Immunohistochemical Study of Cerebral Amyloid Angiopathy. 111. Widespread Alzheimer A4 Peptide in Cerebral Microvessel Walls Colocahes with Gamma Trace in Patients with Leukoencephalopathy Harry V. Vinters, MD,"$ Diana Lenard Secor, MS," William M. Pardridge, MD,tS and Francoise Gray, MD, PhD§ Brain tissue from 11patients with cerebral amyloid angiopathy, changes of Alzheimer's disease, and variable degrees of subcortical leukoencephalopathy was examined by immunohistochemical methods, using primary antibodies to peptide segments representing portions of the Alzheimer A4 (beta-) peptide or gamma-trace peptide (seen most commonly in Icelandic patients with cerebral hemorrhage (hereditary cerebral hemorrhage with amyloidosis [HCHWA-I}). Variable A4 immunostaining was seen within cortical (and rarely white matter) parenchyma in the form of senile plaques (with or without central cores), and within capillary and arteriolar walls. Within individual patients, A4 deposits were often primarily parenchymal or vascular, and when they were vascular they tended to be more prominent in arteriolar than in capillary wall segments. Perivascular A4 deposits were often detected around strongly immunoreactive microvessels. Gamma-trace immunoreactivity was noted in many ACpositive microvessel walls, but staining was always less intense than with the anti-A4 antibody. We conclude that patients with severe cerebral amyloid angiopathy may show wide variation in the severity and topography of A4 deposits within brain parenchyma A4 may colocalize with gamma-trace peptide, suggesting that A4 and gamma-trace forms of cerebral amyloid angiopathy may not be as biochemically distinctive as has been suggested. Other proteases or protease inhibitors may contribute to the pathogenesis of cerebral amyloid angiopathy or cerebral amyloid angiopathy-related stroke syndromes. Vinters HV, Secor DL, Pardridge WM, Gray F. Immunohistochemical study of cerebral amyloid angiopathy. 111. Widespread Alzheimer A4 peptide in cerebral microvessel walls colocalizes with gamma trace in patients with leukoencephalopathy. Ann Neurol 1990;28:34-42
The significance of cerebral amyloid angiopathy (CAA) as a cause of stroke-primarily nontraumatic intracerebral hemorrhage-has been recognized for almost 20 years El1 and CAA has come to be accepted as one of the microscopic hallmarks of neuropathological change in the brains of patients with Alzheimer's disease (AD) o r senile dementia of Alzheimer type (SDAT) {2-41. Amyloidotic lesions of AD or SDAT, especially senile plaques and microvessels affected by CAA, have more recently been characterized by biochemical and immunohistochemical methods {5- 131. Senile plaque cores and amyloid in CAA are composed of a unique amyloid protein with molecular weight of approximately 4,200 Da. Subtle differences
between the plaque amyloid and microvascular a m y loid have been found {14].The Alzheimer A4 or betapeptide originates from a larger protein molecule of 695 amino acids encoded by a gene on chromosome 21 [l5-171. Immunohistochemical methods using primary polyclonal or monoclonal antibodies to the near C-terminal end of the Alzheimer A4 precursor molecule (A4P) can be utilized to study amyloidotic lesions of A D and SDAT in the brain parenchyma {7, 8, 10121. Less common types of CAA include CAA associated with hereditary cerebral hemorrhage with autosomal dominant inheritance in an Icelandic kindred (hereditary cerebral hemorrhage with amyloidosisIcelandic {HCHWA-I)) {18, 19) and familial non-
From the Departments of *Pathology (Neuropathology) and tMedicine, and the $Brain Research Institute, UCLA Medical Center, Los Angeles, CA, and BAnatomie Pathologique, H W d Henri Mondor, Creteil, France.
Received Sep 26, 1989, and in revised form Dec 11. Accepted for publication Jan 16, 1990.
34
~ d correspondence d ~ to~Dr Vinters, ~ ~ Department of Pathology, UCLA School of Medicine, Room 18-170 CHS, Los Angeles, CA 90024-1732.
Copyright 0 1990 by the American Neurological Association
traumatic intraparenchymal hemorrhage associated with CAA in the Netherlands (HCHWA-D) C20f. Whereas the Dutch form of CAA appears biochemically linked to A D and SDAT amyloid 1211, the m y loidotic peptide in affected microvessels of patients with HCHWA-I is related to gamma-trace (cystatin C), an inhibitor of lysosomal cysteine protease, with a molecular weight on the order of 11,000-12,000 Da 122-241. Changes in the white matter may be relatively common in the brains of patients with A D or SDAT 125271.Perhaps the best characterized group of such patients with severe white-matter abnormalities was defined in studies by one of us (with associates) C28, 291. The brain tissue from many of the affected patients described showed microscopic features of A D or SDAT, but an unusually severe degree of leukoencephalopathy was also present in most individuals. The purpose of the present study was to extend the observations previously made by application of established immunohistochemical procedures that allow for detection of Alzheimer A4 and gamma-trace peptides.
Materials and Methods The brain tissue studied was derived from the previously described patient cohort [28], although tissue from 1 patient (Patient 5 ) was unavailable for study. Clinical and neuropathological features of the patients have been presented in detail 128). The diagnosis of CAA in each patient had initially been made using Congo red staining of paraffin sections and examination of the sections under polarized illumination. Routine histological and immunohistochernical procedures were carried out on at least two brain sections from each of the 11 patients previously reported. The representative paraffin blocks in every case included cerebral cortex and subcortical white matter. Usually, the two tissue sections included a section of the hippocampus. Immunohistochemical study of tissues was carried out using one of the two following primary polyclonal antibodies, which were raised in rabbits [7, 81 as previously described: antibodies to a 28-amino-acid peptide fragment representing a portion of the Alzheimer A4 or beta-peptide, or an 18-amino-acid peptide fragment representing amino acids 20 to 37 (inclusive) of the reported gamma-trace-like protein found in the microvessels of patients with HCHWA-I 122, 231. Immunogenic peptides were initially synthesized in the Peptide Synthesis Laboratory at UCLA Medical Center, and the amino acid sequence of each was subsequently verified, as previously described [8]. Peptides were linked to thyroglobulin prior to injection into rabbits, and resultant antibodies were absorbed with thyroglobulin prior to use. The immunohistochemical procedure was a standard avidin-biotin-peroxidase technique, utilizing formic acid for pretreatment of tissue sections to enhance amyloid immunureactivity and immunolabeling [7]. Control sera tested with parallel sections from each tissue block included nonreactive sera breirnmune) and sera preabsorbed with the immunogen to which the antibody was raised [7, 81. As well, in crossabsorption studies each antiserum was first absorbed with
the nonimmunogenic peptide (i.e., anti-A4 absorbed with gamma-trace peptide and anti-gamma-trace absorbed with A4) prior to use on sections. For each run in which the anti28-amino-acid (hereafter referred to as anti-A4) antibody was utilized, a control section of AD brain showing florid parenchymal and microvascular amyloidotic lesions was also stained, to ensure consistency of staining quality between runs. For all runs in which the anti-gamma-trace antibody was utilized, sections from paraffin-embedded brain tissue of patients with HCHWA-I were also stained using the identical procedure (brain tissue generously supplied by D r Olafur Jensson, The National Blood Bank, Reykjavik, Iceland). The data to be presented represent the results of at lcast two, and usually several runs, using both antibodies on all tissue blocks.
