| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
* Department of Microbiology and Immunology, Thomas Jefferson University, Philadelphia, Pennsylvania 19107, USA; and
Department of Pathology and Laboratory Medicine, Boston University School of Medicine, Boston, Massachusetts 02118, USA
1Correspondence: Boston University School of Medicine, 85 East Newton St., Room M1009, Boston, MA 02118, USA. E-mail: jbluszta{at}bu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: PC 12 cells • membrane turnover • phospholipase • CTP:phosphorylcholine cytidylyltransferase • glycerophosphorylcholine
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4
5
6
7)
and
from magnetic resonance spectroscopy on AD patients in vivo
(8
9
10)
is a pattern of abnormalities in membrane
phospholipid turnover. Typically, the reported changes include reduced
levels in AD brain of key phospholipids, including phosphatidylcholine
(2)
phosphatidylethanolamine (2)
,
ethanolamine plasmalogens (11)
, and polyphosphoinositides
(12)
, resulting in impaired phosphoinositide signaling
(13
14
15)
as well as increased concentrations of the
metabolic products of phospholipid breakdown, glycerophosphorylcholine
and glycerophosphorylethanolamine (2
, 5
, 15)
. Alterations
in the levels of phosphorylcholine and phosphorylethanolamine, which
can be considered both as precursors and metabolites of their two
cognate phospholipids, have been reported as well (16
17
18)
(see Fig. 1
for a diagram of phosphatidylcholine metabolism). The overall pattern
of these neurochemical changes suggests an acceleration of phospholipid
turnover in AD, i.e., increased rates of both phospholipid breakdown
and resynthesis.
|
The mechanisms that cause the membrane defect in AD are not known.
Recent studies have shown that energy metabolism and mitochondrial
abnormalities may constitute an important component of the
pathophysiology and/or etiology of AD (for reviews, see refs
19
20
21
). The generalized reduction in brain glucose
utilization that appears early in the disease is likely to be
indicative of profound alterations in ATP metabolism
(22
23
24)
. Studies of cortical biopsy specimens from AD
patients suggested partial uncoupling of the mitochondria
(25)
, and investigations of postmortem AD brain revealed a
reduction in cytochrome aa3 activity
(26
27
28
29)
that is unlikely to be due to mutations in the
mitochondrial genome (30)
. The mitochondrial dysfunction
in AD is not confined to alterations in electron transport. For
example, the activities of other key enzymes of energy metabolism,
pyruvate dehydrogenase (31)
and 2-ketoglutarate
dehydrogenase (21
, 32)
, are also reduced in AD brain. In
addition, elevated concentrations of lactate, succinate, fumarate, and
glutamine were found in the cerebrospinal fluid of AD patients,
consistent with mitochondrial uncoupling (33)
.
Intriguingly, the finding that cultured fibroblasts from AD patients
also exhibit impaired mitochondrial function (34
, 35)
suggests that a systemic mitochondrial defect might play a role in AD.
In the current study we tested the hypothesis that accelerated
phospholipid turnover may be caused by mitochondrial dysfunction, using
as a neuronal model differentiated rat pheochromocytoma PC12 cells
treated with mitochondrial inhibitors. These cells have been used by
many investigators as a model for studies on the pathophysiology of AD
(36
37
38
39
40
41
42
43
44
45
46)
and on the effects of mitochondrial inhibitors on
multiple cellular functions (47
48
49
50
51)
. The data are
consistent with the above hypothesis and show that this model
reproduces the characteristic features of the membrane defect observed
in the AD brain.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell culture
PC12 cells were maintained at 37°C (10%
CO2/90% air) in DMEM containing 10% horse serum
and 5% fetal bovine serum. Cells were plated on 60 or 35 mm dishes
coated with poly-D-lysine and treated with nerve growth factor (NGF)
(50 ng/ml) for 4–6 days prior to each experiment. The medium was
replaced every 2 days.
Incorporation of radiolabeled choline into PC12 cells
Overnight labeling studies
NGF-treated PC12 cells were exposed to DMEM with serum
supplemented with
[Me-14C]choline chloride (0.5
µCi/ml of medium, 3 ml/dish) for 16–18 h. The dishes were then
washed once with DMEM, which contained no serum, and incubated for an
additional 50 min in DMEM without serum. This medium was removed and
replaced with serum-free DMEM containing various concentrations of
carbonyl cyanide m-chlorophenylhydrazone (CCCP) for 4 h. The medium was removed immediately after CCCP treatment and frozen
on dry ice for later analysis. Cells were harvested in 1 ml ice-cold
methanol. Dishes were then rinsed with 1 ml ice-cold water, which was
added to the cell extract.
