|
|
||||||||





* Institute of Normal Morphology, G. dAnnunzio University of Chieti; 66100 Chieti, Italy; and
Department of Morphology and Embryology, Human Anatomy Section, University of Ferrara, 44100 Ferrara, Italy
1Correspondence: Institute of Normal Morphology, G. DAnnunzio University of Chieti; via dei Vestini 6, 66100 Chieti, Italy. E-mail: g.zauli{at}morpho.unich.it
| ABSTRACT |
|---|
|
|
|---|
Key Words: extracellular Tat signal transduction cAMP PC12 HIV-1-associated dementia
| INTRODUCTION |
|---|
|
|
|---|
Tat is actively released by HIV-1-infected cells (2
, 3)
and, when used at concentrations of 0.1 µM or higher, induces massive
and rapid neuronal cell death (4
5
6
7
8
9
10
11
12
13)
. Since microglial
cells and astrocytes are infected with HIV-1 but neurons are not
(14)
, it has been hypothesized that HIV-1 Tat protein
locally released by infected cells may produce neuronal dysfunction
and/or loss. In this respect, Tat protein is expressed in cells
obtained from the central nervous system of patients affected by
encephalitic AIDS (14)
. It has also been shown that, at
concentrations of 0.110 nM, Tat protein activates various signal
transduction pathways in neurons, including phosphatidylinositol 3
kinase (PI-3K) (15
, 16)
, ERK/MAPK (17)
, and
the cyclic adenosine monophosphate (cAMP) -dependent protein kinase A
(PKA) pathway (18)
.
The cyclic nucleotide response element binding protein (CREB) is a
4346 kDa nuclear transcription factor that recognizes the highly
conserved sequence known as cAMP-responsive element (CRE):
5'-TGACGTCA-3' (19)
. CREB activation results from
post-translational modifications, such as phosphorylation of serine-133
(Ser-133) in situ by PKA or calmodulin kinase, following
increases in intracellular cAMP or Ca2+ levels,
respectively (20
, 21)
. More recently it has been shown
that the activation of the ERK/MAPK pathway can also induce CREB
phosphorylation on Ser-133 (22)
. Phosphorylation of CREB
at Ser-133 represents a critical step for trans-activation
of CRE-dependent gene promoters (19)
. Whereas the
expression of the CREB gene is constitutive at low basal levels in many
tissues (23)
, CREB mRNA and protein are synthesized
cyclically at high levels in the testis (24)
. Moreover,
CREB expression depends mainly on the CRE elements present in its
promoter (25)
, suggesting that autoregulation may affect
the transcriptional expression of the CREB gene.
Since it has been clearly established that CREB plays a central role in
promoting the survival/function of neuronal cells in response to
neurotrophic factors (22
, 26
, 27)
, the aim of this study
was to investigate whether HIV-1 Tat protein affects the Ser-133
phosphorylation levels and expression of CREB in neurons. We chose as a
model system the rat PC12 neuronal cell line, which has been shown to
reproduce many of the biological responses to Tat protein observed
in vivo or in cultured primary neurons (4
5
6
7
8
9
10
11
12
13
14
15
16
17
18)
.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Rat pheochromocytoma PC12 cells were obtained from American Type Culture Collection (Rockville, Md.) and cultured in D-MEM (Gibco Laboratories, Grand Island, N.Y.) supplemented with 10% horse serum (HS, Gibco) plus 5% fetal calf serum (FCS, Gibco). PC12 cells were serum-starved in D-MEM plus 0.1% FCS for 40 h before treatment with Tat protein (0.11 nM corresponding to 110 ng/ml), forskolin (10 µM), either alone or in combination, for 5120 min. Equivalent volumes of PBS containing 0.1% BSA were used as negative control. In blocking experiments performed with pharmacological inhibitors, PC12 cells were pretreated for 1 h at 37°C with H89 (1 µM), SB203580 (10 µM), IBMX (0.1 mM), LY294002 (10 µM), wortmannin (0.1 µM), or equal volumes of DMSO.