Results All of the tissues examined from every patient showed a multiplicity of microscopic lesions that were immunoreactive for the A4 peptide. For purposes of microscopic and immunohistochemical characterization, we distinguished capillary from arteriolar amyloidotic lesions based on the size of the vessels (capillaries being defined in general as having a diameter of 10 to 30 Fm), and differentiated senile plaques into two main types: those characterized by a diffuse but relatively circumscribed staining in the brain parenchyma; and those characterized by the more traditionally conceived two-component senile plaque, i.e., a plaque containing a halo and a prominent amyloidotic core
E4, 93. Whereas brain tissue from 4 of the patients showed prominent immunoreactive senile plaque cores and halos, the remaining patients had more diffuse-type senile plaques lacking a prominent central core (Fig 1) or inconspicuous senile plaques. Plaques ranged in diameter up to 75 pm, while plaque cores were 15 to 20 pm in the greatest dimension. In some cortical regions of several patients, the anti-A4 antiserum appeared to decorate only small corelike areas of A4-immunoreactive amyloid within brain cortical parenchyma, i.e., a “neuritic” component was not noted in these plaques. Within the cortex of some patients wherein plaques contained both a core and a halo or neuritic component, the cores were relatively inconspicuous. Two patients showed a complete absence of immunoreactive core or halo plaque components, i.e., the brain failed to show detectable parenchymal A4 immunoreactivity. In these individuals, the immunoreactivity was essentially confined to blood vessel walls. Representative patterns of parenchymal and vascular immunoreactivity are shown in Figure 2, and the immunohistochemical data on all patients are summarized in the Table. In brains that showed vascular immunoreactivity, A4 peptide either was found within capillaries or arterioles, or was approximately equally distributed in the two microvascular components (see Fig 2). In rare
Vinters et al: CAA Immunohistochemistry with Leukoencephalopathy
35
Fig 2. Patterns of neocortical A4 irnmunoreactiz:ity. (A)Patient 2. The only prominent immunoreactioity is within the walls of capillaries and arterioles. (B) Patient 8. In this jield) A4 immunoreactivity is present within parenchymal and leptorneningeal arterioles. and scattered senile plaques. (Ci Patient 12. Prominent immunoreactivity is noted within rnicrmjessel walls and parenchymal (senileplaqlnei deposits. Note peritiascnlar irnmunostaining (arrowheads). Especial& strong immunoreactivity is identified in the deep cortical layers. (A: x 25 before 17% reduction; 8: x 25 before 18% reduction; C: x 2 j before 17% reduction.) Fi g 1. Morphologv of A4-immunoreactioe senile plaques. (A) Patient I. Typical plaque with both a central irnrnunoreactioe core and immnnoreactive halo is shown. (B) Patient I . An A4immunoreactive senile plaque i~ comparable in size to that seen in (A), but lacking a core component. (A and B: X 520.) For this and all subsequent micrographs, unless otherwise stated, tissues were stained using an immunohistochemical procedure with primary antibody t o a synthetic segment of the A4 peptide, as described in {7).
36 Annals of Neurology
Vol 28 No 1 July 1990
b
patients, the A4 immunoreactivity was almost exclusively arteriolar. When parenchymal microvessels showed A4 immunoreactivity, leptomeningeal vessels also showed foci of immunoreactivity in most cases. When parenchymal arterioles and capillaries were immunostained with primary A4 antisera, a prominent perivascular tuft of immunoreactivity was frequently seen around positively stained microvessels (Patients 1, 3, 7 and 12; Fig 3). The predominance of A4 immunostaining, i.e., whether arteriolar or capillary, is summarized in the Table. Staining using primary antibody to the peptide representing a portion of the gamma-trace molecule (antigamma-trace) failed to show significant parenchymal immunoreactivity but frequently showed microvascu-
Summary of A4 und Gamma-Trace Immunohistochemistry in Patients with Cerebral Amyloid Angioputhy (CAA) and Leukoencephalopathy’ A4 Senile Plaques Patient 1 2
NO.^
Core
Halo/Diffuse
Arteriole/Capillary
++
++
0
0
A A A A A A A A A A A
3 4 6
++ + +
7
0
8
9 10 11 12
+
++ + 0
++
++ ++ ++ ++ ++ ++ ++ 0
++
>C = C = C = C >C = C >C >C >C >C = C
Perivascular Immunoreactivity
++ 0
++ 0
+ ++
+ + +
Gamma-Trace‘
0
+
+ * + 0
+ + 0
0
+
-r+
0
“Intensity of staining was graded semiquantitacivelyas 0 (absent) or t (trace) to + + (marked). bTissue from Patient 5 was not available for study. ‘Gamma-trace immunostaining detected only in arterioles and (rarely) capillaries.
A = arteriole; C = capillary.
Vinters et al: CAA Immunohistochernistry with Leukoencephalopathy
37
Fig 3. Focally pronounced zmmunoreectivity is seen amund two microvessels from 2 different patients. IA) Patient I . (BI Putzent 12. Note relatively nomul endathelium in the microvejsel illustruted in (A).(A: x 520 before 27% reduction; B: x 20j before 8% reduction.)
38 Annals of Neurology Vol 2 8 No 1 July 1WI
lar immunoreactivity, with staining of primarily arteriolar and some capillary blood vessel walls (Fig 4). The immunoreactivity, though definitely present, and absent or markedly diminished in parallel sections stained with absorbed antibody, was always much less intense than that noted with the anti-A4 antibody. Cross-absorption studies failed to show evidence of cross-reactivity between the anti-A4 and anti-gammatrace antibodies. Occasionally, hyalinized microvascular segments, frequently appearing to be in continuity with a microaneurysm, were identified. The hyaline material in these cases was consistently devoid of the A4 and gamma-trace peptides (Fig 5). Several of the brain specimens, as previously described [28), showed necrotic foci and foci of old hemorrhage including microhemorrhage. Occasionally, glomerular clusters of A4-immunoreactive microvessels were identified. Subpial staining with primary antibody to A4 was identified in 2 patients and a large leptomeningeal artery was occasionally stained, but usually in the adventitial component (i.e., the media tended to be spared). Rare neurofibrillary tangle “ghosts” showed A4 immunoreactivity, but we stress that these tangles did not show adjacent nuclei, suggesting that the parenchymal neurofibrillary tangles represented residua of tanglebearing neurons rather than tangles within viable neurons. Anti-A4 antibodies have previously been shown to decorate neurofibrillary tangle “ghosts” [301. Although others have demonstrated A4 immunoreactivity of neuronal lipofuscin 1311 and A4 precursor immunoreactivity within astrocytic and neuronal cell bodies in the rat brain 1321, we did not detect astroglial or intraneuronal (cytoplasmic) A4 immunoreactivity in any section. Focal A4 immunoreactivity was occasionally encountered in the white matter, but only when there was massive deposition of cortical amyloid in the brain. In 2 patients, a comparison was made in parallel sections between the number of amyloidotic lesions that could be stained using conventional silver staining (Bielschowsky’s technique) and the A4 immunohistochemical technique. In sections from both patients, the numbers and distribution of senile plaques and microvessels illustrated by both techniques were roughly comparable, although rigorous morphometric assessment of the lesions seen using both methods was not carried out. In patients with an unusually large number of A4-immunoreactive cortical lesions, there was a suggestion of laminar distribution of the immunoreactive lesions, these being seen primarily within the deeper cortical layers (see Fig 2). In Some patients, the white matter-in addition to showing pallordemonstrated prominent vacuolization. The small amount of tissue examined does not allow for a systematic correlation of the degree of neocortical A4-
Fig 4. Colocalization ofA4 and gamma-trace peptides. RougbLy parallel brain sections {Patient 2) were stainedfor A4 (A)and with antiserum to a peptide representing a portion of gammatrace (B). Whereas A4 inmunoreactivity is present within walls of arterioles (arrows in (A))and capillaries, fainter gammatrace reactivity is present only within walls of arterioles (arrows in (Bi). The inset in (B) represents a magnified view oftbe arteride shown by the arrow in the upper rigbt drea of the pdnel. Note the mainly adventitial gamma-trace staining. (A: x 75 before 40% reduction; B: x 7 j before 33% reduction; inset: x 190 &$ore 35% reduction.)
immunoreactive amyloidotic lesions and the degree of pathological change in the subcortical white matter. A review of clinical features in this relatively small number of patients does not allow for optimally detailed clinicopathological correlation; i.e., we cannot relate the degree of cortical A4- or gamma-trace-immunoreactive amyloidosis with specific clinical neurological or psychiatric features in individual patients.
