| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Cell Biology, Center for Experimental Medicine, Institute of Medical Science, University of Tokyo, Tokyo 108-8639, Japan
Address all correspondence and requests for reprints to: Yoichiro Iwakura, D.Sc., Professor, Center for Experimental Medicine, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639, Japan. E-mail: iwakura{at}ims.u-tokyo.ac.jp.
| Abstract |
|---|
|
|
|---|
. We found that a normal febrile response was induced even in IL-1
/ß-deficient mice, indicating that production of IL-1 in the brain is not necessarily required for the response. In contrast, IL-6-deficient mice did not exhibit a febrile response. Cyclooxygenase (Cox)-2 expression in the brain was strongly induced 1.5 h after injection of IL-1
, whereas IL-6 expression was observed 3 h after the injection. Cox-2 expression in the brain was not influenced by IL-6 deficiency, whereas indomethacin, an inhibitor of cyclooxygenases, completely inhibited induction of IL-6. These observations suggest a mechanism of IL-1-induced febrile response in which IL-1 in the blood activates Cox-2, with the resulting prostaglandin E2 inducing IL-6 in the brain, leading to the development of fever. | Introduction |
|---|
|
|
|---|
and IL-1ß, are known as agonists and an antagonist, IL-1 receptor antagonist (IL-1Ra), which binds to the same receptors, is also known. Two IL-1 receptors, the type I IL-1 receptor (IL-1RI) and type II (IL-1RII), exist; only the former transduces IL-1 signaling and the latter rather acts as a decoy. IL-1 is produced by a large variety of cells, including monocytes and macrophages, and IL-1Rs are expressed on a wide range of cells in the immune, neural, and endocrine systems. Because IL-1Rs are induced upon peripheral inflammation in the brain, particularly the hypothalamus, hippocampus, and choroid plexus (2, 3, 4), a role for IL-1 has been suggested in the neuronal system.
Fever is a common response of the body to various stresses such as infection and inflammation. Such peripheral stimuli are transmitted to the brain through the nervous system and also by "endogenous pyrogens" (5). Although it is well known that circulating cytokines such as IL-1 and tumor necrosis factor (TNF)
are important endogenous pyrogens, the precise mechanism by which these cytokines induce fever through activation of the thermoregulatory neurons in the hypothalamus remains to be elucidated. It is especially interesting to elucidate how these cytokines stimulate relevant thermoregulatory brain structures, because these large hydrophilic polypeptides hardly penetrate the blood-brain barrier (BBB) (6, 7).
We showed previously that IL-1 expression was induced in the diencephalon of the brain upon injection with turpentine, and that the febrile response to turpentine was abolished in IL-1
/ß-deficient mice, suggesting involvement of IL-1 in the brain in the development of fever (8). Consistently, IL-1RI-deficient mice also failed to respond to turpentine (9). On the other hand, it has been demonstrated that endogenous hypothalamic IL-1ß is not necessary for the development of IL-1
-, IL-1ß- or lipopolysaccharide (LPS)-induced fever (10). It is not known, however, whether or not a febrile response can be induced in the complete absence of both IL-1
and IL-1ß in the brain.
It is known that prostaglandin (PG)E2 is involved in the development of fever during inflammation, because inhibitors of cyclooxygenases, which catalyze synthesis of PGH2, a precursor of PGE2, can suppress febrile response (11), and mice lacking the EP3 receptor, one of the receptors for PGE2, showed an impaired febrile response during the first hour after IL-1ß injection (12). Although two types of cyclooxygenases (Cox) are known, it was suggested that only Cox-2 is involved in the febrile response upon inflammation (8, 13, 14, 15). Furthermore, endogenously induced IL-6 has also been suggested to be involved in the febrile response induced by IL-1 (16). However, the relationship among IL-1, IL-6, and Cox-2 in the brain has not been established conclusively.
In this report, to elucidate the roles of IL-1, which is endogenously induced in the brain during fever, we examined the febrile response in IL-1
/ß-deficient mice upon peripheral administration of IL-1
. Furthermore, we analyze the signaling cascade in the brain using IL-1
/ß- and IL-6-deficient mice.
| Materials and Methods |
|---|
|
|
|---|
was obtained from Pepro Tech EC LTD (London, UK). The lyophilized protein was dissolved in pyrogen-free 0.9% NaCl (saline) containing 0.1% BSA (A9306; Sigma, St. Louis, MO). Indomethacin was obtained from Sigma.
