(Received for publication, September 16, 1996)
From the Neuroscience Department, Pharmaceutical
Discovery Division, Abbott Laboratories, Abbott Park, Illinois
60064-3500 and the ¶ Department of Psychiatry, University of
Louisville School of Medicine, Louisville, Kentucky 40292
Recent evidence indicates that tumor necrosis
factor- (TNF-
) is up-regulated following brain injury and in
neurodegenerative disorders such as stroke, multiple sclerosis,
Parkinson's disease, and Alzheimer's disease. TNF-
elicits its
biological effects through two distinct TNF receptor (TNFR) subtypes:
p55 TNFR (TNFR1) and p75 TNFR (TNFR2). Studies have demonstrated that
the p55 TNFR contributes to cell death, whereas the role of the p75
TNFR in neuronal viability is unclear. To better understand the role of p75 TNFR, we treated human neuronal SH-SY5Y cells with
phosphorothioate-modified antisense oligonucleotides (ASO) for p75 TNFR
and established that ASO inhibited p75 TNFR expression. Treatment of
SH-SY5Y cells with ASO alone did not affect cell viability, whereas
treatment with both ASO and human TNF-
significantly increased cell
death relative to treatment with TNF-
alone. Moreover, addition of ASO significantly increased the level of cell injury observed following
hypoxic conditions or exposure of
-amyloid peptide. These results
indicate that inhibition of p75 TNFR using ASO increases the
vulnerability of neurotypic cells to insults and suggest that the p75
TNFR may not be required for normal neuronal cell viability but rather
plays a protective role following injury.
Tumor necrosis factor- (TNF-
),1
an inflammatory mediator, is recently reported to be up-regulated
following brain trauma (1, 2) and ischemic injury (3, 4), and in
multiple sclerosis (5, 6), Parkinson's disease (7), and Alzheimer's disease (8). Studies have shown that TNF-
can elicit either a
trophic or toxic effect, which is dependent on the target cell type.
For example, TNF-
is toxic to human oligodendrocytes (9) and
neuronal cells (10, 11), but it is trophic to rat hippocampal neurons
(12). Since TNF-
elicits its biological effects through activation
of two distinct receptors, p55 TNFR and p75 TNFR, both of which have
been cloned (13-15), we hypothesized that the distinct TNF receptor
subtypes might contribute to the multiple biological functions of
TNF-
. Reports indicate that the p55 TNFR may contribute to cell
death (16), whereas the role of the p75 TNFR in neuronal cell function
is not clear. Here we show that inhibition of p75 TNFR by antisense
oligonucleotides (ASO) increased the magnitude of neuronal cell death
induced by TNF-
, hypoxic injury or
-amyloid peptide (A
) as
measured by lactate dehydrogenase (LDH) release. This novel finding
provides evidence that the p75 TNFR might play a protective role
following neuronal insult.
Human neuronal-like cells, SH-SY5Y, were cultured in a 1:1 ratio of minimum Eagle's medium/F-12 medium, 15% fetal calf serum (Life Technologies, Inc.), and 10 µM retinoic acid (RA, Sigma). The cells were seeded at 25,000 cells/well in a 24-well plate for cytotoxicity assays or 3 × 106 cells in Petri dishes for radioimmunoprecipitation study. The medium was replaced every 3 days. Treatments were given to the cells in serum-free medium with 1:100 N2-supplement (Life Technologies, Inc.) 50% minimum Eagle's medium and 50% F-12.
RT-PCRThe PCR technique was used to amplify cDNA
derived from RNA extracted from RA-differentiated SH-SY5Y cells as
described previously (17). The PCR-forward oligonucleotide primer for
p75 TNFR was 5-GGTCACGCAACCTGTCTT-3
, and backward oligonucleotide
primer for p75 TNFR was 5
-GGCTTCATCCCAGCATCA-3
. PCR-forward
oligonucleotides primer for p55 TNFR was 5
-TCGATTTGCTGTACCAAGTG-3
,
and backward oligonucleotides primer for p55 TNFR was
5
-GAAAATGACCAGGGGCAACAG-3
. After an initial denaturation step at
94 °C for 5 min, the cycle was initiated, which consisted of
denaturing for 1 min at 94 °C, annealing for 2 min at 53 °C, and
extending for 2 min at 72 °C. The cycle was repeated 35 times.
Control reactions in the absence of RT with RNA template yielded no
detectable product. A proportion (25%) of the reaction mixture was run
in an agarose gel (1.2%).
Both
sense and antisense p75 TNFR oligonucleotide phosphorothioates were
synthesized on an Applied Biosystems model 392 DNA synthesizer using
phosphoramidite chemistry. Oligonucleotides were purified by
reverse-phase chromatography using Oligo-Pak oligonucleotide
purification columns (Milligen) as reported (18). The p75 TNFR
antisense oligonucleotide sequence was 5-AGCACATCTGAGCTGTCT-3
overlapping to the code for the initiation site of methionine on the
human p75 TNFR gene (19). The p55 TNFR antisense oligonucleotide sequence was 5
-CCACCTCTCCGGGTACGG-3
overlapping to the code for the
initiation site of methionine on the human p55 TNFR gene (20). The
sense oligonucleotides for each receptor subtype were the exact inverse
complement of the antisense oligonucleotides. Lyophilized
oligonucleotides were dissolved in sterile water to prepare stock
solutions, which were stored at
70 °C until use.
