Innate cellular production of IFN-γ is suppressed after repeated exposure to LPS, whereas CpG-containing DNA potentiates IFN-γ production. We compared the modulatory effects of LPS and CpG on specific cellular and cytokine responses necessary for NK-cell dependent IFN-γ synthesis. C3H/HeN mice pretreated with LPS for 2 days generated 5-fold less circulating IL-12 p70 and IFN-γ in response to subsequent LPS challenge than did challenged control mice. In contrast, CpG-pretreated mice produced 10-fold more circulating IFN-γ without similar changes in IL-12 p70 levels, but with 10-fold increases in serum IL-18 relative to LPS-challenged control or endotoxin-tolerant mice. The role of IL-18 in CpG-induced immune potentiation was studied in splenocyte cultures from control, LPS-conditioned, or CpG-conditioned mice. These cultures produced similar amounts of IFN-γ in response to rIL-12 and rIL-18. However, only CpG-conditioned cells produced IFN-γ when cultured with LPS or CpG, and production was ablated in the presence of anti-IL-18R Ab. Anti-IL-18R Ab also reduced in vivo IFN-γ production by >2-fold in CpG-pretreated mice. Finally, combined pretreatment of mice with LPS and CpG suppressed the production of circulating IFN-γ, IL-12 p70, and IL-18 after subsequent LPS challenge. We conclude that CpG potentiates innate IFN-γ production from NK cells by increasing IL-18 availability, but that the suppressive effects of LPS on innate cellular immunity dominate during combined LPS and CpG pretreatment. Multiple Toll-like receptor engagement in vivo during infection can result in functional polarization of innate immunity dominated by a specific Toll-like receptor response.

In the setting of acute microbial infection, the innate cellular immune system is responsible for rapidly initiating cytokine and cellular responses essential for early microbicidal host defenses (1). Within the innate immune system, Toll-like receptors (TLR)3 provide an important sensory mechanism for the detection of infectious threats (2). This family of receptors includes at least 10 diverse surface receptors, each recognizing distinct pathogen-associated molecular patterns (PAMPs) and capable of inducing proinflammatory responses. The recognition of bacterial LPS by TLR4-bearing macrophages and dendritic cells, for example, induces the production of cytokines with important regulatory and inflammatory effects, such as IL-1, TNF-α, chemokines, and IL-12 (3, 4, 5, 6). The bioactive heterodimer form of IL-12, IL-12 p70, promotes host defense by acting through IL-12Rs constitutively expressed on NK cells to induce the synthesis of IFN-γ (7). IFN-γ is then responsible for activating a wide range of cellular responses relevant to microbial killing or inflammatory localization (8).

The importance of IL-12 and IFN-γ as essential factors in the control of infection is well documented. For instance, IL-12- or IFN-γ-deficient mice demonstrate increased susceptibility to a diverse selection of experimental infectious challenges (9, 10, 11, 12). Consistent with an important role for TLRs in initiating IFN-γ-dependent antimicrobial immunity, mice deficient in TLR2 expression, or in the MyD88 adapter protein essential for TLR signaling, underproduce these same cytokines and thereby exhibit increased susceptibility to bacterial infection (13, 14). However, unchecked activation of innate cellular immunity, particularly on a systemic level, can result in serious pathology. This is exemplified by the mouse model of acute endotoxemia, in which circulating cytokines are present at high levels, and acute mortality is largely cytokine mediated (15). The functional capacity of the innate cellular immune response to Toll agonists therefore determines whether a specific insult results in a protective or a pathologic outcome.

There is evidence that innate immunity can be self-regulated by repeated TLR activation, resulting in modulatory effects on cytokine production distinct from those of the acute response. For instance, proinflammatory responses to systemic LPS administration become progressively attenuated after repeated injections with sublethal doses of LPS. This phenomenon of endotoxin tolerance is well described in both rodent and human models and results in loss of IL-12 and IFN-γ synthetic capability to high dose LPS challenge (16, 17, 18). Repeated injection with the TLR2 agonist, bacterial lipopeptide, produces a similar in vivo phenotype, suggesting that at least two TLRs are capable of delayed suppressive effects on innate cellular immune function (19). Acquired IL-12 p70 and IFN-γ deficiency to TLR stimuli is not only characteristic of experimental endotoxin tolerance, but is also observed in severely ill humans (20, 21). This phenomenon of clinical immune paralysis is a common response to trauma- or sepsis-induced systemic inflammation, indicating that suppression is a common outcome of generalized physical or inflammatory injuries.

In contrast, TLR9 activation by oligodeoxynucleotides containing the CpG PAMP motif has been reported to durably increase innate cellular immune function (22). With regard to cytokine production, CpG has short term enhancing effects on the production of IFN-γ in response to LPS (23). Because of these features, CpG treatment strikingly increases host resistance to experimental infection and is a potent vaccine adjuvant (22, 24, 25, 26, 27). As a result, CpG has been proposed as a candidate immunotherapy for the prevention of clinical infection (28). However, little is known about how activation of TLR4 and TLR9 in combination affects host immunity when the modulatory effects of each seem to be functionally opposed. This becomes a practical issue in understanding how host immunity is perturbed after microbial infections that result in exposure to many different types of PAMPs.

We hypothesized that CpG and LPS would mediate their distinct modulatory effects by perturbing specific cytokine and cellular steps in the innate cellular immune response leading up to IFN-γ production by NK cells. The delayed effects of TLR4 and TLR9 activation on innate immunity had not been directly compared in previous studies. IFN-γ synthesis was chosen as the biologic readout for innate cellular immune function in these studies because of its central role in inducing both protective and pathologic responses after Toll stimulation. We report in this study that the opposed tolerizing and potentiating effects of LPS and CpG on LPS-induced IFN-γ production are strongly correlated with IL-18 synthetic capacity. Specifically, IL-18 responses to LPS are uniquely increased in CpG-pretreated mice, and the production of IFN-γ becomes highly IL-18 dependent. However, combined pretreatment with LPS and CpG suppresses both IL-18 and IFN-γ synthetic capacity, suggesting that the mechanisms of endotoxin tolerance are dominant over those supporting CpG immune potentiation. The integrated effects of TLRs with distinct immunologic functions, rather than being additive, can instead be characterized by deviation toward a dominant default phenotype. These findings also indicate that the proposed use of CpG as an immunorestorative treatment in critically ill hosts may be compromised if there is concurrent exposure to circulating LPS.

Salmonella enteriditis LPS was purchased from Sigma-Aldrich (St. Louis, MO). Phosphorothioated CpG and control non-CpG oligonucleotides, 1826 (TCCATGACGTTCCTGACGTT, 5′ - 3′) and 1911 (TCCAGGACTTTCCTCAGGTT), respectively (29), were purchased from Oligos Etc. (Wilsonville, OR) and were proven free of LPS contamination by Limulus lysate assay (Sigma-Aldrich).

Four- to 6-wk-old female C3H/HeN and C57BL/6 ν/ν mice were purchased from Charles River (Lexington, MA) and housed at Case Western Reserve University animal facilities under specific pathogen-free conditions. We chose to use LPS-responsive C3H/HeN mice for these studies because they generate higher concentrations of serum cytokine in response to LPS than do C57BL/6 or BALB/c mice and because we had previously used them extensively to characterize endotoxin tolerance (18, 30). For conditioning with Toll agonists, mice were injected i.p. with 0.2 ml of saline containing 50 μg of S. enteriditis LPS or 50 μg of CpG (1.25 mg/kg) on 2 successive days. Seventy-two hours after the first conditioning dose, the mice were challenged by i.p. injection with 300 μg of LPS (LD90), 150 μg of CpG phosphorothiorate oligodeoxynucleotide (S-ODN) 1846, or 150 μg of control non-CpG S-ODN 1911. To avoid toxicity in experiments in which mice were pretreated with both LPS and CpG, only the first injection consisted of 50 μg each of LPS and CpG. The second injection (at 24 h) combined 25 μg of each. Because CpG preconditioning resulted in rapid lethality to subsequent LPS challenge, mice were routinely euthanized no later than 6 h after challenge. All procedures were approved by the Case Western Reserve University institutional animal care and use committee.