Discussion This paper extends observations previously made on the immunohistochemical nature of Alzheimer A4 peptide to a relatively unique cohort of patients, mos+ of whom showed leukoencephalopathy in addition to clinical and pathological features suggesting AD or SDAT, including abundant neurofibrillary tangles and senile plaques within the neocortex and hippocampus. The previously described neuropathological findings in this cohort, together with the results described here, suggest that the patients may show a neurological disease with features of A D or SDAT, and Binswanger’s subcortical leukoencephalopathy [33, 34}. CAA is sometimes (though by no Seen in brains of patients with Binswmger’s disease, though the kinship of changes in the white matter in A D and SDAT (to be discussed) to those noted in Binswanger’s disease is far from clear [34}.
Fig j,patient 2. A n A4-immunoreactivearteriole i.,incontinuity with a microaneurysm that extend acmjj the pial membrane. The markedly hyalinized microaneurysm wall contains only scattered small amounts of A4-immunoreactive material (arrow). ( x 205 befre 17% reduction.)
Vinters er al: CAA Immunohistochernistry with Leukoencephalopathy
39
Though brain tissue from all patients studied showed parenchymal and microvascular A4-immunoreactive lesions of varying extent, as described by us and others using this technique [S, 9, 11, 12, 351, different patterns of immunoreactivity were seen in different patients. Many patients showed primary or exclusive A4 immunostaining of cortical blood vessel walls (whether in the capillary or arteriolar component), and several showed prominent perivascular staining. Microscopic lesions traditionally accepted as senile plaque cores or plaque halos, o r both, were variably immunostained. Other workers suggest a more extensive subclassification of parenchymal A4-immunoreactive lesions {36], though they indicate that the classification breaks down in an analysis of many parenchymal amyloidotic lesions. An interesting result is that the patients described were substantially indistinguishable by clinical criteria during life despite the wide variation in distribution of A4-immunoreactive lesions within the brain parenchyma and microvasculature. This supports the conclusion, suggested by many, that differential processing of the A4P molecule (possibly at the blood vessel wall) may result in differential distribution (blood vessel versus parenchyma) of the 4,200-molecular weight A4 molecule C3, 4, 71. Local processing of the A4P molecule might determine the exact topography of amyloidotic lesions in a given patient. It is thus of interest, for example, that patients with HCHMA-D show almost exclusively microvascular deposition of an A4-like molecule, yet show clinical features of dementia that closely mimic those experienced by patients with AD or SDAT 120, 21, 37). Phenotypic heterogeneity among patients with A D or SDAT {especially familial AD) is increasingly recognized, though it is unclear whether this represents genetic heterogeneity at molecular or biochemical levels, or both 1381. The heterogeneity of A4 lesions even among patients in our relatively circumscribed study leads us, however, to question whether “sporadic” CAA 112) is a distinct entity separable from A D and SDAT. A further intriguing finding is that significant gamma-trace-like immunoreactivity was frequently observed within microvessel walls (especially arterioles) that also showed more pronounced A4 staining. It is not yet clear what the role of the gamma-trace peptide in the vessel wall or in the microvascular pathological findings might be, but it is known that gammatrace has properties of a protease inhibitor {22, 231. Others have failed to demonstrate A4 and gammatrace immunoreactivity in microvascular lesions of CAA, perhaps reflecting methodological differences {39]. Possibly other as yet unidentified molecules with protease inhibitory or stimulatory activity may be important in the pathogenesis of CAA and CAA-related stroke, though these have not yet emerged in bio-
40 Annals of Neurology
Vol 28 No 1 July 1990
chemical or molecular biological studies. However, researchers at several laboratories have concluded that differential processing of messenger RNA (mRNA) species from the A4P D N A might lead to the generation of molecules that have features of protease inhibitors 140, 411. The serine protease inhibitor alphalantichymotrypsin has been identified in brain amyloid deposits of patients with A D or SDAT {42]. Microglial and astroglial cells have demonstrated monocytelike properties in vitro, including constitutive secretion of lysozyme and cystatin C [43]. A recurring theme presents itself: that of differential preamyloid protein processing or A4 degradation in and adjacent to the blood vessel wall. Although gamma-trace-like immunoreactivity was often seen in strongly A4-immunolabeled microvessels, gamma-trace-like peptide was not detected within parenchymal amyloidotic lesions in any patient. The finding of a relative absence of A4-immunoreactive material in the walls of microaneurysms emerging from amyloidotic microvessels supports the theory that, although A4 may initiate degenerative changes within the microvessel wall that are subsequently important in the causation of cerebral hemorrhage or infarcts, the amyloidotic peptide sets in motion reparative and fibrotic responses in the blood vessel wall that then may lead to cerebral hemorrhage o r necrosis 171. This may explain why such hyalinized vascular segments themselves fail to show prominent A4 immunoreactivity. The relative absence of A4 or gamma-trace immunoreactivity within the parenchymal white matter also suggests that local deposition of A4 or gamma-trace peptides within the white matter is not responsible for the previously described leukoencephalopathy in most of these patients. Whether the degree of overlying neocortical amyloidosis (A4) correlates with leukoencephalopathy cannot specifically be determined from this study, though the hypothesis is an intriguing one 1441. An extension of the hypothesis is that cortical CAA or simply A4 peptide may cause other forms of white matter injury in patients with A D or SDAT or with less common disorders in which leukoencephalopathy and CAA are seen in combination [25-27, 451. In brain tissue from the patients described, we also found that silver staining using Bielschowsky’s technique and the anti-A4 immunohistochemical method showed a comparable topography and quantity of m y loidotic lesions in the brain parenchyma. The effectiveness of the two techniques (and others) has been compared in several laboratories, with varying results {36, 46, 471. The laminar distribution of A4-immunoreactive lesions within the cortex of some individuals with especially extensive neocortical A4 amyloidosis is of interest, given the results from recent studies on the distribution of senile plaques in selected regions of the
cortex, which show that there may be a discrepancy between amyloidotic and neuritic plaques in many areas of the neocortex C4Sl. Addendum Since acceptance of this article for publication, a recent report 1491 has characterized cerebrovascular amyloid immunohistochemically in 46 patients with Alzheimer’s disease, Down’s syndrome, and intracranial hemorrhage and infarction as well as in 10 elderly patients with no neurological disease. Colocdzation of A4 or beta-protein and cystatin C , with a pattern similar to that described in our study, was described, thus indicating independent confirmation of our results. Supported by Public Health Service grant NS 26312-02 and a John Douglas FrenchiWilson Foundation Fellowship (to H. V. V.), and the State of California Department of Health Services and Alzheimer’s Disease and Related Disorders Association (to W. M. P.). Expert assistance in completion of the study and preparation of the manuscript was provided by Scott D. Brooks, Carol Appleton, and Greg Nishimura.
