close

Вход

Забыли?

вход по аккаунту

?

Chromatin structure in dementia.

код для вставкиСкачать
Chromatin Structure in Dementia
D. R. Crapper McLachlan, MD, FRCP(C),* P. N. Lewis, PhD,t W. J. Lukiw, BSc," A. Sima, MD, FRCP(C),S
C. Bergeron, MD, FRCP(C),$ and U. De Boni, PhD"
~~
~~~~~-
Nuclei extracted from neocortex of patients with Alzheimer's disease and treated with micrococcal nuclease release a
population of dinucleosomes that contain an increase in the linker histones H1° and Hl". Five other degenerative
brain diseases that clinically resemble Alzheimer's disease do not result in these changes, although Pick's disease is
associated with an increase in H1 on dinucleosomes. Histones from nuclei of patients with Alzheimer's disease are also
more resistant to salt-induced release from chromatin than are those from age-matched control subjects. These results
support the hypothesis that an alteration in chromatin structure is a marker for Alzheimer's disease.
McLachlan DRC, Lewis PN, Lukiw WJ, Sima A, Bergeron C, De Boni U: Chromatin structure in dementia.
A n n Neurol 15:329-334, 1984
Normal brain function requires availability of genetic
information for transcription into messenger ribonucleic acid (RNA). In brain tissue of patients who died
with advanced Alzheimer's disease, the proportion of
the total deoxyribonucleic acid (DNA) extracted from
neocortex found in chromatin that is highly condensed
is greater than in age-matched controls. Two independent techniques have revealed altered chromatin conformation: mechanical shearing with ultrasound followed by separation of chromatin constituents by
centrifugation [4f and analysis of the kinetics of
chromatin digestion utilizing the linker D N A specificity of the enzyme micrococcal nuclease IS}. Both
studies revealed a reduction in the fraction of D N A
(i.e., euchromatin) that is usually considered accessible
to RNA polymerase. Probably less than 5% of brain
D N A is transcribed into structural proteins, and it has
not yet been established whether any functionally important genetic information actually resides within the
segments of D N A that have become inaccessible in
Alzheimer's disease. Nevertheless, it is likely that transcription processes are altered in Alzheimer's disease,
because both FWA content and nucleolar size [9, 151
are reduced in neurons from brains of patients affected
with Alzheimer's disease. Neurons with neurofibrillary
degeneration are reduced particularly, compared with
those of age-matched controls [b].
Brief digestion of bulk chromatin by micrococcal
nuclease has been reported to digest preferentially regions of chromatin containing active genes [ l f . Because this enzyme cuts D N A at the linker region between nucleosomes, a population of oligonucleosomes
is released. Previous work has revealed a difference in
the linker region histones of dinucleosomes in Alzheimer-affected and control neocortex [8). The present work reports on a complement of linker proteins
associated with dinucleosomes that is unique to Alzheimer's disease. Five other degenerative brain conditions that clinically resemble Alzheimer's disease do
not result in these changes. In addition, the affinity of
linker histones for chromatin was examined and was
found to be higher in preparations of neocortical nuclei
from patients with Alzheimer's disease than in preparations from age-matched controls. The results indicate
that Alzheimer's disease is associated with alterations in
chromatin structure.
Methods
Tissue
Human brains were bisected in the sagittal plane; one-half of
each brain was fixed in formalin, and the other was frozen at
-90°C. O n the basis of the clinical history and extensive
histopathological examination, the brains were divided into
three groups. The first, a control group, was composed of
nine brains of patients whose average age at death was 67
years (range, 47 to 88 years). Control patients had no clinical
history of intellectual impairment or neurological disease,
and the neocortex of each was free of neurofibrillary degeneration. An Alzheimer group consisted of nine brains of patients whose average age at death was 72 years (range, 66 to
88 years). These brains had senile plaques and numerous
tangles in the neocortex together with marked cortical atrophy, and had been obtained from patients with a history
consistent with the diagnosis of Alzheimer's disease. The
brains of a third group of eight patients with dementia caused
-
From the Departments of *Physiology and Medicine, tBiochemistry,
and $Pathology, Medical Sciences Building, University of Toronto,
Toronto, Canada MSS IA8.