Animals
IL-1
/ß-doubly deficient mice were produced as described (8), and IL-6-deficient mice were kindly provided by Dr. Manfred Kopf (17). These mice were backcrossed to C57BL/6J mice for eight generations, and C57BL/6J mice were used as controls. Mice were housed individually from weaning at 4 wk of age, and sex- and age-matched adult (915 wk of age) male mice were used for each experiment. Mice were kept under specific, pathogen-free conditions in an environmentally controlled clean room at the Center for Experimental Medicine, Institute of Medical Science, University of Tokyo. They were housed at an ambient temperature of 24 C and a daily cycle of 12-h light/dark (08002000 h light). All experiments were carried out both according to the institutional ethical guidelines for animal experiments and the safety guidelines for gene manipulation experiments.
Measurement of body temperature
Intraperitoneal temperature of mice was measured using an electric thermometer and tips (ELAMS system; BioMedic Data System, Inc., Maywood, NJ) with an accuracy of 0.1 C. All the tips were tested and adjusted before use. Mice were anesthetized with Nembutal, and the tip was implanted chronically into their peritoneal cavity and ligated to the peritoneum. The position of the tips was verified by postmortem examination. These mice were used for experiments 18 d after the thermometer implantation. IL-1
[1 µg/kg body weight (BW)] was injected iv at 1100 h, and the temperature was measured every 15 min by a person who was accustomed to using the system.
Northern blot hybridization analysis
Northern blot hybridization was performed as described previously using mouse IL-1
, IL-1ß, Cox-2, IL-6, and ß-actin cDNA as probes (8). Wild-type, IL-1
/ß-deficient, and IL-6-deficient mice were injected with IL-1
(1 µg/kg BW). Mice were killed 1.5 and 3 h after injection, and poly-A+ RNA was isolated from the diencephalons. Samples from four mice were pooled for each genotype. Poly-A+ RNA (811 µg) was electrophoresed on a denatured agarose gel and hybridized with specific probes; ß-actin was used as a control. Relative radioactivities of the IL-1
, IL-1ß, IL-6, and Cox-2 bands were compared after normalization with the intensity of the ß-actin band.
Indomethacin treatment
Mice were injected with indomethacin (10 mg/kg BW, ip injection) dissolved in a buffer containing 0.9% NaCl, 5% ethanol, and 4% sodium hydrogen carbonate, or the buffer only 30 min before IL-1 (1 µg/kg BW, iv) injection or saline injection, and IL-6 and Cox-2 mRNA levels were measured 3 h after the treatment by Northern blot hybridization.
Statistical analysis
Averages ± SD are shown. In Fig. 1
, Students t test was used to evaluate statistical significance. Animals with consecutive missing temperature recordings, due to failure of the telemetry system, were excluded from the statistical analysis. In Fig. 2
, Students paired t test was performed to compare before (basal) and after (1.5 h or 3 h after IL-1 injection) data for each genotype.
|
|
| Results |
|---|
|
|
|---|
in IL-1
/ß-deficient and IL-6-deficient mice
/ß-deficient, and IL-6-deficient mice were injected iv with IL-1
, and the resulting febrile responses were examined. The measurement of the temperature started at 0930 h, and IL-1 was injected at 1100 h. The body temperature of untreated mice gradually decreased from 0930 to 1900 h and again rose after that, reflecting the circadian temperature rhythm. The preinjection body temperatures for each experimental group were the same: IL-1-injected wild-type mice, 37.6 ± 0.2; saline-injected wild-type mice, 37.5 ± 0.4; IL-1-injected IL-1
/ß-deficient mice, 37.6 ± 0.1; saline-injected IL-1
/ß-deficient mice, 37.3 ± 0.3; IL-1-injected IL-6-deficient mice, 37.4 ± 0.3; and saline-injected I IL-6-deficient mice, 37.6 ± 0.3. As shown in Fig. 1A
was administered, wild-type mice showed significantly elevated body temperatures (at 30 min and from 2 h 15 min to 8 h 45 min after injection), compared with saline-injected wild-type mouse controls (**, P < 0.05). IL-1-injected IL-1
/ß-deficient mice also showed a febrile response similar to wild-type mice: significantly elevated temperature was observed at 15, 45 min, and 1 h 45 min to 8 h 45 min after IL-1
injection, compared with saline-injected IL-1
/ß-deficient mice (#, P < 0.05). Interestingly, the febrile response of IL-1
/ß-deficient mice was higher than that of wild-type mice (at 1 h 45 min to 2 h, 2 h 30 min, 4 h to 7 h 15 min, and 8 h to 8 h 15 min; *, P < 0.05). In contrast, IL-1-injected IL-6-deficient mice showed no elevation of body temperature at any time points relative to saline-injected IL-6-deficient mice (Fig. 1B
was injected and body temperature was measured every 90 min. These results indicate that endogenous IL-1 induction in the brain is not necessary for the development of fever, whereas IL-6 is essential.