References 1. Vinters HV. Cerebral amyloid angiopathy. A critical review. Stroke 1987;18:311-324 2. Glenner GG, Henry JH, Fujihara S. Congophilic angiopathy in the pathogenesis of Alzheimer’s degeneration. Ann Pathol 1981;1: 120- 129 3. Castafio EM, Frangione B. Human amyloidosis, Alzheimer disease and related disorders. Lab Invest 1988;58:122-132 4. Vinters HV, Miller BL. Pardridge WM. Brain amyloid and Alzheimer disease. Ann Intern Med 1988;109:41-54 5. Masters CL, Simms G, Weinman NA, et al. Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc Natl Acad Sci USA 1985;82:4245-4249 6. Pardridge WM, Vinters HV, Yang J, er al. Amyloid angiopathy of Alzheimer’s disease: amino acid composition and partial sequence of a 4,200-Dalton peptide isolated from cortical rnicrovessels. J Neurochem 1987;49:1394-1401 7.Vinters HV, Pardridge WM, Secor DL, Ishii N. Immunohistochemical study of cerebral amyloid angiopathy. 11. Enhancement of immunostaining using formic acid pretreatment of tissue sections. Am J Pathol 1988;133:150-162 8. Vinters HV, Pardridge WM, Yang J. Immunohistochemical study of cerebral amyloid angioparhy: use of an antiserum to a synthetic 28-amino-acid peptide fragment of the Alzheimer’s disease amyloid precursor. Hum Pathol 1988;19:214-222 9. Selkoe DJ. Biochemistry of altered brain proteins in Alzheimer’s disease. Ann Rev Neurosci 1989;12:493-520 10. Roberts GW, Lofthouse R, Allsop D, et al. CNS amyloid proteins in neurodegenerative diseases. Neurology 1988;38:15341540 11. Joachim CL, Duffy LK, Morris JH, Selkoe DJ. Protein chemical and immunocytochemical studies of meningovascular betaamyloid protein in Alzheimer’s disease and normal aging. Brain Res 1988;474:100-111 12. Coria F, Prelli F, CastGo EM, et al. Beta-protein deposition: a pathogenetic link between Alzheimer’s disease and cerebral amyloid angiopathies. Brain Res 1988;463: 187-191 13. Glenner GG, Wong CW. Alzheimer’s disease and Down’s syn-
drome: sharing of a unique cerebrovascular amyloid fibril protein. Biochem Biophys Res Commun 1984;122:1131-1135 14. Prelli F, Cast60 E, Glenner GG, Frangione B. Differences between vascular and plaque core amyloid in Alzheimer’s disease. J Neurochem 1988;51:648-651 15. Kang J, Lemaire H-G, Unterbeck A, et al. The precursor of Alzheimer’s disease amyloid A4 protein resembles a cell-surface receptor. Nature 1987;325 :733-7 36 16. Tanti RE, Gusella JF, Watkins PC, et al. Amyloid beta protein gene: cDNA, mRNA distribution, and genetic linkage near the Alzheimer locus. Science 1987;325:880-884 17. Tanzi RE, St. George-Hyslop PH, Gusella JF. Molecular genetic approaches to Alzheimer’s disease. Trends Neurosci 1989; 12:152-158 18. Gudmundsson G, Hallgrimsson J, J6nasson TA, Bjamason 0. Hereditary cerebral haemorrhage with amyloidosis. Brain 1972; 95:387-404 19. Jensson 0, Gudmundsson G, Arnason A, er al. Hereditary cystatin C (gamma-trace) amyloid angiopathy of the CNS causing cerebral hemorrhage. Acta Neurol Scand 1987;76:102-114 20. Luyendijk W, Bots GThAM, Vegrer-van der Vlis M, et al. Hereditary cerebral haemorrhage caused by cortical amyloid angiopathy. J Neurol Sci 1988;85:267-280 21. van Duinen SG, Castaiio EM, P r e h F, et al. Hereditary cerebral hemorrhage with amyloidosis in patients of Dutch origin is related to Altheimer disease. Proc Natl Acad Sci USA 1987; 84:599 1-5994 22. Ghiso J, Jensson 0, Frangione B. Amyloid fibrils in hereditary cerebral hemorrhage with amyloidosis of Icelandic type is a variant of gamma-trace basic protein (cystatin c). Proc Natl Acad Sci USA 1986;83:2974-2978 23. Ghiso J, Pons-Estel B, Frangione B. Hereditary cerebral amyloid angiopathy: the amyloid fibrils contain a protein which is a variant of cystatin c, an inhibitor of lysosomal cysteine proteases. Biochem Biophys Res Commun 1986;136:548-554 24. Lofberg H, Grubb AO, Nilsson EK, et al. Irnmunohistochemical characterization of the amyloid deposits and quantitation of pertinent cerebrospinal fluid proteins in hereditary cerebral hemorrhage with amyloidosis. Stroke 1987;18:43 1-440 25. Hachinski VC, Potter P, Merskey H. Leuko-araiosis. Arch Neurol 1987;44:2 1-23 26. Steingart A, Hachinski VC, Lau C, et al. Cognitive and neurologic findings in subjects with diffuse white matter lucencies on computed tomographic scan (leuko-araiosis). Arch Neurol 1987;44:32-35 27. Englund E, Brun A, Alling C. White matter changes in dementia of Alzheimer’s type. Biochemical and neuropathological correlates. Brain 1988;111:1425-1439 28. Gray F, Dubas F, Roullet E, Escourolle R. Leukoencephalopathy in diffuse hemorrhagic cerebral amyloid angiopathy. Ann Neurol 1985;18:54-59 29. Duhas F, Gray F, Roullet E, Escourolle R. Leucoenc6phalopathies art6riopathiques (17 cas anatomo-cliniques). Rev Neurol (Paris) 1985;141:43-108 30. Hyman BT, Van Hoesen GW, Beyreuther K, Masters CL. A4 amyloid protein immunoreactiviry is present in Alzheimer’s disease neurofibrillary tangles. Neurosci Lett 1989;101:352-355 31. Bancher C, Grundke-Iqbal I, Iqbal K, et al. Immunoreactiviry of neuronal lipofuscin with monoclonal antibodies to the amyloid beta-protein. Neurobiol Aging 1988;lO:1 25-1 32 32. Card JP, Meade RP, Davis LG. Immunocytochemical localization of the precursor protein for beta-amyloid in the rat central nervous system. Neuron 1988;1:835-846 33. Babikian V, Ropper AH. Binswanger’s disease: A review. Stroke 1987;18:2-12 34. Fisher CM. Binswanger’s encephalopathy: a review. J Neurol 1989;236:65-79
Vinters et al: CAA Immunohistochemistry with Leukoencephalopathy
41
35. Ogomori K, Kitamoto T, Tateishi J, et al. Beta-protein amyloid is widely distributed in the cenval nervous system of patients with Alzheimer’s disease. Am J Pathol 1989;134:24%25 1 36. Ikeda S-I, Allsop D, Glenner GG. Morphology and distribution of plaque and related deposits in the brains of Alzheimer’s disease and control cases. An immunohistochemical study using amyloid beta-protein antibody. Lab Invest 1989;60:113122 37. Prelli F, Castafio EM, van Duinen SG, et al. Different processing of Alzheimer’s beta-protein precursor in the vessel wall of patients with hereditary cerebral hemorrhage with amyloidosisDutch type. Biochem Biophys Res Commun 1988;l51:11501155 38. Bird TD, Sumi SM, Nemens EJ, et al. Phenotypic heterogeneity in familial Alzheimer’s disease: a study of 24 kindreds. Ann Neurol 1989;2>:12-25 39. Yamada M, Tsukagoshi H, Wada Y ,et al. Absence of the cystatin C amyloid in the cerebral amyloid angiopathy, senile plaque, and extra-CNS amyloid deposits of aged Japanese. Acta Neurol Scand 1989;79:504-509 40. Ponte P, Gonzalez-DeWhitt P, Schihng J, et al. A new A4 amyloid mRNA contains a domain homologous to serine protease inhibitors. Nature 1988;331:525-527 41. Tanzi RE, McClatchey AI, Lamperti ED, et al. Protease inhibitor domain encoded by an amyloid protein precursor mRNA associated with Alzheimer’s disease. Nature 1988;331: 528-530 42. Abraham CR, Selkoe DJ, Potter H. Immunochemical identification of the serine protease inhibitor alpha,-antichymotrypsin in
42
Annals of Neurology
Vol 28 No 1 July 1990
43.
44.
4 5.
46.
47.
48.
