| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Division of Pulmonary and Critical Care Medicine and Bellevue Chest Service,
Department of Pathology, New York University Medical Center, New York, NY 10016;
Department of Medicine, Sendai Kosei Hospital, Sendai, Japan;
§
Department of Microbiology and Infection, Institute of Medical Science, Tokyo University, Tokyo, Japan; and
¶
Public Health Research Institute, New York, NY 10016
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B upon infection with
M. tuberculosis. These results show that induction of
ISGF-3, expression of the inhibitory 16-kDa C/EBPß, and suppression
of HIV-1 replication via a transcriptional mechanism are
macrophage-specific responses to infection with M.
tuberculosis. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We and others have used monocytic THP-1 cells as an in vitro model to
better understand the mechanisms regulating HIV-1 LTR function.
Proinflammatory stimuli such as TNF-
, LPS, or the mycobacterial cell
wall product lipoarabinomannan stimulate HIV-1 long terminal repeat
(LTR)3 promoter
activity and HIV-1 replication in monocytic cells (6, 7, 8, 9, 10).
This is in part due to NF-B activation and in part due to
stimulation by the C/EBP family of transcription factors, which binds
to the LTR near the NF-B sites. The enhancement of HIV-1 replication
also occurs in primary blood monocytes after infection with M.
tuberculosis or stimulation with mycobacterial products (11, 12).
Surprisingly, when monocytes are differentiated to macrophages,
TNF-, LPS, and infection with M. tuberculosis suppress
viral replication and LTR function (13, 14, 15). The effect of
differentiation may be due to the interaction of proinflammatory
stimuli and the IFN system. The suppression of HIV-1 replication seen
after LPS stimulation closely resembles the suppression seen after
addition of IFNs (13). Type I IFN is required for the
suppressive effects of LPS and TNF- on macrophages (16, 17), and IFN-ß mRNA is induced after LPS stimulation
(14). We have recently observed that in macrophages,
M. tuberculosis infection induces an IFN response, and that
IFN induces the expression of an inhibitory C/EBPß transcription
factor (15).
The C/EBPß transcription factor is an important regulator of inflammation (18, 19). C/EBP binding sites are present in many promoters of proinflammatory cytokines as well as in the HIV-1 LTR. The C/EBP sites in the LTR are required for LTR promoter function and viral replication in monocytes, but not lymphocytes (20, 21). The negative regulatory element is the DNA element responsible for suppressing the HIV-1 LTR in macrophages after LPS stimulation (22). The negative regulatory element contains binding sites for the C/EBP transcription factor family (23). The C/EBPß gene (also called NF-IL6, LAP/LIP, and NF-M) has no introns, but a 37-kDa stimulatory isoform and a 16-kDa inhibitory isoform are produced from the same mRNA (24, 25). The inhibitory isoform may be produced by internal start site utilization by the ribosome leading to de novo synthesis of an N-terminally truncated protein (24, 26) or by proteolytic cleavage of the stimulatory isoform (27, 28). Importantly, the 16-kDa C/EBPß isoform is a dominant negative transcription factor, since, if present at 20% the level of the stimulatory 37-kDa C/EBPß, it inhibits transcription (24). Resting alveolar macrophages are like differentiated THP-1 cells treated with IFN-ß; both strongly express inhibitory C/EBPß and inhibit HIV-1 replication (3, 15)
IFNs, originally identified as cytokines that mediate antiviral
immunity, also play a role in protection against bacterial infections
(29). Type I IFNs (IFN- and IFN-ß) constitute an
essential arm of the innate immune response that is activated by
invariant characteristics of the pathogen (30, 31). Innate
immunity is activated by and provides protection from pathogens in the
time window before Ag-specific immunity can occur (32).
The molecular mechanisms by which cells respond to IFNs have begun to
be understood (33).
Signal transduction in the type I IFN system is mediated by a high
affinity transmembrane receptor composed of two subunits
(34). Knockout mice deficient in the p100 subunit are
exquisitely susceptible to viral infection (35). Upon
binding of a type I IFN to its receptor, JAK-1 and TYK-2 protein
tyrosine kinases are activated, resulting in activation of STAT
proteins by tyrosine phosphorylation. Formation of IFN-stimulated gene
factor-3 (ISGF-3) and STAT-1 homodimers can then occur without the need
for protein synthesis. ISGF-3 is a heterotrimer of STAT-1, STAT-2, and
IFN regulatory factor (IRF)-9 (previously called p48 or ISGF-3
) that
rapidly translocates to the nucleus and binds promoter sequences, named
IFN-stimulated response elements (ISRE) to activate gene expression.
STAT-1 homodimers bind to a IFN--activated site (GAS) and activate
an overlapping, but distinct, set of genes (36, 37).
STAT-1 homodimers can also interact with IRF-9 and bind ISRE-type
enhancer elements (36). It has become clear that
additional regulation, often cell type specific, is superimposed on
this framework (33, 37). For example, differentiation of
monocytic cell lines in response to phorbol ester or
all-trans-retinoic acid leads to increased levels of latent
STAT-1, STAT-2, and IRF-9 (38, 39). Serine kinases,
protein tyrosine phosphatases, tyrosine kinase inhibitor proteins, and
activated STAT inhibitor proteins also participate in regulating ISGF-3
and STAT-1 homodimer activity (37, 40, 41, 42, 43, 44, 45, 46, 47).
