- Email: [email protected]
The relationship between formation of senile plaques and neurofibrillary tangles and changes in nerve cell metabolism in Alzheimer type dementia
Mechanisms of Ageing and Development, 17 (1981) 395-401 © Elsevier Sequoia S.A., Lausanne - Printed in The Netherlands
395
THE RELATIONSHIP BETWEEN FORMATION OF SENILE PLAQUES AND NEUROFIBRILLARY TANGLES AND CHANGES IN NERVE CELL METABOLISM IN ALZHEIMER TYPE DEMENTIA
DAVID M. A. MANN* and PETER O. YATES Department of Pathology, University of Manchester, ManchesterM13 9PT (GreatBritain) (Received May 6, 1981 ; in revised form August 27, 1981)
SUMMARY Nerve cell nucleolar volume is reduced, in senile dementia of Alzheimer type, by 15-25% in nerve cells not containing neurofibrillary tangles and by over 35% in those which do contain such changes, in a wide variety of brain regions, when compared to similar cells from non-demented control cases, suggesting that interference with production of proteins may be an early consequence of the pathogenic process. The extent to which nucleolar volume is decreased in the non-tangle-bearing cells is related to frequencies of both neurofibrillary tangle and senile-plaque formation within that region, but the reduction in volume in the tangle-bearing cells correlated with neurofibrillary changes only. It seems, therefore, that the severity with which the dementing process affects an area of brain is initially shown by alterations in cell metabolism, which may invoke reductions in protein synthesis in non-tangle-bearing cells, and is later marked by the proportion of these affected cells which go on to form neurofibrillary tangles. Changes in nerve cell function do not seem to be as well indicated by the density of senile plaques within that area.
INTRODUCTION Although the pathological hallmarks of dementia of the Alzheimer type (AD), namely senile plaques (SPs), neurofibrillary tangles (NFTs) and granulovacuolar degeneration (GVD), have been reported on numerous occasions [1-4] to occur in significantly greater numbers in the brains of demented people than in non-demented people of similar age, there is still little evidence relating their presence to actual changes in nerve cell function, and so their precise contribution to the dementing process is uncertain. *All correspondence and reprint requests to: Dr. D. M. A. Mann, Department of Pathology, University of Manchester, Oxford Road, Manchester M13 9PT, Great Britain.
396 GVD is related to a progressive nerve cell degeneration in non-demented persons [5], but as this change in AD is mainly confined to pyramidal cells of the hippocampus, its relevance to the overall clinical picture is not clear. Abnormalities of dendritic and synaptic ultrastructure within SPs are well documented in AD [ 6 - 8 ] , but as yet it has not been possible to assess either the number of nerve cells affected by the plaque or the manner of their involvement. By contrast, the NFT occurs widely throughout the central nervous system in AD, and its clear intracellular location makes those cells affected by this type of change readily identifiable. The nucleolus has been shown [9] in nerve cells to vary in volume according to the level of metabolism within that cell, and may thus indicate changes in protein synthesis. It is measured in tangle-bearing and non-tangle-bearing nerve cells, in both cases of AD and aged controls, in regions of the brain where both SP and NFT formation are abundant and also in other areas where SP and/or NFT formation are less frequent. In this way, the relationship between the presence of these pathological features and changes in nerve cell metabolism is investigated and their relevance to the underlying disease process discussed.
METHODS Brains were obtained at autopsy from five female patients in the age range 7 7 - 8 5 years (mean age 81 years), who were all, in life, free from overt neurological or psychiatric illness, and also from three other female patients in the age range 7 9 - 8 3 years (mean age 81 years) diagnosed as suffering from AD. All eight brains were fixed in 10% neutral formalin and blocks were cut from standard areas of frontal, temporal, parietal, insular, motor, entorhinal and occipital cortex, amygdala, hypothalamus and mid-brain, at the level of the oculomotor nerve nucleus. Blocks were routinely processed and paraffin sections cut at 5/lm. These were stained for histological appraisal by conventional neuropathological techniques, including G r o s Bielchowsky (GB) silver staining. Measurements of SP, NFT and total nerve cell frequencies were made on the GB-stained sections in the AD cases using a "nearest neighbour" method [10], as we have described elsewhere [11]. Frequency of nerve cells, SP and NFTs is expressed as feature number per unit area of 1 mm 2. Measurements of nerve cell nucleolar diameter were made also on the same GBstained sections, again as described elsewhere [11]. A cell containing NFT was located and its nucleolar diameter measured. The diameter of the nearest nerve aell not bearing a tangle was then measured. From these values nucleolar volumes (V) were calculated from V -- zr/6D 3 (where D = diameter) and the difference in nucleolar volume between nontangle-bearing and tangle-bearing cells was determined for a minimum of 30 pairs of cells in each area. From these 30 values the overall mean difference between non-tangle-bearing and tangle-bearing cells was estimated and the significance of this mean difference tested by reference to t distribution.