49.
the brain amyloid deposits of Alzheimer‘s disease. Cell 1988; 52:487-501 Zucker-Franklin D, Warfel A, Grusky G, et al. Novel monocyte-like properties of microgliaVastroglial cells. Constitutive secretion of lysozyme and cystatin-C. Lab Invest 1987;57:176185 Vinters HV, Pardridge WM, Secor DL, et al. Vascular dementia: immunocytochemical study of brain microvascular and parenchymal lesions associated with Alzheimer’s disease. In: Ginsberg MD, Dietrich WD, eds. Cerebrovascular diseases. New York: Raven Press, 1989:213-220 Heffner RR Jr, Porro RS, Olson ME, Earle KM. A demyelinating disorder associated with cerebrovascular amyloid angiopathy. Arch Neurol 1976;33:501-506 Wisniewski HM, Wen GY, Kim KS. Comparison of four staining methods on the detection of neuritic plaques. Acca Neuropathol 1989;78:22-27 Davies L, Wolska B, Hilbich C, et al. A4 amyloid protein deposition and the diagnosis of Alzheimer’s disease: prevalence in aged brains determined by immunocytochemistry compared with conventional neuropathologic techniques. Neurology 1988;38:1688-1693 Braak H, Braak E, Kalus P. Alzheimer‘s disease: areal and laminar pathology in the occipital isocortex. Acta Neuropathol 1989;77:494-506 Maruyama K, Ikeda S-I, Ishikara T, et al. Immunohistochemical characterization of cerebrovascular amyloid in 46 autopsied cases using antibodies to beta protein and cystatin C . Srrokc 1990;21:397-403
The significance of cerebral amyloid angiopathy (CAA) as a cause of stroke-primarily nontraumatic intracerebral hemorrhage-has been recognized for almost 20 years El1 and CAA has come to be accepted as one of the microscopic hallmarks of neuropathological change in the brains of patients with Alzheimer's disease (AD) o r senile dementia of Alzheimer type (SDAT) {2-41. Amyloidotic lesions of AD or SDAT, especially senile plaques and microvessels affected by CAA, have more recently been characterized by biochemical and immunohistochemical methods {5- 131. Senile plaque cores and amyloid in CAA are composed of a unique amyloid protein with molecular weight of approximately 4,200 Da. Subtle differences
between the plaque amyloid and microvascular a m y loid have been found {14].The Alzheimer A4 or betapeptide originates from a larger protein molecule of 695 amino acids encoded by a gene on chromosome 21 [l5-171. Immunohistochemical methods using primary polyclonal or monoclonal antibodies to the near C-terminal end of the Alzheimer A4 precursor molecule (A4P) can be utilized to study amyloidotic lesions of A D and SDAT in the brain parenchyma {7, 8, 10121. Less common types of CAA include CAA associated with hereditary cerebral hemorrhage with autosomal dominant inheritance in an Icelandic kindred (hereditary cerebral hemorrhage with amyloidosisIcelandic {HCHWA-I)) {18, 19) and familial non-
From the Departments of *Pathology (Neuropathology) and tMedicine, and the $Brain Research Institute, UCLA Medical Center, Los Angeles, CA, and BAnatomie Pathologique, H W d Henri Mondor, Creteil, France.
Received Sep 26, 1989, and in revised form Dec 11. Accepted for publication Jan 16, 1990.
34
~ d correspondence d ~ to~Dr Vinters, ~ ~ Department of Pathology, UCLA School of Medicine, Room 18-170 CHS, Los Angeles, CA 90024-1732.
Copyright 0 1990 by the American Neurological Association
traumatic intraparenchymal hemorrhage associated with CAA in the Netherlands (HCHWA-D) C20f. Whereas the Dutch form of CAA appears biochemically linked to A D and SDAT amyloid 1211, the m y loidotic peptide in affected microvessels of patients with HCHWA-I is related to gamma-trace (cystatin C), an inhibitor of lysosomal cysteine protease, with a molecular weight on the order of 11,000-12,000 Da 122-241. Changes in the white matter may be relatively common in the brains of patients with A D or SDAT 125271.Perhaps the best characterized group of such patients with severe white-matter abnormalities was defined in studies by one of us (with associates) C28, 291. The brain tissue from many of the affected patients described showed microscopic features of A D or SDAT, but an unusually severe degree of leukoencephalopathy was also present in most individuals. The purpose of the present study was to extend the observations previously made by application of established immunohistochemical procedures that allow for detection of Alzheimer A4 and gamma-trace peptides.
Materials and Methods The brain tissue studied was derived from the previously described patient cohort [28], although tissue from 1 patient (Patient 5 ) was unavailable for study. Clinical and neuropathological features of the patients have been presented in detail 128). The diagnosis of CAA in each patient had initially been made using Congo red staining of paraffin sections and examination of the sections under polarized illumination. Routine histological and immunohistochernical procedures were carried out on at least two brain sections from each of the 11 patients previously reported. The representative paraffin blocks in every case included cerebral cortex and subcortical white matter. Usually, the two tissue sections included a section of the hippocampus. Immunohistochemical study of tissues was carried out using one of the two following primary polyclonal antibodies, which were raised in rabbits [7, 81 as previously described: antibodies to a 28-amino-acid peptide fragment representing a portion of the Alzheimer A4 or beta-peptide, or an 18-amino-acid peptide fragment representing amino acids 20 to 37 (inclusive) of the reported gamma-trace-like protein found in the microvessels of patients with HCHWA-I 122, 231. Immunogenic peptides were initially synthesized in the Peptide Synthesis Laboratory at UCLA Medical Center, and the amino acid sequence of each was subsequently verified, as previously described [8]. Peptides were linked to thyroglobulin prior to injection into rabbits, and resultant antibodies were absorbed with thyroglobulin prior to use. The immunohistochemical procedure was a standard avidin-biotin-peroxidase technique, utilizing formic acid for pretreatment of tissue sections to enhance amyloid immunureactivity and immunolabeling [7]. Control sera tested with parallel sections from each tissue block included nonreactive sera breirnmune) and sera preabsorbed with the immunogen to which the antibody was raised [7, 81. As well, in crossabsorption studies each antiserum was first absorbed with
the nonimmunogenic peptide (i.e., anti-A4 absorbed with gamma-trace peptide and anti-gamma-trace absorbed with A4) prior to use on sections. For each run in which the anti28-amino-acid (hereafter referred to as anti-A4) antibody was utilized, a control section of AD brain showing florid parenchymal and microvascular amyloidotic lesions was also stained, to ensure consistency of staining quality between runs. For all runs in which the anti-gamma-trace antibody was utilized, sections from paraffin-embedded brain tissue of patients with HCHWA-I were also stained using the identical procedure (brain tissue generously supplied by D r Olafur Jensson, The National Blood Bank, Reykjavik, Iceland). The data to be presented represent the results of at lcast two, and usually several runs, using both antibodies on all tissue blocks.
Results All of the tissues examined from every patient showed a multiplicity of microscopic lesions that were immunoreactive for the A4 peptide. For purposes of microscopic and immunohistochemical characterization, we distinguished capillary from arteriolar amyloidotic lesions based on the size of the vessels (capillaries being defined in general as having a diameter of 10 to 30 Fm), and differentiated senile plaques into two main types: those characterized by a diffuse but relatively circumscribed staining in the brain parenchyma; and those characterized by the more traditionally conceived two-component senile plaque, i.e., a plaque containing a halo and a prominent amyloidotic core
E4, 93. Whereas brain tissue from 4 of the patients showed prominent immunoreactive senile plaque cores and halos, the remaining patients had more diffuse-type senile plaques lacking a prominent central core (Fig 1) or inconspicuous senile plaques. Plaques ranged in diameter up to 75 pm, while plaque cores were 15 to 20 pm in the greatest dimension. In some cortical regions of several patients, the anti-A4 antiserum appeared to decorate only small corelike areas of A4-immunoreactive amyloid within brain cortical parenchyma, i.e., a “neuritic” component was not noted in these plaques. Within the cortex of some patients wherein plaques contained both a core and a halo or neuritic component, the cores were relatively inconspicuous. Two patients showed a complete absence of immunoreactive core or halo plaque components, i.e., the brain failed to show detectable parenchymal A4 immunoreactivity. In these individuals, the immunoreactivity was essentially confined to blood vessel walls. Representative patterns of parenchymal and vascular immunoreactivity are shown in Figure 2, and the immunohistochemical data on all patients are summarized in the Table. In brains that showed vascular immunoreactivity, A4 peptide either was found within capillaries or arterioles, or was approximately equally distributed in the two microvascular components (see Fig 2). In rare
Vinters et al: CAA Immunohistochemistry with Leukoencephalopathy
35
Fig 2. Patterns of neocortical A4 irnmunoreactiz:ity. (A)Patient 2. The only prominent immunoreactioity is within the walls of capillaries and arterioles. (B) Patient 8. In this jield) A4 immunoreactivity is present within parenchymal and leptorneningeal arterioles. and scattered senile plaques. (Ci Patient 12. Prominent immunoreactivity is noted within rnicrmjessel walls and parenchymal (senileplaqlnei deposits. Note peritiascnlar irnmunostaining (arrowheads). Especial& strong immunoreactivity is identified in the deep cortical layers. (A: x 25 before 17% reduction; 8: x 25 before 18% reduction; C: x 2 j before 17% reduction.) Fi g 1. Morphologv of A4-immunoreactioe senile plaques. (A) Patient I. Typical plaque with both a central irnrnunoreactioe core and immnnoreactive halo is shown. (B) Patient I . An A4immunoreactive senile plaque i~ comparable in size to that seen in (A), but lacking a core component. (A and B: X 520.) For this and all subsequent micrographs, unless otherwise stated, tissues were stained using an immunohistochemical procedure with primary antibody t o a synthetic segment of the A4 peptide, as described in {7).