Since type I IFNs are present in tissue macrophages without any identified infection (48), resting tissue macrophages may be primed by low dose IFN to inhibit viral replication in the absence of inflammation. Therefore, we investigated further the interaction of IFNs and the production of inhibitory C/EBPß. Here we report that production of inhibitory C/EBPß is an IFN response in macrophages, but not monocytes. Alveolar macrophages express low levels of ISGF-3 and are able to activate ISGF-3 to a significant extent after stimulation with IFN or infection with M. tuberculosis. Monocytes fail to activate ISGF-3 and fail to inhibit HIV-1 replication after IFN stimulation or M. tuberculosis infection. Modulation of the IFN signal transduction pathway by differentiation is therefore closely linked to the production of a C/EBPß transcriptional repressor that is likely to promote viral latency in resting alveolar macrophages.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
THP-1 cells were cultured in RPMI 1640 with 10% FBS. Cells were
differentiated with 20 ng of PMA/ml for 24 h or as indicated in
each experiment. For experiments to examine the effect of IFN on
C/EBPß, THP-1 cells were either left untreated or were pretreated
with PMA for 24 h and then treated with IFN-ß as indicated. For
experiments examining IFN signaling, THP-1 cells were either left
untreated or were pretreated with PMA for 24 h and then treated
with 500 U of IFN-/ml or 10 U of IFN-ß/ml for 30 min unless
otherwise indicated. The human embryonic kidney cell line 293T was
cultured in DMEM with 10% FBS. HeLa S3 cells were cultured in DMEM
with 10% CBS.
Cytopathic effect assay
Cytopathic effect assays were conducted essentially as
previously described (49). HeLa S3 cells were seeded in
96-well plates and allowed to attach overnight. Conditioned media
obtained from PMA-treated THP-1 cells 3 days after infection with
M. tuberculosis (see below) and from parallel uninfected
cultures were sterilely filtered. These media were not or were
supplemented with 1/500 dilutions each of anti-IFN and
anti-IFNß antisera (gifts from Jan Vilcek, New York University
School of Medicine, New York, NY), and then used to replace the media
for the HeLa cells. Each of these media or fresh media plus IFN- and
IFN-ß, without or with antisera, was used in two wells. The
monolayers were grown in these media 1 or 2 days more, until nearly
confluent, then infected with various dilutions of vesicular stomatitis
virus (VSV). Approximately 24 h later, cells were stained with
crystal violet (
max = 588 nm) and washed with water. After the
wells were dry, dye was solubilized in 50% ethanol, and absorbance at
590 nm was measured with an ELISA plate reader. OD values were
corrected for stained empty wells. The duplicate wells were averaged,
and the averages were normalized to the average corrected absorbance of
uninfected monolayers. Results are shown for virus multiplicity of
infection (MOI) that produced 30–60% cytopathic effect in the absence
of anti-IFN antisera.
Purification of alveolar macrophages, blood monocytes, and macrophages
Alveolar macrophages were obtained by bronchoalveolar lavage, which was performed with a flexible fiberoptic bronchoscope with local xylocaine anesthesia. We instilled six 50-ml aliquots (total, 300 ml) of normal saline and suctioned at 20–30 mm Hg. The lavage fluid was filtered through sterile gauze to remove mucus, and total cells were counted in a hemocytometer. Cell differentials were performed on cytospin slides stained with Wright-Giemsa. For purification of alveolar macrophages, bronchoalveolar lavage cells were pelleted, suspended in RPMI 1640, allowed to adhere to plastic plates, then washed. PBMC were obtained from blood by Ficoll-Hypaque sedimentation, and buffy coats were washed four times in PBS. PBMC were suspended in RPMI 1640 with 10% FCS and allowed to adhere to plastic plates for 3 h and washed with PBS. Alveolar macrophages and monocytes were removed from plastic by gentle scraping with a rubber policeman. Alveolar macrophages and monocytes were 95% pure by morphology and nonspecific esterase staining. Bronchoalveolar lavage was approved by the New York University Medical Center institutional review board.
HIV-1 and M. tuberculosis infection
Purified blood monocytes were incubated for 7 days with 1000 U/ml of M-CSF. Then cells were infected with BaL or ADA strains of HIV (1 ng of p24/ml). After overnight incubation, cells were coinfected with M. bovis bacillus Calmette-Guérin at various MOI for 2 h at 37°C. They were then washed twice and incubated with RPMI 1640/10% FCS containing M-CSF. p24 was assayed by ELISA 7 days after infection. Experiments with M. tuberculosis were performed with strain TN913, a prevalent clinical isolate obtained from the PHRI TB Center, except, for comparison to previous studies (15), strain H37Ra was used for the experiment to examine the effect of infection on C/EBPß isoforms. All M. tuberculosis infections were conducted at an MOI of 1 for 24 h unless otherwise stated. Control uninfected cells were cultured for 24 h in parallel before extract preparation. IFN-treated samples were stimulated for the last 30 min of the experiment. The extent of infection was determined by acid fast staining of cells washed twice in PBS, before extract preparation. Acid fast bacilli were detected by light microscopy in 20–30% of monocytes and 40–50% of macrophages. Viable cell count and CFU recovery were determined for aliquots of washed cells. The average for individual experiments ranged from one to four bacilli per infected cell, after correction for the percentage of cells infected, consistent with microscopic examination of the stained cells. HIV and M. tuberculosis infections were performed in a BSL-3 laboratory.
Whole-cell and nuclear extract preparation
Cells were washed twice in PBS. Whole-cell extracts were prepared for immunoblot analysis by incubation in RIPA buffer (PBS, 1% Nonidet P-40, and 0.5% deoxycholate) containing 3 µg/ml aprotinin, 1 mM PMSF, and 1 mM sodium orthovanadate for 30 min and disruption by passage through a 21-gauge needle. Whole-cell extracts for EMSA were prepared by Nonidet P-40 lysis and salt extraction with 3 µl of extraction buffer/105 cells, as previously described (49). Nuclear extracts were prepared by Nonidet P-40 lysis and incubation of recovered nuclei in high salt buffer, as previously described (50). Transfection of an IRF-9 expression construct into 293T cells was performed as previously described (36) to obtain recombinant IRF-9 in Nonidet P-40 whole cell extracts prepared 2 days post-transfection. Production of recombinant IRF-9 was assessed by immunoblot and EMSA. Immunodepleted 293T/IRF-9 extracts were prepared by incubation with antiserum (nonspecific or anti-IRF-9) and protein A-agarose beads (Schleicher & Schuell, Keene NH), followed by removal of the beads. Extracts of cells infected with M. tuberculosis were prepared in a BSL-3 laboratory and sterilized by filtration before removal. Pierce bicinchoninic acid or Bio-Rad Bradford reagents (Hercules, CA) were used to determine extract protein concentrations.
