Distribution of precursor amyloid-beta-protein messenger RNA in ...

3 downloads 0 Views 2MB Size Report
Martins, R. N. & Beyreuther, K. (1985) EMBO J. 4,. 2757-2763. 5. Guiroy, D. C., Miyazaki, M., Multhaup, G., Fischer,P.,. Garruto, R. M., Beyreuther, K., Masters, ...
Proc. Natl. Acad. Sci. USA Vol. 85, pp. 1691-1695, March 1988 Neurobiology

Distribution of precursor amyloid-f8-protein messenger RNA in human cerebral cortex: Relationship to neurofibrillary tangles and neuritic plaques (Alzheimer disease/neocortex)

DAVID A. LEWIS*t, GERALD A. HIGGINSt, WARREN G. YOUNG*, DMITRY GOLDGABER§, D. CARLETON GAJDUSEK§, MICHAEL C. WILSON*, AND JOHN H. MORRISON*¶ Departments of *Basic and Clinical Research and tMolecular Biology, Research Institute of Scripps Clinic, La Jolla, CA 92037; and §Laboratory of Central Nervous System Studies, National Institute of Neurological and Communicative Disorders and Stroke, Bethesda, MD 20892

Communicated by Floyd E. Bloom, November 9, 1987

ABSTRACT Neurofibrillary tangles (NFT) and neuritic plaques (NP), two neuropathological markers of Alzheimer disease, may both contain peptide fragments derived from the human amyloid (3 protein. However, the nature of the relationship between NFT and NP and the source of the amyloid 13 proteins found in each have remained unclear. We used in situ hybridization techniques to map the anatomical distribution of precursor amyloid-fi-protein mRNA in the neocortex of brains from three subjects with no known neurologic disease and from five patients with Alzheimer disease. In brains from control subjects, positively hybridizing neurons were present in cortical regions and layers that contain a high density of neuropathological markers in Alzheimer disease, as well as in those loci that contain NP but few NFT. Quantitative analyses of in situ hybridization patterns within layers HI and V of the superior frontal cortex revealed that the presence of high numbers of NFT in Alzheimer-diseased brains was associated with a decrease in the number of positively hybridizing neurons compared to controls and Alzheimer-diseased brains with few NFT. In contrast, no correlation was found between the densities of NP and neurons containing precursor amyloid13-protein mRNA transcripts. These findings suggest that the expression of precursor amyloid-18-protein mRNA may be a necessary but is clearly not a sufficient prerequisite for NFW formation. In addition, these results may indicate that the amyloid (3 protein, present in NP in a given region or layer of cortex, is not derived from the resident neuronal cell bodies that express the mRNA for the precursor protein. The major neuropathological markers of Alzheimer disease (AD), neurofibrillary tangles (NFT), neuritic plaques (NP), and cerebrovascular amyloid deposits, have been reported to contain 4-kDa,8 proteins that share antigenic epitopes and a common amino acid sequence (1-5), although in the case of NFT these findings have been disputed (2, 6, 7). Synthetic oligonucleotide probes based on the amino acid sequence of the 4-kDa,8 protein have been used to identify cDNA clones for the 3.5-kilobase (kb) mRNA that encodes the human amyloid-p8-protein precursor (8-11). The importance of amyloid 8 protein to the primary pathological events of AD has been demonstrated by linkage studies indicating that a common restriction site polymorphism associated with the amyloid-pS-protein gene may be present in familial forms of AD (12). In a previous study using in situ hybridization techniques (13), we found that the mRNA encoding amyloid-p-protein precursor was present in neocortical neurons from the brains of both control subjects and patients with AD, suggesting The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