Western blotting
For analysis of Ser-133 CREB, total CREB, tyrosine
phosphorylated and total AKT and ß-tubulin proteins, Western blot was
performed on
10 x 106 cells per
experimental point. Cell pellets were supplemented at 4°C with a
lysis buffer containing 1% deoxycholate, 1 µg/ml aprotinin, 2
µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 1 mM sodium
orthovanadate (without Triton X-100) for 10 min. Cell lysates were
sonicated and either immediately processed by Western blot or
centrifuged at 100,000 g for 1 h at 4°C. In the
latter case, the supernatants were discharged and the pellets,
containing the nuclei, were solubilized in 100 µl of lysis buffer
containing 0.1% Triton X-100. Protein concentrations were estimated by
the Bio-Rad protein assay (Bio-Rad, Richmond, Calif.) according to the
manufacturers protocol. Equivalent (100 µg) amounts of proteins per
sample were subjected to electrophoresis on a 10% sodium dodecyl
sulfate-acrylamide gel. The gel was then electroblotted onto a
nitrocellulose membrane; equal loading of protein in each lane was
confirmed by brief staining of the blot with 0.1% Ponceau S, followed
by destaining prior to reacting with the specific antibodies. Blotted
membranes were blocked for 30 min in a 3% suspension of dried skimmed
milk in PBS and incubated overnight at 4°C with 1) a
rabbit polyclonal anti-CREB serum, 2) a rabbit serum
directed against the phosphorylated Ser-133 form of CREB (both from
Upstate Biotechnology Inc., Lake Placid, N.Y.; 1:1000 dilution),
3) a monoclonal anti-tyrosine phosphorylated AKT,
4) a monoclonal anti-AKT (both from PharMingen/Transduction
Laboratories, San Diego, Calif.; dilution 1:250), or 5) a
monoclonal anti-ß-tubulin (Sigma; dilution 1:500). Filters were
washed and incubated for 1 h at room temperature with
peroxidase-conjugated anti-rabbit immunoglobulin G (IgG) (to detect
Ser-133 phosphorylated and whole CREB, 1:1500 dilution, Sigma) or
anti-mouse IgG (to detect tyrosine phosphorylated or whole AKT or
ß-tubulin, 1:1500 dilution, Sigma) in 0.1% BSA. Specific reactions
were revealed with the ECL Western blotting detection reagent (Amersham
Corp., Arlington Heights, Ill.).
Assay of cAMP
The intracellular cAMP levels were measured using the cAMP assay
kit (Amersham) according to the manufacturer instructions. Briefly,
after 40 h of starvation in D-MEM + 0.1% FCS, 100 mm plates of
PC12 cells were treated with synthetic Tat (1 nM), BSA solution (0.1%)
or forskolin (10 µM) for up to 120 min. The incubation was stopped by
the addition of the lysis buffer provided by the manufacturer. The
intracellular cAMP content of the samples was measured by ELISA and
expressed as pM/100 µg of proteins.
Plasmids and transfection experiments
The following plasmids, wild-type -1264 pCREB-CAT, -1264°CRE
1 + 2 mut pCREB-CAT, mutated in the CRE sites, and poCAT empty vector
(25)
, were a generous gift of Dr. Habener (Massachusetts
General Hospital, Harvard Medical School, Boston, Mass.).
Transient transfection experiments were performed using the PerFect transfection kit reagent (Invitrogen, Carlsbad, Calif.) following manufacturers instructions. Briefly, semi-confluent T35 culture flasks of PC12 were transfected with 510 µg of plasmid DNA. Thirty-six hours post-transfection, cells were treated with synthetic Tat (0.110 nM), forskolin (10 µM) used alone, or in various combinations. Twelve hours after treatment, cell lysates were assayed for chloramphenicol acetyl transferase (CAT) activity, using volumes of extract corresponding to equal protein amounts.
Detection of apoptosis
Apoptosis was evaluated in PC12 cells by using the TdT-mediated
D-UTP-biotin nick end labeling (TUNEL) technique to monitor the DNA
fragmentation in situ, following a previously described
procedure (28)
.