Immunoblots
Proteins were separated by SDS-PAGE. For each gel, equal amounts
of protein were loaded in each lane. From gel to gel, the amount of
protein per lane was 30–100 µg for detection of endogenous proteins,
but 
1 µg for detection of recombinant IRF-9. For detection of
C/EBPß or NF-B, 10–20% linear-gradient gels were used. For
detection of IRF-9, 10% gels were used. After electrophoresis, protein
was electrotransferred to nylon membranes for detection of C/EBPß or
NF-B or to nitrocellulose membranes for detection of IRF-9.
Membranes were blocked with nonfat dry milk, then probed with Abs
against C/EBPß (catalog no. sc-150, lot D168 or E236), NF-B p65
(catalog no. sc-109, lot D119), both from Santa Cruz Biotechnology
(Santa Cruz, CA) or IRF-9 (36), as indicated for each
experiment, then visualized with anti-rabbit HRP and ECL (Amersham,
Arlington Heights, IL). Multiple film exposures were scanned by laser
densitometry and analyzed with ImageQuant software (Molecular Dynamics,
Sunnyvale, CA) for quantitation of immunoreactive C/EBPß or IRF-9
bands. Both the anti-C/EBPß and anti-IRF-9 antisera reacted
with nonspecific bands, which confirmed equal protein loading from lane
to lane.
EMSA
The ISRE probe was
5'-CTCGGGAAAGGGAAACCGAAACTGAAGCC. The GAS probe was
TACAACAGCCTGATTTCCCCGAAATGACGGC. The NF-B probe was
TGGGCTGGGGAATCCCGCTAA. The respective ISRE, GAS, and NF-B
consensus sequences are shown in boldface. Probes were end
labeled, and binding reactions without or with Abs were performed as
previously described (51). Within an experiment, each
binding reaction included a constant amount of extract protein. The
amount varied from 10 to 20 µg among experiments, except that only 1
µg of 293T/IRF-9 extract was used per binding reaction. Protein/DNA
complexes were electrophoresed on 6% polyacrylamide gels at 4°C with
20 mM Tris-borate (pH 8.3)/0.4 mM EDTA buffer. Images were produced by
autoradiography or PhosphorImager (Molecular Dynamics). ImageQuant
software (Molecular Dynamics) was used for analysis of quantitative
data obtained with the PhosphorImager.
Determination of transcription rates
Cells were treated as described in Fig. 6
. Run-on assays were
performed as previously described (51, 52). Radiolabeled
nascent RNA was isolated and hybridized to excess plasmid DNA fixed to
nitrocellulose. A ß-actin clone was used as an internal standard.
Bluescript KS- (Stratagene, La Jolla, CA) and
RcCMV (Invitrogen, San Diego, CA) were used as negative controls and to
determine background signal. The IRF-9 cDNA clone has been previously
described (36). The STAT-1 and STAT-2 cDNA clones were
previously called p91 and p113, respectively (51). The
results were visualized and quantitated with a PhosphorImager and
ImageQuant software (Molecular Dynamics).
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Similar to monocytic THP-1 cells, primary adherent blood monocytes increase HIV-1 replication after infection with M. tuberculosis (7, 11). We previously observed that M. tuberculosis suppresses HIV-1 replication in THP-1 cells differentiated to a macrophage-like state by treatment with PMA (15). To test whether primary macrophages are also able to suppress HIV-1 replication after mycobacterial infection, HIV-1 replication was assayed in macrophages derived from peripheral blood monocytes by differentiation with M-CSF.
Monocyte-derived macrophages were similar to THP-1 cells treated with
PMA; both supported high levels of HIV-1 replication, producing 1–3
ng/ml of p24 Ag 7 days after infection (Fig. 1A). M. bovis
bacillus Calmette-Guérin produced a dose-dependent suppression of
p24 production in monocyte-derived macrophages, using two
macrophage-tropic strains of HIV-1, ADA and BaL. At an MOI of 1, HIV-1
p24 production was diminished by >50% in both viral strains without a
significant reduction in cell viability. At an MOI of 10, viral
production decreased by 80–90% in both HIV-1 strains. Therefore,
macrophages derived from primary blood monocytes by M-CSF
differentiation are similar to THP-1 cells treated with PMA to become
macrophage-like.
|
B; mean ± SEM of three
independent experiments). In these experiments, addition of the
neutralizing antisera did not alter the VSV cytopathic effect when HeLa
cells were incubated in conditioned media from uninfected PMA-treated
THP-1 cells, but the antisera did neutralize exogenously added IFN-
and IFN-ß (data not shown). This result demonstrates that M.
tuberculosis infection causes secretion of type I IFN. Macrophages, but not monocytes, induced an inhibitory 16-kDa C/EBPß isoform after M. tuberculosis infection or IFN-ß treatment
Because C/EBP binding sites are required for HIV-1 LTR-mediated
transcription in macrophages (6, 20, 21), and the C/EBPß
gene can produce a 16-kDa dominant-negative transcription factor
(24), we investigated whether changes in C/EBPß
expression could account for the effect of monocyte differentiation on
the interaction of M. tuberculosis infection and HIV-1
replication. Without infection or PMA treatment, THP-1 cells had barely
detectable levels of the 37- and 16-kDa C/EBPß (Fig. 2A, lanes 1 and
6). A cross-reacting 40-kDa protein was detected in the
immunoblots, but was nonspecific (Fig. 2, labeled ns), since it was
expressed in unstimulated THP-1 cells that had negligible C/EBP
DNA-binding activity (15).