Pulse-chase studies
Dishes (35 mm) containing NGF-treated PC12 cells were incubated
in serum-free DMEM for 50 min. This medium was removed and replaced
with serum-free DMEM containing
[Me-14C]choline chloride (0.5
µCi/ml of medium, 1.5 ml/dish, specific radioactivity 13 Ci/mol) for
30 min. Cells were then washed in serum-free DMEM and treated with CCCP
(30 µM) for various periods. Cells were harvested as described above.
Additional drug studies
The effects of various drugs alone or in conjunction with CCCP
were evaluated using a similar labeling paradigm as described above for
pulse-chase studies. Cells were labeled for 50 min with
[Me-14C]choline chloride,
then incubated in serum-free DMEM for 50 min, after which the medium
was replaced with serum-free DMEM containing various compounds for an
additional 2 h. Cells were harvested as described above.
Analysis of cell homogenates
For the extraction of lipids and choline-containing compounds, 1
ml of homogenate was mixed with 2 ml of chloroform and 1 ml of water
(52)
. To determine the radioactivities of water-soluble
choline-containing compounds, concentrated samples of the aqueous
fraction were separated and counted by high-performance liquid
chromatography (HPLC) using an on-line radioactivity monitor
(53)
.
Phospholipids were separated by thin-layer chromatography
(54)
. The individual phospholipids were scraped off the
plates and assayed for total radioactivity.
The amount of glycerophosphorylcholine was determined by first
incubating an aliquot of aqueous extract in 1M HCl for 1 h at
90°C. Under these conditions, glycerophosphorylcholine is the only
water-soluble choline ester present in PC12 cell extracts that
undergoes quantitative hydrolysis to free choline. Other compounds,
notably phosphorylcholine, are not hydrolyzed (data not shown). Samples
were then dried under a vacuum and assayed for total choline by HPLC as
described previously (2
, 55)
. Samples that were not
incubated in HCl were also subject to HPLC analysis to determine the
free choline pool. Glycerophosphorylcholine levels were determined by
subtracting the free choline level from that obtained after acid
treatment.
ATP, ADP, and AMP levels were determined by HPLC with UV detection
(56)
. Briefly, an aliquot of the aqueous fraction was
dried under vacuum, resuspended in ice-cold water, and filtered
(Millipore, 0.45 µm). Nucleotides were separated by ion-pair HPLC,
using a reverse phase column (Dynamax, 250x4.6 mm, C18, 5 µm; Waters
Associates, Milford, Mass.). The following gradient system was used:
solvent A, 5 mM tetrabutylammonium phosphate, 30 mM
KH2PO4, 4% acetonitrile,
pH 6; solvent B, 50% acetonitrile in solvent A. A 35 min concave
gradient up to 50% solvent B was used (1.5 ml/min) with a 25 min
equilibration delay.
Data analysis
Data are expressed as means ± SE. Data were
analyzed by ANOVA with a post hoc Duncan’s or Tukey’s test
(see Figs. 2
and 4
). Student’s t test was used to compare
the effects of various drugs (see Table 1
).
|
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and, indeed,
the levels of adenine nucleotides were maintained. In control cells the
levels of ATP, ADP, and AMP were 1.48 ± 0.08, 0.30 ± 0.07,
and 0.10 ± 0.05 mM, and in cells treated with 30 µM CCCP for
2 h, they were 1.65 ± 0.19, 0.34 ± 0.12, and 0.09 ± 0.06 mM, respectively. The turnover of phosphatidylcholine was
assessed in these cells by labeling them with
[14C]choline and after its incorporation into
[14C]phosphatidylcholine and its precursors and
metabolites. Two experimental designs were used. In the first, the
cells were incubated in the presence of
[14C]choline for 17 h so that the
radioactivity in the various metabolites reflected their mass; cells
were subsequently treated with CCCP at various concentrations. In the
second, the cells were pulse-labeled briefly (30 min) with
[14C]choline and then ‘chased’ for varying
periods of time in the absence of [14C]choline
in order to estimate the turnover rates of the choline-containing
intermediates in the phosphatidylcholine metabolic pathway (Fig. 1)
.
