Abstract
We investigated possible feedback mechanisms of febrile temperatures on LPS- and staphylococcal enterotoxin B (SEB)-induced cytokine release in human whole blood. LPS-induced IL-1β release was inhibited at temperatures >38°C, whereas intracellular proIL-1β formation as well as the release of other cytokines except IL-18 were only attenuated above 42°C, indicating that febrile temperatures impair the proteolytic processing of proIL-1β. This attenuated processing is not due to either heat inactivation of caspase-1 or structural changes in proIL-1β produced at higher temperatures. Instead, we propose that febrile conditions change cytosolic compartmentation or trafficking, so that synthesized proIL-1β cannot encounter caspase-1. Febrile temperatures also influenced Th1/Th2 cytokine balance. We observed a 3-fold increase in the Th2-cytokines IL-5 and IL-13 and a reduction to 15% of the Th1-cytokine IL-2 when SEB-stimulated whole blood was incubated at 40°C compared with 37°C. These results indicate that fever limits the production of the fever-inducing IL-1β and also influences the adaptive immune response, favoring Th2 cytokine production.
Fever is a common response to infection and is induced in response to inflammatory stimuli, such as invading pathogens or injected LPS. The pathogenesis of fever is not understood in detail, but it is known that the mediators IL-1β, TNF-α, and IL-6 play a crucial role (1, 2, 3). In response to the inflammatory stimuli, peripheral leukocytes and microglia produce these endogenous pyrogens, which induce the formation of cyclooxygenase-2 in cerebral blood vessels (4, 5). The cyclooxygenase-2 product PGE2 crosses the blood-brain barrier and causes an elevation of body temperature in the hypothalamus. In the study presented we investigated whether febrile temperatures influence the LPS- or staphylococcal enterotoxin B (SEB)2-stimulated cytokine release in human whole blood. To identify a possible feedback loop, we incubated LPS-stimulated whole blood at different temperatures ranging from 37–42°C and measured the release of IL-1β, IL-18, IFN-γ, IL-6, and TNF-α and the intracellular levels of proIL-1β. Additionally, we investigated whether febrile temperatures influence the production of Th1 and Th2 cytokines in SEB-stimulated whole blood.
Materials and Methods
Whole blood incubations
To study the LPS-induced cytokine release of whole blood, 800 μl of saline solution was pipetted into a polypropylene reaction tube (Eppendorf, Hamburg, Germany) and 10 μg of endotoxin from Salmonella abortus equi (Sigma-Aldrich, Deisenhofen, Germany) was added. In some experiments, 50 μM of the caspase-1-inhibitor YVAD-cmk (Alexis, Grunberg, Germany) was also added. Finally, 200 μl of heparinized human whole blood (withdrawn in Lithium-Heparin-S-Monovette (Sarstedt, Numbrecht, Germany)) was added, and the tubes were incubated in heat blocks (Eppendorf) at the indicated temperatures. After 3- or 24-h incubation, the tubes were shaken, and blood cells were sedimented by centrifugation (16,000 × g, 2 min). The cell-free supernatants were removed and stored at −80°C until cytokine measurement.
For the measurement of proIL-1β in cell lysates, this procedure was modified. Blood was incubated for 3 h, then the tubes were shaken, and blood cells were sedimented by centrifugation (2,000 × g, 2 min). Supernatant (500 μl) was removed and stored for the measurement of mature IL-1β. To the remaining cell pellet, 1 ml of Aqua ad iniectabilia (Braun, Melsungen, Germany) was added, and the cells were lysed by freezing at −80°C. Before cytokine measurement, the thawed samples were centrifuged at 16,000 × g for 2 min.
To study SEB-induced cytokine release of whole blood (6) at febrile temperatures, 800 μl of RPMI 1640 (BioWhittaker, Verviers, Belgium) supplemented with 2.5 IU/ml heparin (Liquemin; Hoffmann La Roche, Grenzach-Whylen, Germany), 100 IU/ml penicillin/streptomycin (Biochrom, Berlin, Germany), and 1 μg Staphylococcus aureus enterotoxin B (Sigma-Aldrich) were mixed with 200 μl of heparinized human whole blood and incubated at 37 or 40°C for 72 h. Then the cell-free supernatants were removed and stored at −80°C until cytokine measurement. In some experiments rIL-12 and rIL-18 (both from Tebu, Offenbach, Germany) were added to the samples to reach a final concentration of 10 ng/ml each.