1691

that the protein precursor of the pathological amyloid material in AD is a natural constituent of certain cortical cells. In addition, in prefrontal cortex, the mRNA for this precursor protein appeared to be preferentially expressed by the large pyramidal neurons of layers III and V-neurons that are also prone to NFT formation in AD. In the present study, we extended our laminar and regional analyses in normal human cortex and found that although neurons prone to NFT formation generally contain high levels of precursor amyloid-,B-protein mRNA, in the healthy brain certain neuron types thought resistant to pathology also express this mRNA. In addition, we evaluated the anatomical relationships among cells containing precursor amyloid-13protein mRNA transcripts, cells containing NFT, and extracellular NP in AD brains. Our findings suggest that NFT develop in a subpopulation of neurons that contain precursor amyloid-p-protein mRNA in healthy brain and that the presence ofNFT is associated with a decrease in these transcripts. In contrast, no anatomical association between neurons containing precursor amyloid-/3-protein mRNA and NP was apparent in the cases examined.

MATERIALS AND METHODS Tissue Preparation. Brains from three adults (age range: 66-79 years) with no known neuropsychiatric disorders and from five patients (age range: 68-87 years) with a clinical diagnosis of AD were examined. The clinical diagnosis of AD was confirmed in each case by neuropathological evaluations that revealed atrophy and both NFT and NP in multiple regions of neocortex and the hippocampus. Brain samples were removed within 7 hr of death, placed in either 4% paraformaldehyde or 2% (wt/vol) paraformaldehyde/10 mM periodate/37.5 mM lysine for 48 hr, and then washed in a graded series of sucrose solutions (14). Tissue sections (20 Am) were cut in a cryostat, mounted on gelatin-coated slides, air-dried, and stored at either 40C for 1-3 days or at - 70'C for up to 6 weeks before use. Some hybridized sections and other 40-.um thick sections from the same tissue blocks were stained with thioflavin S, a fluorescent dye that visualizes NFT and NP (15), or with cresyl violet. Cytoarchitectonic regions were identified according to published criteria (16). The following Brodmann's areas were examined: area 9, superior frontal gyrus; area 4, precentral gyrus; area 17, primary visual cortex; area 41, primary auditory cortex; area 22, superior temporal gyrus; and area 20, inferior temporal gyrus. Abbreviations: NFT, neurofibrillary tangles; NP, neuritic plaques, AD, Alzheimer disease. tPresent address: Departments of Psychiatry and Behavioral Neuroscience, University of Pittsburgh, Pittsburgh, PA 15213. 1To whom reprint requests should be addressed.

1692

Neurobiology: Lewis et al.

In Situ Hybridization. Transcription of 35S-labeled probes was performed in 40 mM Tris*HCl (pH 7.5)/6 mM MgCl2/2 mM dithiothreitol/5 units of ribonuclease inhibitor (Promega Biotec, Masidon, WI)/400 juM adenosine triphosphate/400 ttM cytidine triphosphate/25 uM guanosine 5'-[a-35S]thiotriphosphate and of uridine 5'-[a-35S]thiotriphosphate (8001000 Ci/mmol, New England Nuclear; 1 Ci = 37 GBq)/1-2 ptg of linearized DNA template plus 5 to 10 units of SP6 RNA polymerase (Boehringer Mannheim) or T7 RNA polymerase (Stratagene Cloning Systems, San Diego, CA). Transcription was done at 370C for 1-2 hr. Incorporation was measured by trichloroacetic acid precipitation, with typical values of 20-50% incorporation, yielding 50-100 ng of probe with a specific activity greater than 2 x 109 cpm/,ug (13, 17). To obtain a probe for detection of precursor amyloid-p-

Proc. Natl. Acad. Sci. USA 85 (1988)

protein mRNA, a 1-kb EcoRI restriction fragment from the human brain cDNA clone AAm4 encoding amyloid ,B protein (8) was subcloned in the pGEM-3 plasmid (Promega Biotec). 17 RNA polymerase was used for generation of the antisense probe, and SP6 RNA polymerase was used for generation of the sense probe. In situ hybridization was done as described (14, 18). Briefly, slide-mounted sections were immersed sequentially in proteinase K (50 ,g/ml) (Boehringer Mannheim) and 0.05 M HCl and then dehydrated in graded alcohols. After treatment with prehybridization buffer [50% formamide/0.75 M NaCl/25 mM 1,4-piperazinediethanesulfonic acid (Pipes) buffer/25 mM EDTA/5 x Denhardt's solution (1 x Denhardt's solution is 0.02% bovine serum albumin, 0.02%

