G., Frangione, B. and Bugiani, O., Preamyloid deposits in the cerebral cortex of patients with Alzheimer's disease and nondemented individuals, Neurosci. Lett., 93 (1988) 191-196. [26] Uchida, K., Miyauchi, Y., Nakayama, H. and Goto, N., Amyloid angiopathy with cerebral hemorrhage and senile plaque in aged dogs, Jpn. J. Vet. Sci., 52 (1990) 605-611. [27] von Braunmtihl A., Kongophile Angiopathie und 'senile Plaques' bei greisen Hunden, Arch. Psychol. Z. Neurol., 194 (1956) 396-414. [28] Wegiel, J. and Wisniewski, H.M., The complex of microglia cells and amyloid star in three-dimensional reconstruction, Acta Neuropathol. (Berl.), 81 (1990) 116-124. [29] Wegiel, J., Wisniewski, H.M., Dziewiatkowski, J., Tarnawski, M., Dziewiatkowski, A., Soltysiak, Z. and Kim, K.S., Fibrillar and non-fibrillar amyloid in the brain of aged dogs. In K. Iqbal, J.A. Mortimer, A. Winblad and H.M. Wisniewski (Eds.), Research Advances in Alzheimer's Disease and Related Disorders, John Wiley and Sons, Chichester, 1995, pp. 703-707. [30] Wegiel, J., Wisniewski, H.M., Dziewiatkowski, J., Tarnawski, M., Nowakowski, J., Dziewiatkowska, A. and Soltysiak, Z., The origin of amyloid in cerebral vessels of aged dogs, Brain Res., 705 (1995) 225-234. [31] Wisniewski, H.M., Johnson, A.B., Raine, C.S., Kay, W.J. and Terry, R.D., Senile plaques and cerebral amyloidosis in aged dogs. A histochemical and ultrastructural study, Lab. Invest., 23 (1970) 287-296.
[32] Wisniewski, H.M., Wegiel, J., Wang, K.C., Kujawa, M. and Lach, B., Ultrastructural studies of the cells forming amyloid fibers in classical plaques, Can. J. Neurol. Sci., 16 (1989) 535-542. [33] Wisniewski, H.M., Wegiel, J., Morys, J., Bancher, C., Soltysiak, Z. and Kim, K.S., Aged dogs: an animal model to study beta-protein amyloidogenesis. In: K. Maurer, P. Riederer and H. Beckmann (Eds.), Alzheimer's Disease. Epidemiology, Neuropathology, Neurochemistry, and Clinics, Springer-Verlag, Vienna, 1990, pp. 151-168. [34] Wisniewski, H.M., Wegiel, J., Wang, K.C. and Lach, B., Ultrastructural studies of the cells forming amyloid in the cortical vessel wall in Alzheimer's disease, Acta Neuropathol. (Berl.), 84 (1992) 117127. [35] Wisniewski, H.M. and Wegiel, J., Migration of perivascular cells into the neuropil and their involvement in 13-amyloid plaque formation, Acta Neuropathol. (Berl.), 85 (1993) 586-595. [36] Wisniewski, HM. and Wegiel, J., 13-amyloidformation by myocytes of leptomeningeal vessels, Acta Neuropathol. (Berl.), 87 (1994) 233-241. [37] Wisniewski, H.M., Wegiel, J. and Popovitch, E., Age associated changes in 13-amyloid plaque fibrillization in Down syndrome, Dev. Brain. Dysfunct., 7 (1994) 330-339. [38] Yamaguchi, H., Hirai, S., Morimatsu, M., Shoji, M. and Ihara, Y.A., Variety of cerebral amyloid deposits in the brains of the Alzheimertype dementia demonstrated by 13 protein immunostaining, Acta Neuropathol. (BerL), 76 (1988) 541-549. [39] Yamaguchi, H., Nakazato, Y., Hirai, S., Shoji, M. and Harigaya, Y., Electron micrograph of diffuse plaques. Initial stage of senile plaque formation in the Alzheimer brain, Am. J. Pathol., 135 (1989) 593-597. [40] Yamaguchi, H., Sugihara, S., Ishiguro, K., Takashima, A. and Hirai, S., Immunohistochemical analysis of COO-termini of amyloid beta protein (A[3) using end-specific antisera for AI340 and AI342 in Alzheimer's disease and normal aging, Amyloid: Int. J. Exp. Clin. Invest., 2 (1995) 7-16. [41] Zar, J.H., Biostatistical Analysis, Prentice Hall, Englewood Cliffs, 1984.