вход по аккаунту


Decreased concentrations of GLUT1 and GLUT3 glucose transporters in the brains of patients with Alzheimer's disease.

код для вставкиСкачать
Decreased Concentrations of GLUT1 and
GLUT3 Glucose Transporters in the Brains
of Patients with Alzheimer's Disease
Ian A. Simpson, PhD,+ Koteswara R. Chundu, MD," Theresa Davies-Hill, BS,*
William G . Honer, M D , t and Peter Davies, P h D t
Glucose metabolism is depressed in the temporal and parietal regions of the cortex in patients with Alzheimer's disease.
We measured the concentrations of two glucose transporters, GLUTl and GLUT.l, in six regions of brains from both
control subjects and patients with Alzheimer's disease. The concentrations of both transporters were reduced in the
cerebral cortex, with larger and highly significant reductions observed for GLUT.1, the putative neuronal glucose
transporter. The reductions in GLUT3 were greater than the loss of synapses, and should be considered as a potential
cause of the deficits in glucose metabolism.
Simpson IA, Chundu KR, Davies-Hill T, Honer WG, Davies P. Decreased concentrations of GLUTl and GLUT3
glucose transporters in the brans of patients with Alzheimer's disease. Ann Neurol 1994,35.546-55 1
Alzheimer's disease (AD) is a major cause of cognitive
dysfunction in the elderly population of developed
countries [l}. An early and consistent feature of this
disease is a decrease in brain glucose metabolism,
which by positron emission tomography (PET) [Z) appears to be prominent in temporal and parietal regions
of the cortex [3,4}.T h e underlying cause of this deficit
is not understood.
There is a family of proteins that facilitate the transport of glucose across ceH membranes, designated
GLUTl through GLUT5 {5]. In the brain, two transporters have been detected, GLUTl and GLUT3
C6-81. High levels of GLUTl are expressed in microvessels and mediate the transport of glucose across the
blood-brain barrier [9-11). GLUTl has also been detected in neuronal and glial preparations, but this isoform is smaller (45 kd) than that found in microvessels
( 5 5 kd) 112). GLUT3 appears to be the protein responsible for the transport of glucose into neurons ClZ161. Utilizing specific antibodies against the transporters, we measured the concentrations of GLUTl and
GLUT3 by immunoblot analysis in six regions (frontal,
parietal, occipital, and temporal cortical areas, caudate
nucleus, and hippocampus; designated 1-6, respectively) of brains obtained at autopsy from both control
individuals and patients with AD.
Materials and Methods
From the "Experimental Diabetes, Metabolism, and Nutrition Section, National Institute of Diabetes and Digestive and Kidney Dis-
Received May 17, 1993, and in revised form Nov 8. Accepted for
oublicarion L>ec 27., 1993.
eases, Berhesda, MD, and the 'Departments of Pathology and Neuroscience, Albert Einstein College of Medicine, Bronx, NY.
Brain tissue was obtained at autopsy from 12 patients with
AD and from 12 individuals dying without evidence of neurological or psychiatric disorders. Autopsies were performed
within 24 hours of death, with the mean delays being 8.7
4 hours for the Alzheimer patients and 12.5 t 6 hours for
the control subjects (difference not statistically significant).
Control subjects were 2 I to 96 years old (mean 2 standard
deviation, 56 i- 22 years). No significant age-related changes
in any of the glucose transporters measured were observed.
Alzheimer patients were 68 to 86 years old (76 5 5 years).
Neuropachological diagnosis of AD was essentially according
to the criteria of Khachaturian [17), although we required
that all patients had significant numbers of senile plaques
and neurofibrillary tangles in the neocortex. Subjects classified as controls had neither significant neuropathological abnormalities nor a clinical history of dementia. Samples of
each brain region were coded to ensure that assays were
performed with the investigator blinded to diagnosis.
Tissue samples (100-150 mg) were homogenized with a
Dounce homogenizer (12 strokes) in a Tris (20 mM), echylenediaminetetraacetic acid (EDTA) (1 mM), sucrose ( 2 5 5
mM) buffer containing the protease inhibitors phenylrnethyl
sulfanyl fluoride (PMSF) (0.5 X 10 M), aprotonin (1 x
10-"M),leupeptin(l X I O - - M ) , andpepstatin (1 X lo-").