Cytokine measurement
7).
ProIL-1β was measured using a proIL-1β ELISA kit (Bender MedSystems) that showed no cross-reactivity when we tested it with mature IL-1β (data not shown). IL-18 was determined using an IL-18 ELISA kit (Beckman Coulter, Krefeld, Germany) according to the manufacturer’s instructions.
Isolation of PBMC
PBMC were prepared in cell preparation tubes (Vacutainer CPT, sodium citrate; BD Biosciences, Heidelberg, Germany) according to the manufacturer’s instructions. After centrifugation (20 min, 1650 × g), the white layer above the gel containing the PBMC was removed, and the cells were washed three times with RPMI 1640 containing 2.5 IU/ml heparin.
Measurement of caspase-1 activity in monocytes
To determine caspase-1 activity in intact monocytes, a method described by Hug et al. (8) was adapted. PBMC were stimulated with 10 μg/ml LPS at 37 or 40°C. As a control, PBMC were also incubated with LPS and 100 μM YVAD-cmk (Alexis) at 37°C. After 4-h incubation 50 μM D2R (Alexis) was added and incubated for another 90 min. Then PBMC were stained with CD14-APC (BD Biosciences) for 10 min at 4°C and measured in the flow cytometer (BD LSR; BD Biosciences). For quantification of caspase-1 activity, at least 5000 CD14-positive monocytes from each blood donor were gated and analyzed for their median fluorescence in fluorescence channel 1.
Temperature-dependence of recombinant caspase-1
The caspase-1 substrate YVAD-pNA (Alexis) was added to 100 μl of reaction buffer (50 mM HEPES, 50 mM NaCl, 0.1% CHAPS, 10 mM EDTA, 5% glycerol, and 10 mM DTT, pH 7.2) to a final concentration of 50 μM. Then 3 U of recombinant caspase-1 (Alexis) was added and incubated for 4 h at 37, 39, or 41°C. As a control, 100 μM YVAD-cmk was added to a sample at 37°C. The cleavage of YVAD-pNA was measured in a photometer at 405 nm.
Processing of natural proIL-1β by recombinant caspase-1
PBMC (107) in RPMI 1640 containing 2.5 IU/ml heparin were stimulated with 10 μg/ml LPS from Salmonella abortus equi for 4 h at 37 or 40°C. Then blood cells were sedimented by centrifugation (2000 × g, 2 min) and washed with PBS twice. To the remaining cell pellet 100 μl of Aqua ad iniectabilia was added, and the cells were lysed by freezing at −80°C.
To measure caspase-1 activity, 10 μl of cell lysate was mixed with 90 μl of reaction buffer (50 mM HEPES, 50 mM NaCl, 0.1% CHAPS, 10 mM EDTA, 5% glycerol, and 10 mM DTT, pH 7.2) and 5 U of recombinant caspase-1 (Alexis). Samples were incubated at 37 or 40°C for 30 min. The processing of proIL-1β was measured in a proIL-1β ELISA that only detected proIL-1β, not processed, mature IL-1β.
Statistical analysis
All data are given as the mean ± SEM of the number of blood donors indicated. Cytokine release was calculated per milliliter of blood, i.e., corrected for the dilution factor of 5, since 20% blood was used. To calculate the temperature at which cytokine release was inhibited by 50% for each donor, an individual curve fit was performed with PRISM 3.0 (GraphPad, San Diego, CA). Statistical analysis was performed using INSTAT (Instat Statistics; GraphPad). For all data, nonparametric testing was performed. For comparison of two groups, a Wilcoxon matched pairs, signed ranks test was performed. For the comparison of three or more groups, a Kruskal-Wallis test with Dunn’s multiple comparisons test as post-test were used.