Induction of IL-6, IL-1
, IL-1ß, and Cox-2 expression in the brain by IL-1
We next examined the time course of mRNA expression for the IL-6, IL-1
, IL-1ß, and Cox-2 genes in the brain after IL-1
injection (Fig. 2
). In wild-type mice, Cox-2 and IL-1ß were already strongly induced 1.5 h after injection of IL-1
, followed by IL-6 and IL-1
induction after 3 h. Similarly, in IL-1
/ß-deficient mice, Cox-2 was strongly induced 1.5 h after the injection and IL-6 induction followed 3 h after the injection, as observed in wild-type mice. In the case of IL-6-deficient mice, Cox-2 and IL-1ß were also induced 1.5 h after the injection and IL-1
was induced 3 h after the injection, as in wild-type mice. Similar results were obtained in two additional independent experiments using mRNAs from the whole brains. These results show that endogenous IL-1 and IL-6 are not necessary for the induction of Cox-2 and that IL-6 induction occurs later than that of Cox-2, suggesting that IL-6 is induced downstream of PGE2.
Suppression of IL-1-induced IL-6 expression by indomethacin
We next analyzed the effect of indomethacin, an inhibitor of cyclooxygenases, on the expression of IL-6 to examine PGE2 dependency of the IL-6 expression. The expression of IL-6 was measured at 0 h (basal level) and 3 h after IL-1 injection with or without indomethacin treatment. The administration of indomethacin 30 min before IL-1 injection strongly suppressed IL-6 transcription (Fig. 3
). By this treatment, the Cox-2 expression was not inhibited and was rather induced by the treatment itself, even in the absence of IL-1
treatment. Under these conditions, fever development was completely suppressed, in agreement with previous reports (data not shown; see Refs. 14 , 18). These results suggest that IL-6 expression is dependent on PGE2 production.
|
| Discussion |
|---|
|
|
|---|
and IL-1ß mRNA levels in the diencephalon of the brain. These observations strongly suggested involvement of IL-1 in the brain in febrile response (8). In this report, however, we have shown that endogenous IL-1 in the brain is not necessary for the development of fever, because a febrile response was observed in IL-1
/ß-deficient mice upon peripheral administration of IL-1
. Although this result does not necessarily exclude the possibility that endogenous IL-1 in the brain plays an important role in the febrile response in wild-type mice, it indicates that peripheral IL-1 can induce fever through mechanisms other than inducing IL-1 in the brain inside the BBB.
We observed rather higher febrile response in IL-1
/ß-deficient mice than in wild-type mice. This result is consistent with the hyperresponsive febrile reaction of IL-1ß-deficient mice by ip injection of IL-1
, IL-1ß, or LPS (10). The increased febrile response could be explained by the number or the affinity of IL-1Rs in IL-1
/ß-deficient mice, although no such differences in IL-1Rs were reported in IL-1ß-deficient mice (10). The hyperresponsive febrile response may also reflect lower levels of IL-1Ra in the brain. Consistent with this notion, we reported that IL-1Ra levels in the brain were decreased in IL-1
- or IL-1ß-deficient mice (8).
Then, we analyzed the signal transduction mechanisms during febrile response in the brain. It has been reported that Cox-2 is induced by IL-1 (19, 20, 21, 22), that Cox-2 null mutant mice failed to show a febrile response to IL-1ß (15), and that various Cox-2 inhibitors including indomethacin suppress the development of fever upon inflammation (15), suggesting involvement of PGE2 in IL-1-induced febrile response. Endogenously induced IL-6 has also been suggested to be involved in the febrile response induced by IL-1, because IL-6 null mutant mice failed to mount a febrile response to peripherally injected IL-1ß (16) or IL-1
(Fig. 1B
). In this report, we have analyzed the relationship among IL-1, IL-6, and Cox-2, and demonstrated that: 1) Cox-2 is strongly induced 1.5 h after iv injection of IL-1
, followed by IL-6 at 3 h after injection; 2) Cox-2 induction in the brain is not affected by deficiencies in IL-1 or IL-6; 3) no fever development is observed in IL-6-deficient mice, although Cox-2 is induced in the brain; and 4) inhibition of Cox activity by indomethacin suppresses both IL-6 induction and fever development. These results suggest the following signaling cascade in the febrile response: peripheral IL-1
Cox-2 activation in the brain
PGE2
IL-6 induction
fever development. It has been well documented that IL-1 can induce Cox-2 expression both in vitro and in vivo (19, 20, 21, 23). The endothelium of the cerebral vasculature may be the site where circulating IL-1 interacts with its receptors. Indeed, we and other investigators (4, 24) observed expression of IL-1RI on cerebral blood vessels after treatment with IL-1 (data not shown) (4, 24). Because Cox-2 is induced in the microvessels of the brain upon IL-1 injection (24, 25, 26), PGE2 may be released inside the brain through the BBB to induce IL-6 inside the brain. It was also suggested that IL-1 might penetrate the fenestrated endothelia of capillaries of organum vasculosum laminae terminalis, bind to their receptors located on astrocytes that tightly surround the vascular network, and trigger the synthesis and release of PGE2 (5, 27).