Spleens were disrupted in HBSS at 37°C for 15 min, RBCs were lysed using hypotonic ACK lysis buffer (150 mM ammonium chloride, 10 mM potassium carbonate, and 0.1 mM EDTA, adjusted to pH 7.4), and nucleated cells were washed and resuspended in HBSS/1% FCS containing 5 mM EDTA. Unlabeled anti-FcRII/III mAb (10 μg/ml 2.4G2; BD PharMingen, San Diego, CA) was added to block nonspecific labeling due to FcR binding. Cells were then incubated with FITC-anti-mouse IgM, FITC-anti-CD3 (2C11), and FITC-anti-CD19 to label T and B cells. NK cells were labeled using PE-anti-DX-5. Irrelevant fluorochrome or biotin-conjugated rat isotype control Igs were used as negative controls for staining. The cells were again washed and then fixed in 1% formalin before analysis by FACSCalibur (BD Immunocytometry Systems, Mountain View, CA). Cells expressing IgM, CD3, and CD19 were excluded by gating within the FL1 channel, and 20,000 T-/B-negative cells were analyzed for NK or accessory cells. Additional Abs included CD25- and CD122-specific reagents (BD PharMingen) or biotinylated anti-mouse IL-15R and IL-18R Abs (R&D Systems, Minneapolis, MN).

Splenocytes were harvested 3 h after LPS challenge and incubated for 2 h in DMEM/10% FBS medium containing brefeldin A (BD PharMingen), 1 ng/ml rIL-12 (BD PharMingen), and 3 ng/ml rIL-18 (BioSource International, Camarillo, CA). The cell surfaces were stained with FITC-anti-CD3 and allophycocyanin-anti-DX5, followed by fixation and permeabilization using reagents from a commercial kit (BD PharMingen). The cells were subsequently stained with 0.05 μg/ml of either PE-anti-IFN-γ or PE-rat IgG2b isotype control.

Splenocytes harvested as described above were suspended at 107 cells/ml in tissue culture medium (TMC)/10% FBS medium (1/1 mix of RPMI 1640 and DMEM with supplemental 10 mM HEPES (pH 7.4), 1 mM l-glutamine, l-arginine, nonessential amino acids, 50 μM 2-ME, 100 μg/ml penicillin/streptomycin, and 10% FBS). Stimuli were used at concentrations previously determined to be optimal for in vitro induction of IFN-γ. These included LPS (1 μg/ml) and CpG (10 μg/ml). Where indicated, 10 μg/ml neutralizing goat polyclonal anti-IL-18R, goat anti-IL-15R Abs (R&D Systems), or rat anti-mouse IL-12 Ab (clone 17.8) were added to culture. Goat polyclonal anti-influenza hemagglutinin IgG was used as an irrelevant, specifically matched control (Santa Cruz Biotechnology, Santa Cruz, CA).

Cytokine concentrations in serum and conditioned culture medium were determined for IFN-γ and IL-12 p40 using Ab pairs from BD PharMingen as previously described (18). Serum IL-12 p70 was measured using ELISA kits from R&D Systems. Serum IL-18 was measured using an ELISA from Medical and Biological Laboratories (Nagoya, Japan) that detects mature mouse IL-18 with minimal cross-reactivity to precursor forms of IL-18.

Tests for significant differences were made using the nonparametric Mann-Whitney U test.

Groups of five LPS-responsive C3H/HeN mice were pretreated i.p. with 50 μg (1.25 mg/kg) of S. enteriditis LPS daily for 2 successive days or with 50 μg of CpG for 2 days. Seventy-two hours after starting pretreatments, mice were challenged by i.p. injection with high dose LPS (300 μg) or CpG-containing oligonucleotides (150 μg). Six hours after challenge, serum was obtained for ELISA of IFN-γ (Fig. 1). We had previously determined the challenge doses used to be optimal for inducing in vivo cytokine production and had identified 6 h as a suitable single time point for comparing in vivo levels of IFN-γ. As expected, challenge of naive control mice with LPS induced circulating IFN-γ, whereas challenge with CpG resulted in lower levels of serum IFN-γ. Challenge with the control, non-CpG-ontaining oligonucleotide provoked marginally detectable cytokine responses at best (data not shown).

FIGURE 1.

Distinct modulatory effects of LPS and CpG pretreatment on IFN-γ and IL-12 p70 serum responses in vivo. A, LPS pretreatment suppresses and CpG pretreatment increases serum IFN-γ responses to subsequent LPS or CpG challenge. Groups of five female C3H/HeN mice were conditioned with 0.2 ml of saline (□), 0.2 ml of saline containing 50 μg of S. enteriditis LPS (▨), or 50 μg of CpG-containing S-ODN daily for 2 days (▪). Seventy-two hours after the first conditioning dose, all mice were challenged with either 300 μg of LPS or 150 μg of CpG ODN by i.p. injection. Shown are the mean ± SEM level of IFN-γ in serum obtained 6 h postchallenge. These findings were reproduced in five separate experiments. In naive mice not receiving LPS challenge, IFN-γ levels were undetectable. N.D., cytokine not detected. ∗, p < 0.01; ∗∗, p < 0.01. B, LPS pretreatment suppresses and CpG pretreatment preserves serum IL-12 p70 responses to LPS and CpG challenges. Mice were pretreated with saline (□), LPS (▪), and CpG (▨) as described above. Shown are the concentrations of serum IL-12 p70 (picograms per milliliter) at 0, 3, 5, and 6 h after subsequent LPS challenge. Values are averaged across eight different experiments of identical design. Error bars represent SDs of the pooled experimental means. The differences between control and CpG-pretreated mice were not significantly different except at 0 h, when CpG-pretreated serum contained 2-fold more IL-12 p70 (p < 0.05). LPS-pretreated mice generated significantly less IL-12 p70 at all time points tested (p < 0.01).

FIGURE 1.

Distinct modulatory effects of LPS and CpG pretreatment on IFN-γ and IL-12 p70 serum responses in vivo. A, LPS pretreatment suppresses and CpG pretreatment increases serum IFN-γ responses to subsequent LPS or CpG challenge. Groups of five female C3H/HeN mice were conditioned with 0.2 ml of saline (□), 0.2 ml of saline containing 50 μg of S. enteriditis LPS (▨), or 50 μg of CpG-containing S-ODN daily for 2 days (▪). Seventy-two hours after the first conditioning dose, all mice were challenged with either 300 μg of LPS or 150 μg of CpG ODN by i.p. injection. Shown are the mean ± SEM level of IFN-γ in serum obtained 6 h postchallenge. These findings were reproduced in five separate experiments. In naive mice not receiving LPS challenge, IFN-γ levels were undetectable. N.D., cytokine not detected. ∗, p < 0.01; ∗∗, p < 0.01. B, LPS pretreatment suppresses and CpG pretreatment preserves serum IL-12 p70 responses to LPS and CpG challenges. Mice were pretreated with saline (□), LPS (▪), and CpG (▨) as described above. Shown are the concentrations of serum IL-12 p70 (picograms per milliliter) at 0, 3, 5, and 6 h after subsequent LPS challenge. Values are averaged across eight different experiments of identical design. Error bars represent SDs of the pooled experimental means. The differences between control and CpG-pretreated mice were not significantly different except at 0 h, when CpG-pretreated serum contained 2-fold more IL-12 p70 (p < 0.05). LPS-pretreated mice generated significantly less IL-12 p70 at all time points tested (p < 0.01).

