(Received for publication, February 28, 1997, and in revised form, May 5, 1997)
From the Department of Pathology and Kaplan Cancer Center, New York University School of Medicine, New York, New York 10016
Daudi B lymphoblastoid cells are highly sensitive
to the anti-growth and anti-viral effects of interferon (IFN). Unlike
many cell lines, these cells show prolonged transcription of
IFN-stimulated genes following treatment with IFN-. This prolonged
response correlated with the continued presence of the activated
transcription factor, IFN-stimulated gene factor 3 (ISGF3). Pulse-chase
labeling experiments indicated that the half-life of the
phosphorylation of signal transducers and activators of transcription
(Stat)1 and Stat2 was short (<2 h) although the turnover of the
proteins themselves was slow (>24 h), indicative of a constitutive
phosphatase activity. The administration of protein-tyrosine kinase
inhibitors at any time point during IFN stimulation led to rapid
inhibition of the response, indicating that tyrosine kinase activity
was continuously required. Catalytic activity of Jak1 and Tyk2 kinases remained elevated for prolonged periods following stimulation. Continuous presence of IFN-
was necessary for maintaining prolonged activation of ISGF3 and of Janus kinases, an activity that was blocked
by antibodies to IFN-
or by cycloheximide. Conditioned medium of
IFN-
-stimulated cells was capable of stimulating STAT activation in
naive cells. Taken together, these results suggest that the response to
IFN-
is controlled by the duration of stimulated Janus kinase
activity over the background of constitutive dephosphorylation and that
this response can be sustained by autocrine secretion of IFN-
.
Interferons (IFNs)1 are
cytokines with a wide variety of functions, including modulation of
immune responses, inhibition of proliferation, and induction of an
anti-viral state and of resistance to bacterial and parasitic
infection. There are two types of IFN: type I IFN, comprised of many
IFN- genes and one IFN-
gene, and type II IFN, consisting of a
single IFN-
species. These two types of IFN produce different but
somewhat overlapping responses in target cells following binding to two
distinct receptors. The receptors for both types of IFN are composed of
at least two transmembrane proteins involved in signal transduction.
The binding of IFNs to their receptors causes receptor dimerization,
triggering signal transduction by activation of intracellular tyrosine
kinases of the Janus kinase (JAK) family.
The IFN-induced signal pathway has been elucidated recently (1, 2).
During the response to IFN-, Jak1 and Tyk2 tyrosine kinases are
activated, leading to activation of their downstream substrates,
signal transducers and
activators of
transcription (Stat)1 and Stat2. Activated Stat1
and Stat2 assemble as a multimeric complex, the
interferon-stimulated gene factor
3 (ISGF3) complex, with a member of interferon regulatory
factor family termed ISGF3
(3), and translocate into the nucleus
where they bind to interferon-stimulated response elements in the
promoters of IFN-
-stimulated genes. The response to IFN-
is
structured similarly; Jak1 and Jak2 are activated at IFN-
receptors,
leading to the phosphorylation of Stat1. Activated Stat1 forms
homodimers as the
-activated factor, translocates into the nucleus,
and binds the
-activated site elements in the promoters of
IFN-
-induced genes (4). The essential role of Stat1 in IFN signaling
has been demonstrated by the IFN-resistant phenotypes of cells and
animals missing the Stat1 gene product (5-7).
Upon IFN- stimulation, Stat1 and Stat2 are activated by tyrosine
phosphorylation within minutes, resulting in rapid production of ISGF3
(8). However, these responses are often transient. Following prolonged
treatment of cells with IFN, the levels of phosphorylated STAT proteins
decline to near pro-treatment levels, leading to a decline in
IFN-stimulated gene expression (9). The waning of induced gene
expression is often accompanied by a refractory state during which
cells remain unresponsive to addition of fresh inducer (10). This
desensitization phenomenon demonstrates the tight regulation of IFN
responses and is likely important to ensure the proper and controlled
action of these potent cytokines. Mechanisms responsible for the
control of IFN-induced responses probably operate on several levels.