397 Because plaques and tangles were infrequent in the "control" cases, no frequency measurements were made. These cases served, therefore, as age controls for values of nucleolar volume, which was measured, as above, with the mean values from each case being pooled and averaged for each area of the brain.
Cytological observations The five control cases showed only minimal amounts of cerebral softening and SPs and cells containing NFTs were rare. The AD cases typically showed widespread cortical distribution of numerous SPs and NFTs and severe GVD in neurones of the hippocampus, .all in the absence of significant vascular disease. However, the distribution of SPs and NFTs throughout the brain was not similar. Areas such as temporal, frontal, insular and entorhinal cortex and amygdala showed large numbers of both SPs and NFTs but both features seemed less frequent in parietal and motor areas. In the occipital (calcarine) cortex, SPs alone were numerous and NFTs only occasionally seen, but the converse was so in the hypothalamus and floor of aqueduct and IV ventricle (mid-brain) where tanglebearing cells were numerous but SPs were rare or even absent.
RESULTS Values of frequency of nerve cells, SPs and NFTs, in the individual AD cases, were pooled by area and averaged as shown in Table I. No significant correlation (r = 0.643, p > 0.05) was noted between these "raw" values of SP and NFT frequency within the various areas of brain where each was observed. This is because the differing packing densities of neurones in the various regions (Table I) leads to "raw" values of SP and NFT frequency in areas such as amygdala, with low neurone density, being "artificially" low, and others in areas with high neurone density, such as temporal cortex, being "artificially" high. Correction for packing density of neurones in each area enables comparisons of SP and NFT frequency, within the brain regions examined, to be made on an equivalent basis. "Adjusted" SP and NFT frequency is therefore expressed as feature number per 100 nucleolated nerve cells and it is these parameters that are later related to percentage changes in nucleolar volume of such nerve cells. With this correction, a significant linear correlation (r = 0.813, p < 0.01) was observed between adjusted SP frequency (x) and NFT frequency (y), the relationship being described by the equation y = 0.86x - 1.01. This kind of proportional relationship between SP and NFT formation has been reported elsewhere to occur within single areas of brain such as temporal cortex [11], frontal cortex, or hippocampus [12], in a series of demented patients, but this is the first time that such a relationship has been shown within all areas of the same brains. Individual mean values of nucleolar volume, of non-tangle-bearing and tanglebearing cells, in the AD cases, were also pooled by area and averaged. These, together with overall mean values of nucleolar volume within the control cases, are shown for each brain area in Table II.
398 TABLE I MEAN, RAW AND ADJUSTED FREQUENCIES OF SENILE PLAQUES, NEUROFIBRILLARY TANGLES AND NERVE CELLS, IN TEN AREAS OF BRAIN, MEASURED IN THREE CASES OF ALZHEIMER DEMENTIA Brain area
Raw frequency {n/rnrn 2) Senile plaques
Frontal cortex
18.2 ± 2.7*
Neurofibrillary tangles
Adjusted frequency (n per 100 nucleolated nerve cells) Nerve cells
Senile plaques
Neurofibrillary tangles
10.6±0.4
135.0± 3.4
13.5±1.7
7.9±0.5
Temporal cortex 51.9 ± 6.4
44.2±4.4
190.9± 7.9
27.0±2.3
23.0±1.5
Occipital cortex
32.3 ± 2.3
11.9±2.9
284.0±10.9
11.5±1.3
4.3±1.2
Parietal cortex
26.7 ± 5.6
11.4±1.6
180.9± 9.7
14.5±2.4
6.2±0.6
Insular cortex
47.0 ± 6.0
21.1±3.8
173.1± 8.6
27.0±2.1
12.2±2.1
Motor cortex
14.1 ± 3.1
8.6±1.5
164.4± 2.9
8.5±1.7
5.2±1.0
Entorhinal cortex
29.0 ± 3.6
19.9±2.2
163.2±11.4
17.7±1.4
12.1±0.6
Amygdala
42.7±3.0
46.8±4.2
28.0 ± 5.2
30.5±4.6
64.9± 7.5
Hypothalamus
0.2 ± 0.1
13.7±0.6
107.0± 3.5
Mid-brain
0
7.5±1.5
36.3± 4.0
0.2±0.04
13.0±1.4
0
20.4±1.9
*±S.E.
