| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Immunology and Bacteriology, University of Glasgow, Glasgow, United Kingdom; and
Institute for Molecular Biosciences and Department of Biochemistry, University of Queensland, Brisbane, Queensland, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
production in
response to LPS but suppressed the same response to CpG DNA. CSF-1 also
regulated cytokine gene expression in response to CpG DNA and LPS; CpG
DNA-induced IL-12 p40, IL-12 p35, and TNF- mRNAs were all suppressed
by CSF-1 pretreatment. CSF-1 pretreatment enhanced LPS-induced IL-12
p40 mRNA but not TNF- and IL-12 p35 mRNAs, suggesting that part of
the priming effect is posttranscriptional. CSF-1 pretreatment also
suppressed CpG DNA-induced nuclear translocation of NF-
B and
phosphorylation of the mitogen-activated protein kinases p38 and
extracellular signal-related kinases-1/2 in BMMs, indicating that early
events in CpG DNA signaling were regulated by CSF-1. Expression of
Toll-like receptor (TLR)9, which is necessary for responses to CpG DNA,
was markedly suppressed by CSF-1 in both BMMs and
thioglycolate-elicited peritoneal macrophages. CSF-1 also
down-regulated expression of TLR1, TLR2, and TLR6, but not the LPS
receptor, TLR4, or TLR5. Hence, CSF-1 may regulate host responses to
pathogens through modulation of TLR expression. Furthermore, these
results suggest that CSF-1 and CSF-1R antagonists may enhance the
efficacy of CpG DNA in vivo. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Toll-like receptors
(TLRs)3 are an
evolutionarily conserved family that share homology with the IL-1R
family in the cytoplasmic domain. Mammalian TLRs are critical in
instigating responses to bacterial products. C3H/HeJ LPS nonresponder
mice contain an inactivating point mutation in the TLR4
gene (6, 7), and TLR4-deficient mice do not produce
inflammatory cytokines in response to LPS (8).
TLR2-deficient mice are still LPS responsive but fail to respond to
bacterial lipoproteins or peptidoglycan (9), and
TLR9-deficient mice are incapable of responding to CpG-containing DNA
(10). Engagement of TLRs triggers signaling through at
least NF-B and the mitogen-activated protein kinase (MAPK) family
members, extracellular signal-related kinase (ERK)-1 and -2, p38, and
c-Jun N-terminal kinase, and results in transcription of
proinflammatory genes (1, 11).
The macrophage response to bacterial products is also regulated by a
variety of endogenous cytokines. Both IFN-
and GM-CSF (primarily T
cell products) can prime the inflammatory response, whereas IL-4,
IL-10, and TGF-
are able to suppress macrophage activation. CSF-1, a
cytokine that regulates growth, differentiation, and function of
macrophages, is readily detectable in peripheral blood in the steady
state and is further induced in vivo after infection (12)
or challenge with LPS (13). Priming of macrophages with
CSF-1 can enhance LPS-induced IL-1, IL-6, and TNF- production
(14). CSF-1 treatment in vivo increased levels of
LPS-induced TNF- and IL-6 (15), and CSF-1-deficient
op/op mice have enhanced resistance to LPS shock
(16). Furthermore, CSF-1 primed human monocytes for
enhanced responses to LPS (17). These effects of CSF-1
might be due to regulation of LPS recognition via the CD14-TLR4-MD2
complex or at subsequent levels (MyD88, TNFR-associated factor 6,
IL-1R-associated kinase) that appear to be shared with other
microbial agonists such as CpG DNA (18, 19). To
distinguish these alternatives, we compared the effect of CSF-1 on the
response to LPS and CpG DNA in primary murine bone marrow-derived
macrophages (BMMs). We report in this work that CSF-1 enhanced
macrophage responses to LPS but suppressed expression of TLR9 and
responses to CpG DNA. These findings provide a mechanism for the
differential cytotoxic effect of LPS and CpG DNA and may lead to novel