Address reprint requests to Dr McLachlan, Neurology Division,
11-2 13 Eaton Wing, Toronto General Hospital, Toronto, Ontario,
Canada MSG IL7.
Received Mav 17. 1983. and in revised form Sept 6. Accepted for
publication Sept 11, 1083.
329
Table 1 . Histopathological Diagnoses in Control Group 2:
Patients with Non-Alzheimer Dementiu
Patient
No.
Age at
Death (yr)
88
68
62
4
67
5
78
6
7
62
65
8
66
Diagnosis
Multi-infarct dementia
Pick's disease
Picks disease and Alzheimer's
disease
15 nm brainstem neurofibrillary
degeneration
15 nm brainstem neurohbrillary
degeneration
Lewy body encephalopathy
Severe atrophy; dementia
unclassified
Severe atrophy; dementia
unclassified
by a disease other than Alzheimer's disease constituted an
important control group. The histopathological diagnoses for
these patients are given in Table 1.
To examine the binding of histones to DNA, aliquots of
isolated neocortical nuclei were exposed to concentrations
of sodium chloride ranging from 0.10 to 0.75 M. Volumes of
nuclei of 100 pl were suspended in 10 mM of Tris hydrochloric acid, 0.25 M sucrose, 1.5 mM of calcium chloride, 1
mM of PMSF at pH 7.3. Each aliquot contained approximately I50 pg of DNA. After centrifugation at 13.5 Kg av
for three minutes, the supernatant was removed and the pellet exposed to 100 pI NaCl for 2 hours at 4°C with repeated
vortexing. The suspension of nuclei was centrifuged at 13.5
Kg av, and the proteins in the supernatant were precipitated
with fifteen volumes of 20% trichloroacetic acid (TCA) and
iced at 0 to 4°C for 20 hours. The proteins were collected by
treating the 13.5 K g av centrifuged TCA precipitates with
two washes of acetone-O.l% hydrochloric acid and two
washes of dry acetone. Resulting pellets were dried and then
dissolved in 50 pI of tracking buffer containing 30% glycerol, 10 mM of Tris hydrochloric acid at p H 7.5, 0.05%
bromophenol blue, lq,SDS, 5 mM of EDTA, 0.5p ( v h ) 6mercaptoethanol, and 1 mM of PMSF. Typically, 25 pl was
analyzed on a 1.5 mm thick 1 8 9 Tris-glycine-SDS slab gel
system. The amount of linker histone in each aliquot was
compared with the maximum amount extractable, which was
found with salt concentrations greater than 0.75 M sodium
chloride.
Biochemical Procedures
Mixed nuclei from both neurons and glia of frozen, unfixed
cerebral cortex were isolated in sucrose solutions by methods
previously reported [ 8 ] . Neuron and glia enriched fractions
were separated from total nuclear preparations 1141. A suspension of nuclei was exposed to micrococcal nuclease 181,
and the digested nuclei were extracted five times with 10 mM
of ethylenediaminetetraacetate (EDTA), 1 mM of phenylmethanesulfonylfluoride (PMSF), at pH 7.2. The extracts
were pooled and concentrated to about 1 ml using an Amicon Diaflow PM 10 ultrafiltration membrane. The D N A content of this concentrate was assayed according to the method
of Burton or the Hoechst 33258 fluorometric method of
Cesarone and colleagues 12). Examination of the dinucleosome proteins was achieved by loading 0.6 A26,, units of
nuclease digest onto a 0.3 cm x 10 cm 3.5% acrylamideTris-cacodylate EDTA tube gel and electrophoresing for 50
minutes at 200 volts and 3 mA. The gel was stained with
ethidium bromide, and the dinucleosome band was excised
and refitted into a 0.3 x 10 cm 18CF Tris-glycine-sodium
dodecyl sulfate (SDS) tube gel and electrophoresed at 150
volts, 2.5 mA per tube for 3.5 hours. The tube gels were
stained with Coomassie brilliant blue R, destained, and
scanned on a Canaico Model G Scanning Densitometer at
530 nm. Peak amplitude ratios of histones H 1 to H4 were
determined. To validate the numerical accuracy of the ratios
reported, the Bio-Rad protein assay was employed to measure the standard amounts of proteins loaded on the gels.