These include down-regulation and degradation of receptors (11),
regulation of the activity of protein-tyrosine kinases and of
protein-tyrosine phosphatases (PTPs) controlling the phosphorylation of
STATs, the degradation of JAK and STAT proteins, for example, by
proteasomes (12), and potentially the regulated nuclear transport of
activated STATs (13). However, the precise mechanisms limiting the
response to IFN remain undefined.
Human Daudi lymphoblastoid cells are highly sensitive to the biological
effects of IFN. This high sensitivity is at least partially due to a
prolonged rather than transient transcriptional response to IFN in
which IFN-stimulated gene expression remains induced for greater than
24 h post-treatment (14). However, the mechanisms for maintaining
this response are still unknown. We have examined the mechanisms
underlying the prolonged response to IFN- in Daudi cells. We found
that the level of phosphorylated STAT protein in Daudi cells is
prolonged following continued treatment with IFN-
, and that this
prolonged phosphorylation was strictly dependent on continued activity
of JAK protein-tyrosine kinases. Moreover, neither tyrosine
dephosphorylation nor protein turnover appeared to be impaired in Daudi
cells, suggesting that the maintenance of phosphorylated STAT protein
is regulated primarily at the level of active phosphorylation.
Interestingly, we found that the continued activity of JAK kinases and
of STAT phosphorylation were dependent on the induced secretion of
autocrine IFN.
Daudi (ATCC), a human
lymphoblastoid cell lines, and FS2, a human fibroblast cell line, were
maintained in RPMI 1640 and Dulbecco's modified Eagle's medium,
respectively, supplemented with 10% fetal calf serum. Polyclonal
rabbit antisera against Jak1, a gift from Dr. Martin Seidel, Ligand
Pharmaceuticals Inc., and Tyk2, a gift from Dr. Sandra Pellegrini,
Pasteur Institute, France (15), were used for immunoprecipitation.
Monoclonal antibodies against Jak1, phosphotyrosine (Py-20) purchased
from Transduction Labs (Lexington, Kentucky), 4G10 purchased from
Upstate Biotechnology Inc. (Lake Placid, NY), and Tyk2 (from Dr. Sandra
Pellegrini) were used in Western blot analysis. Rabbit antisera against
Stat1 and Stat2 were gifts from Dr. Chris Schindler, Columbia
University, New York, NY (16). Rabbit polyclonal antisera against
IFN- (Interferon Sciences Inc., New Brunswick, NJ) was used for
neutralizing IFN-
. Recombinant IFN-
-2a was a gift from
Hoffmann-La Roche and was used at 500 units/ml.
Gel shift assays were performed as described previously (8). In brief, a double-stranded 32P-labeled DNA probe containing interferon-stimulated response element sequence from the human ISG15 gene (17) was incubated with cell extracts, fractionated on a non-denatured polyacrylamide gel, and autoradiographed.
Immunoprecipitation and Western Blot AnalysisTotal cell extracts were prepared by lysing cells in 1% Triton X-100, 300 mM NaCl, 50 mM HEPES, pH 7.6, 1 mM Na3VO4, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and 5 µg/ml aprotinin and leupeptin. Cytoplasmic extracts were prepared by lysing cells in 0.25% Nonidet P-40, 10 mM NaCl, 10 mM Tri-HCl, pH 7.4, 3 mM MgCl2, 10% glycerol, 1 mM Na3VO4, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and 5 µg/ml aprotinin and leupeptin. Nuclear extracts were prepared by incubating nuclei in 20 mM HEPES, pH 7.9, 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 25% glycerol, and protease inhibitors. Immunoprecipitation and immunoblotting were performed as described previously (18).
In Vivo 32P Pulse-Chase LabelingCells were
preincubated in phosphate-free Dulbecco's modified Eagle's medium
containing 5% dialyzed fetal bovine serum for 1 h, followed by
addition of 1 mCi/ml [32P]orthophosphate and incubation
for 1.5 h (18). Labeled cells were treated with IFN-, 500 units/ml, for 30 min, washed three times with cold 1 × phosphate-buffered saline, and incubated for various times in medium
without label.