Ficoll, and 0.02% polyvinylpyrrolidone)/250 mM dithiop

I

11I

I

II III A

lil IVA

III B

IV B

IVC

IIIB

IV V

IV

V

VI

WM

V

VI

WM VI

Wm

FIG. 1. In situ hybridization of precursor amyloid-p-protein mRNA in superior temporal gyrus (area 22) (A), primary auditory cortex (area 41) (B), and primary visual cortex (area 17) (C) of control human brain. Roman numerals denote cortical layers. In area 41 (B), the columnar arrangement of positively hybridizing neurons parallels that of the pyramidal neurons in this region. In area 17 (C), note the low density of labeled neurons (despite a high density of total neurons) in layer IV compared with the adjacent layers. (Calibration bars = 100 ,um.)

Neurobiology: Lewis et al. threitol/0.2% NaDodSO4/10% (wt/vol) dextran sulfate/ denatured yeast RNA at 500 pug/ml/salmon sperm DNA at 500 tkg/ml], 5 ng of probe in 75 A.l of prehybridization buffer was applied to the slides. After overnight hybridization, sections were treated with pancreatic ribonuclease (50 Ag/ml) and then placed in a series of low-salt rinses at elevated temperature. The slides were exposed to x-ray film and then developed for emulsion autoradiography. Microscopy. Tissue sections were examined and photographed on a computer-assisted Zeiss ICM 405 microscope that was also used for quantitative analyses (14, 18). Grain clusters, indicative of cellular localization of the precursor amyloid-f3-protein mRNA, were viewed under dark-field illumination. The presence of grain clusters over neuronal cytoplasm was confirmed by examining hybridized sections counterstained with cresyl violet or by viewing hybridized sections under fluorescent illumination (Zeiss filter pack 487717), which allowed cortical neurons to be visualized due to autofluorescence. NFT and NP were visualized in thioflavine-stained sections with epifluorescence illumination using the Zeiss filter pack 487702. For quantitative studies, the distances from the pial surface to the borders of layers III and IV and layers IV and V were determined from adjacent Nissl-stained sections. These measurements were then used to locate the layer III (bottom boundary, layer III-IV border) and layer V (top boundary, layer IV-V border) fields on hybridized and thioflavine-S stained sections. The number of grain clusters or NFT and NP were then counted in each field (406 ,um x 583 um) at a magnification of x 160. For each case, counts were made in five different layer III and layer V fields, and the results were expressed as the mean (±+ SEM) number per 1.0 mm2 of cortex.

RESULTS Cortical sections from control human brain hybridized with the antisense strand complementary to precursor amyloidP-protein mRNA showed extensive cellular labeling as evidenced by clustered silver grains (Figs. 1 and 2). In counterstained sections, all grain clusters were located over cells with a size and appearance indicative of neurons (13). In contrast, as previously reported (13, 19) virtually no signal was detectable in adjacent sections of neocortex hybridized with the sense-strand probe. The specificity of cellular labeling with the antisense probe was also evident by the relative absence of grain clusters in regions of high neuronal density, such as layer IV of area 17, compared to adjacent layers with an overall lower neuronal density (see Fig. 1C).