Results
Effects of febrile temperatures on LPS-induced cytokine release
To study the effects of febrile temperatures on LPS-induced cytokine release, whole blood was stimulated with 10 μg/ml LPS and incubated at different temperatures between 37–42°C. The temperature at which cytokine release was inhibited to 50% of the levels reached at 37°C was determined for each cytokine (Table I⇓). This comparison showed that there are two groups of cytokines with different sensitivities toward febrile temperatures: IL-1β, IL-18, and IFN-γ were inhibited to 50% at 38.4 ± 0.5, 37.4 ± 0.6, and 37.3 ± 0.4°C, respectively. However, 50% inhibition of both IL-6 and TNF-α was only seen at 42.5°C. A common characteristic of IL-1β and IFN-γ is that their formation depends directly or indirectly on caspase-1; IL-1β is synthesized as an inactive precursor and needs proteolytic cleavage by caspase-1 (9). The formation of IFN-γ is induced by IL-18 (10), which is also synthesized as an inactive precursor and needs proteolytic cleavage by caspase-1 (11). These results suggest that caspase-1 activity might be impaired at febrile temperatures.
Release of various cytokines is inhibited to 50% of control levels at different febrile temperaturesa
Dependence of IL-1β and IL-18 release on caspase-1
To verify that the maturation of IL-1β and IL-18 in LPS-stimulated whole blood depends on caspase-1 activity, we measured IL-1β and IL-18 release in the presence of a caspase-1 inhibitor. The presence of 50 μM YVAD-cmk attenuated LPS-induced IL-1β release by 92 ± 8% (p < 0.05; data not shown), whereas the release of TNF-α was not influenced by YVAD-cmk. This showed that the processing of proIL-1β in LPS-stimulated whole blood occurs mainly via caspase-1. The caspase-1 inhibitor YVAD-cmk was also tested for its ability to block IL-18 release. In the presence of 50 μM YVAD-cmk, LPS-induced IL-18 release was attenuated by 84 ± 2% (p < 0.05). This indicated that the processing of proIL-18 also depends on caspase-1 activity. The formation of IFN-γ is dependent on the release of mature IL-18. Inhibition of caspase-1 activity with 50 μM YVAD-cmk resulted in an inhibition of IFN-γ release by 77 ± 8% (p < 0.05).
Effect of febrile temperatures on processing of proIL-1β
To test whether febrile temperatures interfere with proteolytic maturation of proIL-1β, released IL-1β and intracellular proIL-1β were measured in whole blood stimulated with LPS for 3 h. As shown above, the release of IL-1β was attenuated at febrile temperatures. A significant inhibition was observed from 38°C upward (Fig. 1⇓). The intracellular formation of proIL-1β, measured in cell lysates, was not inhibited by temperatures from 37–41°C (Fig. 1⇓). Instead, increased levels of proIL-1β were found at 40°C. At 42°C an almost complete inhibition of proIL-1β synthesis was observed. These results indicate that the decreased release of mature IL-1β at temperatures >38°C is due not to impaired synthesis, but to reduced processing of proIL-1β.
Attenuation of IL-1β release, but not proIL-1β formation, at febrile temperatures. Twenty percent whole blood in saline was stimulated with 10 μg/ml LPS. After 3-h incubation time at different temperatures, IL-1β release was measured in the cell-free supernatants, and proIL-1β was measured in whole blood lysates by IL-1β or proIL-1β ELISA. Data represent the mean ± SEM of six different donors. ∗, p < 0.05 vs values of IL-1β at 37°C; +, p < 0.05 vs values of proIL-1β at 37°C.
Effect of febrile temperatures on the activity of recombinant caspase-1
To test whether caspase-1 activity is inhibited at febrile temperatures, recombinant caspase-1 was incubated at 37, 39, and 41°C, and the enzymatic cleavage of the substrate YVAD-pNA was monitored photometrically. At 37°C, 1.7 nmol of YVAD-pNA was cleaved; at 39°C, 1.98 nmol was cleaved; and at 41°C, 2 nmol was cleaved (data not shown). Additionally, we examined the cleavage of natural proIL-1β by recombinant caspase-1. To produce proIL-1β, PBMC were stimulated with LPS for 4 h at 37 or 40°C. Then cells were lysed by freezing. These cell lysates, containing high amounts of proIL-1β, were mixed with caspase-1, and the turnover of proIL-1β to mature IL-1β was monitored by a proIL-1β-specific ELISA. ProIL-1β was processed in all samples at comparable rates (Table II⇓) independent of the temperature of proIL-1β synthesis or the temperature during processing.