|
A, lanes 2–4). M. tuberculosis
infection of PMA-treated THP-1 cells produced both inhibitory 16-kDa
C/EBPß and stimulatory 37-kDa C/EBPß 48 h after infection,
with the inhibitory isoform representing 30% of the total C/EBPß
(Fig. 2A, lane 5). Since the induction of 37-kDa
C/EBPß by infection was similar in both untreated and PMA-treated
THP-1 cells, it is unlikely that greater M. tuberculosis
uptake after PMA treatment accounts for the difference in 16-kDa
C/EBPß expression. Furthermore, treatment of THP-1 cells with PMA by
itself did not alter C/EBPß expression (Fig. 2A,
lanes 7–9). Therefore, PMA-treated THP-1 cells gained the
ability to produce inhibitory 16-kDa C/EBPß in response to M.
tuberculosis infection.
To test whether primary blood-derived monocytes and macrophages are
similar to untreated and PMA-treated THP-1 cells, immunoblots were
performed on blood monocytes and monocytes differentiated to
macrophages by stimulation with M-CSF. Similar to untreated THP-1
cells, blood monocytes expressed barely detectable levels of the
inhibitory 16-kDa C/EBPß (Fig. 2B, lane 1).
After differentiation with M-CSF there was some induction of inhibitory
C/EBPß (Fig. 2B, lane 2), which was augmented
2-fold after infection with M. tuberculosis (Fig. 2B, lane 3), while the level of stimulatory
C/EBPß was essentially constant. Thus, only PMA-treated THP-1 cells
and primary macrophages were able to produce inhibitory C/EBPß. In
the absence of a proinflammatory stimulus, primary macrophages had
higher levels of inhibitory C/EBPß than PMA-treated THP-1 cells.
However, M. tuberculosis infection increased the amount of
inhibitory C/EBPß, and the ratio of inhibitory to stimulatory
C/EBPß in primary macrophages.
Since M. tuberculosis infection of macrophages produces a
type I IFN response, and the type I IFNs induce inhibitory C/EBPß
only in macrophages, differentiation may alter C/EBPß expression in
response to type I IFNs. To test this idea, we examined the C/EBPß
response of untreated THP-1 cells to IFN-ß stimulation. IFN-ß had
no effect on the abundance of C/EBPß isoforms in untreated THP-1
cells (Fig. 2C, lanes 1–3), but strongly induced
inhibitory 16-kDa C/EBPß after PMA treatment (Fig. 2C,
lane 4). PMA-treated THP-1 cells are able to suppress HIV-1
replication and LTR function at 1 U/ml IFN-ß (13, 15),
but THP-1 cells treated with only 1 or 10 U/ml IFN-ß did not suppress
HIV-1 replication (data not shown). This suggests that induction of
inhibitory C/EBPß is a type I IFN response in macrophages, but not
monocytes, as modeled by PMA-treated and untreated THP-1 cells, and
that the state of differentiation affects the ability of type I IFN to
inhibit HIV-1 replication.
ISGF-3 was activated only in macrophages
The observation that type I IFN induces inhibitory C/EBPß in THP-1 cells only after PMA treatment suggests that differentiation alters the IFN response. Differentiation could modulate the abundance or function of IFN receptor, intracellular signaling intermediates, or the transcription factors that mediate IFN response. In general, at least two transcription-factor complexes form without the need for protein synthesis due to activation of latent STAT-1 and STAT-2. They are ISGF-3, a STAT-1, STAT-2, and IRF-9 heterotrimer that binds to ISRE sequences, and STAT-1 homodimer, which binds to GAS sequences (33). Activation of latent transcription factors after IFN stimulation would demonstrate intact upstream signal transduction.
We have previously observed that PMA-treated THP-1 cells activate
ISGF-3 after infection with M. tuberculosis
(15). To test whether differentiation was necessary for
this response to infection, THP-1 cells with and without PMA treatment
were infected with M. tuberculosis in a time-course
experiment. Four hours after M. tuberculosis infection,
ISGF-3 was activated in PMA-treated (Fig. 3A, lane 4), but
not untreated (Fig. 3A, lane 2), THP-1 cells. The
amount of ISGF-3 activated by M. tuberculosis infection was
greater at 14 and 24 h after infection (Fig. 3A,
lanes 8 and 12). Therefore, M.
tuberculosis infection was able to elicit an obvious IFN response
only after PMA treatment. Since monocytes produce type I IFN in
response to proinflammatory stimuli (14), the inability of
infected monocytes to form ISGF-3 probably reflects a partial defect in
their response to IFN.
|
(Fig. 3B, lane 2). In contrast, IFN-
stimulation strongly activated ISGF-3 in PMA-treated THP-1 cells (Fig. 3B, lane 5). Infection of PMA-treated THP-1 cells
with M. tuberculosis induced less ISGF-3 than did IFN-
treatment (Fig. 3B, lane 6).
To test whether THP-1 cells are an accurate model for primary cells,
EMSA was repeated with extracts from blood monocytes
and alveolar macrophages. Blood monocytes did not induce ISGF-3
after infection with M. tuberculosis (Fig. 3B,
lane 8). However, in alveolar macrophages, IFN-
stimulation induced a 23-fold increase in ISGF-3 (Fig. 3B,
lane 10), and M. tuberculosis infection induced a
4-fold increase in ISGF-3 (Fig. 3B, lane 11).