From these values, overall percentage reductions in nucleolar volume in non-tanglebearing and tangle-bearing cells o f the AD cases and the ratio of volumes between nontangle-bearing and tangle-bearing cells are calculated. The significance of differences in value o f overall mean nucleolar volume between non-tangle-bearing and tangle-bearing cells within the AD cases alone, and also between both non-tangle-bearing and tanglebearing cells in AD cases and the control cases, for each area, were tested by reference to t distribution. In all these comparisons, differences were found to be significant at, at least, the 5% probability level. That is to say, for each brain area considered, the mean nucleolar volume o f non-tangle-bearing cells in AD cases was significantly reduced when compared to similar cells in the aged control series; and, furthermore, in the AD cases alone, nucleolar volume of tangle-bearing cells was significantly reduced further when compared to their non-tangle-bearing counterparts. Percentage reduction in nucleolar volume in both non-tangle-bearing and tanglebearing cells, in all areas, was examined for correlation with either SP of NFT frequency. Reduction in nucleolar volume in non-tangle-bearing cells correlated significantly with both NFT (r = 0 . 8 9 9 , p ( 0 . 0 0 1 ) and SP (r = 0 . 7 2 8 , p ( 0 . 0 5 ) frequencies (Fig. 1). However, reduction in nucleolar volume in tangle-bearing cells correlated only with NFT frequency (Fig. 1) and even then less significantly (r = 0.862, p ( 0 . 0 5 ) . (r = 0.474, p > 0.05 for correlation with SP frequency.)
399
TABLE II MEAN V A L U E S O F N U C L E O L A R V O L U M E OF NON-TANGLE-BEARING AND TANGLE-BEARING N E R V E CELLS IN TEN A R E A S OF BRAIN, MEASURED IN T H R E E CASES OF AD AND IN FIVE C O N T R O L CASES (NON-TANGLE ONLY)
Brain area
Nucleolar volu me (#m 3)
Percentage reduction
Control
Alzheimer
Alzheimer Non-tangle
Tangle
Non-tangle
Tangle
Ratio of nucleolar volume
Frontal cortex
16.0 ± 0.9*
12.6 ± 0.9
9.2 ± 0.7
21.3
42.5
1.37
Temporal cortex
14.5 ± 0.8
9.8 ± 0.6
8.0 ± 0.6
32.4
44.8
1.23
4.9 ± 0.3
3.6 ± 0.2
3.0 ± 0.2
26.5
38.8
1.20
16.1 ± 1.2
13.1 ± 1.0
9.7 ± 0.6
18.6
39.8
1.35
Occipital cortex Parietal cortex Insular cortex
10.5 ± 0.7
8.2 ± 0.7
6.1 ± 0.4
22.9
41.9
1.34
Motor cortex
15.9 ± 1.0
13.4 ± 0.8
10.0 ± 0.7
15.7
37.1
1.34
Entorhinal cortex
13.9 ± 0.9
10.9 ± 0.8
8.6 ± 0.5
21.6
38.1
1.27
Amygdala
35.6 ± 2.1
21.3 ± 1.8
18.6 ± 1.5
40.2
47.8
1.15
H y p o t h a l a m u s - 29.3 ± 1.8
22.0 ± 2.1
15.9 ± 1.2
24.9
45.7
1.38
Mid-brain
30.8 ± 1.6
21.2 ± 1.4
29.0
51.2
1.45
43.4 ± 2.0
Values o f percentage reduction in nucleolar volume, in b o t h tangle- and non-tangle-bearing neurones o f the AD cases, together with ratio of volumes (non-tangle to tangle), are also shown. * ± S.E.
Neurofibrillary tangles
i° .S
•
| Senile plaques
'e i20,
Frequency
Fig. 1. Percentage reduction in nucleolar volume in non-tangle-bearing (o) and tangle-bearing (e) nerve cells in ten areas o f brain, plotted against frequency o f neurofibrillary tangles (top) or senile plaques (bottom), in three cases of AD examined at autopsy. Linear regressions, where significant, are shown.