36 Annals of Neurology
Vol 28 No 1 July 1990
b
patients, the A4 immunoreactivity was almost exclusively arteriolar. When parenchymal microvessels showed A4 immunoreactivity, leptomeningeal vessels also showed foci of immunoreactivity in most cases. When parenchymal arterioles and capillaries were immunostained with primary A4 antisera, a prominent perivascular tuft of immunoreactivity was frequently seen around positively stained microvessels (Patients 1, 3, 7 and 12; Fig 3). The predominance of A4 immunostaining, i.e., whether arteriolar or capillary, is summarized in the Table. Staining using primary antibody to the peptide representing a portion of the gamma-trace molecule (antigamma-trace) failed to show significant parenchymal immunoreactivity but frequently showed microvascu-
Summary of A4 und Gamma-Trace Immunohistochemistry in Patients with Cerebral Amyloid Angioputhy (CAA) and Leukoencephalopathy’ A4 Senile Plaques Patient 1 2
NO.^
Core
Halo/Diffuse
Arteriole/Capillary
++
++
0
0
A A A A A A A A A A A
3 4 6
++ + +
7
0
8
9 10 11 12
+
++ + 0
++
++ ++ ++ ++ ++ ++ ++ 0
++
>C = C = C = C >C = C >C >C >C >C = C
Perivascular Immunoreactivity
++ 0
++ 0
+ ++
+ + +
Gamma-Trace‘
0
+
+ * + 0
+ + 0
0
+
-r+
0
“Intensity of staining was graded semiquantitacivelyas 0 (absent) or t (trace) to + + (marked). bTissue from Patient 5 was not available for study. ‘Gamma-trace immunostaining detected only in arterioles and (rarely) capillaries.
A = arteriole; C = capillary.
Vinters et al: CAA Immunohistochernistry with Leukoencephalopathy
37
Fig 3. Focally pronounced zmmunoreectivity is seen amund two microvessels from 2 different patients. IA) Patient I . (BI Putzent 12. Note relatively nomul endathelium in the microvejsel illustruted in (A).(A: x 520 before 27% reduction; B: x 20j before 8% reduction.)
38 Annals of Neurology Vol 2 8 No 1 July 1WI
lar immunoreactivity, with staining of primarily arteriolar and some capillary blood vessel walls (Fig 4). The immunoreactivity, though definitely present, and absent or markedly diminished in parallel sections stained with absorbed antibody, was always much less intense than that noted with the anti-A4 antibody. Cross-absorption studies failed to show evidence of cross-reactivity between the anti-A4 and anti-gammatrace antibodies. Occasionally, hyalinized microvascular segments, frequently appearing to be in continuity with a microaneurysm, were identified. The hyaline material in these cases was consistently devoid of the A4 and gamma-trace peptides (Fig 5). Several of the brain specimens, as previously described [28), showed necrotic foci and foci of old hemorrhage including microhemorrhage. Occasionally, glomerular clusters of A4-immunoreactive microvessels were identified. Subpial staining with primary antibody to A4 was identified in 2 patients and a large leptomeningeal artery was occasionally stained, but usually in the adventitial component (i.e., the media tended to be spared). Rare neurofibrillary tangle “ghosts” showed A4 immunoreactivity, but we stress that these tangles did not show adjacent nuclei, suggesting that the parenchymal neurofibrillary tangles represented residua of tanglebearing neurons rather than tangles within viable neurons. Anti-A4 antibodies have previously been shown to decorate neurofibrillary tangle “ghosts” [301. Although others have demonstrated A4 immunoreactivity of neuronal lipofuscin 1311 and A4 precursor immunoreactivity within astrocytic and neuronal cell bodies in the rat brain 1321, we did not detect astroglial or intraneuronal (cytoplasmic) A4 immunoreactivity in any section. Focal A4 immunoreactivity was occasionally encountered in the white matter, but only when there was massive deposition of cortical amyloid in the brain. In 2 patients, a comparison was made in parallel sections between the number of amyloidotic lesions that could be stained using conventional silver staining (Bielschowsky’s technique) and the A4 immunohistochemical technique. In sections from both patients, the numbers and distribution of senile plaques and microvessels illustrated by both techniques were roughly comparable, although rigorous morphometric assessment of the lesions seen using both methods was not carried out. In patients with an unusually large number of A4-immunoreactive cortical lesions, there was a suggestion of laminar distribution of the immunoreactive lesions, these being seen primarily within the deeper cortical layers (see Fig 2). In Some patients, the white matter-in addition to showing pallordemonstrated prominent vacuolization. The small amount of tissue examined does not allow for a systematic correlation of the degree of neocortical A4-
Fig 4. Colocalization ofA4 and gamma-trace peptides. RougbLy parallel brain sections {Patient 2) were stainedfor A4 (A)and with antiserum to a peptide representing a portion of gammatrace (B). Whereas A4 inmunoreactivity is present within walls of arterioles (arrows in (A))and capillaries, fainter gammatrace reactivity is present only within walls of arterioles (arrows in (Bi). The inset in (B) represents a magnified view oftbe arteride shown by the arrow in the upper rigbt drea of the pdnel. Note the mainly adventitial gamma-trace staining. (A: x 75 before 40% reduction; B: x 7 j before 33% reduction; inset: x 190 &$ore 35% reduction.)
immunoreactive amyloidotic lesions and the degree of pathological change in the subcortical white matter. A review of clinical features in this relatively small number of patients does not allow for optimally detailed clinicopathological correlation; i.e., we cannot relate the degree of cortical A4- or gamma-trace-immunoreactive amyloidosis with specific clinical neurological or psychiatric features in individual patients.
Discussion This paper extends observations previously made on the immunohistochemical nature of Alzheimer A4 peptide to a relatively unique cohort of patients, mos+ of whom showed leukoencephalopathy in addition to clinical and pathological features suggesting AD or SDAT, including abundant neurofibrillary tangles and senile plaques within the neocortex and hippocampus. The previously described neuropathological findings in this cohort, together with the results described here, suggest that the patients may show a neurological disease with features of A D or SDAT, and Binswanger’s subcortical leukoencephalopathy [33, 34}. CAA is sometimes (though by no Seen in brains of patients with Binswmger’s disease, though the kinship of changes in the white matter in A D and SDAT (to be discussed) to those noted in Binswanger’s disease is far from clear [34}.
Fig j,patient 2. A n A4-immunoreactivearteriole i.,incontinuity with a microaneurysm that extend acmjj the pial membrane. The markedly hyalinized microaneurysm wall contains only scattered small amounts of A4-immunoreactive material (arrow). ( x 205 befre 17% reduction.)