therapeutic strategies against bacteria-induced shock.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RPMI 1640 medium (Life Technologies, Paisley, U.K.) containing 10% FCS, penicillin/streptomycin, and glutamine (complete medium) was used for culture of BMMs. BMMs were derived from the femurs of adult BALB/c mice (Harlan Olac, Bichester, U.K.). In some experiments, adult CD1 outbred mice were used for preparation of BMMs with similar results. Briefly, femurs were flushed with complete medium and bone marrow cells were plated out in complete medium containing 104 U/ml (100 ng/ml) recombinant human CSF-1 (a gift from Chiron, Emeryville, CA) on 10-cm bacteriological plastic plates (Bibby Sterilin, Staffordshire, U.K.) for 7 days in a 37°C incubator containing 5% CO2. Thioglycolate-elicited peritoneal macrophages (TEPMs) were obtained by injecting BALB/c mice i.p. with 1 ml of 10% thioglycolate broth followed by peritoneal lavage with 10 ml of PBS 5 days later. LPS from Salmonella minnesota (Sigma-Aldrich, St. Louis, MO) was used at a final concentration of 100 ng/ml in all cell culture experiments. The synthetic tripalmitoylated lipopeptide Pam3CysSerLys4 (Roche, Basel, Switzerland) was used at a final concentration of 10 ng/ml in cell culture. Phosphodiester oligodeoxynucleotides (Sigma-Genosys, Cambridge, U.K.) were used at a final concentration of 3 µM in cell culture. Oligodeoxynucleotides used were activating oligonucleotide-1 (AO-1) (5'-GCTCATGACGTTCCTGATGCTG-3') and nonactivating oligonucleotide-1 (NAO-1) (5'-GCTCATGAGCTTCCTGATGCTG-3') (20). IL-3 (a gift from Dr. A. Hapel, Australian National University, Canberra, Australia) stored at a concentration of 104 U/ml at -20°C was used at 103 U/ml in cell culture. PMA (Sigma-Aldrich) was stored as a stock solution (10 mg/ml) in DMSO at -70°C and used at a final concentration of 100 ng/ml.
In vitro treatment of cells and ELISAs
For all in vitro experiments, BMMs were plated out in 24-well plates at 5 x 105 cells per well in 1 ml of complete medium with or without CSF-1 (104 U/ml) overnight. The next morning, cells were stimulated with 100 ng/ml LPS, 3 µM CpG-containing oligonucleotide (AO-1), 3 µM control oligonucleotide (NAO-1), 10 ng/ml Pam3CysSerLys4, or medium. After 24 h (unless otherwise stated), supernatants were collected and stored at -20°C until ELISAs were performed. ELISAs were conducted using paired Abs (BD PharMingen, San Diego, CA).
Immunoblotting
BMMs (2 x 106) were plated on 60-mm dishes (Corning Life Sciences, Acton, MA) in 5 ml of medium or 5 ml of medium plus CSF-1 (105 U/ml) for 18 h. Culture medium was reduced to 2 ml and cells were treated as described in the figures. Cell monolayers were lysed with boiling 66 mM Tris-Cl (pH 7.4)/2% SDS/1 mM sodium vanadate/1 mM sodium pyrophosphate/1 mM sodium molybdate/10 mM sodium fluoride. Equal amounts of total protein in cell extracts were resolved by SDS-PAGE with 10% polyacrylamide seperating gels, transferred to Immobilon-P (Millipore, North Ryde, New South Wales, Australia), blocked, and probed with the anti-phospho p42/p44 MAPK rabbit polyclonal Ab (1:1000) (New England Biolabs, Beverly, MA), washed, and incubated with HRP-linked anti-rabbit IgG (1:2000) (New England Biolabs). Blots were washed and detected using ECL Plus reagents (Amersham Pharmacia Biotech, Piscataway, NJ) and Hyperfilm-ECL (Amersham Pharmacia Biotech). Membranes were then sequentially stripped with 63 mM Tris-Cl (pH 6.7)/2% SDS/100 mM 2-ME and reprobed with rabbit anti-phospho p38 (New England Biolabs), rabbit anti-p42/p44 MAPK (New England Biolabs), and rabbit anti-p38 (New England Biolabs).
Nuclear extract preparation and gel shift assays
The methods used for preparation of nuclear extracts and gel
shift assays have been described previously (21). For
nuclear extract production, BMMs were pretreated overnight with medium
or CSF-1 (5 x 104 U/ml) and then stimulated
for 45 min with AO-1 (0.3 µM) or LPS (10 ng/ml). For gel shift
assays, a double-stranded oligonucleotide corresponding to an NF-B
site from the murine TNF- promoter
(5'-CAAACAGGGGGCTTTCCCTCCTC-3') (21) was end-labeled
with [-32P]ATP using polynucleotide kinase
and separated on a NAP-5 column (Amersham Pharmacia Biotech).