Cleavage of proIL-1β by caspase-1 at 37 or 40°Ca
These results show that the activity of recombinant caspase-1 is not impaired at febrile temperatures, and that proIL-1β produced at 40°C can still be cleaved by caspase-1.
Effect of febrile temperatures on caspase-1 activity in monocytes
We also measured caspase-1 activity in intact monocytes, since caspase-1 might be blocked in whole cells by mechanisms that do not function in lysates. PBMC were stimulated for 4 h with LPS to activate caspase-1, then the cells were incubated with the cell-permeable, general caspase-substrate D2R (Asp2-rhodamine) for another 90 min. During this time the caspase substrate D2R was cleaved, resulting in an increased fluorescence in fluorescence channel 1 in the flow cytometer. Additional staining with anti-CD14 Abs allowed us to selectively observe the substrate turnover in monocytes. To test the caspase-1 specificity of this assay, PBMC were stimulated with LPS in the presence of the caspase-1 inhibitor YVAD-cmk. Since 100 μM YVAD-cmk blocked D2R cleavage in monocytes, we considered this assay suitable for measuring caspase-1 activity. Febrile temperatures again had no influence on D2R cleavage. At 37°C D2R cleavage reached a median fluorescence of 51 ± 12, and at 40°C a fluorescence of 50 ± 11 was reached (data not shown). These results indicate that caspase-1 activity in intact monocytes is not inhibited at febrile temperatures. Furthermore, this experiment excludes that febrile temperatures affect cell viability.
Reversibility of temperature effects
Next, we determined whether IL-1β inhibition was reversible. To study this, blood was incubated for 3 h at 37 or 41°C, stimulated with 10 μg/ml LPS, and then incubated for another 20 h at 37 or 41°C. Incubation at 37°C during both incubation periods resulted in the release of 20.3 ± 2.8 ng/ml IL-1β (Fig. 2⇓). When blood was first incubated at 37°C and then, after the addition of LPS, was incubated at 41°C, the LPS-induced release of IL-1β was reduced to 1.6 ± 0.5 ng/ml, i.e., a reduction of >90% (Fig. 2⇓). However, when blood was first incubated at 41°C and then at 37°C, IL-1β release was not inhibited (Fig. 2⇓). This result showed that monocytes have no memory for febrile temperatures and that the heat-sensitive phase starts with LPS stimulation.
Reversibility of temperature effects. Twenty percent whole blood in saline was incubated at 37 or 41°C for 3 h. Then blood was stimulated with 10 μg/ml LPS. After another 20-h incubation at 37 or 41°C, IL-1β release was measured in the cell-free supernatants by ELISA. Data represent the mean ± SEM of five different donors. ∗, p < 0.05 vs values of IL-1β in the samples incubated at 37°C during both incubation periods.
Effect of febrile temperatures on Th1/Th2 cytokine production
IL-18 plays a crucial role in the development of a Th1 response. To study the effects of febrile temperatures on Th1/Th2 cytokine release, whole blood was stimulated with 1 μg/ml SEB and incubated at 37 or 40°C for 72 h, and the secreted cytokines in the supernatants were quantified by ELISA. At 40°C the secretion of IL-5 and IL-13 increased 3-fold, whereas the secretion of IL-2 was reduced to 15% of the respective levels reached at 37°C (Fig. 3⇓). The production of IL-4 was not influenced by temperature (118 ± 26 pg/ml at 37°C; 114 ± 32 pg/ml at 40°C). To test whether substitution of IL-18 could reverse the effects of febrile temperatures on Th1/Th2 cytokine production, whole blood was stimulated with 1 μg/ml SEB in the presence or the absence of 10 ng/ml IL-12 and IL-18 and incubated at 37 or 40°C for 72 h. The presence of 10 ng/ml IL-12 and IL-18 effectively increased IFN-γ release. At 40°C controls produced only 3.6 ± 0.8 ng/ml; in the presence of 10 ng/ml IL-12 and IL-18, IFN-γ release increased to 10.1 ± 2 ng/ml. The increased release of IL-5 and IL-13 at 40°C was reversed in the presence of 10 ng/ml IL-12 and IL-18, whereas the reduced release of IL-2 at 40°C was not restored by IL-12/IL-18 (Table III⇓).