400 DISCUSSION Nucleolar diameter has been reported [13] as being reduced in tangle-bearing cells of the temporal cortex relative to their non-tangle-bearing neighbours, in a series of 33 patients of whom some had suffered from AD. However, because the AD cases were not considered separately from their aged controls, these authors were unable to show that cells unaffected by tangles in AD were, in fact, "truly normal" and had not also been affected by the pathological process. In this. report we confirm that nucleolar size (volume) is reduced in tangle-bearing cells, relative to their non-tangle-bearing counterparts, and have shown that, in those nerve cells apparently unaffected by tangle formation, nucleolar volume is decreased by about 15-40% (as compared t o 40-50% shown by tanglebearing cells) and these cells must be considered to have been affected also by the disease process. The magnitude of the reduction in nucleolar volume in these non-tangle-bearing cells correlates strongly with both NFT and SP frequencies in that area of brain, whereas the reduction in tangle-bearing cells correlates only to a much lesser extent when related to NFT frequency, and not at all with SP frequency. Perhaps many of these latter cells affected by SP and NFT formation have reached a minimal value of nucleolar volume necessary to their survival and would, therefore, show no further reduction despite increased formation of NFTs and SPs in that region. Findings presented here, therefore, indicate that a widespread reduction in nucleolar volume in nerve cells is an early feature of the pathogenic process in AD which may result in alterations in the level of protein synthesis. Moreover, the severity with which this process affects a particular area is marked initially by the magnitude of the reduction in nucleolar volume in those nerve cells apparently unaffected by NFT or SP formation, but is later indicated by the percentage of those cells which undergo change in metabolism resulting in tangle formation (i.e. NFT density). Correlations between reduced nucleolar volume and frequency of SPs and NFTs may also imply causal pathogenic relationships. Changes in structure at synaptic level [5, 7, 14] involving neurotransmitter deficits could result in decreased requirements for protein in both tangle-bearing and non-tangle-bearing neurones, and displacement and disruption of intracellular membranes and organelles due to NFT accumulation may cause restrictions in protein synthesis or transport, or both. In this context findings of reductions in protein synthesis, of the order reported here, in Purkinje cells of the cerebellum and in neurones of the dentate and inferior olivary nuclei [15] (cell types which in AD never apparently form NFTs, with SPs being only occasionally seen in cerebellar cortex) are hard to reconcile unless it is assumed that such changes are secondary to others elsewhere in the central nervous system, resulting, perhaps, from a transsynaptic degeneration of nerve terminals. Thus any direct linkage between an early disordered protein synthesis with production of that abnormal protein that characterises NFT formation, must, at present, remain speculative. ACKNOWLEDGMENTS We wish to thank Miss A. Hoyle for the preparation of the manuscript.
401 REFERENCES 1 A. D. Dayan, Quantitative histological studies on the human aged brain. II. Senile plaques and neurofibrillary tangles in senile dementia. Acta Neuropathol., 16 (1970) 95-102. 2 B. E. Tomlinson, G. Blessed and M. Roth, Observations on the brains of demented old people. J. Neurol. Sci., 11 (1970) 205-242. 3 M. J. Ball, Neurofibrillary tangles and the pathogenesis of dementia. A quantitative study. Neuropathol. Appl. Neurobiol., 2 (1976) 395-410. 4 M. J. Ball, Topographic distribution of neurofibrillary tangles and granulovacuolar degeneration in hippocampal cortex of aging and demented patients. A quantitative study. Acta NeuropathoL, 42 (1978) 7 3 - 8 0 . 5 D. M. A. Mann, Granulovacuolar degeneration in pyramidal cells of the hippocampus. Acta Neuropathol., 42 (1978) 149-151. 6 M. Kidd, Alzheimer's disease--an electron microscope study. Brain, 87 (1964) 307-327. 7 R. D. Terry, Neuronal fibrous protein in human pathology. J. Neuropathol. Exp. Neurol., 30 (1971) 8 - 1 9 . 8 H. M. Wisniewski, H. K. Narang and R. D. Terry, Neurofibrillary tangles of paired helical filaments. J. Neurol. Sci., 27 (1976) 173-181. 9 W. E. Watson, Observations on the nucleolar and total cell body nucleic acid of injured nerve cells; J. Physiol. London, 196 (1968) 655-676. 10 W. A. Aherne and P. J. Diggle, The estimation of neuronal population density by a robust distance method. J. Microsc., 114 (1978) 285-293. 11 D. M. A. Mann, D. Neary, P. O. Yates, J. Lincoln, J. S. Snowden and P. Stanworth, NeurofibriUary pathology and protein synthetic capability in nerve cells in Alzheimer's disease. Neuropathol. Appl. NetirobioL, 7 (1981) 3 7 - 4 7 . 12 P. M. Farmer, A. Peck and R. D. Terry, Correlation among numbers of neuritic plaques, neurofibrillary tangles and severity of senile dementia. J. Neuropathol. Exp. Neurol., 35 (1976) 367. 13 A. D. Dayan and M. J. Ball, Histometric observations on the metabolism of tangle bearing neurones. J. Neurol. ScL, 19 ( 1 9 7 3 ) 4 3 3 - 4 3 6 . 14 N. H. Gonatas, A. Anderson and 1. Evangelista, The contribution of altered synapses in the senile plaque. An electron microscopy study in Alzheimer's disease. J. Neuropathol. Exp. Neurol., 25 (1967) 2 5 - 3 9 . 15 D. M. A. Mann and K. G. A. Sinclair, The quantitative assessment of lipofuscin pigment, cytoplasmic RNA and nucleolar volume in senile dementia. Neuropathol. Appl. NeurobioL, 4 (1978) 129-135.