Proc. Natl. Acad. Sci. USA 85 (1988)

1693

Cells labeled with the antisense probe were present in every cortical region examined, including primary sensory, sensory association, and motor regions. For example, the superior frontal (prefrontal association), superior temporal (auditory association), and inferior temporal (visual association) gyri, regions known to contain a high density of neuropathological markers in AD (14, 20, 21) had a large number of grain clusters in control brains (Figs. LA and 2A). In addition, primary motor and sensory regions, such as primary auditory (Fig. 1B) and primary visual (Fig. 1C) cortex, areas that contain NP but few NFT in AD (14, 20, 21), also had a high density of labeled cells in control brains. The laminar pattern was similar in all regions with minor shifts in the density of labeled neurons that correlated with the general cytoarchitectonic characteristics of the region. Few labeled neurons were present in the cell-poor layer I of every cortical region. In most regions, layer II had a particularly high density of small clusters of grains (Fig. 1). Layer III also had a high density of labeled neurons. In the superior and inferior portions of the temporal lobe, layer III was marked by a striking radial array of labeled neurons; this pattern was particularly prominent in primary auditory cortex (Fig. 1B), a region noted for the radial arrangement of its pyramidal neurons. With the exception of layer I, layer IV contained the lowest density of positively hybridizing neurons. In primary sensory regions, which have a highly developed layer IV, the relative absence of grain clusters in layer IV was particularly conspicuous in comparison to the adjacent layers III and V (Fig. 1 B and C). Layers V and VI had overall densities of labeled neurons similar to that of layer III. The superficial portion of layer V, like the deep portion of layer III, did have a high density of large grain clusters that presumably indicate the presence of precursor amyloid-f3-protein mRNA in the large pyramidal neurons typically found in these layers. The densities of positively hybridizing neurons, NFT, and NP were quantitatively analyzed in the superior frontal gyrus of both control and AD brains (Fig. 2 and Table 1). In the superior frontal gyrus (Brodmann's area 9), labeled cells were present in all cortical layers, and in layers III and V, the density of labeled cells was relatively high and uniform across the three control cases examined (cases 1-3 in Table 1). In the two AD cases in which the number of NFT was low (cases 4 and S in Table 1), the density of grain clusters in layers III and V was very similar to that of the controls. A high density of labeled cells was present even in case 4, which contained a very high density of NP. In contrast, the average number of labeled cells per mm2 of tissue (layer III, 85.9; layer V, 93.5) in the three AD brains with a high number of NFT (cases 6-8) was only 25% and 30%6, in layers

FIG. 2. In situ hybridization of precursor amyloid-o-protein mRNA (A and B) and thioflavine localization of NP and NFT (C) in layer III of normal (case 3) and AD (case 5) superior frontal gyrus. Note the higher density of grain clusters in the section from normal brain (A) as compared with AD brain (B). As is apparent in C, this AD case had numerous NP (filled arrow) and NFT (open arrow) in the superior frontal cortex. All three photographs are examples of material used for quantitative analysis (see Table 1). Calibration bar = 100 gm.

1694

Neurobiology: Lewis et al.

Proc. Natl. Acad. Sci. USA 85 (1988)

Table 1. Quantitative analysis of labeled cell density

Layer III Diag- Age, Case nosis yr AMY NFT NP 1 N 78 381.1

(24.5) 2

Layer V AMY 370.9

NFT

NP

(19.4)

317.7 319.4 (22.0) (9.3) 3 N 79 305.9 326.2 (12.7) (13.1) 4 AD 86 360.8 0.8 75.2 348.1 2.5 48.2 (27.9) (0.8) (6.3) (19.4) (0.8) (3.8) 87 322.8 5 AD 0.8 4.2 215.5 4.2 18.6 (32.1) (0.8) (1.3) (23.7) (1.7) (3.0) 71 103.9 131.0 38.0 173.2 123.4 22.0 6 AD (3.4) (14.4) (6.3) (13.9) (12.7) (3.0) 7 AD 82 83.7 87.9 10.1 44.8 152.1 10.1 (7.6) (5.5) (3.8) (7.2) (12.3) (2.5) 8 AD 68 70.1 30.4 44.8 62.5 77.7 21.1 (8.4) (7.2) (8.0) (5.1) (4.6) (3.8) Mean (± SEM) number per mm2 of neurons containing precursor amyloid-,3-protein mRNA transcripts (AMY) of NFT and of NP in layers III and V of human superior frontal cortex. Each entry represents the mean of five sections. Note that AMY counts were made in sections 20 jm thick, whereas NFT and NP were counted in sections 40 jim thick. NFT were not seen in cases 1-3, and NP were not present in cases 2 and 3 and were only rarely present in case 1; therefore, NFT and NP were not counted in the controls. N, neurologically normal. The postmortem intervals ranged from 1.3 to 7.0 hr. N