SEB-induced cytokine release at febrile temperatures. Twenty percent whole blood in RPMI 1640 was stimulated with 1 μg/ml SEB. After 72-h incubation at 37 or 40°C, released cytokines were measured in the cell-free supernatants by ELISA. Data represent the mean ± SEM of 10 different donors. ∗∗, p < 0.01 vs values at 37°C.
Substitution of IL-18a
Discussion
In the presented study we compared the effects of elevated temperatures on the release of various cytokines in LPS- or SEB-stimulated whole blood. Experiments at temperatures ranging from 37–42°C revealed that there are two categories of LPS-inducible cytokines regarding susceptibility to febrile temperatures. The release of IL-1β, IL-18, and IFN-γ was already inhibited by 50% at temperatures >38°C, whereas the release of IL-6 and TNF-α was only inhibited by 50% at a temperature of 42°C. This indicates that the reduced release of IL-1β, IL-18, and IFN-γ at 38/39°C is not due to a general inhibition of protein synthesis, but is based on a regulatory process. At temperatures >42°C the formation of all cytokines was impaired, probably due to denaturation processes and cell death.
IL-1 was identified as the endogenous pyrogen, which is responsible for the generation of a fever response in the body. Since temperatures >42°C are lethal to humans, the fever response has to be kept under strict control, and therefore the release of IL-1β must also be carefully regulated. Astonishingly, the effects of febrile temperatures on cytokine release have not been studied in detail. One report showed that at 39°C the release of IL-1β from activated PBMC was inhibited (12). In LPS-stimulated THP-1 cells the formation of proIL-1β was inhibited at 39°C accompanied by an increase in heat shock proteins 70 and 90 (13). For adherent PBMC stimulated with LPS, a decrease in proIL-1β formation and the induction of heat shock proteins were only seen at temperatures >41°C (13). Our results are in line with the in vivo finding that the survival rate of LPS shock in rats was increased when the body temperature was increased to 41°C by heating the animals in a neonatal incubator. The elevated temperatures also led to a decrease in IL-1β plasma concentrations (14).
Elevated temperatures before LPS stimulation did not influence IL-1β release, whereas febrile temperatures after LPS stimulation strongly inhibited IL-1β release. This result showed that the heat-sensitive phase only starts with LPS stimulation, indicating that it is not a general heat shock reaction.
IL-1β and IL-18 are both synthesized as inactive precursors that require proteolytic cleavage for maturation. Caspase-1 was identified as the protease responsible for this processing (8, 10), but other proteases, such as cathepsin G and elastase or caspase-3 and proteinase-3, can cleave proIL-1β or proIL-18, respectively (15). Therefore, we studied whether the formation of IL-1β and IL-18 in LPS-stimulated whole blood depends on cleavage by caspase-1 or whether alternative pathways for cytokine maturation are used. Since in the presence of the caspase-1 inhibitor YVAD-cmk the release of IL-1β and IL-18 was almost completely inhibited, we concluded that the processing of proIL-1β and proIL-18 in LPS-stimulated whole blood depends on caspase-1 activity.
The observation that IL-1β, IL-18, and IFN-γ release was more sensitive to febrile temperatures than that of other cytokines led to the hypothesis that the proteolytic processing of proIL-1β and proIL-18 is impaired at these temperatures. This hypothesis was supported by the finding that at temperatures from 38°C upward, IL-1β release was significantly inhibited, whereas at these temperatures the intracellular levels of proIL-1β tended to accumulate and only decreased at 42°C. At 40°C the levels of proIL-1β were significantly elevated compared with the levels reached at 37°C. These results showed that at temperatures between 38–41°C, proIL-1β was still produced inside the cell, but the processing of proIL-1β was disturbed, which resulted in an accumulation of intracellular proIL-1β.