Total RNA isolation and quantitative PCR
Total RNA was prepared using RNAzol B (Biogenesis, Poole, U.K.)
according to the manufacturer’s instructions. RNA was treated with
DNase 1 (Ambion, Austin, TX) and reverse transcribed to cDNA using
Superscript reverse transcriptase (Life Technologies). Negative control
samples (no first strand synthesis) were prepared by performing reverse
transcription reactions in the absence of reverse transcriptase. cDNA
levels of murine IL-12 (p40), IL-12 (p35), TNF-, hypoxanthine
phosphoribosyl transferase (HPRT), TLR1, TLR2, TLR4, TLR5, TLR6, and
TLR9 were quantitated by real-time PCR using an ABI Prism 7700 sequence
detector (Applied Biosystems, Foster City, CA) according to the
manufacturer’s instructions. Amplification was achieved using an
initial cycle of 50°C for 2 min and 95°C for 10 min, followed by 40
cycles of 95°C for 15 s and 50°C for 1 min. cDNA levels during
the linear phase of amplification were normalized against HPRT
controls. Determinations were made in triplicate and mean ± SD
was determined. Primers (f, forward; r, reverse) and
5'-6-carboxy-fluorescein-labeled/3'-6-carboxy-tetramethyl-rhodamine-labeled
probes (p) used to detect expression of the corresponding murine genes
were as follows: IL-12 (p40) (f, 5'-GGAATTTGGTCCACTGAAATTTTAAA-3';
r, 5'-CACGTGAACCGTCCGGAGTA-3'; p,
5'-ACAAGACTTTCCTGAAGTGTGAAGCACCAAAT-3'); IL-12 (p35) (f,
5'-AAGACATCACACGGGACCAAA-3'; r,
5'-CAGGCAACTCTCGTTCTTGTGTA-3'; p,
5'-CAGCACATTGAAGACCTGTTTACCACTGGA-3'); TNF- (Applied Biosystems);
TLR1 (f, 5'-TGGATGTGTCCGTCAGCACTA-3'; r, 5'-AGAGCAGCCCTGGTCTTCAA-3'; p,
5'-CACACACTTGATGTTAGACAGTTCCAAACCGAT-3'); TLR2 (f,
5'-AAGATGCGCTTCCTGAATTTG-3'; r, 5'-TCCAGCGTCTGAGGAATGC-3'; p,
5'-CGTTTTTACCACCCGGATCCCTGTACTG-3'); TLR4 (f,
5'-AGGAAGTTTCTCTGGACTAACAAGTTTAGA-3'; r,
5'-AAATTGTGAGCCACATTGAGTTTC-3'; p, 5'-GCCAATTTTGTCTCCACAGCCACCA-3');
TLR5 (f, 5'-GCACGAGGCTTCTGCTTCA-3'; r,
5'-GCATCCAGGTGTTTGAGCAA-3'; p,
5'-CATTCTGTGCCCATTCAAAGTCTTTGCTG-3'); TLR6 (f,
5'-CTCGGAGACAGCACTGAAGTCA-3'; r, 5'-CGAGTATAGCGCCTCCTTTGAA-3'; p,
5'-ATGATAGAGCACGTCAAAAACCAAGTGTTCCTC-3'); TLR9 (f,
5'-AGGCTGTCAATGGCTCTCAGTT-3'; r, 5'-TGAACGATTTCCAGTGGTACAAGT-3'; p,
5'-TGCCGCTGACTAATCTGCAGGTGCT-3'); HPRT (f,
5'-GCAGTACAGCCCCAAAATGG-3'; r, 5'-AACAAAGTCTGGCCTGTATCCAA-3'; p,
5'-TAAGGTTGCAAGCTTGCTGGTGAAAAGGA-3').