Statistical analysis
Statistical analysis was performed using the two-tailed
Students t test.
| RESULTS |
|---|
|
|
|---|
|
|
Bimodal changes of the intracellular cAMP levels in Tat-treated
PC12 cells
Among several pathways that can induce CREB phosphorylation on
Ser-133 in neuronal cells (20
21
22
23
, 26
, 27
, 29)
, a major
role is played by PKA. Thus, to ascertain whether PKA was involved in
Tat-mediated CREB phosphorylation in PC12 cells, we used the
pharmacological compounds H89, a specific inhibitor for the PKA
pathway, and SB203580, an inhibitor of the p38 MAPK pathway, used as
control. Western blot analysis of PC12 nuclear cell lysates showed that
H89, but not SB203580, significantly (P<0.05) inhibited the
Tat-mediated Ser-133 CREB phosphorylation (Fig. 2
and Table 1
). This suggested that activation of the cAMP/PKA pathway
was required for the Tat-induced CREB phosphorylation in these cells.
|
To confirm this hypothesis, the intracellular levels of cAMP were
analyzed after treatment with Tat, used alone (Fig. 3A
) or in combination with forskolin (Fig. 3B
).
Consistent with a role for the cAMP/PKA pathway in inducing CREB
phosphorylation in PC12 cells, Tat stimulated a rapid (peak at 5 min)
elevation of intracellular cAMP over the basal levels observed in
control (BSA-treated) cells (Fig. 3A
). At later time points,
the intracellular levels of cAMP in Tat-treated cells progressively
declined and were significantly (P<0.01, at 60120 min)
lower than those found in BSA-treated control cells (Fig. 3A
).
|
To further investigate this biphasic effect of Tat on the intracellular
levels of cAMP, particularly the secondary inhibitory phase,
experiments were performed by treating PC12 cells with a combination of
Tat plus forskolin. As expected, treatment of PC12 cells with forskolin
induced a rapid and persistent elevation of intracellular cAMP (Fig. 3B
). The simultaneous addition of Tat and forskolin to PC12
cells resulted in an additive (P<0.05) effect of the two
agonists on the cAMP levels at early time points (5 min). On the other
hand, at later time points (60120 min) the intracellular levels of
cAMP were significantly (P<0.01) lower in cells treated
with Tat + forskolin than in cells treated with forskolin alone (Fig. 3B
). Moreover, pretreatment of PC12 cells with IBMX, a broad
inhibitor of cyclic nucleoside phosphodiesterases (PDE), completely
abrogated the ability of Tat to decrease the levels of cAMP in PC12
cells costimulated with forskolin (Fig. 3C
). Taken together,
these findings suggest that the decline of intracellular cAMP levels
observed in Tat-treated cells was mediated by PDE activity.
In parallel, aliquots of the same samples examined for intracellular
cAMP were subjected to Western blot to analyze the levels of CREB
Ser-133 phosphorylation (Fig. 4
). At 510 min, the combination of Tat plus forskolin showed an
additive effect (P<0.05) on CREB Ser-133 phosphorylation,
as demonstrated by the densitometric analysis of the Western blot bands
(Fig. 4
and Table 1
). At 60120 min, the same combination induced
levels of CREB Ser-133 phosphorylation significantly
(P<0.05) lower than forskolin alone (Fig. 4
and Table 1
).
These changes in Ser-133 phosphorylation took place in the absence of
significant modifications of total nuclear CREB (Fig. 4)
. Changes in
CREB Ser-133 phosphorylation levels closely paralleled those in
intracellular cAMP (Fig. 3A
, B
). In this and other
experiments, a doublet of immunoreactive proteins instead of a single
band was sometimes observed. The faster migrating band likely
represents a different CREB isoform (21
, 26)
. Due to the
proximity of the two phosphorylated bands, stripping, and reprobing
with the anti-CREB serum did not allow us to discriminate which band
represented CREB. Densitometric analysis was then calculated
considering an area comprising both phosphoproteins.
|
Tat stimulates PDE through a PI-3K/AKT pathway
We next investigated how extracellular Tat induced the PDE
activity responsible for the secondary decrease in the intracellular
cAMP levels and in CREB Ser-133 phosphorylation. We took advantage of
the fact that it had been shown that some PDE isoforms can lie
downstream the serine-threonine kinase AKT (30
31
32)
, which
in turn is activated by PI 3-K (33
, 34)
. Having previously
demonstrated that extracellular Tat activates PI 3-K in PC12 cells
(15)
through p125FAK tyrosine kinase
(16)
, we next asked whether Tat also activates AKT in
these cells and whether AKT may be an intermediate in Tat-induced
stimulation of PDE. As shown in Fig. 5A
, synthetic Tat (0.11 nM) rapidly stimulated AKT.