The inhibited processing at febrile temperatures could have several reasons. Caspase-1 activity could be impaired, or proIL-1β could undergo structural changes that render it uncleavable for caspase-1. These two possibilities could be ruled out, since we saw no impairment of recombinant or natural caspase-1 activity at febrile temperatures, and proIL-1β produced at 40°C was cleaved as efficiently as proIL-1β produced at 37°C. We therefore suggest a third alternative: febrile temperatures could influence cellular compartmentation or trafficking, so that proIL-1β does not encounter caspase-1. IL-1β is a secretory protein that lacks a signal peptide and therefore does not follow the classical endoplasmic reticulum-to-Golgi pathway of secretion. It has been proposed that IL-1β is transported out of the cell in endolysosome-related vesicles (16). Until now, it is not understood where and how processing of proIL-1β takes place. Febrile temperatures could interfere with this unknown secretory pathway and prevent the encounter of caspase-1 and proIL-1β. Similar phenomena could cause the inhibition of IL-18 release. This cytokine is closely related to IL-1β, also lacks a signal peptide, and is probably secreted via a similar pathway.
Febrile temperatures inhibit IL-18 release and therefore also the formation of IFN-γ, which is induced by IL-18 (10). The Th1 cytokine IFN-γ is a central mediator for initiating cell-mediated immune responses against intracellular pathogens such as viruses. During a fever response, which is more pronounced in the case of extracellular pathogens in the body, the attenuation of IFN-γ could result in a favorable predominance of Th2 cytokines favoring the humoral defense system. We therefore studied the effects of febrile temperatures on SEB-induced cytokine release. Actually, febrile temperatures increased the secretion of IL-13 and IL-5, both Th2 cytokines, and inhibited the secretion of IL-2, a Th1 cytokine. IL-4, another typical Th2 cytokine, was not influenced by temperature, but this cytokine was produced only in small quantities after SEB stimulation. Perhaps an optimal stimulation of IL-4 secretion could also show an influence of febrile temperatures on IL-4 production.
The increased release of IL-5 and IL-13 at febrile temperatures was reversed when IL-12 and IL-18 were supplemented. This indicates that the limited amounts of IL-18 and the resulting low amounts of IFN-γ are responsible for the Th1/Th2 shift observed at febrile temperatures. However, the reduced release of IL-2 at febrile temperatures could not be restored with substitution of IL-18, suggesting that the release of IL-2 is not dependent of IFN-γ. Since IL-2 is produced during lymphocyte proliferation, this result could indicate a reduced proliferative capacity at febrile temperatures.
In conclusion, there might be an association of extracellular pathogens, fever, and Th2 response, while intracellular pathogens induce less fever and sustain Th1 responses.
Concluding from these results we propose a novel feedback mechanism of elevated body temperature on the production of the endogenous pyrogen IL-1β. An increase in body temperature above 38°C inhibits the processing of proIL-1β and therefore restricts the formation of mature IL-1β. This could protect the body from an excessive temperature increase with fatal consequences.
The concomitant inhibition of IL-18 release and the resultant decrease in IFN-γ secretion could lead to a predominance of Th2 cytokines, which shapes the type of immune response initiated. This mechanism would have interesting corollaries for the effects of both hyperthermic therapy, i.e., reinforced Th2 response, as well as the abrogation of fever by antipyretic drugs, i.e., reduced Th2 response, in infectious diseases.
Acknowledgments
The excellent technical assistance of Anke Biedermann and Ina Seuffert is greatly appreciated. We thank Stefan Fennrich, Astrid Leja, Sonja von Aulock, and Pascal Renner for their great support.
Footnotes
- Received September 19, 2002.
- Accepted May 5, 2003.
- Copyright © 2003 by The American Association of Immunologists
References
In this issue