A total membrane pellet was obtained by centrifugation at
200,OOOg for 1 hour. The pellet was resuspended in the same
buffer aiid protein content was determined. Aliquots (30 or
Address correspondence to Dr Davies, Departments of Parhology
and Neuroscience, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461
546 Copyright 0 I994 by the American Neurological Association
60 p,g of protein) of each sample were solubilized and analyzed by immunoblotting. GLUTl was detected with a rabbit
polyclonal antibody raised against purified human erythrocyte GLUTl transporter and iodine 125 ( 1251)-labeled protein A (New England Nuclear). The relative concentrations
of three species of GLUTl (apparent molecular masses of
5 5 . 45, and 38 kd) were determined by autoradiographic
visualization of antibody binding. The individual bands were
excised and radioactivity measured by scintillation counting.
g membrane protein with
GLUT3 was detected in 30 ( ~ . of
a rabbit polyclonal antibody raised against the I5-amino acid
C-terminal sequence of the human GLUT3 protein and '251labeled protein A, and visualized by autoradiography. Quantitation was performed by phosphorimage analysis (Molecular Dynamics, Sunnyvale, CA). The level of the synaptic
protein SP14 in the same tissue samples was determined
using a mouse monoclonal antibody and "51-labeled sheep
anti-mouse IgG second antibody (Amersham, Arlington
Heights, IL), and quantitated by phosphorimaging analysis.
Statistical analysis was performed using one- and two-way
analyses of variance.
In Figure 1, representative immunoblots using the
GLUTl antibody are presented, illustrating the three
forms of this transporter that were quantitated. A significant overall deficit in both the 55-kd and 45-kd
bands and an additional (38-kd) form of GLUTl were
observed in the total cortex including the hippocamFig 1 . Typical immunoblot.r of a nomal .subject and u putient
with Alzheimer's disease. showing the three major fornu of' the
GLUTl transporter detected by the antibody used. Brain regions are (1)midfrontal cortexx,(2) inferior parietal cortex,
( 3 ) occipitul pole, (4) midtemporal gyrus, ( 5 ) hippocampus, and
( 6 ) caudate nucleus. In a/l cases, the indiz'idual bands were cat
fmm the gels follozchg autoradiography, and radioactivity UJUS
quantitated by .rcintillation counting.
92 -
c 55K +
27 -
1 2 3 4 5 6
Brain Region
1 2 3 4 5 6
Brain Region
F i g 2. Quuntitatiw data d4rizred us described in Figure 1 and
"Materials and Methods" (A)Pooled data from all cortical regions (including the hippocampus). The asterisks indicate statajticulbp significunt differences (p < 0.01). (B) Quantitative data
from the cuudate nucleus, with samples from 12 normal subjects
and 12 patients with Alzheimer? disease. Asterisks indicate
statisticallj significant diflertnces (p < 0 . 0 1 ) . S E N =
stundard error of mean.
pus. However, while apparent deficits were noted in
individual cortical areas 1 through 4, the interpatient
variation was too great for these differences to be statistically significant in any single region. In Figure 2A,
data from the neocortical regions and hippocampus
have been pooled. In Figure 2B, the distribution of
the three forms of GLUTl in the caudate nucleus is
compared. The 55-kd and 45-kd GLUTl isoforms in
the caudate nucleus were significantly reduced by approximately 45% ( p < 0.01).
Representative GLUT3 immunoblots of the same
brain regions (1-6) obtained from 6 normal subjects
and 6 Alzheimer patients are shown in Figure 3, and
the quantitative analysis for all patients (n = 12igroup)
is presented in Figure 4. It is notable that individual
cases showed some variability within brain regions, and
there was also variability as to which regions of the
Alzheimer brains had the largest reductions in GLUT3.
However, concentrations of GLUT3 were significantly
reduced in all regions of the brains from patients with
A D except the frontal cortex (see Fig 4A).
Simpson et al: Decreased Glucose Transporters in A D
Frontal Cortex
Parietal Cortex
45 4
Temporal Cortex
Caudate Nucleus
F i g 3. Typical GLUT3 immunoblots of samples of the six
brain regions from 6 control subjects and 6 patients with
Alzheimev’s disease.