PCR/Southern hybridization
Total RNA was treated with DNase 1 (Ambion), reverse transcribed to cDNA using Superscript reverse transcriptase (Life Technologies), and used as a template for semiquantitative PCR. No reverse-transcriptase negative controls were performed for all samples. Primers used were as follows: TLR9 (f, 5'-CTACAACAGCCAGCCCTTTA-3', r, 5'-GCTGAGGTTGACCTCTTTCA-3'); TLR4 (f, 5'-AGAGAATCTGGTGGCTGTGG-3'; r, 5'-TCAACCGATGGACGTGTAAA-3'); and HPRT (f, 5'-GTTGGATACAGGCCAGACTTTGTTG-3'; r, 5'-GAGGGTAGGCTGGCCTATAGGCT-3'). PCR cycling conditions were as follows: 94°C for 30 s, 54°C for 30 s, 72°C for 60 s (21 cycles for HPRT, 21 cycles for TLR4, 21 cycles for TLR9 with BMM cDNA, or 24 cycles for TLR9 with TEPM cDNA). PCR products were separated on 1.8% agarose gels, transferred to zetaprobe nylon membrane (Bio-Rad, Hercules, CA), and subjected to Southern hybridization using cDNA probes for TLR9, TLR4, and HPRT. Probes were labeled by random priming (Amersham Pharmacia Biotech).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To compare the effects of CSF-1 on macrophage responses to LPS and
CpG DNA, we used primary BMMs, which respond well to both agents
(20), and measured production of inflammatory cytokines by
ELISA. Fig. 1
demonstrates that overnight
pretreatment of BMMs with CSF-1 enhanced levels of LPS-induced IL-6,
IL-12, and TNF- protein release into the medium by 15-, 72-, and
6-fold, respectively. In direct contrast, CSF-1 pretreatment suppressed
CpG DNA-induced IL-6, IL-12, and TNF- by 10-, 8-, and 7-fold,
respectively. We analyzed the effect of CSF-1 on CpG DNA responses in
more detail. Fig. 2A shows
that CSF-1 down-regulated induction of IL-6 by CpG DNA in BMMs over a
concentration range, whereas Fig. 2B shows that this effect
was apparent over the length of a 24-h time course. Because IFN- can
prime macrophage responses to both LPS and CpG DNA (21, 22), we assessed whether IFN- could overcome the suppressive
effect of CSF-1 on the DNA response. As expected, IFN- pretreatment
enhanced levels of LPS-induced IL-6 and IL-12 in BMMs (Fig. 3). This priming effect was not as
striking in CSF-1-pretreated macrophages, suggesting that CSF-1 and
IFN- might provide similar priming signals. Levels of CpG
DNA-induced IL-6 and IL-12 were also enhanced by IFN- priming, but
priming with IFN- did not overcome the suppressive effect of CSF-1
on the CpG DNA response (Fig. 3).
|
|
|
shows that only CSF-1 was able to
enhance the LPS response and suppress the CpG DNA response. IL-3
pretreatment primed responses to both LPS and CpG DNA, whereas PMA did
not affect the LPS response but suppressed the CpG DNA response
slightly (1.5- to 2-fold). Hence, the ability of CSF-1 to enhance LPS
responses and suppress CpG DNA responses is unlikely to be related to
its ability to trigger macrophage proliferation.
|
and IL-12 mRNA by
CSF-1
To determine the level at which CSF-1 differentially regulates the
LPS and CpG DNA response, we assessed mRNA levels of IL-12 p40 and p35
and TNF- in response to LPS and CpG DNA with or without CSF-1
priming (Fig. 5). Priming with CSF-1
enhanced LPS-induced IL-12 (p40) mRNA levels at 4 h by 
10 fold
but did not alter LPS-induced IL-12 (p35) and TNF- mRNAs. Whereas
CSF-1 was selective in priming LPS-induced cytokine mRNAs, levels of
IL-12 (p40), IL-12 (p35), and TNF- after 4 h of CpG DNA were
all suppressed by priming with CSF-1 (6-, 4.5-, and 3-fold,
respectively). This suppressive effect was also apparent at 2 h
post-CpG DNA treatment (data not shown).
|
B, p38,
and ERK-1/2 MAPK activation
NF-B and p38 MAPK activation are early events in triggering
both LPS and CpG DNA-induced gene expression (21, 27, 28, 29).
Nuclear translocation of NF-B in response to CpG DNA but not LPS was
inhibited by CSF-1 pretreatment (Fig. 6A). Similarly,
phosphorylation of p38 in response to CpG DNA was suppressed by
pretreatment with CSF-1 from the earliest time point examined (Fig. 6B). To determine whether CSF-1 altered the ligand
dose-response curve (sensitivity) or the maximal response, the effect
of CSF-1 pretreatment on p38 phosphorylation over a range of CpG DNA
doses was determined at 30 min poststimulation. Fig. 7A demonstrates that 10-fold
higher concentrations of CpG DNA were required to induce p38
phosphorylation in CSF-1-pretreated cells. In contrast,
LPS-induced p38 phosphorylation was not affected by CSF-1 pretreatment
(Fig. 7B), indicating that enhancement of LPS responses by
CSF-1 occurred independently of p38 activation. Phosphorylation of
ERK-1/2 is also triggered by LPS and CpG DNA in BMMs (20),
and we assessed the effect of CSF-1 pretreatment on CpG DNA- and
LPS-triggered ERK-1 and -2 phosphorylation (Fig. 7). Because CSF-1
itself triggers sustained phosphorylation of ERK-1 and -2 in BMMs
(30), basal levels of phosphorylated ERK-1 and -2 were
much higher in CSF-1-pretreated BMMs than in untreated BMMs.