When assessed by the first experimental approach, the total amount of
[14C]choline radioactivity was not affected by
CCCP. The lipid fraction contained approximately twice the
radioactivity found in the aqueous fraction and consisted of 90%
phosphatidylcholine, 7% sphingomyelin, and 2.7%
lysophosphatidylcholine. CCCP (30 µM) had no effect on either the
total radioactivity observed in the lipid fraction (388,000±126,000
and 326,000±88,000 DPM/mg protein in control and treated cells
respectively) or on the individual choline-containing lipids. However,
CCCP modified the distribution of the label among the water-soluble
choline metabolites in a concentration-dependent, saturable fashion
(Fig. 2
). The proportion of total radiolabel associated with
[14C]choline and
[14C]phosphorylcholine was reduced
(n=3–5, P<0.01) whereas the percentage of
radiolabel in [14C]CDPcholine and
[14C]glycerophosphorylcholine was increased
(n=3–5, P<0.01) by CCCP. These changes closely
resemble those observed in AD brain, namely, a reduction in the amount
of choline and an increase in the amount of glycerophosphorylcholine.
Specifically, the ratio of glycerophosphorylcholine/choline was 
4
and 11, respectively, in control and AD brain (2)
; in the
current study, the ratio was 4.6 in control and 18 in CCCP-treated
cells.
The shift in choline metabolism and the accumulation of labeled
glycerophosphorylcholine induced by CCCP suggested that the uncoupler
enhanced membrane phosphatidylcholine turnover. To test this
hypothesis, a pulse-chase paradigm was used employing CCCP (30 µM)
during the chase period. After a 30 min labeling period, the majority
of the [14C]choline radioactivity was
recovered in [14C]phosphorylcholine and
[14C]choline. The amount of label in these
pools declined with time during the chase period whereas the
radioactivity in [14C]phosphatidylcholine and
[14C]glycerophosphorylcholine rose, consistent
with the precursor-product relationships depicted in Fig. 1
. CCCP
accelerated the rate of decline of radioactivity present in
[14C]choline (Fig. 3
) (as evidenced by the reduction of its half-life from 58 min to 32 min)
and in [14C]phosphorylcholine (as evidenced by
the reduction of its half-life from 90 min to 55 min). Moreover, CCCP
increased [14C]CDPcholine radioactivity over
sixfold above the control levels at 40, 60, and 120 min. Together,
these data indicate that CCCP accelerated the first two enzymatic steps
of phosphatidylcholine synthesis, i.e., those catalyzed by choline
kinase and CTP:phosphorylcholine cytidylyltransferase (CT),
respectively (Fig. 1)
. Although the net accumulation of radioactivity
in [14C]phosphatidylcholine was not affected
(data not shown), CCCP enhanced the appearance of the breakdown product
[14C]glycerophosphorylcholine radioactivity
relative to control cells. Moreover, CCCP doubled the specific
radioactivity of [14C]glycerophosphorylcholine
(measured at 60 min of the ‘chase’ period; Fig. 3D
,
inset). The latter result supports the conclusion that CCCP accelerated
phosphatidylcholine turnover.
|
The accumulation of CDPcholine in the CCCP-treated cells suggested that
although phosphatidylcholine synthesis was accelerated, the utilization
of CDPcholine was not maximal, limited perhaps by the availability of
diacylglycerol, the second substrate for cholinephosphotransferase, the
enzyme that uses CDPcholine and diacylglycerol for phosphatidylcholine
synthesis (Fig. 1)
. A synthetic diacylglycerol (dioctanoylglycerol,
DC8) was therefore used to stimulate
phosphatidylcholine synthesis in the presence and absence of CCCP in a
30 min pulse, 1 h chase paradigm. In addition to its effect as
substrate for phosphatidylcholine synthesis, DC8
activates CT, which catalyzes the conversion of phosphorylcholine to
CDPcholine (58
, 59)
. Indeed, in PC12 cells labeled as
described above and treated for 1 h with
DC8, there was a significant reduction
(P<0.05) in the radioactivity recovered in
[14C]phosphorylcholine (20%) to a degree
similar to that observed in cells treated with CCCP (Fig. 4A
), consistent with increased utilization of this precursor.
Moreover, in cells treated with both CCCP and
DC8, there was a synergistic decline in
[14C]phosphorylcholine radioactivity that was
reduced by greater than 60% (from 78.1±4.3 to 18.2±8.0% of total
aqueous [14C]choline radioactivity;
P<0.01).