Various ratios of H 1 to H 4 were separated on 18% polyacrylamide gels and stained with Coomassie blue R. In the
range of the 5 pg H1 loading routinely employed in this
study, photometric densitometer readings revealed staining
to be linear in relation to the amount of HI loaded. At H I /
H4 ratios of 0.5, assays based on densitometer readings were
within 6 p of the amount as measured by the Bio-Rad
method.
330 Annals of Neurology
Vol 15 No 4
April 1984
Results
Dinucleosome Linker Proteins
Proteins separated from dinucleosomes released by micrococcal nuclease digestion of nuclei extracted from
human neocortex are shown in Figures 1, 2, and 3. Five
bands were usually observed, with apparent molecular
weights on 18% SDS polyacrylamide gels of 31.6,
30.0,28.9,26.6, and 2 5 3 K daltons. These proteins are
labeled H l a , H l b , Hl", HloO,and H l D , respectively
[S]. All of these 5% perchloric acid-soluble basic proteins are included in the H 1 family of histones because
amino acid analyses have revealed high lysine and
alanine contents and an overall composition closely resembling that reported for mammalian H1 proteins C3,
7f. Histone H1" is the only linker histone known to
contain methionine C11, 133. Peptide chains containing
methionine are cleaved at that residue following exposure to cyanogen bromide. As shown in Figure 2, band
3 is selectively degraded following cyanogen bromide
treatment, which indicates that band 3 contains Hl".
Based on mobility, band 4 is tentatively considered to
be Hl"" [ 101; this peptide was not cleaved by cyanogen
bromide.
To study possible laboratory degradation, nuclei
were incubated for periods of up to 100 hours at 37,
20, and 4°C. Bands 3 and 4 revealed no degradation
even when protease inhibitors were omitted; bands 1
and 2 decreased and band 5 increased in stain intensity,
indicating that band 5 (HlD) is a degradation product
of H1.
In preparations from seven control brains, the ratios
of H1, Hl", and Hl"" to the core histone H4, when
1.
- CNBr
+ CNBr
F i g I . Eighteen percent Tris-glycine-sodium dodecyl suFate
po(yacqdamide gel ehctrophoresis gels. (A)Lane 1:low molecular
weight markers; lane 2: dinucleosome proteins from neocortex of
control subject, aged 88 years; lane 3: dinucleosome proteinsfrom
brain of patient with Alzheimer? diseuse, aged 81 years. iBi
Lane 4: whole nucleoproteins; lane 5: glia enriched fraction; lane
6: neuron enrichedfraction; lane 7: high molecular weight markers. Note the increased density of H I region histones in lane 3
and H l a in lane 5 .
Fig 2. Photometric profile of H I proteins extracted with 5 % perchloric acid from dinucieosomes of neocortex of patients with Alzheimer's diseaJe. Left scan made prior to treatment; right scan
made foliowing suspension of proteins in 70% formic acid and
exposure to cyunogen bromide (CNBr) at 3 7°C for 24 hours.
Note selectioe cleavage of methionine-containing band designated
H l ' , D = HID.
compared with the average for the group, exhibited a
small standard deviation of the mean (Table 2). The
ratios of these proteins were independent of age in the
range from 47 to 88 years. In contrast, ten dinucleosome preparations from six Alzheimer-affected brains
contained a statistically significant increase in linker region proteins, particularly H1" and Hl"" (analysis of
variance; p < 0.01). Although the overall mean value
of H1 extracted from Altheimer-affected brain was the
same as from control brain, analysis of variance revealed that the two populations differed at the 0.02
significance level. Overall, linker proteins were increased, on the average to 150% of the control value,
with a twofold increase in H1" and a threefold increase
in Hl"". The increase in these two proteins was not a
postmortem artifact. The amounts of the dinucleosome
proteins were examined at postmortem intervals
shorter and longer than 12 hours. In control brain, only
H1 was reduced in material obtained between 12 and
24 hours postmortem. In Alzheimer material all H1
proteins were reduced when studied more than 12
hours after death. The mean interval for control brain
was 19 hours and for Alzheimer brain 14 hours. Therefore, the H1" and H 1"" accumulations are not postmortem artifacts.