395
THE RELATIONSHIP BETWEEN FORMATION OF SENILE PLAQUES AND NEUROFIBRILLARY TANGLES AND CHANGES IN NERVE CELL METABOLISM IN ALZHEIMER TYPE DEMENTIA
DAVID M. A. MANN* and PETER O. YATES Department of Pathology, University of Manchester, ManchesterM13 9PT (GreatBritain) (Received May 6, 1981 ; in revised form August 27, 1981)
SUMMARY Nerve cell nucleolar volume is reduced, in senile dementia of Alzheimer type, by 15-25% in nerve cells not containing neurofibrillary tangles and by over 35% in those which do contain such changes, in a wide variety of brain regions, when compared to similar cells from non-demented control cases, suggesting that interference with production of proteins may be an early consequence of the pathogenic process. The extent to which nucleolar volume is decreased in the non-tangle-bearing cells is related to frequencies of both neurofibrillary tangle and senile-plaque formation within that region, but the reduction in volume in the tangle-bearing cells correlated with neurofibrillary changes only. It seems, therefore, that the severity with which the dementing process affects an area of brain is initially shown by alterations in cell metabolism, which may invoke reductions in protein synthesis in non-tangle-bearing cells, and is later marked by the proportion of these affected cells which go on to form neurofibrillary tangles. Changes in nerve cell function do not seem to be as well indicated by the density of senile plaques within that area.
INTRODUCTION Although the pathological hallmarks of dementia of the Alzheimer type (AD), namely senile plaques (SPs), neurofibrillary tangles (NFTs) and granulovacuolar degeneration (GVD), have been reported on numerous occasions [1-4] to occur in significantly greater numbers in the brains of demented people than in non-demented people of similar age, there is still little evidence relating their presence to actual changes in nerve cell function, and so their precise contribution to the dementing process is uncertain. *All correspondence and reprint requests to: Dr. D. M. A. Mann, Department of Pathology, University of Manchester, Oxford Road, Manchester M13 9PT, Great Britain.
396 GVD is related to a progressive nerve cell degeneration in non-demented persons [5], but as this change in AD is mainly confined to pyramidal cells of the hippocampus, its relevance to the overall clinical picture is not clear. Abnormalities of dendritic and synaptic ultrastructure within SPs are well documented in AD [ 6 - 8 ] , but as yet it has not been possible to assess either the number of nerve cells affected by the plaque or the manner of their involvement. By contrast, the NFT occurs widely throughout the central nervous system in AD, and its clear intracellular location makes those cells affected by this type of change readily identifiable. The nucleolus has been shown [9] in nerve cells to vary in volume according to the level of metabolism within that cell, and may thus indicate changes in protein synthesis. It is measured in tangle-bearing and non-tangle-bearing nerve cells, in both cases of AD and aged controls, in regions of the brain where both SP and NFT formation are abundant and also in other areas where SP and/or NFT formation are less frequent. In this way, the relationship between the presence of these pathological features and changes in nerve cell metabolism is investigated and their relevance to the underlying disease process discussed.
METHODS Brains were obtained at autopsy from five female patients in the age range 7 7 - 8 5 years (mean age 81 years), who were all, in life, free from overt neurological or psychiatric illness, and also from three other female patients in the age range 7 9 - 8 3 years (mean age 81 years) diagnosed as suffering from AD. All eight brains were fixed in 10% neutral formalin and blocks were cut from standard areas of frontal, temporal, parietal, insular, motor, entorhinal and occipital cortex, amygdala, hypothalamus and mid-brain, at the level of the oculomotor nerve nucleus. Blocks were routinely processed and paraffin sections cut at 5/lm. These were stained for histological appraisal by conventional neuropathological techniques, including G r o s Bielchowsky (GB) silver staining. Measurements of SP, NFT and total nerve cell frequencies were made on the GB-stained sections in the AD cases using a "nearest neighbour" method [10], as we have described elsewhere [11]. Frequency of nerve cells, SP and NFTs is expressed as feature number per unit area of 1 mm 2. Measurements of nerve cell nucleolar diameter were made also on the same GBstained sections, again as described elsewhere [11]. A cell containing NFT was located and its nucleolar diameter measured. The diameter of the nearest nerve aell not bearing a tangle was then measured. From these values nucleolar volumes (V) were calculated from V -- zr/6D 3 (where D = diameter) and the difference in nucleolar volume between nontangle-bearing and tangle-bearing cells was determined for a minimum of 30 pairs of cells in each area. From these 30 values the overall mean difference between non-tangle-bearing and tangle-bearing cells was estimated and the significance of this mean difference tested by reference to t distribution.