Cells
were preincubated in methionine-free Dulbecco's modified Eagle's
medium containing 5% dialyzed fetal bovine serum for 1 h,
followed by addition of 1 mCi/ml [35S]methionine and
incubation for 2 h (13). Labeled cells were washed, incubated in
label-free medium, and then treated with IFN- (500 units/ml) for
different times.
Jak1 or Tyk2 immunoprecipitated from total cell extracts were incubated with 15 µl of in vitro kinase buffer (10 mM HEPES, pH 7.4, 50 mM NaCl, 5 mM MgCl2, 0.1 mM MnCl2, 0.1 mM Na3VO4) containing 20 µM ATP for 30 min at room temperature. The reaction was stopped by adding protein sample buffer and heated at 95 °C for 5 min. The reaction mixture was fractionated on 8.5% SDS-polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane, and phosphorylation was detected using anti-phosphotyrosine antibodies (19). As a control, reaction mixtures lacking ATP were processed in parallel.
Since Daudi cells
maintain transcription of IFN-stimulated genes during prolonged
treatment with IFN, we studied the kinetics of activation of ISGF3 and
of Stat1 and Stat2 phosphorylation. Cells were treated with recombinant
IFN--2a for 30 min, 2 h, 4 h, and 24 h, and the
kinetics of ISGF3 activation were determined using EMSA. For
comparison, ISGF3 was also analyzed in FS2 cells, a human fibroblast
cell line that displays a typically transient response. ISGF3 was
detected after stimulation for 30 min in FS2, and the complex
dramatically decreased after 4 h (Fig.
1A, upper panel),
indicative of a transient response to IFN-
. Following 24 h of
continuous treatment with IFN, no active ISGF3 was detected in either
cytoplasmic or nuclear extracts (data not shown). A similar phenomenon
was seen in other human cell lines, such as HeLa S3 and 2fTGH (data not
shown). In contrast, Daudi cells showed prolonged activation of ISGF3.
Activated ISGF3 was detected after 30 min of IFN-
treatment and was
maintained for greater than 24 h (Fig. 1A, lower
panel, and data not shown). The level of active ISGF3 in extracts
of Daudi cells treated with IFN for 24 h was ~70% of the level
observed following 30 min of treatment. As previously reported (20), a
basal activation of ISGF3 was also observed in Daudi cells in the
absence of exogenous IFN-
stimulation (Fig. 1A,
lower panel, lane 1).
To confirm that Stat1 and Stat2 phosphorylation was prolonged in
response to IFN-, the kinetics of tyrosine phosphorylation of these
two proteins were followed by immunoprecipitating with anti-Stat1 (Fig.
1B, left panel) or Stat2 (Fig. 1B,
right panel) antibody followed by blotting with anti-Tyr(P)
antibody. Stat1 and Stat2 were tyrosine-phosphorylated after
stimulation for 10 min, and the phosphorylation was maintained for at
least 24 h of treatment (Fig. 1B, upper
panel), consistent with the formation of ISGF3. Since activated
Stat1 and Stat2 formed heterodimers, antibody to either Stat1 or Stat2
also co-immunoprecipitated its counterpart. By reblotting with
anti-Stat1 and Stat2 antibodies, the protein levels were shown to be
comparable after 24 h of treatment (Fig. 1B,
lower panel). These results indicated that Daudi cells have
prolonged tyrosine phosphorylation of Stat1 and Stat2, which results in
prolonged activation of ISGF3 in response to IFN-
.