66

III and V, respectively, of the average number (layer III, 338.0; layer V, 315.6) in cases 1-5. Consistent with these observations, hybridized sections counterstained with thioflavine S revealed almost a complete dissociation between the location of grain clusters and NFT. That is, cells containing both precursor amyloid-f3-protein mRNA transcripts and NFT were rarely seen in AD brains. In fact, neurons with thioflavine S-positive NFT were conspicuous by their relative lack of grains.

DISCUSSION NFT and NP, two of the histological hallmarks of AD, have distinctive and consistent regional and laminar patterns of distribution in the neocortex of AD brains (14, 20, 21), suggesting that they are related to specific elements of cortical circuitry (14, 22). In addition, both NFT and NP may contain peptide fragments derived from the human precursor amyloid f protein (2-5) that is the product of a gene located on chromosome 21 (8-12). However, the nature of the relationship between NFT and NP and the source of the amyloid f3 proteins found in each have remained unclear. The mRNA encoding amyloid-3-protein precursor is present in cortical neurons in both control and AD brains. These cells are present in all regions of the neocortex, and their patterns of distribution in control brains exhibit both similarities and dissimilarities to those of NFT and NP in AD brains. For example, in the auditory and visual association cortices of the superior and inferior temporal lobes of normal human brain, mRNA for the amyloid-13-protein precursor is particularly abundant in the large pyramidal neurons of layers III and V, although the signal is also high in layers II and VI. In these same regions of AD brains, NFT are also predominantly located in layers III and V. However, in primary auditory and visual cortical regions, which rarely contain NTr in AD (14, 20), there are large numbers of neurons that contain amyloid-p3-protein precursor mRNA transcripts.

Primary and association visual cortices have comparable numbers of NP in AD and, on a qualitative basis, similar densities of amyloid-j3-protein precursor mRNA-containing cells. However, on a laminar basis, the distribution of NP and cells expressing amyloid-,f-protein precursor do not match. In primary visual cortex, for example, layer IV contains the greatest number of NP, whereas in control brains, this layer has a relatively low density of neurons expressing precursor amyloid-,8-protein mRNA, compared to layers of III, V, and VI. In the superior frontal gyrus (Brodmann's area 9) of control brains, neurons containing precursor amyloid-13protein mRNA are also present in high numbers in layers III and V. In these laminae in AD brains the number of labeled neurons appears inversely related to NFT number (see Table 1); that is, AD brains with a high number of NFT in layers III and V of superior frontal cortex (cases 6-8) have a low number of labeled neurons, and AD brains with a relatively low number of NFT (cases 4 and 5) have numbers of labeled neurons similar to those of controls (cases 1-3). This presence of NFT in association with a decrease in the number of neurons containing the mRNA transcripts suggests that NFT develop in neurons that normally express the precursor amyloid-f3-protein mRNA and that the development of NFT impairs further transcription of the precursor amyloid-,3protein gene, at least as is detectable by in situ hybridization. This interpretation is supported by the virtually complete absence in all AD cases examined of cells containing both grain clusters and NFT in hybridized sections counterstained with thioflavine S. However, whether this decrease in the number of neurons expressing precursor amyloid-f3protein mRNA in some AD brains reflects specific downregulation of this gene or a general decrease in neuronal viability remains to be determined. Thus, it appears that the expression of precursor amyloid(3-protein mRNA may be a necessary, but is clearly not a sufficient, prerequisite for NFT formation. Thus, although neurons that contain NFT in AD may have expressed this mRNA prior to the formation of NFT, many neurons that express this mRNA in the healthy brain never develop NFT during AD. Some other factor, or combination of factors, such as the presence of high intracellular levels of nonphosphorylated neurofilament protein (22, 23), a long axonal projection terminating in neocortex (14, 23), or a distinctive pattern of corticocortical connections (14, 20) may render a neuron vulnerable to deposition of amyloid (3 protein as a

pathological product.