The cells were first transferred into a serum-free medium and then placed in a modular incubator containing 5% carbon dioxide and 95% nitrogen at 37 °C. Following desired exposure (6 h), cultures were returned to a normoxic atmosphere of 5% carbon dioxide and 95% room air at 37 °C.
Preparations of Human Tumor Necrosis Factor-Human tumor necrosis factor- (R & D Systems) was
aliquoted into 1 µg/ml of medium and then diluted into 20 ng/ml with
serum-free N2 medium. A
-(1-42) was dissolved in dry
Me2SO at 6.5 mM and then diluted with
serum-free N2 medium.
RA-differentiated SH-SY5Y cells (3 × 106) were cultured with the medium without methionine and cysteine for 2 h followed by incubation with [35S]methionine and [35S]cysteine (50 µCi/ml, DuPont NEN) in methionine- and cysteine-free medium at 37 °C. Fifteen hours later, cells were washed twice for 15 min each. Cell lysates were harvested and centrifuged at 45,000 rpm at 4 °C for 30 min, incubated with p75 TNFR antibody (polyclonal detection antibody, R & D Systems, 1:500) for 16 h at 4 °C, and precipitated with Staphylococcus protein A (Boehringer Mannheim). The immunoprecipitates were resolved in a 10% SDS-polyacrylamide gel electrophoresis. After electrophoresis, gels were fixed in 10% acetic acid and 50% methanol solution and then incubated in Enlightning (DuPont) for 30 min with rocking. Bands in dried gels were detected by autoradiography.
Western Blotting ProcedureRA-differentiated SH-SY5Y cells were cultured without fetal calf serum for 24 h. Cells were then lysed with deoxycholate (5 mg/ml), phenylmethylsulfonyl fluoride (0.175 mg/ml), and aprotinin (100 kallikrein-inactivating units/ml). The lysate was analyzed by SDS-polyacrylamide gel electrophoresis (Enprotech) 10-20% Tris/glycine gradient gel and then electrotransferred onto Immobilon-P (Millipore). Blots were saturated in a 5% nonfat dry milk phosphate-buffered saline with Tween 20 for 60 min at room temperature and then incubated with p55 TNFR antibody 1:1000 (polyclonal detection antibody, R & D Systems) for 60 min. After 3 washings with 0.1% Tween 20/phosphate-buffered saline, horseradish peroxidase-conjugated antibody 1:1000 (Amersham) was added to the blots for 2 h at room temperature. Finally, the blots were washed with Tween 20/phosphate-buffered saline and developed with the ECL Western blotting detection system (Amersham).
Cytotoxicity AssayFor quantitative assessment of neuronal cell damage, the release of LDH from degenerating neurons was measured using a CytoTox 96 non-radioactive cytotoxicity assay kit (Promega) as described previously (17). This biochemical index correlated with the morphological estimate. The percentage of LDH release was calculated as the portion of LDH in supernatant over total LDH from both supernatant and cell lysate.
We used RA-differentiated SH-SY5Y
cells, considered to represent human neuronal-type cells, as our
in vitro model. Results from RT-PCR demonstrated that both
the p55 and p75 TNFRs were detected at the message level (Fig.
1) in RA-differentiated SH-SY5Y cells. The sequences of
two excised DNA bands of both TNFRs are identical to published ones
(data not shown). Western blotting or immunoprecipitation studies have
shown that these receptors were expressed at the protein level (Fig.
2a, p75 TNFR; Fig. 2b, p55 TNFR).
These results are consistent with a report that another human
neuronal-like cell line (SKNBE) expresses endogenous TNFRs (21). To
investigate a potential role of p75 TNFR in neuronal cell survival, we
synthesized phosphorothioate-modified ASO for p75 TNFR in which the
nucleotide sequence overlaps the code for the methionine initiation
site on the human p75 TNFR gene (14). We confirmed by
immunoprecipitation that treatment of RA-differentiated SH-SY5Y cells
with ASO (5-10 µM) directed at the p75 TNFR blocked expression of p75 TNFR (Fig. 2a, ASO-1 and ASO-2).
Expression of p75 TNFR was almost completely inhibited by ASO at 10 µM (Fig. 2a, ASO-2) compared with that at 5 µM (Fig. 2a, ASO-1). Moreover, a number of
studies were conducted to establish the specificity of the ASO
treatment (22, 23). Control experiments in RA-differentiated SH-SY5Y
cells showed that expression of p75 TNFR protein was not affected in
cells treated with hypoxia or A-(1-42) alone, sense oligonucleotides (SO) alone, or co-treated with SO and hypoxia or
A
-(1-42) (Fig. 2a). Furthermore, protein expression of
the p55 TNFR was not affected in RA-differentiated SH-SY5Y cells
following treatment with the p75-directed ASO (Fig. 2b).