Thus, the change in IFN signaling after differentiation of THP-1 cells
is similar to the difference between primary blood monocytes and
alveolar macrophages. Unlike differentiated THP-1 cells, resting
alveolar macrophages expressed a low level of ISGF-3 (Fig. 3B, lane 9). This may be due to the low level of
IFN observed in tissue macrophages (48).
The identity of ISGF-3 in extracts of alveolar macrophages was
confirmed using competition with excess unlabeled ISRE oligonucleotide
and supershifts with Abs against STAT-1, STAT-2, and IRF-9. Competition
demonstrated that the ISGF-3 complexes formed after infection (Fig. 3C, compare lanes 1 and 2) or
treatment with type I IFN (Fig. 3C, compare lanes
3 and 4) were specific for the ISRE sequence. The
ISGF-3 complexes induced by IFN were disrupted by Abs against IRF-9,
STAT-1, and STAT-2 (Fig. 3C, lanes 6,
8, and 9). ISRE/protein complexes that contain
IRF-9 and STAT-2 and migrate slightly more rapidly than similar
complexes that contain STAT-1, STAT-2, and IRF-9 became apparent after
supershift with STAT-1 Ab (Fig. 3C, lane 8). A
nonspecific Ab and an Ab against IRF-1 did not affect any of these
ISRE/protein complexes (Fig. 3C, lanes 5 and
7). Additional ISRE/protein complexes that contain IRF-9 are
apparent in the alveolar macrophage extracts (Fig. 3C,
compare lanes 5 and 6). These data demonstrate
that differentiation alters IFN responsiveness. Macrophages gain the
ability to form ISGF-3 after treatment with type I IFN or infection
with M. tuberculosis. At least in PMA-treated THP-1 cells,
this effect of infection presumably includes an autocrine response to
the type I IFN secreted upon infection.
STAT-1 was activated by IFN in both monocytes and macrophages
Formation of STAT-1 homodimer was examined by EMSA with a GAS
probe to assess whether the lack of ISGF-3 formation was due to failure
to transduce a type I IFN signal. Both THP-1 cells and PMA-treated
THP-1 cells produced STAT-1 homodimers after stimulation with 500 U/ml
IFN- (Fig. 4, lanes 2 and
7) or 100 U/ml IFN-ß (data not shown). The identity of
STAT-1 was confirmed by its reaction with specific, but not
nonspecific, antisera (Fig. 4, compare lane 5 with
lane 4 and lane 10 with lane 9).
M. tuberculosis infection did not activate STAT-1 homodimer
formation in untreated or PMA-treated THP-1 cells (Fig. 4, lanes
3 and 8). The results of the EMSA demonstrate that in
this model of monocytes and macrophages, both cell types have intact
signal transduction pathways up to the point of STAT-1
phosphorylation.
|
Failure to form ISGF-3 after type I IFN stimulation or infection
with M. tuberculosis might be due to limiting levels of
IRF-9. To test this possibility, EMSA was used to detect IRF-9 activity
with an ISRE probe. Undifferentiated THP-1 cell extracts did not
produce an IRF-9/ISRE complex (Fig. 5A, lane 1). When
recombinant IRF-9 derived from transiently transfected 293T cells was
added to these THP-1 extracts, an IRF-9/ISRE complex was detected (Fig. 5A, lane 2). PMA treatment induced high level
IRF-9 activity in THP-1 cells (Fig. 5A, lane 3).
Similar to untreated THP-1 cells, an extract of blood monocytes did not
produce an IRF-9/ISRE complex, but exogenous IRF-9 added to that
extract did bind the ISRE (Fig. 5A, lanes 4 and
5). Similar to PMA-treated THP-1 cells, alveolar macrophages
had high levels of IRF-9 activity, as judged by comigration with an
IRF-9/ISRE complex (Fig. 5A, lane 6) and reaction
with anti-IRF-9 Ab in an EMSA (data not shown). The similarities
between transformed and primary cells further supports using THP-1
cells and PMA-treated THP-1 cells as a model for primary monocytes and
macrophages. Therefore, differentiation of THP-1 cells and primary
cells up-regulated IRF-9 activity.
|
-stimulated
monocytes. Recombinant IRF-9 did reconstitute ISGF-3 in extracts of
IFN-treated THP-1 cells (Fig. 5B, compare lanes 3
and 4). This demonstrated that THP-1 cells were able to
activate STAT-1 and STAT-2 after IFN stimulation. Complementation of
monocyte extracts was not due to nonspecific effects. ISGF-3 was not
reconstituted when control extracts from untransfected 293T cells were
mixed with extracts of untreated or IFN-treated monocytes (Fig. 5B, lane 1, and data not shown), and recombinant
IRF-9 did not reconstitute ISGF-3 when mixed with extract from
untreated THP-1 cells (Fig. 5B, lane 2).
We confirmed that IRF-9 was responsible for reconstitution of ISGF-3 by
immunodepletion experiments. ISGF-3 was reconstituted after mock
immunodepletion of IRF-9 from 293T extracts by nonspecific antiserum
(Fig. 5B, lane 5); immunodepletion with
anti-IRF-9 antiserum abrogated ISGF-3 formation (Fig. 5B, lane 6). To corroborate the identity of
ISGF-3 we performed supershift reactions. The reconstituted
ISRE/protein complexes were not affected by control Ab (Fig. 5B, lane 7), but both complexes reacted with Abs
against each subunit of ISGF-3 (Fig. 5B, lanes
8–10). Similar to supershift experiments with IFN-treated
alveolar macrophages, some residual ISRE/protein complex remained after
addition of anti-STAT-1 Ab (Fig. 5B, lane 9).
These data indicate that IFN treatment of undifferentiated THP-1 cells
activates latent STAT-1 and STAT-2, but that ISGF-3 is not produced
because of inadequate IRF-9.