Close modal

LPS-and CpG-conditioned mice generated very different patterns of serum IFN-γ production. LPS-conditioned mice were confirmed to be tolerant to LPS challenge, as they generated ∼5-fold less IFN-γ than LPS-challenged control mice. The low levels of IFN-γ induced by CpG challenge in control mice were reduced another 50% in LPS-pretreated mice. In contrast, CpG-pretreated mice produced 10-fold greater serum IFN-γ levels after challenge with LPS and CpG than did control mice (Fig. 1 A). This pattern of opposing modulatory effects of LPS and CpG on IFN-γ synthetic capacity was reproduced in at least eight experiments of similar design.

Serum IL-12 p70 responses after LPS challenge were also suppressed in LPS-pretreated mice. Compared with controls, LPS-conditioned mice consistently produced ∼5-fold less IL-12 p70 after LPS (Fig. 1B) and 10-fold less after CpG challenge (1059 ± 235 vs 92 ± 45 pg/ml, respectively; p < 0.01). Decreased LPS-induced cytokine responses are defining characteristics of endotoxin tolerance (18, 31), but deficient IFN-γ and IL-12 p70 responses after CpG challenge are new findings suggesting that endotoxin-tolerant mice are also cross-tolerant to TLR9 activation. The IL-12 p70 responses of CpG-conditioned mice were consistently greater than those of endotoxin-tolerant mice, but varied by as much as 2-fold compared with those of control mice. However, when the results of eight studies were averaged for serum IL-12 p70 levels at 3, 5, and 6 h after LPS challenge (Fig. 1B), no significant differences between control and CpG-pretreated mice were observed (p > 0.05).

Because IL-12 p70 levels did not increase proportionately with circulating IFN-γ in CpG-pretreated mice, we hypothesized that increased levels of other IFN-γ-inducing cytokines, such as IL-18, critically contributed to the greatly augmented IFN-γ response seen in these mice. We therefore compared circulating levels of IL-18 present before or 5 h after LPS challenge in control, LPS-conditioned, or CpG-conditioned C3H/HeN mice (Fig. 2). LPS challenge acutely induced serum IL-18 levels in CpG-pretreated mice that were 10- and 7-fold greater than LPS-induced serum IL-18 levels in control and LPS-conditioned mice, respectively. Consistent with previous reports of constitutive cytokine release after CpG treatment (24), serum IFN-γ and IL-12 levels in CpG-pretreated mice were elevated before LPS challenge, although LPS challenge resulted in at least a doubling of their concentration. In contrast, IL-18 was entirely dependent on LPS exposure for production in the current studies. We also examined concentrations of these cytokines 3 h after challenge, which more closely corresponds to known peak values of IL-12 and IL-18 during endotoxemia. Serum levels of IL-18 were again >10-fold higher (p < 0.01) in the CpG pretreatment group than in controls or LPS-pretreated mice (15.3 ± 3.6 compared with 0.97 ± 0.22 and 1.46 ± 0.39 ng/ml, respectively). LPS challenge induced circulating IL-12 p70 in both naive controls (1.10 ± 0.15 ng/ml) and CpG-pretreated mice (0.51 ± 0.10 ng/ml; p < 0.01), but this cytokine was undetectable in LPS-pretreated mice.

FIGURE 2.

CpG-pretreated, but not LPS-pretreated, mice generate 10-fold increased serum levels of IL-18 in response to LPS challenge. Groups of five female C3H/HeN mice were conditioned with S. enteriditis LPS or CpG DNA daily for 2 days as described above. Seventy-two hours after the first conditioning dose, mice were either unchallenged (□) or challenged with 300 μg of LPS by i.p. injection (▪). Shown are the mean ± SEM for IFN-γ, IL-12 p70, and IL-18 as determined in 5 h serum by specific ELISA. Asterisks indicate statistical significance (p < 0.01, by Mann-Whitney U test) for cytokine levels, comparing LPS-challenged groups only. N.D., cytokine not detected. Asterisks indicate significant differences at p < 0.01, where exact p value is not provided.

FIGURE 2.

CpG-pretreated, but not LPS-pretreated, mice generate 10-fold increased serum levels of IL-18 in response to LPS challenge. Groups of five female C3H/HeN mice were conditioned with S. enteriditis LPS or CpG DNA daily for 2 days as described above. Seventy-two hours after the first conditioning dose, mice were either unchallenged (□) or challenged with 300 μg of LPS by i.p. injection (▪). Shown are the mean ± SEM for IFN-γ, IL-12 p70, and IL-18 as determined in 5 h serum by specific ELISA. Asterisks indicate statistical significance (p < 0.01, by Mann-Whitney U test) for cytokine levels, comparing LPS-challenged groups only. N.D., cytokine not detected. Asterisks indicate significant differences at p < 0.01, where exact p value is not provided.

Close modal

To prove that the modulatory effects of LPS and CpG on IFN-γ were mediated through the innate cellular immune system and did not involve participation of B or T cells as cytokine sources, we repeated these studies using C57BL/6 SCID mice (Fig. 3). Relative to control SCID mice, LPS- and CpG-conditioned SCID mice generated the same polarized pattern of serum IFN-γ responses as that observed in our studies on wild-type C3H/HeN mice. Specifically, LPS-induced IFN-γ and IL-12 p70 responses that were present in control mice were ablated in endotoxin-tolerant mice. In contrast, IFN-γ and IL-18 levels were increased >5-fold in CpG-pretreated mice. In this study CpG conditioning resulted in an ∼50% increase in serum IL-12 p70 (p = 0.04). This confirmed that CpG pretreatment augments serum IFN-γ production in association with increased IL-18 levels and that these cytokine changes are similarly T and B cell independent.

FIGURE 3.

B and T cells are not required for the immune-potentiating effects of CpG pretreatment on serum IFN-γ and IL-18 responses. Groups of five female C57BL/6 SCID mice were conditioned with LPS or CpG and subsequently challenged with high dose LPS as described in the previous experiments. Sera were obtained at 5 h, and the circulating levels of IFN-γ, IL-12 p70, and IL-18 were determined by specific ELISA. Shown are the mean ± SEM serum cytokine levels. Asterisks demonstrate statistical significance (p < 0.01, by Mann-Whitney U test) for the comparisons indicated by the brackets.

FIGURE 3.

B and T cells are not required for the immune-potentiating effects of CpG pretreatment on serum IFN-γ and IL-18 responses. Groups of five female C57BL/6 SCID mice were conditioned with LPS or CpG and subsequently challenged with high dose LPS as described in the previous experiments. Sera were obtained at 5 h, and the circulating levels of IFN-γ, IL-12 p70, and IL-18 were determined by specific ELISA. Shown are the mean ± SEM serum cytokine levels. Asterisks demonstrate statistical significance (p < 0.01, by Mann-Whitney U test) for the comparisons indicated by the brackets.

Close modal

Spleen cells from control, LPS-conditioned, or CpG-conditioned mice were cultured with rIL-12 and/or rIL-18 (1 and 3 ng/ml, respectively) in concentrations previously determined to be optimal for IFN-γ stimulation by normal spleen cells. Although rIL-12 supplementation alone supported IFN-γ production by normal spleen cells, CpG-conditioned cells produced 5-fold less IFN-γ (Fig. 4). In contrast, the addition of IL-18 markedly increased steady state levels of IFN-γ in all groups, with only insignificant differences apparent between CpG- and LPS-conditioned mice relative to control spleen. Increased concentrations of rIL-12 and rIL-18 or the addition of IL-15 to these cultures did not further enhance IFN-γ production (data not shown). Based on these findings, the increased in vivo production of IFN-γ by CpG-conditioned mice did not simply reflect greater responsiveness of spleen cells to combined stimulation with rIL-12 or rIL-18 in vitro.