DC8 prevented the accumulation of
[14C]CDPcholine caused by CCCP alone (Fig. 4B
), suggesting that CDPcholine was now efficiently used for
the synthesis of phosphatidylcholine. At the same time, as reported
previously (59)
, DC8 increased the
radioactivity recovered in
[14C]glycerophosphorylcholine (by twofold)
(Fig. 4C
). CCCP caused a similar increase in
[14C]glycerophosphorylcholine levels; in cells
treated with both DC8 and CCCP, the accumulation
of [14C]glycerophosphorylcholine increased
sevenfold (Fig. 4C
). The latter result shows that when the
limits in supply of diacylglycerol were overcome by the addition of
exogenous DC8, a remarkable acceleration of
phosphatidylcholine turnover by CCCP became apparent.
A separate series of experiments was designed to examine whether other
inhibitors of mitochondrial function produce similar actions on
phosphatidylcholine turnover. We used another uncoupler of oxidative
phosphorylation, 2,4-dinitrophenol, and an inhibitor of electron
transport, sodium azide. As a control we used the calcium ionophore
ionomycin, which causes elevations in intracellular calcium
concentrations—one of the consequences of mitochondrial hypofunction.
PC12 cells were briefly labeled (30 min) with
[14C]choline, washed, and exposed to treatments
for 1 h. Both dinitrophenol (100 µM) and sodium azide (4 mM)
significantly elevated
[14C]glycerophosphorylcholine and
[14C]CDPcholine and decreased
[14C]phosphorylcholine radioactivities
(Table 1
) as did CCCP. In contrast, ionomycin (1 µM) had
no effect on the distribution of radioactivity among the choline
metabolites (data not shown), suggesting that the actions of the
mitochondrial toxins were not mediated by calcium.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
), the concomitant enhancement of the synthetic pathway
is a novel and unexpected finding. The results indicate that
mitochondrial inhibitors increase phosphatidylcholine breakdown by
activating a phospholipase A2
(PLA2) (61
, 62)
and a
lysophospholipase (as evidenced by the accumulation of
glycerophosphorylcholine). It is not clear at this time whether
calcium-activated or calcium-independent PLA2
isoforms are sensitive to these changes in oxidative metabolism
(6)
. Subsequently, mitochondrial inhibitors stimulate
biosynthesis of the phospholipid by activating the rate-limiting enzyme
in the pathway (63)
, CT, as indicated by the striking
accumulation of CDPcholine. The latter effect was possible because the
pools of high-energy phosphates (including ATP) necessary for the
synthesis of CTP were maintained, apparently by glycolysis. The
activation of CT is probably mediated by fatty acids, which are potent
activators of the enzyme (64
65
66
67)
and are liberated from
phosphatidylcholine by the combined actions of
PLA2 (60)
and lysophospholipase. It
is worth noting that CT activity is significantly elevated in
postmortem cortical samples of AD patients relative to age-matched
controls (3)
.
The accumulation of CDPcholine in the CCCP-treated cells indicates that
whereas in control cells CT activity appears to limit the rate of
phosphatidylcholine synthesis (63)
, in cells treated with
the uncoupler, it is not the enzyme activity, but rather the supply of
diacylglycerol, that is rate-limiting. Indeed, provision of exogenous
diacylglycerol reduced CDPcholine levels to control values. In previous
studies, Araki and Wurtman found that diacylglycerol levels can be
rate-limiting when cells up-regulate membrane synthesis and sprout
neurites (68)
. In CCCP-treated cells, exogenous
diacylglycerol was efficiently used for the synthesis of an apparently
rapidly turning over pool of phosphatidylcholine. The total
phosphatidylcholine pool was not changed by CCCP, suggesting that
inhibitors of mitochondrial bioenergetics do not cause a reduction in
the total amount of membrane. Similarly, in AD brain the total amount
of phosphatidylcholine was only slightly reduced whereas
glycerophosphorylcholine mass was increased by up to 150% (2
, 5)
.
Taken together, the data indicate that mitochondrial dysfunction does
not necessarily lead to net membrane breakdown, because it also allows
for rapid phospholipid synthesis. Our results are consistent therefore
with observations that AD brain is characterized not only by a
degenerative process, but also by neuritic sprouting and remodeling,
processes that require active membrane synthesis (69)
.