The administration of
phosphatase inhibitors such as sodium vanadate or pervanadate to cells
causes the accumulation of tyrosine phosphorylated STATs even in the
absence of IFN stimulation (21-23), suggesting tyrosine phosphatases
are required to maintain STAT proteins in a latent state. To test the
involvement of phosphatases in the prolonged response, the turnover
rate of phosphorylated STATs was measured using in vivo
pulse-chase labeling. Daudi cells were labeled with
[32P]orthophosphate for 1.5 h, followed by treatment
with IFN- for 30 min. Cells were then washed and incubated in the
absence of label for 0, 30 min, 2 h, 4 h, 8 h, and
12 h. Stat1 and Stat2 proteins were immunoprecipitated, and
phosphorylated protein was detected by autoradiography.
32P-Labeled, activated Stat1 (Fig.
2A, left panel) and
Stat2 (right panel) were observed in both cytoplasmic (Fig.
2A, upper panel) and nuclear (lower
panel) extracts at the beginning of the chase process (Fig.
2A, lane 2 in each panel). The
signals, however, disappeared in both compartments after 4 h (Fig.
2A), even though protein levels of Stat1 and Stat2 were
still comparable throughout the experiment (data not shown). This
result indicated that the phosphorylation of activated Stat1 and Stat2
turned over very rapidly, although the proteins themselves were
relatively stable.
The short half-life (<2 h) of phosphorylated STATs suggested that
protein-tyrosine phosphatases constantly counteracted
ligand-dependent tyrosine phosphorylation. Alternatively,
protein degradation could be responsible for the rapid turnover of
activated STAT proteins. To distinguish these possibilities, we
performed [35S]methionine pulse-chase labeling to
determine the half-life of Stat1 and Stat2 proteins following IFN
treatment. Daudi cells were labeled with [35S]methionine
for 2 h. Cells were then washed, stimulated with IFN-, and
incubated in the absence of label for 0 min, 30 min, 2 h, 4 h, and 8 h. Stat1 and Stat2 were immunoprecipitated from cell
extracts and labeled protein was detected by autoradiography. The
levels of Stat1 (Fig. 2B, left panel) and Stat2
(Fig. 2B, right panel) were found to be
comparable during the course of the experiment, indicating that both
Stat1 and Stat2 protein stability is considerably greater than the
relatively short half-life of the tyrosine phosphate (Fig.
2A).
The rapid turnover of phosphorylated Stat1 and Stat2 was in marked
contrast to the near constant levels of active ISGF3 observed throughout the treatment period. This result suggested that new protein
was being constantly phosphorylated to counteract the loss of active
STAT protein due to dephosphorylation. This prompted us to examine the
requirement for continued kinase activity. Cells were treated for 30 min with IFN- to activate ISGF3. Then, the tyrosine kinase inhibitor
genistein was added to block further JAK kinase activity. ISGF3
formation was detected after the initial 30 min of treatment; however,
complex formation was significantly blocked following 2 h of
treatment with genistein and was undetectable following incubation for
24 h (Fig. 2C). A similar phenomenon was observed using
another kinase inhibitor, staurosporine, while the protein kinase C
inhibitor, H7, had no inhibitory effect (data not shown). These results
show that continuous protein tyrosine kinase activity was required for
maintaining a prolonged response to IFN.
The
observation that continuous protein kinase activity was required for
sustaining the IFN- response in Daudi cells prompted us to study
directly the kinase activity of Jak1 and Tyk2. Jak1 (Fig.
3, upper panel) or Tyk2
(lower panel) were immunoprecipitated from cells either left
untreated or treated with IFN-
for 30 min, 4 h, and 24 h,
and kinase activity was measured in vitro. Both Jak1 and
Tyk2 showed similar kinetics of tyrosine phosphorylation in response to
IFN-
. A basal phosphorylation was detected in the absence of IFN-
treatment, which was stimulated to high levels after 30 min of
treatment. Augmented levels of kinase activity were maintained for
24 h (Fig. 3, left panel). Similarly, both Jak1 and
Tyk2 exhibited elevated kinase activity toward exogenous substrates
24 h after treatment of cells with IFN (data not shown), suggesting that the kinase activity of Jak1 and of Tyk2 were prolonged in Daudi cells in response to IFN-
.