Nonetheless, CSF-1 pretreatment blocked the ability of CpG DNA to
enhance levels of phosphorylated ERK-1 and -2 over the concentration
range examined (Fig. 7A), whereas LPS was still able to
activate ERK-1/2 even in the presence of CSF-1 (Fig. 7B).
Furthermore, the extent of the LPS-mediated ERK-1/2 phosphorylation was
not altered by CSF-1 pretreatment, despite the elevated basal
activation state (Fig. 7B). These data suggest that CSF-1
alters an early stage of CpG DNA recognition but acts more distally to
activate LPS responses.
|
|
Mice deficient for TLR9 are unable to respond to CpG-containing
phosphorothioate DNA (10). Whether TLR9 is required for
uptake of DNA, directly recognizes CpG DNA, or lies downstream in the
recognition pathway is yet to be determined (31). The
diminished CpG responsiveness observed above hints at a reduction in
receptor expression or affinity in response to CSF-1. Therefore, we
assessed the effect of CSF-1 on expression of TLR9 mRNA in BMMs.
Indeed, CSF-1 treatment resulted in a 20-fold reduction in TLR9 mRNA
levels in BMMs (Fig. 8A). To
assess the specificity of this response, we examined the effect of
CSF-1 on expression of other TLR family members, because these
receptors are instrumental in triggering cellular responses to other
bacterial products, including LPS, peptidoglycan, and bacterial
lipoproteins. Overnight treatment with CSF-1 did not significantly
affect mRNA levels of the LPS receptor TLR4 (Fig. 8B).
Levels of TLR5 mRNA were also unaffected by CSF-1, but mRNA levels of
TLR1, TLR2, and TLR6 were all suppressed by CSF-1 treatment (2.4-,
4.8-, and 4-fold, respectively). Hence, CSF-1 has a selective effect on
expression of different TLR family members. The effect of CSF-1 was
most pronounced on TLR9 expression (20-fold or greater reduction), but
the moderate suppression of TLR2 and TLR6 mRNA levels by CSF-1 (4- to
5-fold) implies that responses to bacterial lipopeptides,
peptidoglycan, and other TLR2 agonists might also be modulated by
CSF-1. Therefore, we tested the ability of CSF-1 to regulate IL-6
production from BMMs in response to a synthetic lipopeptide,
Pam3CysSerLys4, that is known to act through TLR2 (32).
Fig. 8C shows that although CSF-1 pretreatment clearly
suppressed IL-6 production in response to CpG DNA, it had no effect on
the response to Pam3CysSerLys4. Hence, although CSF-1 can down-regulate
the level of TLR2 mRNA, the effect appears to be insufficient to
modulate IL-6 production in response to a TLR2 agonist.
|
Next, we analyzed the effect of other agents that cause BMM
proliferation on TLR9 expression. Levels of TLR9 mRNA were assessed in
BMMs treated overnight with medium, CSF-1, IL-3, or PMA. Fig. 9A demonstrates that only
CSF-1 was able to dramatically regulate TLR9 expression; IL-3 did not
alter levels of TLR9 mRNA, and PMA, which down-regulated CpG responses
1.5- to 2-fold (Fig. 4
), had a similar effect on TLR9 expression. To
assess the time course of CSF-1 down-regulation of TLR9 expression,
BMMs starved overnight of CSF-1 were treated with CSF-1 over a 20-h
time course. Maximal suppression of TLR9 expression occurred between 4
and 8 h, and an effect was apparent by 1 h post-CSF-1 (Fig. 9B).
|
C). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Even phosphorothioate-stabilized CpG
oligonucleotides have minimal toxicity (4). The relative
nontoxicity of CpG DNA is an attractive feature for its use as a
vaccine adjuvant and for other therapeutic strategies.