397 Because plaques and tangles were infrequent in the "control" cases, no frequency measurements were made. These cases served, therefore, as age controls for values of nucleolar volume, which was measured, as above, with the mean values from each case being pooled and averaged for each area of the brain.
Cytological observations The five control cases showed only minimal amounts of cerebral softening and SPs and cells containing NFTs were rare. The AD cases typically showed widespread cortical distribution of numerous SPs and NFTs and severe GVD in neurones of the hippocampus, .all in the absence of significant vascular disease. However, the distribution of SPs and NFTs throughout the brain was not similar. Areas such as temporal, frontal, insular and entorhinal cortex and amygdala showed large numbers of both SPs and NFTs but both features seemed less frequent in parietal and motor areas. In the occipital (calcarine) cortex, SPs alone were numerous and NFTs only occasionally seen, but the converse was so in the hypothalamus and floor of aqueduct and IV ventricle (mid-brain) where tanglebearing cells were numerous but SPs were rare or even absent.
RESULTS Values of frequency of nerve cells, SPs and NFTs, in the individual AD cases, were pooled by area and averaged as shown in Table I. No significant correlation (r = 0.643, p > 0.05) was noted between these "raw" values of SP and NFT frequency within the various areas of brain where each was observed. This is because the differing packing densities of neurones in the various regions (Table I) leads to "raw" values of SP and NFT frequency in areas such as amygdala, with low neurone density, being "artificially" low, and others in areas with high neurone density, such as temporal cortex, being "artificially" high. Correction for packing density of neurones in each area enables comparisons of SP and NFT frequency, within the brain regions examined, to be made on an equivalent basis. "Adjusted" SP and NFT frequency is therefore expressed as feature number per 100 nucleolated nerve cells and it is these parameters that are later related to percentage changes in nucleolar volume of such nerve cells. With this correction, a significant linear correlation (r = 0.813, p < 0.01) was observed between adjusted SP frequency (x) and NFT frequency (y), the relationship being described by the equation y = 0.86x - 1.01. This kind of proportional relationship between SP and NFT formation has been reported elsewhere to occur within single areas of brain such as temporal cortex [11], frontal cortex, or hippocampus [12], in a series of demented patients, but this is the first time that such a relationship has been shown within all areas of the same brains. Individual mean values of nucleolar volume, of non-tangle-bearing and tanglebearing cells, in the AD cases, were also pooled by area and averaged. These, together with overall mean values of nucleolar volume within the control cases, are shown for each brain area in Table II.
398 TABLE I MEAN, RAW AND ADJUSTED FREQUENCIES OF SENILE PLAQUES, NEUROFIBRILLARY TANGLES AND NERVE CELLS, IN TEN AREAS OF BRAIN, MEASURED IN THREE CASES OF ALZHEIMER DEMENTIA Brain area
Raw frequency {n/rnrn 2) Senile plaques
Frontal cortex
18.2 ± 2.7*
Neurofibrillary tangles
Adjusted frequency (n per 100 nucleolated nerve cells) Nerve cells
Senile plaques
Neurofibrillary tangles
10.6±0.4
135.0± 3.4
13.5±1.7
7.9±0.5
Temporal cortex 51.9 ± 6.4
44.2±4.4
190.9± 7.9
27.0±2.3
23.0±1.5
Occipital cortex
32.3 ± 2.3
11.9±2.9
284.0±10.9
11.5±1.3
4.3±1.2
Parietal cortex
26.7 ± 5.6
11.4±1.6
180.9± 9.7
14.5±2.4
6.2±0.6
Insular cortex
47.0 ± 6.0
21.1±3.8
173.1± 8.6
27.0±2.1
12.2±2.1
Motor cortex
14.1 ± 3.1
8.6±1.5
164.4± 2.9
8.5±1.7
5.2±1.0
Entorhinal cortex
29.0 ± 3.6
19.9±2.2
163.2±11.4
17.7±1.4
12.1±0.6
Amygdala
42.7±3.0
46.8±4.2
28.0 ± 5.2
30.5±4.6
64.9± 7.5
Hypothalamus
0.2 ± 0.1
13.7±0.6
107.0± 3.5
Mid-brain
0
7.5±1.5
36.3± 4.0
0.2±0.04
13.0±1.4
0
20.4±1.9
*±S.E.