FIGURE 4.

Splenocytes from control, LPS-conditioned, and CpG-conditioned mice respond comparably to IL-18 in culture. C3H/HeN mice were pretreated with LPS and CpG as described, and spleens were harvested at 72 h. Cells from spleens were cultured in the medium alone or in medium containing 1 ng/ml rIL-12 or rIL-12 plus 3 ng/ml rIL-18. Conditioned media were harvested 24 h after stimulation, and concentrations of IFN-γ were determined by specific ELISA. Shown are the mean ± SEM for groups of four mice each.

FIGURE 4.

Splenocytes from control, LPS-conditioned, and CpG-conditioned mice respond comparably to IL-18 in culture. C3H/HeN mice were pretreated with LPS and CpG as described, and spleens were harvested at 72 h. Cells from spleens were cultured in the medium alone or in medium containing 1 ng/ml rIL-12 or rIL-12 plus 3 ng/ml rIL-18. Conditioned media were harvested 24 h after stimulation, and concentrations of IFN-γ were determined by specific ELISA. Shown are the mean ± SEM for groups of four mice each.

Close modal

Spleen cells from CpG-conditioned mice, but not those from control or LPS-conditioned mice, produced increased levels of IFN-γ when cultured with LPS or CpG (Fig. 5 A), suggesting that spleen cultures provided a valid model for further analysis of cytokine interactions underlying the CpG-enhanced IFN-γ response. The addition of neutralizing anti-IL-12 Ab and neutralizing anti-IL-15R Ab to culture reduced LPS- or CpG-induced IFN-γ levels by only 20–40%. However, neutralizing anti-IL-18R Ab essentially ablated IFN-γ production. Control polyclonal goat Ab (anti-hemagglutinin) had no effect on IFN-γ production (data not shown). LPS- or CpG-stimulated splenocytes from control and LPS-conditioned mice produced small amounts of IFN-γ that were reduced in the presence of anti-IL-18R.

FIGURE 5.

IL-18 is necessary for CpG-enhanced IFN-γ production in vitro and in vivo. A, IL-18 is necessary for IFN-γ production by CpG-preconditioned spleen cells in response to LPS or CpG. C3H/HeN mice (n = 4) were pretreated with saline (□), LPS (▨), or CpG (▪) as described, and spleens were harvested at 72 h. Pooled cells were cultured in medium alone or in medium containing stimulatory concentrations of LPS (1 μg/ml) or CpG (10 μg/ml). Replicate wells included 10 μg/ml neutralizing monoclonal anti-mouse IL-12 (17.5), polyclonal anti-mouse IL-15R or polyclonal anti-mouse IL-18R Abs. Conditioned media were harvested at 24 h, and concentrations of IFN-γ were determined by specific ELISA. Shown are the mean values of triplicate cultures. Asterisks indicate decreases from control of <2-fold (∗, p < 0.01) or changes of >20-fold (∗∗, p < 0.001). B, Anti-IL-18R Ab inhibits LPS-triggered serum IFN-γ production in CpG-pretreated mice. Groups of five mice were pretreated as described above and then challenged with 300 μg of LPS. Where indicated, one group of mice was also treated with 200 μg of goat anti-IL-18R IgG at the time of LPS challenge. Sera were obtained at 6 h for ELISA determination of IFN-γ levels. The decline in IFN-γ for anti-IL-18R-treated mice was >2-fold and statistically significant (p < 0.01).

FIGURE 5.

IL-18 is necessary for CpG-enhanced IFN-γ production in vitro and in vivo. A, IL-18 is necessary for IFN-γ production by CpG-preconditioned spleen cells in response to LPS or CpG. C3H/HeN mice (n = 4) were pretreated with saline (□), LPS (▨), or CpG (▪) as described, and spleens were harvested at 72 h. Pooled cells were cultured in medium alone or in medium containing stimulatory concentrations of LPS (1 μg/ml) or CpG (10 μg/ml). Replicate wells included 10 μg/ml neutralizing monoclonal anti-mouse IL-12 (17.5), polyclonal anti-mouse IL-15R or polyclonal anti-mouse IL-18R Abs. Conditioned media were harvested at 24 h, and concentrations of IFN-γ were determined by specific ELISA. Shown are the mean values of triplicate cultures. Asterisks indicate decreases from control of <2-fold (∗, p < 0.01) or changes of >20-fold (∗∗, p < 0.001). B, Anti-IL-18R Ab inhibits LPS-triggered serum IFN-γ production in CpG-pretreated mice. Groups of five mice were pretreated as described above and then challenged with 300 μg of LPS. Where indicated, one group of mice was also treated with 200 μg of goat anti-IL-18R IgG at the time of LPS challenge. Sera were obtained at 6 h for ELISA determination of IFN-γ levels. The decline in IFN-γ for anti-IL-18R-treated mice was >2-fold and statistically significant (p < 0.01).

Close modal

Compared with control C3H/HeN mice, CpG-pretreated animals increased circulating IFN-γ by >10-fold at 6 h after LPS challenge (Fig. 5 B), but coinjection with neutralizing anti-IL-18R Ab decreased the peak serum IFN-γ levels by more than half (p < 0.01). The suppressive effect of goat anti-IL-18R IgG was reproduced in another experiment in which control goat IgG failed to reduce levels of LPS-induced IFN-γ in CpG-conditioned mice (39.6 ± 2.6, 1.1 ± 1.0, and 38.6 ± 3.3 ng/ml for no Ab, anti-IL-18R, and goat IgG, respectively). These data, obtained in vitro and in vivo, are consistent with a model of CpG-enhanced IFN-γ production in which LPS-induced IL-18 synthesis is markedly increased and mechanistically critical for subsequent IFN-γ synthesis.

In response to either CpG or LPS, DX5+ NK cells developed an activated phenotype consisting of increased surface expression of CD69 and IL-18R compared with control NK cells. The total number of NK cells per spleen was unchanged for CpG-conditioned mice and was increased 2-fold for LPS-pretreated mice (Table I). CpG-conditioned NK cells expressed IL-18R levels twice those in normal control mice (mean fluorescence intensity, 190 vs 81, respectively) and 50% greater than those in LPS-conditioned mice (mean fluorescence intensity, 124). Otherwise, all groups expressed similar levels of CD122 (IL-2R/IL-15Rβ subunit) and IL-12Rβ1. Surface CD25 and IL-15R were not detected (data not shown). DX5+ NK cells were also shown to be the source of IFN-γ in both control and CpG-conditioned SCID mice after LPS challenge, as determined by intracellular staining for IFN-γ (Fig. 6). DX5+ cells were also the only source of LPS-induced IFN-γ in the spleens of wild-type C3H/HeN mice after CpG conditioning (data not shown).

Table I.

Increased surface expression of CD69 and IL-18R on DX5+ NK cells 3 days after initial pretreatment with LPS or CpG

GroupMFIaAverage Number of NK Cells per Spleenb
CD69IL-18R
Control 81 3.2 × 106 
LPS pretreated 37 124 6.7 × 106 
CpG pretreated 93 190 4.0 × 106 
GroupMFIaAverage Number of NK Cells per Spleenb
CD69IL-18R
Control 81 3.2 × 106 
LPS pretreated 37 124 6.7 × 106 
CpG pretreated 93 190 4.0 × 106 
a

Cells were stained with PE-anti-CD69 or PE-anti-IL-18R. MFI, mean fluorescence intensity for FL2 channel gated on APC-DX5+ and FITC-CD3/CD19 cells.

b

Mean number of cells (percentage of DX5+ times total number of cells, n = 4 spleens pooled): data are representative of three experiments.