IRF-9 expression is post-transcriptionally regulated
Since increased IRF-9 activity is essential for ISGF-3 induction, we sought to better understand the regulation of IRF-9 expression during differentiation. A nuclear run-on assay was used to measure transcription of the IRF-9 gene after treatment with type I IFN in THP-1 cells with and without PMA differentiation. For comparison, the transcription of the STAT-1 and STAT-2 genes was also measured. The amount of radioactivity incorporated into nascent RNA was normalized to the signal for actin transcription, so that different nuclear preparations could be directly compared. The actin signal did not vary >20% among experimental conditions, indicating that actin is an appropriate internal standard.
Surprisingly, PMA treatment by itself did not significantly increase
IRF-9 transcription (Fig. 6A),
while it did produce a small increase in STAT-1 transcription. After a
4-h IFN-ß stimulation of the untreated or PMA-treated cells, there
was little or no change in IRF-9 transcription, while STAT-1
transcription was strongly induced. These data provide further evidence
of a functional response to type I IFN in THP-1 cells with and without
PMA treatment and show that regulation of IRF-9 DNA-binding activity is
post-transcriptional.
We next determined whether changes in IRF-9 mRNA could account for the
increased IRF-9 DNA-binding activity observed upon differentiation. By
Northern blot analysis, IRF-9 mRNA was barely detectable in untreated
THP-1 cells (Fig. 6B, lane 1). PMA treatment
markedly increased IRF-9 mRNA (Fig. 6B, lane 2).
Type I IFN treatment led to a further increase in IRF-9 expression
(Fig. 6B, lanes 2 and 4). Finally, we
assayed IRF-9 protein abundance by immunoblot. Little or no IRF-9
protein was detected in THP-1 cells and primary monocytes (Fig. 6C, lanes 1 and 4). IRF-9 protein was
markedly up-regulated in PMA-treated THP-1 cells and primary
macrophages (Fig. 6C, lanes 2 and 6).
Importantly, IRF-9 was not induced in monocytes by infection with
M. tuberculosis (Fig. 6C, compare lanes
4 and 5) and was only slightly induced in macrophages
after M. tuberculosis infection (Fig. 6C, compare
lanes 6 and 7). Recombinant IRF-9 was used to
confirm the authenticity of the bands observed in monocytes and
macrophages (Fig. 6C, lane 3). These data
demonstrate that differentiation increased IRF-9 activity by increasing
IRF-9 protein. Furthermore, the occurrence of a significant change in
IRF-9 mRNA and protein, but not transcription, supports the conclusion
that induction of IRF-9 upon differentiation is post-transcriptionally
regulated.
Differentiation did not alter NF-B activation after
proinflammatory stimuli
It is possible that the effect of differentiation on viral
replication after a proinflammatory stimulus is mediated by more than
one transcription factor. NF-B is another transcription factor that
is activated by M. tuberculosis infection or other
proinflammatory stimuli such as LPS. The HIV-1 LTR is strongly
stimulated by NF-B, and inhibition of NF-B activation could add
to the effect of differentiation on LTR function and viral replication
after a proinflammatory stimulus. To test whether the down-regulation
of the HIV-1 LTR observed in differentiated THP-1 cells after
proinflammatory stimulation is due in part to altered regulation of
NF-B, the effects of M. tuberculosis infection or LPS
treatment were examined. EMSA with an NF-B probe was performed with
whole-cell or nuclear extracts of untreated and PMA-treated THP-1 cells
prepared 3 days after infection with M. tuberculosis. M.
tuberculosis infection increased NF-B activation 4-fold in
differentiated and undifferentiated THP-1 cells (Fig. 7A), as judged by formation of
a protein/DNA complex that was supershifted with Ab to NF-B or
competed with excess unlabeled NF-B oligonucleotide (data not
shown). In time-course experiments using LPS as the proinflammatory
stimulus, p65 Rel A translocated to the nucleus within 3 h with
and without PMA treatment (Fig. 7B). There was some
variation from sample to sample, but nuclear localization of NF-B,
indicating activation, remained high over the 72-h time course in both
cell types. These data demonstrate that differentiation does not limit
NF-B activation after a proinflammatory stimulus and suggest that
the inhibitory effect of the 16-kDa C/EBPß factor is dominant over
the stimulatory effect of NF-B.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this paper we examined primary monocytes and macrophages and, as a model, the THP-1 cell line before or after PMA treatment, for the effect of differentiation on HIV-1 replication. We observed that macrophages and PMA-treated THP-1 cells gained the ability to respond to proinflammatory stimuli by induction of the type I IFN-specific transcription factor ISGF-3. Up-regulation of the IRF-9 component of ISGF-3 during differentiation was a critical regulator of the IFN response. Importantly, M. tuberculosis infection of PMA-treated THP-1 cells led to secretion of type I IFN. This probably accounts entirely for the observed activation of ISGF-3, since no other activator is known. However, alternative mechanisms cannot be formally ruled out. The ability to produce ISGF-3 was associated with suppressed HIV-1 replication and induction of an inhibitory C/EBPß transcription factor after an inflammatory stimulus. In the absence of IRF-9, monocytes may not be able to counter the proinflammatory effect of M. tuberculosis infection on HIV-1 replication.
In vitro infection of primary macrophages or the model PMA-treated THP-1 cells with M. tuberculosis did not reproduce the state of activation observed in alveolar macrophages during the cellular immune response to active tuberculosis. Macrophages in lung segments involved with tuberculosis lose C/EBPß expression, express another C/EBP site-binding activity, and support high level HIV-1 replication (4, 5, 15). As expected, PMA-treated THP-1 cells or primary macrophages infected in vitro with M. tuberculosis exhibited a type I IFN response characteristic of resting alveolar macrophages, which strongly expressed the inhibitory C/EBPß and suppressed HIV-1 replication (15).
The interaction of C/EBP transcription factors with the LTR is
particularly important for control of viral replication in macrophages.