Vinters er al: CAA Immunohistochernistry with Leukoencephalopathy
39
Though brain tissue from all patients studied showed parenchymal and microvascular A4-immunoreactive lesions of varying extent, as described by us and others using this technique [S, 9, 11, 12, 351, different patterns of immunoreactivity were seen in different patients. Many patients showed primary or exclusive A4 immunostaining of cortical blood vessel walls (whether in the capillary or arteriolar component), and several showed prominent perivascular staining. Microscopic lesions traditionally accepted as senile plaque cores or plaque halos, o r both, were variably immunostained. Other workers suggest a more extensive subclassification of parenchymal A4-immunoreactive lesions {36], though they indicate that the classification breaks down in an analysis of many parenchymal amyloidotic lesions. An interesting result is that the patients described were substantially indistinguishable by clinical criteria during life despite the wide variation in distribution of A4-immunoreactive lesions within the brain parenchyma and microvasculature. This supports the conclusion, suggested by many, that differential processing of the A4P molecule (possibly at the blood vessel wall) may result in differential distribution (blood vessel versus parenchyma) of the 4,200-molecular weight A4 molecule C3, 4, 71. Local processing of the A4P molecule might determine the exact topography of amyloidotic lesions in a given patient. It is thus of interest, for example, that patients with HCHMA-D show almost exclusively microvascular deposition of an A4-like molecule, yet show clinical features of dementia that closely mimic those experienced by patients with AD or SDAT 120, 21, 37). Phenotypic heterogeneity among patients with A D or SDAT {especially familial AD) is increasingly recognized, though it is unclear whether this represents genetic heterogeneity at molecular or biochemical levels, or both 1381. The heterogeneity of A4 lesions even among patients in our relatively circumscribed study leads us, however, to question whether “sporadic” CAA 112) is a distinct entity separable from A D and SDAT. A further intriguing finding is that significant gamma-trace-like immunoreactivity was frequently observed within microvessel walls (especially arterioles) that also showed more pronounced A4 staining. It is not yet clear what the role of the gamma-trace peptide in the vessel wall or in the microvascular pathological findings might be, but it is known that gammatrace has properties of a protease inhibitor {22, 231. Others have failed to demonstrate A4 and gammatrace immunoreactivity in microvascular lesions of CAA, perhaps reflecting methodological differences {39]. Possibly other as yet unidentified molecules with protease inhibitory or stimulatory activity may be important in the pathogenesis of CAA and CAA-related stroke, though these have not yet emerged in bio-
40 Annals of Neurology
Vol 28 No 1 July 1990
chemical or molecular biological studies. However, researchers at several laboratories have concluded that differential processing of messenger RNA (mRNA) species from the A4P D N A might lead to the generation of molecules that have features of protease inhibitors 140, 411. The serine protease inhibitor alphalantichymotrypsin has been identified in brain amyloid deposits of patients with A D or SDAT {42]. Microglial and astroglial cells have demonstrated monocytelike properties in vitro, including constitutive secretion of lysozyme and cystatin C [43]. A recurring theme presents itself: that of differential preamyloid protein processing or A4 degradation in and adjacent to the blood vessel wall. Although gamma-trace-like immunoreactivity was often seen in strongly A4-immunolabeled microvessels, gamma-trace-like peptide was not detected within parenchymal amyloidotic lesions in any patient. The finding of a relative absence of A4-immunoreactive material in the walls of microaneurysms emerging from amyloidotic microvessels supports the theory that, although A4 may initiate degenerative changes within the microvessel wall that are subsequently important in the causation of cerebral hemorrhage or infarcts, the amyloidotic peptide sets in motion reparative and fibrotic responses in the blood vessel wall that then may lead to cerebral hemorrhage o r necrosis 171. This may explain why such hyalinized vascular segments themselves fail to show prominent A4 immunoreactivity. The relative absence of A4 or gamma-trace immunoreactivity within the parenchymal white matter also suggests that local deposition of A4 or gamma-trace peptides within the white matter is not responsible for the previously described leukoencephalopathy in most of these patients. Whether the degree of overlying neocortical amyloidosis (A4) correlates with leukoencephalopathy cannot specifically be determined from this study, though the hypothesis is an intriguing one 1441. An extension of the hypothesis is that cortical CAA or simply A4 peptide may cause other forms of white matter injury in patients with A D or SDAT or with less common disorders in which leukoencephalopathy and CAA are seen in combination [25-27, 451. In brain tissue from the patients described, we also found that silver staining using Bielschowsky’s technique and the anti-A4 immunohistochemical method showed a comparable topography and quantity of m y loidotic lesions in the brain parenchyma. The effectiveness of the two techniques (and others) has been compared in several laboratories, with varying results {36, 46, 471. The laminar distribution of A4-immunoreactive lesions within the cortex of some individuals with especially extensive neocortical A4 amyloidosis is of interest, given the results from recent studies on the distribution of senile plaques in selected regions of the
cortex, which show that there may be a discrepancy between amyloidotic and neuritic plaques in many areas of the neocortex C4Sl. Addendum Since acceptance of this article for publication, a recent report 1491 has characterized cerebrovascular amyloid immunohistochemically in 46 patients with Alzheimer’s disease, Down’s syndrome, and intracranial hemorrhage and infarction as well as in 10 elderly patients with no neurological disease. Colocdzation of A4 or beta-protein and cystatin C , with a pattern similar to that described in our study, was described, thus indicating independent confirmation of our results. Supported by Public Health Service grant NS 26312-02 and a John Douglas FrenchiWilson Foundation Fellowship (to H. V. V.), and the State of California Department of Health Services and Alzheimer’s Disease and Related Disorders Association (to W. M. P.). Expert assistance in completion of the study and preparation of the manuscript was provided by Scott D. Brooks, Carol Appleton, and Greg Nishimura.