FIGURE 6.

DX5+ NK cells are the source of LPS-induced IFN-γ in control and CpG-pretreated mice. Groups of C57BL/6 SCID mice (n = 4 each) were pretreated with LPS or CpG as described above. Seventy-two hours after the first preconditioning dose, mice were challenged with high dose LPS. Spleens were harvested 4 h later and cultured for another 4 h in medium containing brefeldin A, rIL-12 (1 ng/ml), and rIL-18 (3 ng/ml). Cells were washed and labeled with allophycocyanin-DX5 anti-mouse NK Ab (rat IgM). Cells were then fixed and permeabilized using a commercial kit (BD PharMingen) and incubated with 0.05 μg/ml PE-anti-mouse IFN-γ or PE-labeled isotype control IgG. A, Two representative color dot plots demonstrating DX5+ NK cells (x-axis) with intracellular staining of IFN-γ (y-axis) in CpG-conditioned DX5+ cells relative to LPS-preconditioned cells. CpG-preconditioned cells were also stained with a PE-labeled rat IgG2a as an intracellular isotype control to confirm the specificity of IFN-γ staining. B, Comparison of fluorescence histograms of intracellular IFN-γ staining within gated DX5+ NK cell populations from the designated group of mice and including one isotype Ab control. Similar findings were obtained using wild-type C3H/HeN mice, without evidence of IFN-γ production by non-DX5+ cell populations.

FIGURE 6.

DX5+ NK cells are the source of LPS-induced IFN-γ in control and CpG-pretreated mice. Groups of C57BL/6 SCID mice (n = 4 each) were pretreated with LPS or CpG as described above. Seventy-two hours after the first preconditioning dose, mice were challenged with high dose LPS. Spleens were harvested 4 h later and cultured for another 4 h in medium containing brefeldin A, rIL-12 (1 ng/ml), and rIL-18 (3 ng/ml). Cells were washed and labeled with allophycocyanin-DX5 anti-mouse NK Ab (rat IgM). Cells were then fixed and permeabilized using a commercial kit (BD PharMingen) and incubated with 0.05 μg/ml PE-anti-mouse IFN-γ or PE-labeled isotype control IgG. A, Two representative color dot plots demonstrating DX5+ NK cells (x-axis) with intracellular staining of IFN-γ (y-axis) in CpG-conditioned DX5+ cells relative to LPS-preconditioned cells. CpG-preconditioned cells were also stained with a PE-labeled rat IgG2a as an intracellular isotype control to confirm the specificity of IFN-γ staining. B, Comparison of fluorescence histograms of intracellular IFN-γ staining within gated DX5+ NK cell populations from the designated group of mice and including one isotype Ab control. Similar findings were obtained using wild-type C3H/HeN mice, without evidence of IFN-γ production by non-DX5+ cell populations.

Close modal

We tested how the opposed modulatory effects of CpG and LPS were functionally integrated when both were used as pretreatments. Mice were injected i.p. on 2 consecutive days with low dose CpG and LPS, either alone or combined, followed by challenge with high dose LPS (Fig. 7). For mice conditioned with LPS or CpG alone, we observed the same pattern of LPS-induced tolerance and CpG-induced potentiation of IFN-γ as that previously shown in these studies. However, mice pretreated with both LPS and CpG developed an endotoxin-tolerant phenotype, with suppression of LPS-inducible IL-12 p70 and IFN-γ levels to nearly undetectable levels. Similarly, the expansion of LPS-inducible IL-18 production in CpG-pretreated mice was suppressed >10-fold when LPS was added to the conditioning regimen. Comparable findings were obtained in a separate study of C57BL/6 SCID mice (data not shown).

FIGURE 7.

Combined pretreatment with both LPS and CpG suppresses the potentiating effect of CpG on IFN-γ and IL-18 serum responses to subsequent LPS challenge. As designated on the x-axis, groups of five female C3H/HeN mice were untreated (None) or were pretreated with S. enteriditis LPS (LPS), CpG-containing S-ODN (CpG), or both LPS and CpG administered by i.p. injection (LPS/CpG) given daily for 2 days. Doses were as described above, except that the amounts of LPS and CpG given in combination were halved in the second injection to 25 μg each. Seventy-two hours after the first conditioning dose, all mice were challenged with 300 μg of LPS by i.p. injection. Shown are the mean ± SEM for the indicated cytokines in serum obtained 6 h postchallenge. Asterisks indicate significant differences between samples indicated by the brackets (p < 0.01, by Mann-Whitney U test).

FIGURE 7.

Combined pretreatment with both LPS and CpG suppresses the potentiating effect of CpG on IFN-γ and IL-18 serum responses to subsequent LPS challenge. As designated on the x-axis, groups of five female C3H/HeN mice were untreated (None) or were pretreated with S. enteriditis LPS (LPS), CpG-containing S-ODN (CpG), or both LPS and CpG administered by i.p. injection (LPS/CpG) given daily for 2 days. Doses were as described above, except that the amounts of LPS and CpG given in combination were halved in the second injection to 25 μg each. Seventy-two hours after the first conditioning dose, all mice were challenged with 300 μg of LPS by i.p. injection. Shown are the mean ± SEM for the indicated cytokines in serum obtained 6 h postchallenge. Asterisks indicate significant differences between samples indicated by the brackets (p < 0.01, by Mann-Whitney U test).

Close modal

Systemic endotoxin has been shown to deplete dendritic cells in multiple tissues, including spleen, through apoptotic mechanisms (32, 33, 34). We confirmed that the LPS pretreatment protocol used in these studies also depleted by 3-fold the percentage of CD11c+CD11b+ dendritic cells present in C3H/HeN spleen 3 days later (Table II). Similar findings were obtained when using CD11c in combination with either CD40 or MHC II as alternative DC surface phenotypes. Pretreatment with CpG alone failed to deplete dendritic cells, but the combination of LPS and CpG pretreatment again resulted in 3-fold reductions in this cell population. This indicates that different bacterial PAMPs have distinct effects on dendritic cell survival in vivo, possibly due to opposed regulatory effects mediated through their respective TLRs. Consistent with TLR4-dependent suppression of innate cellular immunity, endotoxin-unresponsive C3H/HeJ mice pretreated with LPS failed to show FACS evidence of DC depletion and generated normal increases in serum IL-12 p70 after systemic challenge with CpG 3 days later (data not shown). Because dendritic cells are critical sources of IL-12 p70 and are known to also generate IL-18, the dominant negative effects of LPS on dendritic cell survival may be mechanistically relevant to the dominant suppression of innate cellular immunity induced by LPS in these studies.

Table II.

Systemic injection with LPS, but not CpG, depletes splenic dendritic cells (n = 4 mice/group).