Note also that differences in the expression of precursor

amyloid-,B-protein mRNA may exist among neuronal popu-

lations in AD. Although we saw a decrease in this mRNA in neurons of the superior frontal gyrus in some AD cases, other neurons, such as those in the parasubiculum (19), may contain increased levels of precursor amyloid-(3-protein

mRNA. In contrast with NFT, there is no apparent association between the densities of NP and neurons containing precursor amyloid-(3-protein mRNA transcripts. For example, both low (cases 6 and 8) and apparently normal (case 4) numbers of labeled neurons may be present in AD brains with a relatively high density of NP. Furthermore, AD brains with a relatively low density of NP may have either low (case 7) or normal (case 5) numbers of precursor amyloid-(3-protein mRNA-containing neurons. In addition to this dissociation of NP and labeled neurons in superior frontal gyrus, note also that the laminar distribution of NP in area 17 (14) differs markedly from that of precursor amyloid-(3-protein mRNA-

containing cells. These findings givenAregion suggest that within a or layer of cortex the amyloid (3 protein present in NP is not derived from the cells of that region or layer that express the mRNA

Neurobiology: Lewis et al. for the precursor protein. Instead, these findings are consistent with reports suggesting that NP are formed in the terminal axonal fields of neurons that develop NFT (4, 24, 25), such as long corticocortical projection neurons (14, 22) or cortically projecting brainstem neurons (26, 27), which, by inference, possessed the cellular machinery for producing amyloid 8 protein and transporting it down the axon. Alternatively, the amyloid 1 protein of NP may be derived from nonneuronal sources, such as plasma proteins (28, 29). In contrast, should amyloid p protein prove a major constituent of NFT, it probably accumulated from intraneuronal sources as a result of abnormalities in the synthesis, degradation, or transport of the protein or its precursor. We thank F. E. Bloom for consultation during the course of these studies, the Institute for Biogerontology Research (Sun City, AZ), Dr. C. Bouras (Geneva, Switzerland) for the donation of human postmortem tissue, and Nancy Delaney for manuscript preparation. This work was supported by National Institute of Mental Health Research Scientist Development Award MH00519, The Alzheimer Disease and Related Disorders Association, The MacArthur Foundation, the Hereditary Disease Foundation, and National Institutes of Health Grants AG06647, NS22347, NS23038, and CA33730. 1. Glenner, G. G. & Wong, C. W. (1984) Biochem. Biophys. Res. Commun. 120, 885-890. 2. Wong, C. W., Quaranta, V. & Glenner, G. G. (1985) Proc. Natl. Acad. Sci. USA 82, 8729-8732. 3. Masters, C. L., Simms, G., Weinman, N. A., Multhaup, G., McDonald, B. L. & Beyreuther, K. (1985) Proc. Natl. Acad. Sci. USA 82, 4245-4249. 4. Masters, C. L., Multhaup, G., Simms, G., Pottgiesser, J., Martins, R. N. & Beyreuther, K. (1985) EMBO J. 4, 2757-2763. 5. Guiroy, D. C., Miyazaki, M., Multhaup, G., Fischer, P., Garruto, R. M., Beyreuther, K., Masters, C. L., Simms, G., Gibbs, C. J. & Gajdusek, D. C. (1987) Proc. Natl. Acad. Sci. USA 84, 2073-2077. 6. Kosik, K. S., Joachim, C. L. & Selkoe, D. J. (1986) Proc. Natl. Acad. Sci. USA 83, 4044-4048. 7. Selkoe, D. J., Abraham, C. R., Podlisny, M. B. & Duffy, L. K. (1986) J. Neurochem. 146, 1820-1834. 8. Goldgaber, D., Lerman, M. I., McBride, W., Saffiotti, U. & Gajdusek, D. C. (1987) Science 235, 877-880. 9. Tanzi, R. E., Gusella, J. F., Watkins, P. C., Bruns, G. A. P., St. George-Hyslop, P., Van Keuren, M. L., Patterson, D., Pagan, S., Kurnit, D. M. & Neve, R. L. (1987) Science 235,