Analysis of macrophage gene expression in vitro has suggested that the
differing toxicities of LPS and CpG DNA may be due to both qualitative
and quantitative differences in cytokine gene induction. For example,
CpG DNA was a relatively poor stimulus for IL-1 (21),
and LPS but not CpG DNA stimulated NO production from macrophages
without IFN- priming (21, 22). However, the situation
in vivo is likely to be more complex due to the presence of many other
cytokines, which will modify the responses to LPS and CpG DNA. Here, we
have found that CSF-1, which is present constitutively in vivo and is a
macrophage growth and survival factor, differentially affects the
responses to LPS and CpG DNA. In the presence of CSF-1, the LPS
response was elevated and the CpG DNA response suppressed so that LPS
became far more effective than CpG DNA at stimulating IL-6,
IL-12, and TNF- production from BMMs (62-, 27-, and 23-fold,
respectively).
The ability of CSF-1 to selectively enhance the LPS and suppress the
CpG DNA response of BMMs is unlikely to be related to its activity as a
growth factor. IL-3 and PMA did not have selective effects on LPS and
CpG DNA responses; IL-3 pretreatment primed BMMs for enhanced IL-6
production in response to both LPS and CpG DNA, whereas PMA
pretreatment did not have significant effects on LPS-induced IL-6 and
modestly suppressed CpG DNA-induced IL-6 synthesis. Furthermore, IL-3
and PMA had little effect on TLR9 expression, whereas CSF-1 markedly
suppressed expression of this molecule. IL-3 was actually more
effective at priming CpG DNA responses than LPS responses. Although the
ability of IL-3 to synergize with LPS for macrophage activation has
been documented (25, 26), its ability to regulate
responses to CpG DNA has not been reported. Given that CpG DNA drives
strong Th1 responses in vivo and that IL-3 is a product of activated T
cells, IL-3 may be involved in amplification of responses to CpG DNA in
vivo, as has been suggested for IFN- (22).
Because concentrations of CSF-1 are markedly and rapidly enhanced in serum, spleen, liver, lung, and kidney after LPS administration (13) and during infection (12, 34), macrophages recruited to the site of infection during a bacterial challenge would be expected to have an impaired response to CpG DNA. The implications of this are not obvious, because the role of bacterial DNA in an infection is not clear and will remain so until experiments using bacterial infection models in TLR9 gene-targeted mice are performed. Intact bacterial pathogens do not display their DNA, and detection of CpG DNA during a bacterial challenge may imply that the host has successfully destroyed the invading organism. In this case, CSF-1 may be important in dampening down inappropriate inflammatory responses to bacterial DNA. In contrast, both LPS and CpG DNA rapidly down-regulate cell surface expression of the CSF-1R in BMMs (20). Hence, macrophages present at the site of infection will have already encountered bacterial cell wall products such as LPS and are unlikely to be CSF-1 responsive. Such cells therefore may be hypersensitive to the effects of CpG DNA. Consistent with this model, we have found that LPS and CpG DNA can synergize for IL-6 and IL-12 production from BMMs (data not shown).
TLR9 is an essential component of the DNA response (10),
and the ability of human and murine TLR9 to confer species-specific
responses to different CpG-containing phosphorothioate-stabilized CpG
oligonucleotides is suggestive of a role in direct recognition
(35). The expression of TLR9 was profoundly reduced by
CSF-1, which provides a clear explanation for the suppression of
bacterial DNA responses by CSF-1. CSF-1 also moderately suppressed
TLR1, 2, and 6 mRNA levels in BMMs, but surprisingly did not regulate
IL-6 production in response to the TLR2 ligand Pam3CysSerLys4 (Fig. 8C). This may be because the level of down-regulation of
these TLR family members by CSF-1 is not sufficient to regulate
responses to their ligands. Another possibility is that stimulation
with TLR2 agonists may up-regulate TLR2 expression, which can in turn
overcome the suppressive effect of CSF-1. In this respect, we have
found that CpG DNA does up-regulate TLR9 mRNA levels in BMMs, but only
in the presence of CSF-1 (data not shown), although this autocrine
regulation is apparently insufficent to overcome the suppressive effect
of CSF-1 on CpG DNA responses. It would be interesting to assess the
effect of TLR2 ligands on TLR2 expression in macrophages in the
presence and absence of CSF-1. Indeed, a variety of inflammatory
stimuli, including LPS, selectively up-regulated TLR2 but not TLR4
expression (36, 37). The differential regulation of TLR
members implies that, as in drosophila, different kinds of pathogens
could elicit different outcomes. In keeping with this view, Sing et al.
have reported that Gram-negative organisms induce IFN-, whereas
Gram-positive bacteria lack this activity (38).