Moreover, pretreatment of PC12 cells with two unrelated pharmacological
inhibitors of PI 3-K (10 µM LY294002 and 0.1 µM wortmannin) before
Tat addition significantly increased CREB Ser-133 phosphorylation in
PC12 cells with respect to cells treated with vehicle containing DMSO
(Fig. 5B
and Table 1
). These data strongly suggested that
Tat promoted PDE activity through a PI 3-K/AKT-dependent pathway.
|
Inhibition of CREB promoter activity and CREB protein expression
after prolonged treatment of PC12 cells with Tat
It has been shown that steady-state levels of CREB are extremely
sensitive to variations of intracellular cAMP, due to the presence of
cAMP response elements in its promoter (25)
. Thus, in the
next group of experiments we investigated whether the ability of
extracellular Tat to modulate the levels of intracellular cAMP and CREB
Ser-133 phosphorylation could affect CREB transcription. PC12 cells
were transfected with a plasmid containing the intact CREB promoter
(wild-type -1264 pCREB-CAT) and a CREB promoter mutated in the CRE
sites (-1264°CRE 1+2 mut pCREB-CAT), fused to CAT (25)
.
The basal transcriptional activity of wild-type pCREB-CAT was low but
clearly detectable in BSA-treated PC12 cells, and was significantly
(P<0.05) reduced by the addition of Tat for the last
12 h of culture (Fig. 6A
). On the other hand, no CAT activity was detected when PC12
cells were transfected with either mut pCREB-CAT or poCAT empty vector
(data not shown). This suggested that the Tat-mediated activation of
PDE (Fig. 3A
) was able to down-modulate the CREB promoter
activity in PC12 cells.
|
This hypothesis was corroborated in next experiments, in which PC12
cells were transfected with pCREB-CAT and then treated with forskolin
plus Tat (Fig. 6B
). As expected, due to the ability of
forskolin to elevate the cAMP levels (Fig. 3B
), when
pCREB-CAT transfected PC12 cells were treated with forskolin, a
pronounced increase of CREB promoter activity was noticed. The addition
of Tat resulted in a dose-dependent inhibition of the forskolin-induced
pCREB-CAT activity with respect to cells treated with forskolin plus
BSA (Fig. 6B
). This inhibitory effect of Tat on
forskolin-induced pCREB-CAT activation was also consistently observed
when transfected PC12 cells were treated with Tat 1 h before
adding forskolin (Fig. 6C
). On the other hand, no
significant variations were noticed when Tat was added 1 h after
forskolin (Fig. 6C
). The transcriptional activity of mut
pCREB-CAT was also undetectable in cultures stimulated with forskolin
used alone or in combination with Tat, confirming that the CRE sites
played a central role in driving CREB transcription (data not shown).
To verify the relevance of Tat-induced down-modulation of CREB
promoter, in the next group of experiments we analyzed the total amount
of CREB protein in PC12 cells after prolonged Tat treatment (Fig. 7
). For this purpose, cell lysates obtained from whole PC12 cells and
nuclei were analyzed for the content of CREB and its levels of
phosphorylation on Ser-133 after 2448 h from the addition in culture
of 1 nM Tat. As shown in Fig. 7
and Table 2
, a significant decline in both Ser-133-phosphorylated CREB and whole
CREB protein was reproducibly observed between 24 and 48 h. This
decrease could not be ascribed to a nonspecific loss of proteins since
equal loading of proteins in each lane was confirmed by staining the
blot with Ponceau (not shown) and by densitometric analysis of
ß-tubulin (Table 2)
. The possibility that the loss in CREB protein
expression was due to a nonspecific cytotoxicity of Tat was ruled out
by an analysis of apoptosis at different time points in Tat- and
BSA-treated cells (Table 3
). In fact, in four separate experiments the percentage of apoptosis,
analyzed by using the TUNEL technique, was lower (P<0.05)
in PC12 cells treated with Tat then in cells treated with BSA (0.1%)
up to 24 h post-treatment, whereas no significant differences were
observed after 48 h (Table 3)
. At later time points (7296 h
post-treatment), however, a moderate but significant
(P<0.01) increase of apoptosis was noticed in Tat-treated
with respect to BSA-treated cultures. Thus, the decrease of CREB
protein expression anticipated the loss in cell viability.