Conditioning RegimenDendritic Cells (CD11b+/CD11c+)a% Positive for CD11c and
% PositivebTotal numbers in spleen (×106)MHCIICD40
Control 3.05 ± 0.4% 2.4 ± 0.36 4.51 4.42 
LPS pretreated 1.01 ± 0.05% (p = 0.012) 1.27 ± 0.06 1.35 1.30 
CpG pretreated 3.04 ± 0.19% (p = 0.99) 2.53 ± 0.01 3.11 3.81 
LPS/CpG pretreated 1.09 ± 0.23% (p = 0.01) 1.27 ± 0.29 1.53 1.18 
Conditioning RegimenDendritic Cells (CD11b+/CD11c+)a% Positive for CD11c and
% PositivebTotal numbers in spleen (×106)MHCIICD40
Control 3.05 ± 0.4% 2.4 ± 0.36 4.51 4.42 
LPS pretreated 1.01 ± 0.05% (p = 0.012) 1.27 ± 0.06 1.35 1.30 
CpG pretreated 3.04 ± 0.19% (p = 0.99) 2.53 ± 0.01 3.11 3.81 
LPS/CpG pretreated 1.09 ± 0.23% (p = 0.01) 1.27 ± 0.29 1.53 1.18 
a

Defined by expression of CD11c, and high or intermediate levels of CD11b with exclusion of CD3/CD19/sIgM-positive cells.

b

Gated to exclude FITC-CD3, FITC-CD19, and FITC-mouse IgM-positive cells from analysis. The p values were determined by two sample t test and are vs the control value.

Because most microbial pathogens are sources of diverse PAMPs, infection probably activates multiple TLRs early in the course of disease. Although most of these receptors are well characterized for their acute contributions to proinflammatory responses, our findings indicate that the longer term effects of specific TLR activation will polarize cytokine responses to subsequent challenges. As delayed effects on innate immunity may critically determine the ability of the host to resist infection or tissue damage, it is important to understand the manner in which opposing modulatory TLR signals functionally integrate, either at the level of a single cell or at the level of the whole organism. Among our central findings, we show that repeated treatments with well-characterized TLR4 or TLR9 agonists respectively suppress or enhance IFN-γ synthetic capacity within the innate cellular immune response. Furthermore, LPS-induced endotoxin tolerance was exhibited against both a homologous challenge with LPS and a heterologous challenge with CpG. Endotoxin-induced cross-tolerance for cytokine induction by TLR2 agonists has been previously reported (19, 35). Our data support an even broader inhibition of TLR function in endotoxin-tolerant mice. In contrast, CpG pretreatment had the opposite effect of cross-potentiating IFN-γ production in response to either LPS or CpG challenges. However, pretreatment with both LPS and CpG suppressed cytokine synthesis. This demonstrates that the presence of a dominant negative pattern of integration between two TLRs with opposing modulatory functions. We speculate that the example reported in this study may predict similar patterns of dominant signal integration for other combinations of PAMP receptor.

In these studies we used IFN-γ and IL-12 p70 serum levels as outcomes for assessing the delayed effects of TLR activation on innate cellular immunity function. IFN-γ is physiologically important due to its ability to recruit antimicrobial host defenses, but its production also signifies the integrity of specific cellular and cytokine responses within the innate cellular immunity system. Production of IL-12 p70 is normally essential for IFN-γ production during endotoxemia, and we confirmed that the IFN-γ deficiency of endotoxin-tolerant mice closely associates with decreased serum IL-12 after LPS challenge (31, 33). We also newly observed the presence of cross-tolerance to the IL-12 p70-inducing effects of CpG. In contrast, CpG pretreatment strikingly enhanced serum IFN-γ levels in response to either LPS or CpG challenge (by 10-fold) without corresponding increases in serum IL-12 p70. The disproportion existing between decreased IL-12 p70 levels and an average 6-fold increase in IFN-γ suggested that other factors were acting on NK cells in CpG-preconditioned mice.

In consequence, we observed a marked increase in LPS-inducible serum IL-18 in CpG-pretreated mice. The levels of IL-18 observed were 10-fold increased compared with those in control or endotoxin-tolerant mice and were essentially proportionate to the increase in IFN-γ synthetic capacity in these mice. There is a precedent for increased IL-18 production in response to infectious and/or inflammatory stimuli. For instance, prior infection with Propionobacterium acnes leads to marked increases in serum IL-18 levels after subsequent LPS challenge (36). Sublethal endotoxemia also triggers prolonged increases in IL-18 production, accounting for the low levels of constitutive IL-18 we detected in endotoxin-tolerant mice before LPS challenge (37). However, our current observations specifically identify CpG as a more potent and better defined regulator of IL-18 synthetic capacity in vivo. As such, this might represent a model for identifying potentially novel mechanisms for regulation of the IL-18 response during infection. The synthesis of bioactive IL-18 release is complexly regulated at both transcriptional and translational steps. The release of bioactive IL-18 requires further processing of pro-IL-18 by caspase-1, which is activated by LPS in vivo and is necessary for IFN-γ production during endotoxemia (37, 38). We speculate that CpG recruits new cellular sources of pro-IL-18 synthesis that then require an LPS trigger for processing and release of mature cytokine.

The presence of increased serum IL-18 in CpG-pretreated mice suggests a mechanism accounting for the enhanced IFN-γ production. Although IL-18 supports IFN-γ synthesis from both T and NK cells, IL-18R is constitutively expressed only on NK cells (39). We confirmed that NK cells were the source of IFN-γ in these studies. Specifically, the same patterns of CpG- and LPS-manipulated serum IFN-γ responses were reproduced in SCID mice. Furthermore, NK cells stained strongly for intracellular IFN-γ after stimulation of control and CpG-pretreated mice or their splenocytes. Although CpG-pretreated NK cells expressed more IL-18R than did control or LPS-pretreated cells, the increase in IFN-γ synthetic response attributable to rIL-18 in culture was not limited by levels of IL-18R expression. However, IFN-γ synthesis in splenic culture was entirely blocked by anti-IL-18R Ab and was only modestly affected by anti-IL-12 Abs. We also showed that anti-IL-18R Ab significantly reduced LPS-induced production of serum IFN-γ in CpG-pretreated mice. From these findings we conclude that the increased availability of IL-18 mediates expanded IFN-γ production in CpG-pretreated animals. On the average, IL-12 p70 levels were preserved in these studies and presumably contribute equally to the distinct IFN-γ responses of control and CpG-pretreated mice.

In these same studies, LPS and CpG mediated opposing effects on dendritic cell survival in the spleen. Endotoxemia triggers dendritic cell depletion from multiple tissues through apoptotic mechanisms and within hours of exposure (32, 34, 40), which we confirmed using FACS analysis of spleen cells bearing classic DC phenotypes (41). In contrast, no depletion of DC was evident after CpG pretreatment. As dendritic cells are well characterized sources of IL-12 production in response to either LPS or CpG, their reduced number in endotoxin-tolerant mice provides a reasonable explanation for IL-12 deficiency in this setting (6, 42). It is also possible that dendritic cells were a CpG-recruitable source of IL-18 synthesis in these studies (43). However, we cannot exclude important contributions by other cells capable of IL-12 and IL-18 synthesis during endotoxemia. In this regard, a second potential mechanism for LPS-induced cytokine deficiency might be mediated through IL-IR-associated kinase M (IRAK-M)-dependent desensitization of macrophages to TLR signaling (44).

A related and important finding in these studies is that simultaneous administration of TLR4 and TLR9 agonists strikingly deviates the innate cellular immune response toward an endotoxin-tolerant phenotype. This included marked decreases in LPS-inducible IFN-γ, IL-12 p70, and IL-18 production as well as 3-fold depletions in dendritic cells that were not observed after CpG pretreatment alone. Although TLR9 activation can actively protect dendritic cells from apoptotic death in vivo (45), the cytokine-suppressive and dendritic cell-depleting effects of LPS clearly predominated over the potentiating effects of CpG. This again suggests a possible mechanistic connection between dendritic cell survival and the amounts of IL-12 and IL-18 produced upon LPS challenge. Otherwise, combined LPS and CpG did not significantly impair NK cell function, as spleen cells from coconditioned mice were as responsive to IL-12 and IL-18 in culture as cells from mice conditioned with LPS or CpG alone (data not shown). Further studies may determine whether the suppressive effects occur by cross-talk between TLR4 and TLR9 expressed on the same cell (cis-interacting) or reflect communication between cells with distinct TLR repertoires (trans-interacting).