From these values, overall percentage reductions in nucleolar volume in non-tanglebearing and tangle-bearing cells o f the AD cases and the ratio of volumes between nontangle-bearing and tangle-bearing cells are calculated. The significance of differences in value o f overall mean nucleolar volume between non-tangle-bearing and tangle-bearing cells within the AD cases alone, and also between both non-tangle-bearing and tanglebearing cells in AD cases and the control cases, for each area, were tested by reference to t distribution. In all these comparisons, differences were found to be significant at, at least, the 5% probability level. That is to say, for each brain area considered, the mean nucleolar volume o f non-tangle-bearing cells in AD cases was significantly reduced when compared to similar cells in the aged control series; and, furthermore, in the AD cases alone, nucleolar volume of tangle-bearing cells was significantly reduced further when compared to their non-tangle-bearing counterparts. Percentage reduction in nucleolar volume in both non-tangle-bearing and tanglebearing cells, in all areas, was examined for correlation with either SP of NFT frequency. Reduction in nucleolar volume in non-tangle-bearing cells correlated significantly with both NFT (r = 0 . 8 9 9 , p ( 0 . 0 0 1 ) and SP (r = 0 . 7 2 8 , p ( 0 . 0 5 ) frequencies (Fig. 1). However, reduction in nucleolar volume in tangle-bearing cells correlated only with NFT frequency (Fig. 1) and even then less significantly (r = 0.862, p ( 0 . 0 5 ) . (r = 0.474, p > 0.05 for correlation with SP frequency.)
399
TABLE II MEAN V A L U E S O F N U C L E O L A R V O L U M E OF NON-TANGLE-BEARING AND TANGLE-BEARING N E R V E CELLS IN TEN A R E A S OF BRAIN, MEASURED IN T H R E E CASES OF AD AND IN FIVE C O N T R O L CASES (NON-TANGLE ONLY)
Brain area
Nucleolar volu me (#m 3)
Percentage reduction
Control
Alzheimer
Alzheimer Non-tangle
Tangle
Non-tangle
Tangle
Ratio of nucleolar volume
Frontal cortex
16.0 ± 0.9*
12.6 ± 0.9
9.2 ± 0.7
21.3
42.5
1.37
Temporal cortex
14.5 ± 0.8
9.8 ± 0.6
8.0 ± 0.6
32.4
44.8
1.23
4.9 ± 0.3
3.6 ± 0.2
3.0 ± 0.2
26.5
38.8
1.20
16.1 ± 1.2
13.1 ± 1.0
9.7 ± 0.6
18.6
39.8
1.35
Occipital cortex Parietal cortex Insular cortex
10.5 ± 0.7
8.2 ± 0.7
6.1 ± 0.4
22.9
41.9
1.34
Motor cortex
15.9 ± 1.0
13.4 ± 0.8
10.0 ± 0.7
15.7
37.1
1.34
Entorhinal cortex
13.9 ± 0.9
10.9 ± 0.8
8.6 ± 0.5
21.6
38.1
1.27
Amygdala
35.6 ± 2.1
21.3 ± 1.8
18.6 ± 1.5
40.2
47.8
1.15
H y p o t h a l a m u s - 29.3 ± 1.8
22.0 ± 2.1
15.9 ± 1.2
24.9
45.7
1.38
Mid-brain
30.8 ± 1.6
21.2 ± 1.4
29.0
51.2
1.45
43.4 ± 2.0
Values o f percentage reduction in nucleolar volume, in b o t h tangle- and non-tangle-bearing neurones o f the AD cases, together with ratio of volumes (non-tangle to tangle), are also shown. * ± S.E.
Neurofibrillary tangles
i° .S
•
| Senile plaques
'e i20,
Frequency
Fig. 1. Percentage reduction in nucleolar volume in non-tangle-bearing (o) and tangle-bearing (e) nerve cells in ten areas o f brain, plotted against frequency o f neurofibrillary tangles (top) or senile plaques (bottom), in three cases of AD examined at autopsy. Linear regressions, where significant, are shown.