|
|
|
| DISCUSSION |
|---|
|
|
|---|
At the neuropathological level, a range of abnormalities has been
reported in persons with HIV-1 dementia, including extensive loss of
neurons within certain regions of the brain (37
, 38)
and
high levels of neuronal apoptosis (39
, 40)
. The induction
of neuronal dysfunction is believed to occur as a result of the
production and release of a number of neurotoxic factors, which include
HIV-1 Tat protein, that can be released locally by infected microglial
cells or across the bloodbrain barrier (41)
. Detectable
levels of Tat protein have been demonstrated to be present in the sera
of HIV-1-seropositive patients (42)
, and Tat can easily
across the bloodbrain barrier (43)
. Thus, it is possible
that also systemic Tat may reach the central nervous system. How
extracellular Tat elicits its biological effects on neuronal cell
survival/growth is still incompletely understood. In fact, Tat protein
can be taken up by intact neurons and localizes to the nucleus quite
rapidly both in vitro (4)
and in
vivo (44)
. In addition, Tat interacts with a variety
of surface receptors, including receptors for integrins
(45)
, members of the vascular endothelial growth factor
receptor family (46)
, and receptors for chemokines
(47)
.
Previous studies have shown that at micromolar concentrations, Tat can
be neurotoxic by inducing rapid and massive apoptotic cell death of
neuronal cells both in vitro and in vivo
(4
5
6
7
8
9
10
11
12
13)
. These high concentrations of Tat cause
neurotoxicity by direct neuronal depolarization (9)
, by
increasing levels of intracellular calcium (8
, 11)
, and by
activating excitatory amino acid receptors (6
, 10)
.
In this study, we have demonstrated for the first time that synthetic extracellular Tat also shows subtler, but potentially noxious effects on neuronal cell survival at concentrations as low as 0.11 nM. In fact, after an initial rapid induction of CREB phosphorylation on Ser-133, Tat induced a secondary, prolonged down-regulation of CREB expression. These changes in CREB phosphorylation and expression anticipated and correlated with the bimodal effect of low Tat concentrations on PC12 cell viability: early protection from apoptosis induced by serum withdrawal (up to 24 h), followed by secondary increase of apoptosis (7296 h) with respect to BSA-treated cells.
CREB is encoded by a gene that contains at least 11 exons spread over
more than 40 kilobases of DNA located on the long arm of human
chromosome 2 (48
, 49)
. CREB is the prototype of the
CREB/CREM/ATF superfamily of bZIP or leucine zipper transcription
factors, including similar factors with a basic DNA binding domain that
recognize the CRE motif. Due to its essential role for neuronal cell
survival (26
, 27)
and as a mediator of long-term memory
(50)
, the Tat-induced down-regulation of CREB expression
is expected to have profound detrimental effects on neuronal cell
survival and/or function. What is particularly remarkable is that these
effects were obtained at relatively low Tat concentrations.
In considering the possible levels of Tat that may be present in
vivo, concentrations of 0.010.1 nM of Tat in the sera of
HIV-1-infected individuals have been reported (42)
. Tat
can be expected to be present at higher concentrations in tissues with
active HIV-1 replication or in proximity to cells with productive HIV-1
infection, such as in regions containing inflammatory infiltrates and
productively infected microglia and macrophages (51)
.
Therefore, although Tat within the brain has never been measured, it is
not unreasonable to expect that concentrations of around 0.11 nM may
be present in close proximity to virus-positive microglia or brain
macrophages. However, in our opinion the neurotoxic effects associated
with higher (µM) concentrations of Tat (4
5
6
7
8
9
10
11
12
13)
are very
unlikely to occur in vivo in the brain of HIV-1 infected
individuals.