Continuous Presence of IFN-
The preceding results indicated that the
prolonged response to IFN- was maintained by continued JAK kinase
activity. We next considered whether this continued kinase activity
required continuous ligand stimulation. Cells were pulsed with IFN-
for 30 min, then incubated in the absence of IFN-
for 30 min, 2 h, 4 h, 8 h, and 24 h. Cytoplasmic extracts were
analyzed by EMSA. ISGF3 was found to be maintained for 24 h in the
absence of exogenous IFN. However, there was a transient decrease
followed by a reaccumulation of ISGF3 (Fig.
4A). This profile suggested
that the transient loss of ISGF3 might be due to the turnover of
activated STAT proteins as observed during pulse-chase labeling (Fig.
2A). However, active ISGF3 was restored at later times,
presumably through the action of induced factors.
To test if the synthesis of new proteins might account for these
induced factors, cells were treated with IFN in the presence of CHX.
Daudi cells were first pretreated with CHX (50 µg/ml) for 15 min
followed by treatment with IFN- for 15 min, 30 min, 2 h, 4 h, and 8 h. ISGF3 formation in both cytoplasm (Fig. 4B, upper panel) and nucleus (lower panel) was
gradually reduced and eventually disappeared after 8 h of CHX
treatment. Tyrosine phosphorylation of Stat1 and its associated Stat2
was also diminished after CHX treatment for 8 h (Fig.
4C, upper panel), correlating with the loss of
ISGF3 in EMSA. However, no significant change in the levels of Stat1
protein was detected during treatment with CHX (Fig. 4C,
lower panel), indicating that degradation of Stat1 is not a
significant mechanism for removal of activated STATs and that its
synthesis does not account for CHX sensitivity. Taken together, these
results suggested that continuous presence of IFN-
is required to
maintain the prolonged activation of STAT proteins, maintaining JAK
kinase activity through the action of induced factors.
It has been shown that Daudi cells secrete low levels
of IFN- (20). To test if secretion of IFNs was induced in response to IFN treatment and thus contributed to the reactivation of ISGF3 during the IFN-
withdrawal experiment, anti-IFN-
antisera were employed. Daudi cells were first pulsed with IFN-
for 30 min, and
then incubated in the absence of exogenous IFN-
but in the presence
of anti-IFN-
antisera for 30 min, 2 h, 4 h, 8 h,
24 h. ISGF3 formation was significantly reduced after 4 h,
and was totally abolished after 8 h of antibody treatment (Fig.
5A), indicating that secretion
of endogenous IFN-
is required for maintaining a prolonged response
in the absence of exogenous IFN-
.
To further confirm that secretion of IFN- indeed was induced in
Daudi cells by IFN-
treatment, conditioned medium collected from
IFN-
-pulsed cells was used to treat naive Daudi cells. Cells were
treated with IFN-
for 30 min and then incubated in the absence of
IFN for an additional 24 h. Conditioned medium from these cultures was used to treat fresh cultures of Daudi cells for 30 min, 2 h,
4 h, and 24 h, which were subsequently analyzed by EMSA.
Conditioned medium was able to activate ISGF3 formation at 30 min, and
this activation was maintained for 24 h (Fig. 5B).
Conditioned medium from cells that had never been exposed to IFN-
showed no such activity (Fig. 5B, lane 1). ISGF3
complex formation was used as a bioassay to quantitate the amount of
IFN-
secreted in IFN-
-stimulated Daudi cells. Activity equivalent
to about 0.5 unit/ml recombinant human IFN-
-2a was produced from
1 × 107/ml cells previously pulsed with IFN-
(data
not shown).
Since anti-IFN- antibody blocked prolonged activation
of ISGF3 in Daudi cells, the effect of this antibody on the activation of Stat1 and Stat2 was examined. Daudi cells were pulsed with IFN-
and treated with anti-IFN-
antibody for 4 and 24 h. As a
control, cells were treated with IFN-
for 10 min in the absence of
neutralizing antibody. Activated Stat1 and Stat2 were monitored by
using antibodies for immunoprecipitation followed by blotting with
anti-Tyr(P) antibody. The results showed that tyrosine phosphorylation of Stat1 and Stat2 appeared after 10 min of stimulation and was readily
blocked by treatment with antibody for 4 h (Fig.