Phospholipid synthesis in normal brain is thought to consume at least
10–15% of the total ATP pool (70)
. If one were to assume
that the amount of glycerophosphorylcholine may be used as an index of
phosphatidylcholine synthesis and breakdown, then the total turnover of
phosphatidylcholine was approximately doubled in the CCCP-treated cells
relative to controls. This would indicate that when mitochondrial
function is inhibited, the cells use a larger proportion of their ATP
for maintenance of membrane integrity, perhaps at the expense of other
key functions. In addition to endogenous phospholipid synthesis, brain
neurons take up lipids transported by the apolipoprotein E (ApoE)
-containing lipoproteins. The quantitative significance of this process
may increase in neurons whose membrane turnover is accelerated and in
cells undergoing neuritic sprouting. There is strong evidence that the
APOE 
4 allele is a risk factor for AD, and it has been hypothesized
that ApoE4-containing lipoproteins may be less efficient in supplying
lipids to neurons (71)
. Thus, mitochondrial dysfunction in
individuals who are carriers of the APOE 4 allele may cause a
membrane abnormality that cannot be sufficiently compensated for by
lipid delivery to the cells. Indeed, changes of phospholipid turnover
were more pronounced in samples from autopsied brains of AD patients
expressing the ApoE3/4 protein isoforms than in ApoE3/3 homozygotes
(17)
.
Evidence that mitochondrial dysfunction is a feature of the
pathophysiology, and possibly the etiology, of AD is growing and
includes data obtained from various brain regions as well as from
peripheral tissues such as fibroblasts and platelets (19
, 20)
. Phospholipid abnormalities have been observed in AD brain
in regions, including the cortex and hippocampus, that are typically
associated with the histopathological manifestations of the diseases
(i.e., amyloid plaques and neurofibrillary tangles), as well as in
other brain areas devoid of overt histopathology (caudate and
cerebellar cortex) (5)
and even in non-neuronal cells such
as platelets (72
, 73)
. These data, our current
observations, and studies showing that mitochondrial inhibitors
stimulate the production of potentially amyloidogenic peptides
(74)
support the notion that the mitochondrial defect may
be a primary cause of the pathophysiology of AD. However, it has also
been suggested that down-regulation of mitochondrial oxidative
metabolism is secondary to reduced demand for ATP resulting from other
causes such as reduced synaptic activity (24
, 75)
.
Given that phosphatidylcholine is synthesized in the endoplasmic
reticulum and that mitochondrial inhibitors stimulate the production of
amyloidogenic peptides in this organelle (74)
, it is
possible that local alterations of phospholipid turnover, producing
changes in membrane composition and in the concentrations of
phospholipid-derived signaling molecules (e.g., lysophospholipids or
fatty acids), are responsible for the altered regulation of amyloid
precursor protein processing and the generation of the amyloidogenic
peptides by proteases sensitive to these perturbations of the membrane.
In contrast, no abnormalities of phosphatidylcholine turnover have been
detected in Down’s syndrome (trisomy 21), another condition
characterized by brain beta amyloid deposition in which the
accumulation of amyloid plaques is attributable to the overexpression
of the amyloid precursor protein (whose gene resides on chromosome 21),
though oxidative stress is also thought to be a feature of this
disease.
Mitochondrial dysfunction may be the primary cause of beta amyloid
generation (74
, 76)
, the abnormal pattern of tau protein
phosphorylation (77
78
79)
, and oxidative damage
(19)
in AD. Our data show that inhibitors of mitochondrial
bioenergetics mimic the membrane abnormalities of AD, and point to the
mitochondria as the etiologic factor for the membrane defect in this
disease.
|
ACKNOWLEDGMENTS |
|---|
Received for publication November 23, 1999.
Revision received May 8, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. G. Bundy, N. G. Iyer, M. S. Gentile, D.-E. Hu, M. Kettunen, A.-T. Maia, N. P. Thorne, J. D. Brenton, C. Caldas, and K. M. Brindle Metabolic Consequences of p300 Gene Deletion in Human Colon Cancer Cells. Cancer Res., August 1, 2006; 66(15): 7606 - 7614. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. R. Clapp, G. M. Hirschfield, C. Storry, J. R. Gallimore, R. P. Stidwill, M. Singer, J. E. Deanfield, R. J. MacAllister, M. B. Pepys, P. Vallance, et al. Inflammation and Endothelial Function: Direct Vascular Effects of Human C-Reactive Protein on Nitric Oxide Bioavailability Circulation, March 29, 2005; 111(12): 1530 - 1536. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Dager, S. D. Friedman, A. Parow, C. Demopulos, A. L. Stoll, I. K. Lyoo, D. L. Dunner, and P. F. Renshaw Brain Metabolic Alterations in Medication-Free Patients With Bipolar Disorder Arch Gen Psychiatry, May 1, 2004; 61(5): 450 - 458. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