The increased amounts of linker histones associated
with dinucleosomes appear to be restricted to brain. In
liver the Hl/H4, Hl"/H4, Hlo"/H4 ratios were 0.34,
0.15, and 0.11, respectively, for control patients and
0.35, 0.13, and 0.11, respectively, for patients who
died in the advanced stages of Alzheimer's disease.
To study the possible effects of chronic malnutrition,
neuroleptic drugs, and agonal processes, and the
nonspecific effects of central nervous system degenerative disease, on chromatin organization, dinucleosomes
were prepared from the brains of eight patients who
died with a dementia process clinically resembling
Alzheimer's disease. Employing HIo/H4 and H 1""/H4
ratios that exceeded the mean plus two standard deviations of the mean for control preparations as an index
of abnormality, we found none of these degenerative
conditions to meet the criterion. Taken as a group, the
ratios of HI" and Hl"" were not significantly different
in affected brains and controls (analysis of variance and
Student t test). There was, however, a significant increase in H1 histone, with the two brains from patients
with Pick's disease showing the largest increase. The
brain that evidenced the histological abnormalities of
both Pick's and Alzheimer's diseases had elevated H1"
and HloO(see Table 2 and Fig 3) supporting the supposition that the Alzheimer component of the illness
contributed to the elevation of these proteins.
McLachlan et al: Chromatin in Dementia
331
Fig 3. Eighteen percent Tris-&cine-sodium dodecyl sulf;lte
polyacrylamidegel electrophoresi3 gels of proteins extracted from
dinucleosomesfrom chromatin of cerebral cortex nuclei. Note the
increused linker regioti hiitones in fane -3.Lane 1:supranuclear
pals-y,c u e 1 Ifrontul fobel: lane 2: supranuclearpalsy, case 1
(parietal lobe); lane -3: Pick? ditease with Alzheimer's disease
frontal lobe);lane 4: suprunucleur pa1s.y. case 2 tparietul lobe);
lane 5: Pick's disease (parietal lobe,; Iune 6: supranuclearpalsy,
case 2 (parietal lobe);Iane 7: cerebral cortex histone H I (55k perchloric acid-joluble frai.tio~i);lune 8: iitu thymus histone H 1
(5% perchlorii acid-solrrble fractioni; fune 9: Bio-Rad low
molecular weight markers (14.2K. 21K. N K , 43K, 68K, 94K).
To check for a possible overall increase in linker
proteins in total nuclear chromatin, neuron and glia
enriched fractions of bulk chromatin were examined.
As shown in Table 3, no such increase was detected.
Salt-Induced Releae of Linker Htstones from Chromatin
The fact that the levels of H1" and Hl"" from total
neuronal and glial chromatin do not change with the
onset of Alzheimer's disease suggests that the release
of dinucleosomes enriched in these proteins from
nuclease-digested Alzheimer chromatin results from a
change in the property of chromatin. This alteration
may be related to the spatial distribution of these
proteins throughout chromatin. Alternately, the observed changes may be caused by differences in posttranslational linker histone modifications, such as phosphorylation, that are known to alter chromatin
structure.
332
Annals of Neurology
Vol 15 N o 4
April 1984
To determine whether there are changes in the binding of these linker proteins to control and Alzheimer
chromatin, a salt extraction study of brain nuclei was
carried out. Concentrations of sodium chloride between 0.15 M and 0.75 M were employed, and the
amount of HI and core histones removed was compared with the total amount extractable. Table 4 gives
the average percentages of the total extractable linker
histone observed at salt concentrations between 0.3 M
and 0.6 M for six control and four Alzheimer-affected
brains. O n the average, control nuclei linker histones
begin to dissociate at 0.3 M sodium chloride and rcach
half maximal dissociation at 0.45 M sodium chloride
concentration. In contrast, linker histones from Altheimer-affected cortical nuclei do not dissociate until a
concentration of 0.45 M sodium chloride is reached,
and then only 1 4 q of the total is extracted. At molar
concentration of 0.6 there is virtually no difference in
the total amount extracted. Examination of Table 3
indicates that the dissociation characteristics of H 1
and Hl"" are very similar, whereas the methioninecontaining histone H1" has a higher affinity for DNA.