6A, upper panel)
even though protein levels were comparable during the 24 h of treatment
(Fig. 6A, lower panel). The kinase activity of Jak1 was also transient under these conditions. Jak1 was no longer autophosphorylated after 4 h of antibody treatment (Fig.
6B, left panel). Further, the ability to
phosphorylate exogenous substrate was also abrogated (data not shown),
indicating that neutralization of secreted IFN-
inhibited the
prolonged response to IFN. These results suggested that autocrine
production and secretion of IFNs were required for maintaining STAT
phosphorylation.
Regulation of tyrosine phosphorylation (activation) and
dephosphorylation (deactivation) of STAT proteins controls IFN-mediated signaling and biological activities. Transient activation of STAT proteins results in rapid responses, whereas prolonged activation leads
to prolonged levels of gene induction and enhanced biological activities. The regulation of tyrosine phosphorylation and
dephosphorylation is orchestrated by the coordinated action of PTPs and
protein-tyrosine kinases (24, 25). Modulation of the function of either
PTPs or protein tyrosine kinases results in imbalanced activity,
leading to altered signal transduction and response. For example, an
EPO receptor-associated phosphatase SH-PTP1 (HCP, PTP1C, SHP-1)
regulates EPO signaling (26). Cells expressing mutant EPO-R lacking
SH-PTP1 docking sites are hypersensitive to EPO stimulation and display a prolonged activation of the receptor-associated protein-tyrosine kinase, Jak2. Similarly, macrophages derived from motheaten mice, which
lack SH-PTP1, show increased tyrosine phosphorylation of Jak1 and Stat1
upon IFN- stimulation (27). The region of STAT proteins involved in
the interaction with phosphatases is further suggested by the
constitutive activation and enhanced tyrosine phosphorylation of an
amino-terminally deleted version of Stat1 when overexpressed in cells.
While this result suggests that the amino terminus is crucial for
modulating STAT activity through tyrosine dephosphorylation (28),
whether tyrosine dephosphorylation is constitutive or
ligand-dependent has not been determined.
Data presented in this report demonstrate that the unusually prolonged
IFN- response of Daudi cells requires continued protein-tyrosine kinase activity. Jak1 and Tyk2 remain active for prolonged periods following IFN treatment, and addition of protein-tyrosine kinase inhibitors at any point during the response leads to a rapid cessation of STAT activity. The need for continued kinase activity appears to be
due to the opposing action of one or more PTPs. The half-life of STAT
phosphorylation was found to be short (<2 h), reflecting the
constitutive activity of PTPs. This rapid turnover of phosphorylation is sufficient to explain the loss of activated STAT proteins observed in the presence of tyrosine kinase inhibitors. It would also account for the limited duration of STAT activation observed in more typical cell lines that do not maintain prolonged kinase activity
(e.g. FS2). Therefore, it is likely that the regulation of
JAK activity, rather than changes in the rate of dephosphorylation, is
the primary mode for controlling the duration of IFN-
responses.
An additional mechanism for the deactivation of STAT proteins besides
dephosphorylation could be protein degradation. Given the relatively
stable nature of Stat1 and Stat2 protein, such degradation would need
to be targeted specifically for phosphorylated protein to explain the
loss of activated STATs in the absence of significant loss of total
protein. Indeed, specific degradation of phosphorylated Stat1 in
IFN- treated HeLa cells through a ubiquitin-proteasome-mediated
pathway has been recently reported (12). However, protein degradation
does not appear to be responsible for the observed turnover of
phosphorylated Stat1 or Stat2 in IFN-
-treated Daudi cells. First,
treatment of cells with proteasome inhibitors did not prevent the
turnover of phosphorylated STATs (data not shown). Second, we estimate
that approximately 50% of total cellular Stat2 and 15-20% of
cellular Stat1 are phosphorylated in Daudi cells treated with IFN-
for 30 min (Stat1 is more abundant than Stat2). This estimate was based
on the fraction of STAT protein that is translocated to the nucleus and
the relative intensities of anti-phosphotyrosine signals in cytoplasmic
and nuclear extracts (data not shown). Given this large fraction of the
cellular pool of STAT protein that is activated, significant loss of
protein would be expected to be observed following 4 h of
treatment if protein degradation played a major role in turnover. In
contrast, little or no turnover of total Stat1 or Stat2 was observed by [35S]methionine pulse-chase labeling (Fig. 2B)
while significant turnover of phosphorylation was observed (Fig.