Neuron and synapse loss are common in the brains
of A D patients, and may account for several of the
neurochemical deficits observed in such samples [ 18211. T o determine if such losses were responsible for
the decrease in GLUT3 concentration, we measured
synapse density using an antibody to a synaptic protein,
SP14 [22). Consistent with earlier observations [18211, synaptic density was significantly reduced in frontal, parietal, and occipital regions of the cortex (see Fig
4B) in patients with AD. To account for these deficits,
we presented the concentrations of GLUT3 relative to
the concentration of the SP14 antigen in individual
samples (see Fig 4C). Such analysis revealed that in the
parietal cortex, temporal cortex, caudate nucleus, and
hippocampus, the deficits in GLUT3 cannot be accounted for simply by synaptic loss and thus reflect a
specific decrease in the expression of the GLUT3 protein in these regions.
The results presented appear to show reduced concentrations of both GLUTl and GLUT3 in the brain tissue
from patients with AD. Both the microvessel and nonvascular forms of GLUTl were reduced in concentration, as was the apparently neuron-specific GLUT3.
These results would argue that the capacity to transport
548 Annals of Neurology Vol 35 No 5 May 1994
glucose into the brain and into neurons is reduced in
this condition.
The decreased concentration of GLUTl that we observed in cerebral cortex is generally consistent with
previous findings. Kalaria and Harik {23] reported decreased cytochalasin B (CB) binding in neocortical and
hippocampal tissues from subjects with AD. Microvessel preparations from corresponding tissue also showed
a significant decrease in CB binding and thus the tissue
deficits were interpreted as a reduction in the concentration of the blood-brain barrier glucose transporter.
We also observed a significant decrease in CB binding
in ssnples of temporal cortex from the AD patients
compared to normal patients (7.5 -t 2.5 pmol/mg
[n = 4) compared to 14.4 5 1.8 pmolimg [n = 61).
We concur that the microvessel glucose transporter
(55-kd GLUT1) is indeed reduced in the cortex of
patients with AD. However, CB has been shown to
bind to all forms of GLUTl and GLUT3 [24, 251,
and thus the observed decreases in CB binding do not
simply reflect a loss of blood-brain barrier glucose
transporters. The concentrations of nonvascular forms
of GLUTl as well as GLUT3 were also reduced in our
samples from patients with AD.
The decreased concentrations of both the 55-kd and
the 45-kd forms of GLUTl and of GLUT3 in the caudate nucleus were unexpected findings, as this brain
region is generally not affected by AD. It might be
significant that the concentrations of both isoforms of
B. SP14
0 Alzheimer
C. GLUT3/SP-14
3 r
Fig 4. (A) The quantitation of the GLUT3 Western blotting
data of all patients including those shown in Figure 3. (Bi The
cowesponding levels of the synaptic protein SP14. (C) The data
expressed in (A) have been nomlized for the synaptic protein
losses seen in (B). Asterisks indicate statistzcally signijbant dz$
ferences (p < 0.01). SEM = standard error of the mean.
GLUTl and of GLUT3 were considerably higher in
the caudate nucleus than in the other brain regions
studied (see Fig 3A), even in samples from patients
with AD, such that reductions in the level of GLUT3
may not be reflected in deficits in glucose utilization.
Glucose metabolism in the hippocampus is not usually
determined by PET because of the limited anatomical
resolution currently available; however, the loss of
GLUT3 was to be expected given the neuronal damage
commonly observed in this structure.