Increased Cell Degeneration Induced by hTNF-
To assess
whether oligonucleotide treatments altered cell viability under basal
conditions, the levels of LDH released from RA-differentiated SH-SY5Y
cells treated with ASO and from cells treated with the corresponding SO
for p75 TNFR were compared to non-treated control levels. No
significant differences in LDH levels were detected following either
ASO or SO for p75 TNFR treatment (10 µM) compared with
the no treatment control (Fig. 3). Furthermore, no
significant differences in LDH levels were observed in the cells
treated with either ASO or SO for p55 TNFR treatment (10 µM) compared with the no treatment control (data not
shown). Moreover, co-treatment of RA-differentiated SH-SY5Y cells with
ASO for p75 TNFR (10 µM) and hTNF- (20 ng/ml) produced
significantly higher levels of LDH release compared to hTNF-
treatment alone (Fig. 3). Co-treatment with SO and hTNF-
did not
affect the level of LDH release relative to hTNF-
treatment alone
(Fig. 3). To support the findings that inhibition of p75 TNFR
expression increases vulnerability of the cell to hTNF-
, we added
anti-human p75 TNFR monoclonal neutralizing antibody into the cells to
neutralize the biological activity of human p75 TNFR but not p55 TNFR
(21, 22) and then treated the cells with hTNF-
(20 ng/ml). We found a similar increase in LDH release relative to that observed following treatment with ASO for p75 TNFR (Fig. 3).
Increased Cell Degeneration Induced by Hypoxic Injury and
RA-differentiated SH-SY5Y cells exposed to oxygen
deprivation for 6 h displayed morphological changes,
e.g. swollen cell bodies (data not shown), and produced
increased levels of LDH release relative to controls (Fig. 3).
Treatment of ASO for p75 TNFR during the 6 h of hypoxic exposure
resulted in a significant increase in LDH release compared with hypoxia
alone or treatment with SO under the identical hypoxic conditions (Fig.
3). We found a similar increase in LDH release from hypoxic cells
treated with anti-human p75 TNFR antibody (Fig. 3). We also extended
our studies by using A-(1-42). Co-treatment of RA-differentiated
SH-SY5Y cells with the ASO (10 µM) and A
(5 µM) produced significantly higher levels of LDH release
compared with A
treatment alone (Fig. 4). Treatment of RA-differentiated SH-SY5Y cells with a combination of SO and A
did not affect the level of LDH release relative to A
treatment alone (Fig. 4).
The observation that ASO directed for p75 TNFR does not alter cell
viability despite inhibition of p75 TNFR protein expression suggested
that p75 TNFR may not be required for normal neuronal functioning under
basal conditions. Our data also demonstrated that inhibition of p75
TNFR using ASO increased vulnerability of neurotypic cells toward
injury induced by hTNF-, hypoxia, or A
. These results were
further supported by similar results obtained using anti-human p75
TNFR-neutralizing antibody, suggesting that p75 TNFR may play a
critical role in the ability of neurons to respond to insults. In
addition, treating human neurotypic cells with ASO in vitro
has proven to be a useful technique to suppress specific gene
expression and to enable study of the functional significance of the
corresponding protein (24, 25).
Neurons, akin to a number of other cell types, can express both p55 and
p75 TNFRs (19, 26, 27); the genes for these receptors are located on
human chromosomes 12 and 1, respectively. Based on amino acid
sequencing, there is a 28% overall homology between p55 and p75 TNFRs.
The amino acid sequences in the extracellular domains, transmembrane
region, and intracellular domains between these two TNFR subtypes share
22, 28, and 10% homologies, respectively. The low level of homology
suggests that distinct functions may be associated with each of these
two receptors. Thus, TNF- can elicit either a trophic or toxic
response in neurons, which could be dependent on which TNF receptor
subtype is activated. Recent studies indicate that the p55 TNFR is
responsible for cell death signaling through activation of
phospholipase A2 and NF-
B (28). Recent studies reported
that knocking out both p55 and p75 TNF receptors increased
neurodegeneration induced by ischemia in mouse brain (29). Our results
are the first to indicate that the p75 TNFR may protect human neurons
against insults, e.g. exposure to hypoxia or
-amyloid.
Thus when a neuron is insulted, TNF-
may be released to activate the
p75 TNFR (30, 31), which in turn may trigger a series of cellular
events, e.g. secretion of trophic factors (32), attenuation
of calcium disruption, or free radical formation (12), that enables the
cell to defend itself against the insult. Therefore, p75 TNFR may be
necessary for neuronal cell survival under pathological conditions
following injury, and conditions that attenuate levels or functioning
of the p75 TNFR may render cells more vulnerable to insult and
disease.
We thank Tom Kavanaugh for providing
phosphorothioate-modified oligonucleotides and Richard Perner for
providing A-(1-42). We also appreciate very much Drs. Mike
Williams, James Kerwin, and Art Hancock for their helpful discussion,
critical comments, and support.