This finding is in keeping with the properties of H1" in
other tissues [ l l , 131. Core histones were also not
eluted from Alzheimer-affected nuclei as readily as
from control nuclei.
Discussion
Degenerative brain disorders associated with atrophy
and neuron loss, but without neurofibrillary degeneration and senile plaques, do not exhibit an increase in
dinucleosome-associated linker histones H 1" and
H 1"". Cachexia, terminal illness, drug ingestion, neuron loss, gliosis, and the postmortem interval do not
appear to influence chromatin structure or the concentrations of these proteins.
Whereas the methionine-containing histone H 1" of
human cerebral cortex resembles H 1" of other tissues,
H 1"" resembles H 1 in elution profile following treatment with NaCl (see Table 4 ) , in resistance to cyanogen bromide cleavage (see Fig 2), and in elution profile
with ion exchange and gel exclusion chromatography
(C. Mizzin, P. Lewis, D. R. McLachlan, unpublished
results, 1983).Because no evidence indicates that Hl""
is a postmortem degradation artifact, the increase
found in dinucleosomes of patients with Alzheimer's
disease may represent an accumulation of a catabolic
product of H1 in vivo.
There are several possible explanations for the observed increase in the amount of histones HI" and
HI"" in dinucleosomes liberated from a micrococcal
nuclease digestion of neuronal chromatin from patients
with Alzheimer's disease. The linker histones may have
become redistributed on the chromatin, perhaps as a
result of changes in their metabolism, thus creating
condensed domains rich in the HI" and H 1"" proteins.
Table 2. Ratio of Linker Histones t o H4 on Dinucfeosomes Extracted from Brains of Normal Controls,
Patients with Alzheimer's Disease, and Patients with Non-A fzheimer Type Dementia
Patient
Group
No. of
Brains
No. of
Assays
Age
(yr)
H llH4
H 1"lH4
H 1""/H4
Total
Control"
Alzheimer's
disease"
Multi-infarct
dementia
Picks disease
Supranuclear
palsy
Lewy body
encephalopathy
Atrophy
unclassified
Pick's and
Alzheimer's
diseases
7
6
8
10
72 +- 16
79 +- 7
0.44 & 0.04
0.40 +- 0.21
0.18 2 0.04
0.38 2 0.17
0.08 2 0.03
0.25 t 0.11
0.71
1.12
1
2
88
0.47
0.18
0.14
0.81
1
2
4
62
2
67, 78
0.77
0.55
0.26
0.14
0.11
0.06
1.14
0.75
1
2
62
0.48
0.22
0.09
0.79
2
4
65, 66
0.59
0.20
0.06
0.85
1
2
62
0.66
0.24
0.27
1.17
"Ages and ratios expressed as means
standard deviations.
?
Table 3. Ratio of Linker Histones to Core Histone H4 in Bulk Chromatin from Neuron and Glia EnrichEd Neocorticd Frastions
Glia Enriched Fractions
Neuron Enriched Fractions
Patient
Group
H1
~
H1"
~~
Control
SD
of total
Alzheimer's
disease
SD
of total
Hl""
~
Total
~
H1
H1"
H 1""
Total
~
1.12
0.1
70
1.04
0.33
0.1
21
0.35
0.14
0.02
9
0.16
1.59
0.18
...
1.55
1.31
0.17
76
1.15
0.34
0.1
20
0.42
0.07
0.006
4
0.16
1.72
0.26
...
1.73
0.06
67
0.04
23
0.04
10
0.13
...
0.1
66
0.06
24
0.04
9
0.12
...