Although the ability of CSF-1 to prime murine macrophage responses to
LPS has previously been reported for IL-6 and TNF- production
(14, 15), its effect on LPS-induced IL-12 production has
not been documented. In the case of IL-6, synergy between CSF-1 and LPS
might be partially due to CSF-1-induced GM-CSF production
(39). We have not fully investigated the mechanism by
which CSF-1 augments LPS responses of BMMs. We did find that CSF-1 did
not affect levels of LPS-induced TNF- mRNA, despite having a marked
effect on TNF- protein secretion, implying that posttranscriptional
mechanisms are responsible for this effect. In support of this, p38
phosphorylation and nuclear translocation of NF-B in response to LPS
were not altered by CSF-1 pretreatment. In the case of IL-12,
LPS-induced p40, but not p35, mRNA was enhanced by CSF-1 priming.
Whether the effect of CSF-1 on LPS-induced IL-12 (p40) mRNA is due to
enhanced transcription or enhanced mRNA stability has not yet been
investigated. One possibility is that CSF-1 and LPS synergize at the
level of transcription because CSF-1 triggers sustained phosphorylation
and activation of Ets-2 (30) and Ets-2 is necessary for
full activation of the IL-12 promoter in response to LPS (40, 41).
The majority of our findings used BMMs as a primary macrophage model.
We have found that TEPMs cultured ex vivo retain responsiveness to LPS
but lose responsiveness to CpG DNA. Because of this decline in CpG DNA
responsiveness ex vivo, we have been unable to address the effect of
CSF-1 pretreatment on CpG responses in TEPMs. Nonetheless, the low
basal expression of TLR9 in TEPMs was further down-regulated by
overnight treatment with CSF-1 (Fig. 9), implying that this phenomenon
is likely to occur with all CSF-1-responsive macrophage populations. In
summary, CSF-1 reprograms macrophage responses to different microbial
stimuli. To our knowledge, this is the first report of a cytokine or
growth factor that has differential effects on macrophage responses to
LPS and CpG DNA, and it highlights the importance of regulated
expression of TLRs. Discordant regulation of TLRs may underlie
different toxicities of TLR agonists in vivo and may have relevance for
the role of CSF-1 during bacterial infections. Furthermore, CSF-1 and
CSF-1R antagonists may enhance the efficacy of CpG DNA in therapeutic
strategies and/or increase its toxicity in vivo.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Matthew J. Sweet, Institute for Molecular Biosciences and Department of Biochemistry, University of Queensland, Brisbane, Queensland 4072, Australia. E-mail address: M.Sweet{at}imb.uq.edu.au
3 Abbreviations used in this paper: TLR, Toll-like receptor; MAPK, mitogen-activated protein kinase; ERK, extracellular signal-related kinase; BMM, bone marrow-derived macrophage; TEPM, thioglycolate-elicited peritoneal macrophage; AO-1, activating oligonucleotide-1; NAO-1, nonactivating oligonucleotide-1; HPRT, hypoxanthine phosphoribosyl transferase; f, forward; r, reverse; p, probe.