Although we cannot exclude that the down-regulation of CREB requires
internalization and nuclear localization of extracellular Tat protein
(4)
, our data strongly suggest that Tat down-modulates
CREB gene expression, altering the intracellular levels of cAMP. In
fact, even though the early and rapid elevation of cAMP is likely due
to the activation of adenylate cyclase, the signal transduction pathway
involved in the secondary CREB down-regulation requires activation of a
PI 3-K/AKT/PDE pathway. It has previously been shown that the steady
state, intracellular levels of cAMP are controlled predominantly by PDE
(52)
. It is noteworthy that CREB expression is extremely
sensitive to variations of the intracellular cAMP levels due to the
presence of critical CRE elements in its promoter (25)
.
Although this study was performed using a neuronal cell line in
vitro and there may be important differences with in
vivo systems, our data strengthen the notion that Tat may be an
important cause of neurocognitive dysfunction in HIV-1-seropositive
individuals. Moreover, we have provided a molecular mechanism to
explain how low concentrations of extracellular Tat can impair the
survival/function of neuronal cells by decreasing the levels of cAMP
and the expression of CREB protein. Several authors, including
ourselves, have previously proposed and shown the potential usefulness
of a Tat vaccination to slow or even arrest the progression toward AIDS
in both nonhuman primates and human beings infected with simian
immunodeficiency virus or HIV-1, respectively (3
, 53
54
55
56)
.
If our present findings were confirmed with in vivo studies,
they might suggest that a vaccination of HIV-1-seropositive individuals
with Tat protein would be beneficial to counteract the occurrence of
neurological abnormalities, frequently observed in these
patients.
| ACKNOWLEDGMENTS |
|---|
Received for publication June 1, 2000.
Revision received August 2, 2000.
| REFERENCES |
|---|
|
|
|---|
and activation of non-N-methyl-D-aspartate receptors by a NF
B-independent mechanism. J. Biol. Chem. 273,17852-17858This article has been cited by other articles:
![]() |
C. V. Garat, D. Fankell, P. F. Erickson, J. E.-B. Reusch, N. N. Bauer, I. F. McMurtry, and D. J. Klemm Platelet-Derived Growth Factor BB Induces Nuclear Export and Proteasomal Degradation of CREB via Phosphatidylinositol 3-Kinase/Akt Signaling in Pulmonary Artery Smooth Muscle Cells. Mol. Cell. Biol., July 1, 2006; 26(13): 4934 - 4948. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. H. Dworet and J. L. Meinkoth Interference with 3',5'-Cyclic Adenosine Monophosphate Response Element Binding Protein Stimulates Apoptosis through Aberrant Cell Cycle Progression and Checkpoint Activation Mol. Endocrinol., May 1, 2006; 20(5): 1112 - 1120. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Wong, A. Sharma, S. Awasthi, E. F. Matlock, L. Rogers, C. Van Lint, D. J. Skiest, D. K. Burns, and R. Harrod HIV-1 Tat Interactions with p300 and PCAF Transcriptional Coactivators Inhibit Histone Acetylation and Neurotrophin Signaling through CREB J. Biol. Chem., March 11, 2005; 280(10): 9390 - 9399. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Taguchi, T. Kondo, M. Watanabe, M. Miyaji, H. Umehara, Y. Kozutsumi, and T. Okazaki Interleukin-2-induced survival of natural killer (NK) cells involving phosphatidylinositol-3 kinase-dependent reduction of ceramide through acid sphingomyelinase, sphingomyelin synthase, and glucosylceramide synthase Blood, November 15, 2004; 104(10): 3285 - 3293. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. C. Deregibus, V. Cantaluppi, S. Doublier, M. F. Brizzi, I. Deambrosis, A. Albini, and G. Camussi HIV-1-Tat Protein Activates Phosphatidylinositol 3-Kinase/ AKT-dependent Survival Pathways in Kaposi's Sarcoma Cells J. Biol. Chem., July 5, 2002; 277(28): 25195 - 25202. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Hay and G. Kannourakis A time to kill: viral manipulation of the cell death program J. Gen. Virol., June 1, 2002; 83(7): 1547 - 1564. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. B. Reusch and D. J. Klemm Inhibition of cAMP-response Element-binding Protein Activity Decreases Protein Kinase B/Akt Expression in 3T3-L1 Adipocytes and Induces Apoptosis J. Biol. Chem., January 4, 2002; 277(2): 1426 - 1432. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||