Table 4. Percentages of Maximum Extractable H 1 Proteins at Variouj Concentrations of Sodium Chloride
Molar Concentration of Sodium Chloride
Patient
Group
Control
Alzheimer's
disease
0.30
0.35
0.40
0.45
0.50
0.55
0.60
4
13
0
28
0
54
14
77
48
73
0
77
81
82
2
0
9
0
19
0
46
8
56
45
66
67
78
75
32
57
14
57
70
66
83
76
82
HI0
Control
Alzheimer's
disease
~
Control
Alzheimer's
disease
~~
~
5
0
~
-
~
17
0
0
McLachlan et al: Chromatin in Dementia
333
Alternately, the chromatin structure itself may have
changed as a result of the expression of certain histone
variants, posttranslational modifications of the histones
(such as those from phosphorylation, acetylation, or
poly adenosine diphosphate ribosylation), changes in
the level of the complement of nonhistone proteins, or
metal ion binding. The latter group of possibilities is
supported by the results of linker histone salt binding
we have described. Metal ions, including aluminum,
may alter protein-DNA binding, an observation particularly important because aluminum has been detected in the nucleus of neurons with neurofibrillary
degeneration by two independent techniques: electron
probe microanalysis [ 121 and atomic absorption spectroscopy { 5 ] . A systematic examination of each of the
possible histone modifications will be necessary before
the defect in Alzheimer’s disease can be characterized.
The relation of this defect to the primary pathogenic
events of the disease remains to be determined, but the
observed changes in chromatin structure may be useful
in evaluating, in model systems, the effects of such
possible primary pathogenic agents as slow viruses, genetic mutations, and trophic facrors.
Supported by the Ontario Mental Health Foundation and the Canadian Geriatric Research Foundation. Tissue was obtained through
the Canadian Brain Tissue Bank.
References
1. Bloom KS, Anderson J N : Chromatin conformation of ovalbu-
men and globin genes in chromatin during differential gene expression. J Biol Chem 254:10532-10539, 1979
334 Annals of Neurology Vol 15 N o 4 April 1984
2. Cesarone CF, Bolognesi C, Santi L Improved microfluorometric D N A determination in biological marerial using 332513
Hoechst. Anal Biochem 100:188-197, 1979
3. Chrismar AM Von, Brito M, Burzio LO: Isolation and characterization of histone H1” from human placenta. Biochem Int
5~263-269, 1982
4. Crapper DR, Quirtkat S, De Boni U: Altered Chromatin conformation in Alzheimer’s disease. Brain 102:483-495, 1979
5. Crapper DR, Quitrkar S, Krishnan SS, et al: Intranuclear aluminum content in Alzheimer’s disease, dialysis encephalopathy and
experimental aluminum encephalopathy. Actd Neuropathol
50~19-24, 1980
6. Dayan AD, Ball MJ: Histiometric observarions o n the merabolism of tangle bearing neurons. J Neurol Sci 19:433-436, 1973
7. Lewis PN, Lukiw WJ, De Boni U, Crapper McLachlan DR:
Changes in chromatin structure associated with Alzheimer’s disease (Abstract No. 665), Can Fed Biol Soc 24: 1981
8. Lewis PN, Lukiw WJ, De Boni U, Crapper McLachlan DR:
Changes in chromatin structure associated with Alzheimer’s disease. J Neurochem 37:1193-1202, 1981
9. Mann DMA, Neary D, Yates PO, er al: Alterations in protein
synthetic capacity of nerve cells in Alzheimer’s disease. J Neurol
Neurosurg Psychiatry 4497- 102, 198 I
10. Medvedev ZLA, Medvedeva M N , Huschtscha LI: Age changes
of the pattern of F1 histone subfractions in rat liver and spleen
chromatin. Gerontology 23:334-341, 1977
11. Medvedev ZLA, Medvedeva MN, Robson L: Tissue specificity
and age changes of the pattern of H 1 group of histones in
chromatin from mouse tissues. Gerontology 24:286-292, 1978
12. Per1 DP, Brody AR: Alzheimer’s disease: x-ray spectrometric
evidence oEdwninum and accurnularion in neurcihbrillary tangle
bearing neurons. Science 208:297-299, 1980
13. Pokier GG, Savard P: ADP-ribosylation of pancreatic histone
H1 and other histones. Can J Biochem 58:509-515, 1980
14. Thompson RJ: Studies on R N A synthesis in two populations of
nuclei from the mammalian cerebral cortex. J Neurochem
21:19-40, 1973
15. Uemura E, Hartmann HA: Quantitative studies of neurotial
R N A o n the subiculum of demented old indivlduals. Brain Res
Bull 41301-305, 1979
Документ
Категория
Без категории
Просмотров
0
Размер файла
912 Кб
Теги
structure, chromatin, dementia
1/--страниц
Пожаловаться на содержимое документа