400 DISCUSSION Nucleolar diameter has been reported [13] as being reduced in tangle-bearing cells of the temporal cortex relative to their non-tangle-bearing neighbours, in a series of 33 patients of whom some had suffered from AD. However, because the AD cases were not considered separately from their aged controls, these authors were unable to show that cells unaffected by tangles in AD were, in fact, "truly normal" and had not also been affected by the pathological process. In this. report we confirm that nucleolar size (volume) is reduced in tangle-bearing cells, relative to their non-tangle-bearing counterparts, and have shown that, in those nerve cells apparently unaffected by tangle formation, nucleolar volume is decreased by about 15-40% (as compared t o 40-50% shown by tanglebearing cells) and these cells must be considered to have been affected also by the disease process. The magnitude of the reduction in nucleolar volume in these non-tangle-bearing cells correlates strongly with both NFT and SP frequencies in that area of brain, whereas the reduction in tangle-bearing cells correlates only to a much lesser extent when related to NFT frequency, and not at all with SP frequency. Perhaps many of these latter cells affected by SP and NFT formation have reached a minimal value of nucleolar volume necessary to their survival and would, therefore, show no further reduction despite increased formation of NFTs and SPs in that region. Findings presented here, therefore, indicate that a widespread reduction in nucleolar volume in nerve cells is an early feature of the pathogenic process in AD which may result in alterations in the level of protein synthesis. Moreover, the severity with which this process affects a particular area is marked initially by the magnitude of the reduction in nucleolar volume in those nerve cells apparently unaffected by NFT or SP formation, but is later indicated by the percentage of those cells which undergo change in metabolism resulting in tangle formation (i.e. NFT density). Correlations between reduced nucleolar volume and frequency of SPs and NFTs may also imply causal pathogenic relationships. Changes in structure at synaptic level [5, 7, 14] involving neurotransmitter deficits could result in decreased requirements for protein in both tangle-bearing and non-tangle-bearing neurones, and displacement and disruption of intracellular membranes and organelles due to NFT accumulation may cause restrictions in protein synthesis or transport, or both. In this context findings of reductions in protein synthesis, of the order reported here, in Purkinje cells of the cerebellum and in neurones of the dentate and inferior olivary nuclei [15] (cell types which in AD never apparently form NFTs, with SPs being only occasionally seen in cerebellar cortex) are hard to reconcile unless it is assumed that such changes are secondary to others elsewhere in the central nervous system, resulting, perhaps, from a transsynaptic degeneration of nerve terminals. Thus any direct linkage between an early disordered protein synthesis with production of that abnormal protein that characterises NFT formation, must, at present, remain speculative. ACKNOWLEDGMENTS We wish to thank Miss A. Hoyle for the preparation of the manuscript.
401 REFERENCES 1 A. D. Dayan, Quantitative histological studies on the human aged brain. II. Senile plaques and neurofibrillary tangles in senile dementia. Acta Neuropathol., 16 (1970) 95-102. 2 B. E. Tomlinson, G. Blessed and M. Roth, Observations on the brains of demented old people. J. Neurol. Sci., 11 (1970) 205-242. 3 M. J. Ball, Neurofibrillary tangles and the pathogenesis of dementia. A quantitative study. Neuropathol. Appl. Neurobiol., 2 (1976) 395-410. 4 M. J. Ball, Topographic distribution of neurofibrillary tangles and granulovacuolar degeneration in hippocampal cortex of aging and demented patients. A quantitative study. Acta NeuropathoL, 42 (1978) 7 3 - 8 0 . 5 D. M. A. Mann, Granulovacuolar degeneration in pyramidal cells of the hippocampus. Acta Neuropathol., 42 (1978) 149-151. 6 M. Kidd, Alzheimer's disease--an electron microscope study. Brain, 87 (1964) 307-327. 7 R. D. Terry, Neuronal fibrous protein in human pathology. J. Neuropathol. Exp. Neurol., 30 (1971) 8 - 1 9 . 8 H. M. Wisniewski, H. K. Narang and R. D. Terry, Neurofibrillary tangles of paired helical filaments. J. Neurol. Sci., 27 (1976) 173-181. 9 W. E. Watson, Observations on the nucleolar and total cell body nucleic acid of injured nerve cells; J. Physiol. London, 196 (1968) 655-676. 10 W. A. Aherne and P. J. Diggle, The estimation of neuronal population density by a robust distance method. J. Microsc., 114 (1978) 285-293. 11 D. M. A. Mann, D. Neary, P. O. Yates, J. Lincoln, J. S. Snowden and P. Stanworth, NeurofibriUary pathology and protein synthetic capability in nerve cells in Alzheimer's disease. Neuropathol. Appl. NetirobioL, 7 (1981) 3 7 - 4 7 . 12 P. M. Farmer, A. Peck and R. D. Terry, Correlation among numbers of neuritic plaques, neurofibrillary tangles and severity of senile dementia. J. Neuropathol. Exp. Neurol., 35 (1976) 367. 13 A. D. Dayan and M. J. Ball, Histometric observations on the metabolism of tangle bearing neurones. J. Neurol. ScL, 19 ( 1 9 7 3 ) 4 3 3 - 4 3 6 . 14 N. H. Gonatas, A. Anderson and 1. Evangelista, The contribution of altered synapses in the senile plaque. An electron microscopy study in Alzheimer's disease. J. Neuropathol. Exp. Neurol., 25 (1967) 2 5 - 3 9 . 15 D. M. A. Mann and K. G. A. Sinclair, The quantitative assessment of lipofuscin pigment, cytoplasmic RNA and nucleolar volume in senile dementia. Neuropathol. Appl. NeurobioL, 4 (1978) 129-135.