In summary, these findings show that LPS and CpG exposures mediate delayed and opposing effects on the cytokine synthetic capacity of innate cellular immunity in response to repeated TLR stimulation. The suppressive effects of LPS were dominant, in that endotoxin-tolerant mice were unresponsive to challenge with either CpG or LPS and because pretreatment with both LPS and CpG gave only an endotoxin-tolerant phenotype. Because CpG has been proposed as immunotherapy to restore innate immune function in the setting of injury or sepsis (25), these findings are clinically relevant in that they predict that such therapy will be ineffective if endotoxemia is present. This may also indicate a similar lack of efficacy in the setting of clinical immune paralysis, a syndrome that is phenotypically and mechanistically similar to endotoxin tolerance (46, 47). We do not know whether findings obtained in comparisons of LPS and CpG can be extended to other TLR-specific agonists, although repeated activation of TLR2 by bacterial lipopeptides or lipotechoic acid induces a cytokine hyporesponsive state similar to that of endotoxin tolerance (19). If so, the immune-enhancing effects of CpG may be dominantly suppressed by coexposures to either TLR2- or TLR4-reactive PAMPs. Ultimately, these findings suggest that the net effect of systemic bacterial infection will be biased towards suppression of innate immunity. Although multiple Toll agonists with diverse modulatory effects are released in this setting, activation of TLR4 (and/or TLR2)-dependent modulatory effects will result in an overall suppression of the IFN-γ synthetic capacity, a speculation that agrees well with the clinical immunology observed in survivors of sepsis (48).

We thank Eric Pearlman and Christopher King for critical reading of the manuscript. We also gratefully acknowledge Gopal Yadavalli, Jeffrey Auletta, and Lopamudra Das for valuable discussions.

1

This work was supported by National Institute of Allergy and Infectious Disease Grants AI45602 and AI35979 and by Merit Review funding from the Veterans Affairs Medical Research Center.

3

Abbreviations used in this paper: TLR, Toll-like receptor; PAMP, pathogen-associated molecular pattern; S-ODN, phosphorothiorate oligodeoxynucleotide.