The deficit in glucose metabolism in the temporal
and parietal cortex of patients with AD has been well
established by PET, and one report clearly suggests
this occurs early in the course of the disease 1261. In
PET studies, rates of glucose transport are represented
by two terms, K, and k,, which represent inward and
outward glucose flux, respectively. However, the ratelimiting step is usually assumed to be the activity of
hexokinase, k, [27}. Both K, and k2 terms describe
rates of passage of glucose across microvessel endothelial cells, diffusion through the interstitial space, and
transport into or out of neurons. Movement of glucose
in either direction involves both GLUTl and GLUT3,
and changes in the concentration of either of these
molecules would be expected to alter both the K, and
k, terms. A recent PET study by Jagust and coworkers
[261 suggested that deficits in glucose transport into
the brain are found in living patients with AD. They
observed significant changes in K," values, while no
changes in k," were apparent. As indicated above, it is
generally assumed that hexolunase is the rate-limiting
enzyme for the metabolism of glucose in the central
nervous system. This assumption is based on the original observation of Crane and Sols 1281 that hexokinase
is inhibited by one of its substrates (ATP) and one of
its products (glucose-6-phosphate). Under most circumstances, the concentrations of these two compounds in brain are high enough to result in almost
complete inhibition of hexokinase activity, thus controlling the flow of glucose down the glycolytic pathway. This may explain why the basal rate of glucose
metabolism in brain (0.6 ~mollgmimin{291) is far
lower than the maximum (uninhibited) capacity of hexokinase (27.5 pmol/gm/min), according to Lowry and
colleagues {30). More recently, it was suggested that
the regulation of hexokinase activity may be much
more complex than was originally thought, with the
extent to which the enzyme is inhibited being dependent on whether or not the enzyme is cytoplasmic or
bound to organelles such as the endoplasmic reticulum
or mitochondria [31).
None of these earlier studies seriously considered
the possibility that neurons may differ from glia in the
regulation of hexokinase activity, or that subpopulations of neurons may exhibit different regulatory
mechanisms, perhaps dependent on the level of neur o d activity, through effects o n concentrations of glucose-6-phosphate and ATP. Neurons appear to use
GLUT3 as the major glucose transporter, while glial
cells probably rely more on GLUT1. In this study, it
is obvious that there are regional variations in GLUT3
concentration in the brain (see Fig 3C). There is also
evidence for regional variations in type 1 hexokinase
activity in the brain f321.
In peripheral tissues, glucose transporter activity is
rate limiting for glucose metabolism despite the presence of hexokinase activities that are equal in their
sensitivity to inhibition to those in brain. In adipocytes
in the absence of insulin, glucose transport is rate limiting, while in the presence of insulin, glycolysis becomes Limiting. The change in rate-limiting steps occurs because of a fivefold increase in the glucose
transporter concentration in the adipocyte plasma
membrane in response to insulin 133, 341. More recent
Sirnpson et al: Decreased Glucose Transporters in AD
studies in muscle indicated that the impaired glucose
metabolism associated with diabetes may result from a
glucose transporter deficit [ 3 5 , 361.
Two situations in which the rates of glucose transport become limiting for cerebral glucose utilization
are hypoglycemia and seizures. In hypoglycemia, a reduction in plasma glucose concentrations from 5 to 2.5
mM or less results in widespread neuronal dysfunction
137). Although detailed comparisons are extremely difficult to make at this time, conceptually the reductions
in the concentrations of both GLUTl and GLUT3 glucose transporters could produce similar deficits in glucose availability, as the passage of glucose across the
blood-brain barrier and into individual neural cells is
directly dependent both on the concentration of glucose and on the concentration of transporters. In some
seizure states, it seems clear that elevations in regional
rates of glucose utilization produce a situation in which
transport cannot match demand for the substrate 138).
O n considering the possibility that AD may be an
additional example of a situation in which glucose
transport becomes rate limiting for metabolism, we
thought it appropriate to calculate the potential maximum glucose transport capacity. Assuming a transporter concentration of 14 pmolimg of membrane protein (1231 and the CB-binding data described above),
3 K,of 10 mM { 2 4 , 2 5 ] ,a turnover number of approximately 30,000imin (determined for GLUTl and
GLUT4 [24, 2 5 ) ) , and a plasma glucose concentration
of 5 mM, we calculated a transport capacity in the normal brain of about 10.5 pmoligmimin. This figure is
of similar magnitude to the operational hexokinase activity (cited above). Clearly, given caveats regarding
cellular heterogeneity and the possibility of differential
regulation in different cell types for hexokinase,
GLUT1, and GLUT3, it is possible that any one of
these could be rate limiting.
In this study we observed regional differences in the
concentrations of GLUTl and GLUT3, and significant
deficits in their concentrations in cerebral tissues from
patients with AD. These deficits remain even after correction for neuronal loss. A decrease in the ability to
transport glucose across the blood-brain barrier will
probably reduce the rate of entry of glucose into the
brain. The decrease in GLUT3 concentrations in certain neurons might further compromise glucose availability, and result in the characteristic deficits in glucose
metabolism evident in PET studies of these patients.