Mutation of the C/EBP sites in the LTR produces a virus that continues
to replicate in lymphocytes, but is no longer able to replicate in
macrophages (21). Monocytic THP-1 cells and primary
monocytes are able to up-regulate HIV-1 replication after infection
with M. tuberculosis or stimulation with LPS (11, 22). The enhancing effects of proinflammatory stimuli require
intact C/EBP and NF-B sites in the HIV-1 LTR (6, 7).
Surprisingly, the opposite effect was observed when THP-1 cells were
differentiated to macrophage-like cells by PMA treatment. Repression
requires intact C/EBP sites and is associated with induction of an
inhibitory 16-kDa C/EBPß (15, 22). In the present study
primary monocytes differentiated to macrophages by M-CSF also
suppressed HIV-1 replication after infection with mycobacteria and were
able to induce the inhibitory C/EBPß. Therefore, THP-1 cells are an
appropriate model for investigating the mechanisms underlying the
production of transcriptional repressors during an IFN response.
After infection with M. tuberculosis, monocyte-like THP-1 cells induced only stimulatory 37-kDa C/EBPß. In contrast, infection of PMA-treated THP-1 cells or primary macrophages also induced a dominant-negative 16-kDa C/EBPß transcription factor. This result is at most partly due to an autocrine response to type I IFN, since IFN-independent mechanisms can also lead to induction of 16-kDa C/EBPß in macrophages (M. Weida and N. Tanaka, unpublished observations). Thus, differentiation enables the cell to produce short form C/EBPß either by regulating translation start site utilization during protein synthesis or by inducing a protease (24, 26, 27, 28). Both stimulatory and inhibitory C/EBPß isoforms were expressed in macrophages, but repression occurred because the 16-kDa C/EBPß is a dominant negative factor that represses transcription when expressed at 20% the level of the stimulatory 37-kDa C/EBPß (24). Expression of the inhibitory C/EBPß represses promoters containing C/EBP sites such as the HIV-1 LTR (6) and also disrupts the function of the ubiquitously expressed serum response factor (55). Therefore, cells that express high levels of inhibitory C/EBPß may have global inhibition of multiple promoters. Modulation of C/EBPß expression by differentiation is likely to have a profound effect on viral replication and expression of proinflammatory genes.
The dominant negative C/EBPß transcription factor was induced by type I IFN in THP-1 cells only if they were first treated with PMA. Since without PMA treatment, THP-1 cells also fail to produce ISGF-3 after IFN stimulation, a possible explanation for the failure to alter C/EBPß expression in the monocyte-like cells after IFN stimulation is that ISRE-mediated transcription is required, either directly or indirectly, to up-regulate inhibitory C/EBPß expression. The IFN-induced transcriptional repressor would inhibit viral replication only after the provirus had integrated into genomic DNA. The observation that type I IFNs or proinflammatory stimuli such as LPS inhibit viral replication at a stage of the viral life cycle before viral integration (56, 57, 58) supports the idea that IFNs can protect both monocytes and macrophages from viral infection at a stage before LTR promoter function is required. Macrophages, however, gain the ability to inhibit LTR-mediated transcription after type I IFN stimulation. The induction of inhibitory transcription factors by IFNs is therefore particularly important for the production of latent infection and may provide a selective advantage to the virus by creating a reservoir of infected macrophages able to avoid destruction by HIV-1-specific cytolytic T cells.
Only PMA-treated, macrophage-like THP-1 cells or alveolar macrophages had significant activation of ISGF-3, the type I IFN-specific transcription factor complex, in response to type I IFN stimulation or mycobacterial infection. This was associated with increased IRF-9 activity in both the cell line and primary macrophages. Infection with M. tuberculosis did not up-regulate IRF-9 expression in primary monocytes. This probably contributes to their inability to control HIV-1 replication after M. tuberculosis infection.
Regulation of the abundance of IRF-9 is an important mechanism for modulating IFN responsiveness (59). Further support for this conclusion comes from the observations that induction of ISGF-3 in undifferentiated monocytic cell lines is delayed and requires protein synthesis (60), while cells differentiated with PMA or all-trans-retinoic acid have increased levels of IRF-9 and exhibit rapid activation of ISGF-3 after IFN stimulation without the need for protein synthesis (38, 39). The molecular role of IRF-9 in formation of ISGF-3 has functional significance, since transfection of IRF-9 into IRF-9-deficient cells restores the antiviral response to type I IFN (61). The fact that differentiation did not alter the transcription rate of the IRF-9 gene while both mRNA and protein abundance were increased demonstrates that the effect of differentiation on IRF-9 levels is post-transcriptionally mediated. Furthermore, induction of IRF-9 by type I IFN was post-transcriptionally regulated in THP-1 cells both before and after PMA treatment.
Insufficient IRF-9 may account for the inability of monocyte-like THP-1 cells to form ISGF-3, since the addition of IRF-9 to extracts of type I IFN-treated cells reconstituted ISGF-3. The ability to reconstitute ISGF-3 with IRF-9 shows that both latent STAT-1 and STAT-2 were activated by type I IFN in monocyte-like and macrophage-like cells. These results are consistent with the observation that JAK-1 and TYK-2, the enzymes responsible for STAT-1 and STAT-2 phosphorylation in response to type I IFN, are activated after IFN stimulation of undifferentiated THP-1 cells (62). Further, at least in fibroblastic cells, STAT-2 phosphorylation is required for activation of STAT-1 and formation of STAT-1 homodimers in response to type I IFN (63, 64, 65). Thus, THP-1 cells have intact signal transduction through the type I IFN receptor.