Proc. Natl. Acad. Sci. USA 85 (1988)

1695

880-884. 10. Kang, J., Lemaire, H. G., Unterbeck, A., Salbaum, J. M., Masters, C. L., Grzeschik, K. H., Beyreuther, K. & MullerHill, B. (1987) Nature (London) 325, 733-736. 11. Robakis, N. K., Ramakrishna, N.', Wolfe, G. & Wisniewski, H. M. (1987) Proc. Natl. Acad. Sci. USA'84, 4190-4194. 12. St. George-Hyslop, P., Tanzi, R. E., Polinsky, R. J., Haines, J. L., Nee, L., Watkins, P. C., Meyers, R. H., Feldman, R. G., Pollen, D., Drachman, D., Growdon, J., Bruni, A., Foncin, J. F., Salmon, D., Frommelt, P., Amaducci, L., Sorbi, S., Piacentini, S., Stewart, G. D., Hobbs, W. J., Con-

neally, P. M. & Gusella, J. F. (1987) Science 235, 885-890. 13. Bahmanyar, S., Higgins, G. H., Goldgaber, D., Lewis, D. A., Morrison, J. H., Wilson, M.'C., Shankar, S. K. & Gajdusek, D. C. (1987) Science 237, 77-80. 14. Lewis, D. A., Campbell, M. J., Terry, R. D. & Morrison, J. H. (1987) J. Neurosci. 7, 1799-1808. 15. Schwartz, P. (1972) Gerontologia 18, 321-362. 16. Brodmann, K. (1909) Lokalisationslehre Der Grosshirnrinde (Barth, Leipsiz, G.D.R.). 17. Higgins, G. A. & Wilson, M. C. (1987) in In Situ Hybridization: Applications to Neurobiology, eds. Valentino, K., Eberwine, J. & Barchas, J. (Oxford Univ. Press, New York), pp. 146-162. 18. Young, W. G., Morrison, J. H. & Bloom, F. E. (1985) Soc. Neurosci. Abstr. 11, 679. 19. Higgins, G. A., Lewis, D. A., Bahmanyar, S., Goldgaber, D., Gajdusek, D. C., Young, W. G., Morrison, J. H. & Wilson, M. C. (1988) Proc. Natl. Acad. Sci. USA 85, 1297-1301. 20. Pearson, R. C. A., Esiri, M. M., Hiorns, R. W., Wilcock, G. K. & Powell, T. P. S. (1985) Proc. Natl. Acad. Sci. USA 82, 4531-4534. 21. Kemper, T. (1984) in Clinical Neurology of Aging, ed. Albert, M. L. (Oxford Univ. Press, New York), pp. 9-52. 22. Morrison, J. H., Lewis, D. A. & Campbell, M. J. (1987) in Neurochemistry of Aging, eds. Daies, P. & Finch, C. (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY), in press. 23. Morrison, J. H., Lewis, D. A., Campbell, M. J., Huntley, G. W., Benson, D. L. & Bouras, C. (1987) Brain Res. 416, 331-336. 24. Suzuki, K. & Terry, R. D. (1967) Acta Neuropathol. 8, 276-284. 25. Gajdusek, D. C. (1985) N. Engl. J. Med. 312, 714-719. 26. Nakashima, S. & Ikuta, F. (1985) Acta Neuropathol. 66, 37-41. 27. Ishii, T. (1966) Acta Neuropathol. 6, 181-187. 28. Glenner, G. G., Wong, C. W., Quaranta, V. & Eanes, E. D. (1984) Appl. Pathol. 2, 357-369. 29. Selkoe, D. J. (1986) Neurobiol. Aging 7, 425-431.