1
Janeway, C. A., Jr, R. Medzhitov.
2002
. Innate immune recognition.
Annu. Rev. Immunol.
20
:
197
.
2
Takeda, K., T. Kaisho, S. Akira.
2003
. Toll-like receptors.
Annu. Rev. Immunol.
21
:
335
.
3
Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. Van Huffel, X. Du, D. Birdwell, E. Alejos, M. Silva, C. Galanos, et al  
1998
. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene.
Science
282
:
2085
.
4
Huang, Q., D. Liu, P. Majewski, L. C. Schulte, J. M. Korn, R. A. Young, E. S. Lander, N. Hacohen.
2001
. The plasticity of dendritic cell responses to pathogens and their components.
Science
294
:
870
.
5
Lien, E., T. K. Means, H. Heine, A. Yoshimura, S. Kusumoto, K. Fukase, M. J. Fenton, M. Oikawa, N. Qureshi, B. Monks, et al  
2000
. Toll-like receptor 4 imparts ligand-specific recognition of bacterial lipopolysaccharide.
J. Clin. Invest.
105
:
497
.
6
Heufler, C., F. Koch, U. Stanzl, G. Topar, M. Wysocka, G. Trinchieri, A. Enk, R. M. Steinman, N. Romani, G. Schuler.
1996
. Interleukin-12 is produced by dendritic cells and mediates T helper 1 development as well as interferon-gamma production by T helper 1 cells.
Eur. J. Immunol.
26
:
659
.
7
Trinchieri, G..
1995
. Interleukin-12: a proinflammatory cytokine with immunoregulatory functions that bridge innate resistance and antigen-specific adaptive immunity.
Annu. Rev. Immunol.
13
:
251
.
8
Boehm, U., T. Klamp, M. Groot, J. C. Howard.
1997
. Cellular responses to interferon-γ.
Annu. Rev. Immunol.
15
:
749
.
9
Way, S. S., A. C. Borczuk, R. Dominitz, M. B. Goldberg.
1998
. An essential role for gamma interferon in innate resistance to Shigella flexneri infection.
Infect. Immun.
66
:
1342
.
10
Zantl, N., A. Uebe, B. Neumann, H. Wagner, J. R. Siewert, B. Holzmann, C. D. Heidecke, K. Pfeffer.
1998
. Essential role of γ interferon in survival of colon ascendens stent peritonitis, a novel murine model of abdominal sepsis.
Infect. Immun.
66
:
2300
.
11
Rubins, J. B., C. Pomeroy.
1997
. Role of γ interferon in the pathogenesis of bacteremic pneumococcal pneumonia.
Infect. Immun.
65
:
2975
.
12
Orange, J. S., B. Wang, C. Terhorst, C. A. Biron.
1995
. Requirement for natural killer cell-produced interferon γ in defense against murine cytomegalovirus infection and enhancement of this defense pathway by interleukin 12 administration.
J. Exp. Med.
182
:
1045
.
13
Takeuchi, O., K. Hoshino, S. Akira.
2000
. Cutting edge: TLR2-deficient and MyD88-deficient mice are highly susceptible to Staphylococcus aureus infection.
J. Immunol.
165
:
5392
.
14
Schnare, M., G. M. Barton, A. C. Holt, K. Takeda, S. Akira, R. Medzhitov.
2001
. Toll-like receptors control activation of adaptive immune responses.
Nat. Immunol.
2
:
947
.
15
Heinzel, F. P..
1990
. The role of IFN-γ in the pathology of experimental endotoxemia.
J. Immunol.
145
:
2920
.
16
Granowitz, E. V., R. Porat, J. W. Mier, S. F. Orencole, G. Kaplanski, E. A. Lynch, K. Ye, E. Vannier, S. M. Wolff, C. A. Dinarello.
1993
. Intravenous endotoxin suppresses the cytokine response of peripheral blood mononuclear cells of healthy humans.
J. Immunol.
151
:
1637
.
17
Beeson, P. B..
1947
. Tolerance to bacterial pyrogens. I. Factors influencing its development.
J. Exp. Med.
86
:
29
.
18
Balkhy, H. H., F. P. Heinzel.
1999
. Endotoxin fails to induce IFN-γ in endotoxin-tolerant mice: deficiencies in both IL-12 heterodimer production and IL-12 responsiveness.
J. Immunol.
162
:
3633
.
19
Wang, J. H., M. Doyle, B. J. Manning, S. Blankson, Q. D. Wu, C. Power, R. Cahill, H. P. Redmond.
2003
. Cutting edge: bacterial lipoprotein induces endotoxin-independent tolerance to septic shock.
J. Immunol.
170
:
14
.
20
Weijer, S., F. N. Lauw, J. Branger, B. van den Blink, T. van der Poll.
2002
. Diminished interferon-γ production and responsiveness after endotoxin administration to healthy humans.
J. Infect. Dis.
186
:
1748
.
21
Ertel, W., M. Keel, R. Neidhardt, U. Steckholzer, J. P. Kremer, U. Ungethuem, O. Trentz.
1997
. Inhibition of the defense system stimulating interleukin-12 interferon-γ pathway during critical Illness.
Blood
89
:
1612
.
22
Wagner, H..
2001
. Toll meets bacterial CpG-DNA.
Immunity
14
:
499
.
23
Cowdery, J. S., J. H. Chace, A. K. Yi, A. M. Krieg.
1996
. Bacterial DNA induces NK cells to produce IFN-γ in vivo and increases the toxicity of lipopolysaccharides.
J. Immunol.
156
:
4570
.
24
Krieg, A. M., L. Love-Homan, A. K. Yi, J. T. Harty.
1998
. CpG DNA induces sustained IL-12 expression in vivo and resistance to Listeria monocytogenes challenge.
J. Immunol.
161
:
2428
.
25
Weighardt, H., C. Feterowski, M. Veit, M. Rump, H. Wagner, B. Holzmann.
2000
. Increased resistance against acute polymicrobial sepsis in mice challenged with immunostimulatory CpG oligodeoxynucleotides is related to an enhanced innate effector cell response.
J. Immunol.
165
:
4537
.
26
Zimmermann, S., O. Egeter, S. Hausmann, G. B. Lipford, M. Rocken, H. Wagner, K. Heeg.
1998
. CpG oligodeoxynucleotides trigger protective and curative Th1 responses in lethal murine leishmaniasis.
J. Immunol.
160
:
3627
.
27
Rhee, E. G., S. Mendez, J. A. Shah, C. Y. Wu, J. R. Kirman, T. N. Turon, D. F. Davey, H. Davis, D. M. Klinman, R. N. Coler, et al  
2002
. Vaccination with heat-killed Leishmania antigen or recombinant leishmanial protein and CpG oligodeoxynucleotides induces long-term memory CD4+ and CD8+ T cell responses and protection against Leishmania major infection.
J. Exp. Med.
195
:
1565
.
28
Dalpke, A., S. Zimmermann, K. Heeg.
2002
. CpG DNA in the prevention and treatment of infections.
BioDrugs
16
:
419
.
29
Yi, A. K., R. Tuetken, T. Redford, M. Waldschmidt, J. Kirsch, A. M. Krieg.
1998
. CpG motifs in bacterial DNA activate leukocytes through the pH-dependent generation of reactive oxygen species.
J. Immunol.
160
:
4755
.
30
Heinzel, F. P., R. M. Rerko, F. Ahmed, A. M. Hujer.
1996
. IFN-γ-independent production of IL-12 during murine endotoxemia.
J. Immunol.
157
:
4521
.
31
Wysocka, M., S. Robertson, H. Riemann, J. Caamano, C. Hunter, A. Mackiewicz, L. J. Montaner, G. Trinchieri, C. L. Karp.
2001
. IL-12 suppression during experimental endotoxin tolerance: dendritic cell loss and macrophage hyporesponsiveness.
J. Immunol.
166
:
7504
.
32
Roake, J. A., A. S. Rao, P. J. Morris, C. P. Larsen, D. F. Hankins, J. M. Austyn.
1995
. Dendritic cell loss from nonlymphoid tissues after systemic administration of lipopolysaccharide, tumor necrosis factor, and interleukin 1.
J. Exp. Med.
181
:
2237
.
33
Reis e Sousa, C., S. Hieny, T. Scharton-Kersten, D. Jankovic, H. Charest, R. N. Germain, A. Sher.
1997
. In vivo microbial stimulation induces rapid CD40 ligand-independent production of interleukin 12 by dendritic cells and their redistribution to T cell areas.
J. Exp. Med.
186
:
1819
.
34
De Smedt, T., B. Pajak, G. G. Klaus, R. J. Noelle, J. Urbain, O. Leo, M. Moser.
1998
. Antigen-specific T lymphocytes regulate lipopolysaccharide-induced apoptosis of dendritic cells in vivo.
J. Immunol.
161
:
4476
.
35
Lehner, M. D., S. Morath, K. S. Michelsen, R. R. Schumann, T. Hartung.
2001
. Induction of cross-tolerance by lipopolysaccharide and highly purified lipoteichoic acid via different Toll-like receptors independent of paracrine mediators.
J. Immunol.
166
:
5161
.
36
Okamura, H., H. Tsutsi, T. Komatsu, M. Yutsudo, A. Hakura, T. Tanimoto, K. Torigoe, T. Okura, Y. Nukada, K. Hattori, et al  
1995
. Cloning of a new cytokine that induces IFN-γ production by T cells.
Nature
378
:
88
.
37
Joshi, V. D., D. V. Kalvakolanu, J. D. Hasday, R. J. Hebel, A. S. Cross.
2002
. IL-18 levels and the outcome of innate immune response to lipopolysaccharide: importance of a positive feedback loop with caspase-1 in IL-18 expression.
J. Immunol.
169
:
2536
.
38
Nakanishi, K., T. Yoshimoto, H. Tsutsui, H. Okamura.
2001
. Interleukin-18 regulates both Th1 and Th2 responses.
Annu. Rev. Immunol.
19
:
423
.
39
Netea, M. G., G. Fantuzzi, B. J. Kullberg, R. J. Stuyt, E. J. Pulido, R. C. McIntyre, Jr, L. A. Joosten, J. W. Van der Meer, C. A. Dinarello.
2000
. Neutralization of IL-18 reduces neutrophil tissue accumulation and protects mice against lethal Escherichia coli and Salmonella typhimurium endotoxemia.
J. Immunol.
164
:
2644
.
40
Reis e Sousa, C., R. N. Germain.
1999
. Analysis of adjuvant function by direct visualization of antigen presentation in vivo: endotoxin promotes accumulation of antigen-bearing dendritic cells in the T cell areas of lymphoid tissue.
J. Immunol.
162
:
6552
.
41
Banchereau, J., R. M. Steinman.
1998
. Dendritic cells and the control of immunity.
Nature
392
:
245
.
42
Jakob, T., P. S. Walker, A. M. Krieg, M. C. Udey, J. C. Vogel.
1998
. Activation of cutaneous dendritic cells by CpG-containing oligodeoxynucleotides: a role for dendritic cells in the augmentation of Th1 responses by immunostimulatory DNA.
J. Immunol.
161
:
3042
.
43
Stoll, S., H. Jonuleit, E. Schmitt, G. Muller, H. Yamauchi, M. Kurimoto, J. Knop, A. H. Enk.
1998
. Production of functional IL-18 by different subtypes of murine and human dendritic cells (DC): DC-derived IL-18 enhances IL-12-dependent Th1 development.
Eur. J. Immunol.
28
:
3231
.
44
Kobayashi, K., L. D. Hernandez, J. E. Galan, C. A. Janeway, Jr, R. Medzhitov, R. A. Flavell.
2002
. IRAK-M is a negative regulator of Toll-like receptor signaling.
Cell
110
:
191
.
45
Park, Y., S. W. Lee, Y. C. Sung.
2002
. Cutting edge: CpG DNA inhibits dendritic cell apoptosis by up-regulating cellular inhibitor of apoptosis proteins through the phosphatidylinositide-3′-OH kinase pathway.
J. Immunol.
168
:
5
.
46
Ohashi, K., V. Burkart, S. Flohe, H. Kolb.
2000
. Cutting edge: heat shock protein 60 is a putative endogenous ligand of the toll-like receptor-4 complex.
J. Immunol.
164
:
558
.
47
Dybdahl, B., A. Wahba, E. Lien, T. H. Flo, A. Waage, N. Qureshi, O. F. Sellevold, T. Espevik, A. Sundan.
2002
. Inflammatory response after open heart surgery: release of heat-shock protein 70 and signaling through Toll-like receptor-4.
Circulation
105
:
685
.
48
Weighardt, H., C. D. Heidecke, K. Emmanuilidis, S. Maier, H. Bartels, J. R. Siewert, B. Holzmann.
2000
. Sepsis after major visceral surgery is associated with sustained and interferon-γ-resistant defects of monocyte cytokine production.
Surgery
127
:
309
.