This work was supported by National Institutes of Health and National Institute of Mental Health grants AGO6803 and MH38623
and by the British Columbia Health Research Foundation.
W e are grateful to Drs Janet Passoneau and Robert Veech for their
comments on an early version of this manuscript, and to Dr Dennis
W. Dickson for neuropathological evaluation of all the brains used
in this study.
550 Annals of Neurology Vol 35 No 5
May 1994
1. Katzman R. The prevalence and malignancy of Alzheimer's disease. Ann Neurol 1976;33:217-218
2. Benson DF, Kuhl DE, Randall A, et al. The fluorodeoxyglucose
"F scan in Alzheimer's disease and multi-infarct dementia. Arch
Neurol 1981;40:711-714
3. Duara R, Grady C, Haxby J, et al. Positron emission tomography in Alzheimer's disease. Neurology 1986;36:879-887
4. Friedland RP, Jagust WJ, Huesman RH, et al. Regional cerebral
glucose transport and utilization in Alzheimer's disease. Neurology 1989;39: 1427- 1434
5 . Bell GI, Kayano T, Buse JB, et al. Molecular biology of mammalian glucose transporters. Diabetes Care l990;l3: 198-208
6. Mueckler M, Caruso C, Baldwin S, et al. Sequence and structure
of a human glucose transporter. Science 1985;229:941-945
7. Birnbaum MJ, Haspel HC, Rosen OM. Cloning ,and characterization of a cDNA encoding a rat brain glucose transporter protein. Proc Natl Acad Sci USA 1986;83:5784-5788
8. Kayano T, Fukumoto H, Eddy RL, et al. Evidence for a family
of- human glurose rransporter-like proteins: sequence and gene
localization of a protein expressed in fetal skeletal muscle and
other tissues. J Biol Chem 1988;263:15,245-15,248
9. Dick APK, Harik SI, Klip A, Walker DM. Identification and
characterization of the glucose transporter of the blood brain
barrier by cytochalasin B binding and immunological reactivity.
Proc Natl Acad Sci USA 1984;81:7253-7237
0. Boado RJ, Pardridge WM. The brain-type glucose transporter
mRNA is specifically expressed at the blood-brain barrier. Biochem Biophys Res Commun 1990;lhh:174- 179
1. Pardridge WM, Boado RJ, Farrell CR. Brain-type glucose transporter (GLUT1) is specifically expressed at the blood-brain barrier. J Biol Chem 1990;2h5: lX,035-18,040
2 . Sivitz W, DeSaurel S. Walker PS, Pcssin JE. Regulation of glucose transporter in developing rat brain. Endocrinology 1989;
1L'I:1875-- 1880
3. Nagamarsu S, Kornhauser JM, Burant CF, et al. Glucose transporter expression in brain. cDNA sequence of mouse GLUT3,
the brain facilitative glucose transporter isoform, and identihcation of sites of expression by in siru hybridization. J Biol Chem
4. Yano H , Seino Y, Inagaki N, et al. Tissue distribution and
species difference of the brain type glucose transporter
(GLUT?). Biochem Biophys Res Commun 1991;174:470-477
5 . Bondy C, Lee WM, Z h o u J . Ontogeny and cellular distribution
of hrain glucose transporter gene expression. Mol Cell Neurosci
1992;3:305-3 14
16. Maher F, Davies-Hill T, Lysko P, et al. Expression of two glucose transporters, GLUTl and GLUTI, in cultured cerebellar
neurons: evidence for neuronal specific expression of GLLT3.