References 1. Vinters HV. Cerebral amyloid angiopathy. A critical review. Stroke 1987;18:311-324 2. Glenner GG, Henry JH, Fujihara S. Congophilic angiopathy in the pathogenesis of Alzheimer’s degeneration. Ann Pathol 1981;1: 120- 129 3. Castafio EM, Frangione B. Human amyloidosis, Alzheimer disease and related disorders. Lab Invest 1988;58:122-132 4. Vinters HV, Miller BL. Pardridge WM. Brain amyloid and Alzheimer disease. Ann Intern Med 1988;109:41-54 5. Masters CL, Simms G, Weinman NA, et al. Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc Natl Acad Sci USA 1985;82:4245-4249 6. Pardridge WM, Vinters HV, Yang J, er al. Amyloid angiopathy of Alzheimer’s disease: amino acid composition and partial sequence of a 4,200-Dalton peptide isolated from cortical rnicrovessels. J Neurochem 1987;49:1394-1401 7.Vinters HV, Pardridge WM, Secor DL, Ishii N. Immunohistochemical study of cerebral amyloid angiopathy. 11. Enhancement of immunostaining using formic acid pretreatment of tissue sections. Am J Pathol 1988;133:150-162 8. Vinters HV, Pardridge WM, Yang J. Immunohistochemical study of cerebral amyloid angioparhy: use of an antiserum to a synthetic 28-amino-acid peptide fragment of the Alzheimer’s disease amyloid precursor. Hum Pathol 1988;19:214-222 9. Selkoe DJ. Biochemistry of altered brain proteins in Alzheimer’s disease. Ann Rev Neurosci 1989;12:493-520 10. Roberts GW, Lofthouse R, Allsop D, et al. CNS amyloid proteins in neurodegenerative diseases. Neurology 1988;38:15341540 11. Joachim CL, Duffy LK, Morris JH, Selkoe DJ. Protein chemical and immunocytochemical studies of meningovascular betaamyloid protein in Alzheimer’s disease and normal aging. Brain Res 1988;474:100-111 12. Coria F, Prelli F, CastGo EM, et al. Beta-protein deposition: a pathogenetic link between Alzheimer’s disease and cerebral amyloid angiopathies. Brain Res 1988;463: 187-191 13. Glenner GG, Wong CW. Alzheimer’s disease and Down’s syn-
drome: sharing of a unique cerebrovascular amyloid fibril protein. Biochem Biophys Res Commun 1984;122:1131-1135 14. Prelli F, Cast60 E, Glenner GG, Frangione B. Differences between vascular and plaque core amyloid in Alzheimer’s disease. J Neurochem 1988;51:648-651 15. Kang J, Lemaire H-G, Unterbeck A, et al. The precursor of Alzheimer’s disease amyloid A4 protein resembles a cell-surface receptor. Nature 1987;325 :733-7 36 16. Tanti RE, Gusella JF, Watkins PC, et al. Amyloid beta protein gene: cDNA, mRNA distribution, and genetic linkage near the Alzheimer locus. Science 1987;325:880-884 17. Tanzi RE, St. George-Hyslop PH, Gusella JF. Molecular genetic approaches to Alzheimer’s disease. Trends Neurosci 1989; 12:152-158 18. Gudmundsson G, Hallgrimsson J, J6nasson TA, Bjamason 0. Hereditary cerebral haemorrhage with amyloidosis. Brain 1972; 95:387-404 19. Jensson 0, Gudmundsson G, Arnason A, er al. Hereditary cystatin C (gamma-trace) amyloid angiopathy of the CNS causing cerebral hemorrhage. Acta Neurol Scand 1987;76:102-114 20. Luyendijk W, Bots GThAM, Vegrer-van der Vlis M, et al. Hereditary cerebral haemorrhage caused by cortical amyloid angiopathy. J Neurol Sci 1988;85:267-280 21. van Duinen SG, Castaiio EM, P r e h F, et al. Hereditary cerebral hemorrhage with amyloidosis in patients of Dutch origin is related to Altheimer disease. Proc Natl Acad Sci USA 1987; 84:599 1-5994 22. Ghiso J, Jensson 0, Frangione B. Amyloid fibrils in hereditary cerebral hemorrhage with amyloidosis of Icelandic type is a variant of gamma-trace basic protein (cystatin c). Proc Natl Acad Sci USA 1986;83:2974-2978 23. Ghiso J, Pons-Estel B, Frangione B. Hereditary cerebral amyloid angiopathy: the amyloid fibrils contain a protein which is a variant of cystatin c, an inhibitor of lysosomal cysteine proteases. Biochem Biophys Res Commun 1986;136:548-554 24. Lofberg H, Grubb AO, Nilsson EK, et al. Irnmunohistochemical characterization of the amyloid deposits and quantitation of pertinent cerebrospinal fluid proteins in hereditary cerebral hemorrhage with amyloidosis. Stroke 1987;18:43 1-440 25. Hachinski VC, Potter P, Merskey H. Leuko-araiosis. Arch Neurol 1987;44:2 1-23 26. Steingart A, Hachinski VC, Lau C, et al. Cognitive and neurologic findings in subjects with diffuse white matter lucencies on computed tomographic scan (leuko-araiosis). Arch Neurol 1987;44:32-35 27. Englund E, Brun A, Alling C. White matter changes in dementia of Alzheimer’s type. Biochemical and neuropathological correlates. Brain 1988;111:1425-1439 28. Gray F, Dubas F, Roullet E, Escourolle R. Leukoencephalopathy in diffuse hemorrhagic cerebral amyloid angiopathy. Ann Neurol 1985;18:54-59 29. Duhas F, Gray F, Roullet E, Escourolle R. Leucoenc6phalopathies art6riopathiques (17 cas anatomo-cliniques). Rev Neurol (Paris) 1985;141:43-108 30. Hyman BT, Van Hoesen GW, Beyreuther K, Masters CL. A4 amyloid protein immunoreactiviry is present in Alzheimer’s disease neurofibrillary tangles. Neurosci Lett 1989;101:352-355 31. Bancher C, Grundke-Iqbal I, Iqbal K, et al. Immunoreactiviry of neuronal lipofuscin with monoclonal antibodies to the amyloid beta-protein. Neurobiol Aging 1988;lO:1 25-1 32 32. Card JP, Meade RP, Davis LG. Immunocytochemical localization of the precursor protein for beta-amyloid in the rat central nervous system. Neuron 1988;1:835-846 33. Babikian V, Ropper AH. Binswanger’s disease: A review. Stroke 1987;18:2-12 34. Fisher CM. Binswanger’s encephalopathy: a review. J Neurol 1989;236:65-79
Vinters et al: CAA Immunohistochemistry with Leukoencephalopathy
41
35. Ogomori K, Kitamoto T, Tateishi J, et al. Beta-protein amyloid is widely distributed in the cenval nervous system of patients with Alzheimer’s disease. Am J Pathol 1989;134:24%25 1 36. Ikeda S-I, Allsop D, Glenner GG. Morphology and distribution of plaque and related deposits in the brains of Alzheimer’s disease and control cases. An immunohistochemical study using amyloid beta-protein antibody. Lab Invest 1989;60:113122 37. Prelli F, Castafio EM, van Duinen SG, et al. Different processing of Alzheimer’s beta-protein precursor in the vessel wall of patients with hereditary cerebral hemorrhage with amyloidosisDutch type. Biochem Biophys Res Commun 1988;l51:11501155 38. Bird TD, Sumi SM, Nemens EJ, et al. Phenotypic heterogeneity in familial Alzheimer’s disease: a study of 24 kindreds. Ann Neurol 1989;2>:12-25 39. Yamada M, Tsukagoshi H, Wada Y ,et al. Absence of the cystatin C amyloid in the cerebral amyloid angiopathy, senile plaque, and extra-CNS amyloid deposits of aged Japanese. Acta Neurol Scand 1989;79:504-509 40. Ponte P, Gonzalez-DeWhitt P, Schihng J, et al. A new A4 amyloid mRNA contains a domain homologous to serine protease inhibitors. Nature 1988;331:525-527 41. Tanzi RE, McClatchey AI, Lamperti ED, et al. Protease inhibitor domain encoded by an amyloid protein precursor mRNA associated with Alzheimer’s disease. Nature 1988;331: 528-530 42. Abraham CR, Selkoe DJ, Potter H. Immunochemical identification of the serine protease inhibitor alpha,-antichymotrypsin in
42
Annals of Neurology
Vol 28 No 1 July 1990
43.
44.
4 5.
46.
47.
48.
49.
the brain amyloid deposits of Alzheimer‘s disease. Cell 1988; 52:487-501 Zucker-Franklin D, Warfel A, Grusky G, et al. Novel monocyte-like properties of microgliaVastroglial cells. Constitutive secretion of lysozyme and cystatin-C. Lab Invest 1987;57:176185 Vinters HV, Pardridge WM, Secor DL, et al. Vascular dementia: immunocytochemical study of brain microvascular and parenchymal lesions associated with Alzheimer’s disease. In: Ginsberg MD, Dietrich WD, eds. Cerebrovascular diseases. New York: Raven Press, 1989:213-220 Heffner RR Jr, Porro RS, Olson ME, Earle KM. A demyelinating disorder associated with cerebrovascular amyloid angiopathy. Arch Neurol 1976;33:501-506 Wisniewski HM, Wen GY, Kim KS. Comparison of four staining methods on the detection of neuritic plaques. Acca Neuropathol 1989;78:22-27 Davies L, Wolska B, Hilbich C, et al. A4 amyloid protein deposition and the diagnosis of Alzheimer’s disease: prevalence in aged brains determined by immunocytochemistry compared with conventional neuropathologic techniques. Neurology 1988;38:1688-1693 Braak H, Braak E, Kalus P. Alzheimer‘s disease: areal and laminar pathology in the occipital isocortex. Acta Neuropathol 1989;77:494-506 Maruyama K, Ikeda S-I, Ishikara T, et al. Immunohistochemical characterization of cerebrovascular amyloid in 46 autopsied cases using antibodies to beta protein and cystatin C . Srrokc 1990;21:397-403