Infection does not lead to formation of STAT-1 homodimers in THP-1
cells before or after PMA treatment, even though in both cases they are
able to respond to type I IFN stimulation by formation of STAT-1
homodimers and induction of STAT-1 transcription. Moreover,
proinflammatory stimuli induce autocrine type I IFN in both monocytes
and macrophages (14, 15). Infection may up-regulate any of
the mechanisms known to limit activation or DNA-binding activity of
STAT-1. First, suppressor of cytokine signaling-1 inhibits the
activation of JAK-1 and TYK-2, thereby limiting activation of STAT-1 in
response to type I IFN (45, 46). Second, protein inhibitor
of activated STAT-1 can block the DNA-binding activity of STAT-1
homodimers. The differential effects of these two mechanisms on the
STAT-1 homodimer/ISGF-3 ratio is unknown. Third, src homology domain 2
containing protein tyrosine phosphatase limits accumulation of
tyrosine-phosphorylated STAT-1, thus reducing formation of STAT-1
homodimers relative to that of ISGF-3 (40, 44). It is also
possible that differentiation or infection may modulate the serine
kinase ERK2 or the tyrosine phosphatase protein tyrosine phosphatase-1D
systems, which play a role in forming functional ISGF-3 (41, 42). The most likely explanation for the failure to produce
STAT-1 homodimers after infection, however, is the dose-response
characteristics of the type I IFN system. STAT-1 homodimers were
observed after stimulation of THP-1 cells with 100 U/ml of IFN-ß or
500 U/ml of IFN-, but not after stimulation with 10 U/ml of IFN-ß
or 50 U/ml of IFN- (R. Pine and A. Canova, unpublished observations)
before or after PMA treatment. M. tuberculosis induced less
ISGF-3 than 500 U/ml of IFN-. Taken together, these data suggest
that M. tuberculosis infection produces relatively low
levels of type I IFN, inducing ISGF-3 if sufficient IRF-9 is present,
but not STAT-1 homodimer.
Before or after PMA treatment, THP-1 cells activated NF-B to a
similar extent in response to a proinflammatory stimulus. Thus, the
difference in M. tuberculosis uptake did not significantly
alter the extent of cellular activation in either culture. The NF-B
family of proteins is essential for activating type I IFN production
after viral infection (66). Despite this fundamental
response to infection, the set of genes activated by the innate immune
response depends on the state of cellular differentiation. Further, the
high level expression of NF-B when the LTR is repressed suggests
that inhibitory C/EBPß is dominant over the stimulatory effects of
NF-B. This finding reinforces the idea that the LTR is repressed by
strong expression of the inhibitory C/EBPß until the cellular immune
response activates the alveolar macrophage. In patients with pulmonary
tuberculosis, the level of inhibitory C/EBPß is reduced when high
levels of LTR promoter activity are expressed (15).
Therefore, derepression may be needed before promoters containing C/EBP
sites can achieve high levels of expression in tissue macrophages.
Expression of the inhibitory C/EBPß would have the effect of
preventing high levels of proinflammatory cytokine production in the
lung due to activation of NF-B or through other mechanisms after a
brief inhalation of environmental LPS or other proinflammatory
stimuli.
The regulation of HIV-1 replication in monocytes and macrophages bears
a striking similarity to expression of proinflammatory cytokines, such
as IL-6 and TNF-. The high degree of correlation between HIV-1
production in the lung during tuberculosis and the levels of TNF-
and IL-6 may indicate coordinate regulation of cytokine promoters and
the HIV-1 LTR (4). The promoters of TNF- and IL-6 share
C/EBP and NF-B transcription factor binding sites with the LTR
(67). The TNF- promoter, like the HIV-1 LTR, is
strongly repressed by engineered expression of the inhibitory C/EBPß
in monocytes (68). C/EBPß knockout mice, but not mice
deficient in both IL-6 and C/EBPß, develop a lymphoproliferative
disease associated with high levels of IL-6 (69, 70),
suggesting that C/EBPß is an important negative regulator of IL-6
production in vivo. Finally, aerosolized IFN- inhibits IL-6 and
TNF- production in the lungs of patients with tuberculosis
(71). Thus, the innate immune response to proinflammatory
stimuli should have similar effects on cytokine production and LTR
function.
The inclusion of C/EBP sites in the HIV-1 LTR may be another example of HIV-1 usurping normal immune regulation for its own advantage. Pathways that inhibit cytokine production may also promote viral latency, protecting provirally infected macrophages from immune surveillance. Alternately, up-regulation of viral replication by the same transcription factors that up-regulate proinflammatory cytokines during pulmonary tuberculosis leads to enhanced viral production in the context of opportunistic infections. Alteration of C/EBP expression may underlie many circumstances in which immune system activation is associated with enhanced viral production. An improved understanding of the interaction of C/EBP transcription factors and the IFN system may contribute to our ability to blunt the enhanced HIV-1 replication observed in macrophages during opportunistic infection.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Richard Pine, Public Health Research Institute, 455 First Avenue, New York, NY 10016.
3 Abbreviations used in this paper: LTR, long terminal repeat; C/EBP, CCAAT/enhancer binding protein; GAS, IFN--activated site; HAART, highly active antiretroviral therapy; IRF, IFN regulatory factor; ISGF-3, IFN-stimulated gene factor-3; ISRE, IFN-stimulated response element; MOI, multiplicity of infection; VSV, vesicular stomatitis virus.
Received for publication August 20, 1999. Accepted for publication May 25, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. J. Virol. 65:6425. activates human immunodeficiency virus type 1 through induction of nuclear factor binding to the NFB sites in the long terminal repeat. Proc. Natl. Acad. Sci. USA 86:5974. and ß and alters macrophage responses. Proc. Natl. Acad. Sci. USA 92:11284./ß suppresses colony-stimulating factor (CSF-1) stimulated macrophage DNA synthesis and mediates inhibitory effects of LPS and TNF-. J. Immunol. 156:2553.[Abstract]