Mol Cell Neurosci 1991;2:351-360
17. Khachaturian ZD. Diagnosis of Alzheimer's disease. Arch Neurol 1985;42:1097-2002
18. Terry RD, Masliah E, Salmon DP. Physical basis of cognitive
alterations in Alzheimer's disease: synaptic loss is the major corrrlate of cognitive impnirment. Ann Neurol 1991;30:572-580
19. Masliah E, Haiisen L, Albright 'r, et al. Immunoelecrron microscopic study of synaptic pathology in Alzheimer's disease. Acra
Ncuropathol (Bed) 1991;81:428-433
20. Masliah E, Terry RD, Alford M, DeTeresa RM. Quantitative
immunohistochemistry of synaptophysin in human neocortex:
an alternative method to estimate density of presynapric terminals in paraffin sections. J Histochem Cytochem 1990;38:837-
21. Honer WG, Dickson DW, Gleeson J, Davies P. Regional synaptic pathology in Alzheimer's disease. Neurobiol Aging 1992;13:
22. Honer WG, H u L, Davies P. Human synaptic proteins with a
heterogeneous distribution in cerebellum and visual cortex.
Brain Res 1993;609:9-20
23. Kalaria R N , Harik SI. Reduced glucose transporter at the blood
brain barrier and in the cerebral cortex in Alzheimer's disease.
J Neurochem 1989;>3:1083-1088
24. Carruthers A. Facilitated diffusion of glucose. Am Physiol Soc
25. Gould GW, 'Thomas HW, Jess TJ, Bell GI. Expression of human glucose transporters in Xenopus oocytes: kinetic characterization and substrate specificities of the erythrocyte, liver and
brain isoforms. Biochemistry 1991;30:5 139-5 145
26. Jagust WJ, Scab JP, Huesman RH, et d. Diminished glucose
transport in Alzheimer's disease: dynamic PET studies. J Cereb
Blood Flow hletab 1991;11:323-330
27. Sokoloff L, Reivich M, Kennedy C, et al. The "'C deoxyglucose
method for the measurement of the local cerebral glucose utilization: theory, procedure and normal values in the conscious
and anesthetized rat. J Neurochem 1977;28:897-916
28. Crane RK, Sols A. The competitive inhibition of brain hexokinase by glucose-6-phosphate and related compounds. J Biol
Chem 1954;210:597-606
29. Hawkins RA, Miller AL. Cremer JE, Veech KL. Measurement
of the rate of glucose utilization by rat brain in vivo. J Neurochem 1974;23:917-923
10. Lowry OH, Passonncau JV, Hasselberger FX. Schulz DW. The
relationships between substrates and enzymes of glycolysis in
brain. J Biol Chem 1964;239:18-30
31. Parry DM, Pederson PC. Glucose catabolism in brain. Intracel-
Mar localization of hexokinase. J Biol Chem 1990;265:1059-
32. Burgess RW, Wilson JE. Hexokinase levels in discrete regions
of rat brain: evaluation by quantitative immunofluorescence using interactive laser cytometry and comparison with basal rates
of glucose utilization. J Neurochem 1991;57:44 1-446
3.3. Vinren J, Gliemann J, Osterlind J. Exchange of 3-0-methylglucose in isolated fat cells. Concentration dependence and effect of insulin. J Biol Chem 1976;251:794-800
31. Simpson IA, Cushman SW'.Hormonal regulation of mammahan
glucose transporters. Annu Rev Biochem 1986;55:1059- 1089
35. Yki-Jrvinen H , Sahlin K, Ren JM, Koivisto VA. Localization of
rate-limiting defect for glucose disposal in skeletal muscle of
insulin-resistant type I diabetic patients. Diabetes 1990;39:
36. Rothman DL, Shulman RG; Shulman GI. %'Pnuclear magnetic
resonance measurements of muscle glucose 6 phosphate: evidence for reduced insulin dependent muscle glucose transport
or phosphorylation activity in non-insulin dependent diabetes
mellitus. J Clin Invest 1992;89:1064-1075
37. Lewis LD, Ljunggren B:Norberg K, Siesjo BK. Changes in
carbohydrate, amino acids and ammonia in the brain during
insulin-induced hypoglycemia. J Neurochem 1974;23:659-67 1
38. Ingvar M, Siesjo BK. Pathophysiology of epileptic brain damage. In: Wasterlain CG, Vert P, eds. Neonatal seizures. New
York: Raven, 1990:115-122
Simpson e t al: Decreased Glucose Transporters in A D 5 5 1
Без категории
Размер файла
585 Кб
decrease, glut, patients, concentrations, disease, brain, alzheimers, transporters, glucose
Пожаловаться на содержимое документа