2A). Therefore, although we cannot completely exclude a role
for protein degradation in the turnover of phosphorylated STAT
proteins, we conclude that a constitutive PTP contributes the major
activity. Similar conclusions implicating the dephosphorylation of
activated Stat1 in the absence of significant protein degradation have
been recently reported for IFN-
-treated cells (13).
Mechanisms underlying the prolonged kinase activity of Jak1 and Tyk2 in
Daudi cells are still unknown. The continuous requirement of ligand
stimulation in maintaining a prolonged response suggests that
receptor-mediated signaling is involved. Although evidence has
suggested that IFN--induced receptor down-regulation and degradation
occurred rapidly in Daudi cells (11), the retention of
ligand-receptor-kinase complexes within endosomes might contribute to
prolonged kinase activity similar to that seen in the epidermal growth
factor-induced response of liver parenchyma (29). Alternatively, recycling of a small portion of high affinity receptors in Daudi cells
could also maintain the kinase activity. Interestingly, the
administration of neutralizing antibody inhibited this prolonged response (Fig. 4B), and conditioned medium collected from
IFN-
-stimulated Daudi cells contained IFN-
(Fig. 4C),
suggesting the continuous production and secretion of IFN-
. This
autocrine loop of IFN-
not only contributes to a prolonged response,
it also increases the sensitivity to IFN-
owing to the basal
activation of the signaling machinery. It has been shown that specific
IFN-
mRNAs are constitutively present at low levels in organs of
normal humans (30), indicating that the constant production of IFN-
could also provide an important host defense mechanism against invading pathogens. Thus, the involvement of autocrine secretion and production of IFN-
in maintaining a prolonged response may play a significant role in vivo.
Although Daudi cells have been shown to secrete low amounts of IFN-
constitutively (20), we show here that IFN-
secretion is induced to
higher levels in Daudi cells in response to IFN-
. The finding that
CHX treatment blocked the prolonged response (Fig. 4) suggested the
requirement for new protein synthesis. The IFN detected in conditioned
medium was indeed produced by the cells and could not be ascribed to
residual exogenous IFN from the treatment pulse. Neutralization with
anti-IFN monoclonal antibodies demonstrated that the secreted IFN was a
mixture of IFN-
subtypes rather than the recombinant IFN-
-2a used
for exogenous treatment (data not shown).
IFN- also has been shown to induce prolonged activation of Stat1 and
continued transcriptional activation of an IFN-
-inducible gene, such
as GBP (31, 32). However, there are several different aspects that
distinguish the IFN-
-induced prolonged response. First, the
IFN-
-induced response is stable to removal of ligand, while the
response induced by IFN-
is dependent on continuous presence of
IFN-
. Second, the response to IFN-
is sensitive to protein kinase
C inhibitors while the IFN-
-induced response is not affected by
these kinase inhibitors. Third, the response to IFN-
is only
partially affected by CHX treatment, but the continued response of
Daudi cells to IFN-
is totally blocked by CHX. The differences
between IFN-
and IFN-
-induced prolonged response suggest that a
unique mechanism may be involved in the regulation of the IFN-
response in Daudi cells.
We thank M. Seidel for providing the Jak1 antibody, S. Pellegrini for Tyk2 antibodies, and C. Schindler for Stat1 and Stat2 antibodies.