Received for publication July 30, 2001. Accepted for publication October 24, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-deficient, and IL-1-deficient op/op mice to endotoxin. Ann. NY Acad. Sci. 762:499.[Medline]
primes macrophage responses to bacterial DNA. J. Interferon Cytokine Res. 18:263.[Medline]
-hydroxysteroid dehydrogenase in mouse macrophages, hemopoietic cells, and cell lines and its induction by colony-stimulating factors. J. Immunol. 134:2492.[Abstract]
in vivo and increases the toxicity of lipopolysaccharides. J. Immunol. 156:4570.[Abstract]
-interferon in mice is a function of the lipopolysaccharide component. Infect. Immun. 68:1600. in monocytic cells. J. Exp. Med. 183:147.This article has been cited by other articles:
|
|
 |
|
  A. Achuthan, P. Masendycz, J. A. Lopez, T. Nguyen, D. E. James, M. J. Sweet, J. A. Hamilton, and G. M. Scholz Regulation of the Endosomal SNARE Protein Syntaxin 7 by Colony-Stimulating Factor 1 in Macrophages Mol. Cell. Biol., October 15, 2008; 28(20): 6149 - 6159. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Fancke, M. Suter, H. Hochrein, and M. O'Keeffe M-CSF: a novel plasmacytoid and conventional dendritic cell poietin Blood, January 1, 2008; 111(1): 150 - 159. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Schroder, M. Lichtinger, K. M. Irvine, K. Brion, A. Trieu, I. L. Ross, T. Ravasi, K. J. Stacey, M. Rehli, D. A. Hume, et al. PU.1 and ICSBP control constitutive and IFN-{gamma}-regulated Tlr9 gene expression in mouse macrophages J. Leukoc. Biol., June 1, 2007; 81(6): 1577 - 1590. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Fleetwood, T. Lawrence, J. A. Hamilton, and A. D. Cook Granulocyte-Macrophage Colony-Stimulating Factor (CSF) and Macrophage CSF-Dependent Macrophage Phenotypes Display Differences in Cytokine Profiles and Transcription Factor Activities: Implications for CSF Blockade in Inflammation J. Immunol., April 15, 2007; 178(8): 5245 - 5252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Wei, X.-M. Dai, and E. R. Stanley Transgenic expression of CSF-1 in CSF-1 receptor-expressing cells leads to macrophage activation, osteoporosis, and early death J. Leukoc. Biol., December 1, 2006; 80(6): 1445 - 1453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Irvine, C. J. Burns, A. F. Wilks, S. Su, D. A. Hume, and M. J. Sweet A CSF-1 receptor kinase inhibitor targets effector functions and inhibits pro-inflammatory cytokine production from murine macrophage populations FASEB J, September 1, 2006; 20(11): 1921 - 1923. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. T. Aung, K. Schroder, S. R. Himes, K. Brion, W. van Zuylen, A. Trieu, H. Suzuki, Y. Hayashizaki, D. A. Hume, M. J. Sweet, et al. LPS regulates proinflammatory gene expression in macrophages by altering histone deacetylase expression FASEB J, July 1, 2006; 20(9): 1315 - 1327. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Minns, L. C. Menard, D. M. Foureau, S. Darche, C. Ronet, D. W. Mielcarz, D. Buzoni-Gatel, and L. H. Kasper TLR9 Is Required for the Gut-Associated Lymphoid Tissue Response following Oral Infection of Toxoplasma gondii. J. Immunol., June 15, 2006; 176(12): 7589 - 7597. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Banerjee, R. Gugasyan, M. McMahon, and S. Gerondakis Diverse Toll-like receptors utilize Tpl2 to activate extracellular signal-regulated kinase (ERK) in hemopoietic cells PNAS, February 28, 2006; 103(9): 3274 - 3279. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Heinonen, N. Dube, A. Bourdeau, W. S. Lapp, and M. L. Tremblay Protein tyrosine phosphatase 1B negatively regulates macrophage development through CSF-1 signaling PNAS, February 21, 2006; 103(8): 2776 - 2781. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. L. Roberts, J. A. Dunn, T. D. Terry, M. P. Jennings, D. A. Hume, M. J. Sweet, and K. J. Stacey Differences in Macrophage Activation by Bacterial DNA and CpG-Containing Oligonucleotides J. Immunol., September 15, 2005; 175(6): 3569 - 3576. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. I. Singh, S. M. Cruickshank, D. J. Newton, L. Wakenshaw, A. Graham, J. Lan, J. P. A. Lodge, P. J. Felsburg, and S. R. Carding Toll-like receptor-mediated responses of primary intestinal epithelial cells during the development of colitis Am J Physiol Gastrointest Liver Physiol, March 1, 2005; 288(3): G514 - G524. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Bjorkbacka, K. A. Fitzgerald, F. Huet, X. Li, J. A. Gregory, M. A. Lee, C. M. Ordija, N. E. Dowley, D. T. Golenbock, and M. W. Freeman The induction of macrophage gene expression by LPS predominantly utilizes Myd88-independent signaling cascades Physiol Genomics, February 7, 2005; 19(3): 319 - 330. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Rossi, G. Muralimohan, J. R. Maxwell, and A. T. Vella Staphylococcal enterotoxins condition cells of the innate immune system for Toll-like receptor 4 stimulation Int. Immunol., December 1, 2004; 16(12): 1751 - 1760. [Abstract] |