There has been some confusion among data relating to the signaling cascade in the brain that occurs during inflammation. It was reported that intracerebroventricular injection of IL-6 induces an increase of PGE2 levels in the cerebrospinal fluid in parallel with the rise in body temperature in cats (28), suggesting that IL-6 induces Cox-2. Another study has also demonstrated that PGE2 release from rat hypothalamic explants is increased by both IL-1ß and IL-6 in vitro (29). In contrast, this report shows that peripherally injected IL-1 can induce Cox-2 in IL-6-deficient mice without fever development, and indomethacin inhibits IL-6 induction, suggesting that IL-6 is induced downstream of Cox-2. In support of this notion, several reports have demonstrated that IL-6 is induced by PGE2 in activated peritoneal macrophages (30, 31). The expression of IL-6 in the brain has also been suggested to be downstream of PGE2; one such study reported that hyperthermia induced by PGE2 was markedly suppressed by the microinjection of anti-IL-6 directly into the anterior hypothalamic preoptic area (32). Furthermore, PGE2 induces IL-6 in U373 MG human astroglioma cells (33), and inhibition of Cox-2 reduces IL-6 synthesis in human postmortem astrocyte cultures (34). Intravenous administration of IL-6, however, did not induce Cox-2 expression in the rat brain. Thus, IL-6 may induce Cox-2, and vice versa, under artificial conditions, but only IL-6 induction by PGE2 appears to occur in the brain during an in vivo febrile response against IL-1
in mice.
It is known that IL-1
induces other cytokines such as IL-6 or TNF
in the periphery (35). IL-1 and TNF
synergistically increase the production of IL-6 in human synovial fibroblast (36). Thus, it is possible that the febrile response is not caused directly by IL-1
, but rather other cytokines that are induced by IL-1
in the periphery. Regarding this possibility, we could not detect any febrile response by the injection of IL-6 into wild-type mice, consistent with the previous observation (16). On the other hand, administration of TNF
could evoke febrile response in IL-1
-deficient mice as in wild-type mice. Thus, it is possible that the febrile response may actually be caused by other cytokines that are induced by IL-1
in the periphery.
Finally, we have shown that both early febrile response, which occurs during the 1545 min after IL-1
administration, and late response, which occurs after 1 h 45 min, were abolished in IL-6-deficient mice. However, suppression of the early response is rather unexpected, because IL-6 induction was only low during the early phase. It is possible that only low levels of IL-6 are enough for the induction of fever. Consistent with our results, Chai et al. (16) reported that LPS- or IL-1ß-induced fever response was completely abolished in IL-6-deficient mice, although the early response was not clear. However, because the early febrile response against IL-1
in mice was very small compared with that of rabbits or rats against LPS or IL-1ß (5, 37), we might have failed to detect the response. We are now further analyzing the signaling mechanism of the early-phase febrile response.
In conclusion, we have shown that endogenous brain IL-1 is not necessary for IL-1-induced fever development. We suggest that the signaling cascade for the febrile response in the brain is as follows: peripheral IL-1
Cox-2 activation
PGE2
IL-6 induction
fever. We are now analyzing the roles of other cytokines in the brain in the development of fever.
| Footnotes |
|---|
Current address for M.A.: Division of Transgenic Animal Science, Kanazawa University Advanced Science Research Center, 13-1 Takaramachi, Kanazawa 920-8640, Japan.
This work was supported by the grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan, the Ministry of Health, Labor and Welfare of Japan, and Pioneering Research Project in Biotechnology.
Abbreviations: BBB, Blood-brain barrier; BW, body weight; Cox, cyclooxygenase; IL-1R, IL-1 receptor; IL-1Ra, IL-1R antagonist; LPS, lipopolysaccharide; PG, prostaglandin.
Received January 19, 2004.
Accepted for publication July 12, 2004.
| References |
|---|
|
|
|---|
, IL-1ß, IL-1
/ß, and IL-1 receptor antagonist shows that IL-1ß is crucial in turpentine-induced fever development and glucocorticoid secretion. J Exp Med 187:14631475
and IL-1ß, and altered brain cytokine mRNA and serum cytokine levels, in IL-1ß-deficient mice. Proc Natl Acad Sci USA 94:26812686
. Evidence for post-transcriptional regulation. J Biol Chem 269:1176911775
elicits inflammatory cell infiltration in the skin by inducing IFN-
-inducible protein 10 in the elicitation phase of the contact hypersensitivity response. Int Immunol 15:251260
synergistically increase the production of interleukin 6 in human synovial fibroblast. J Clin Lab Immunol 34:107113[Medline]
This article has been cited by other articles:
![]() |
R. Goldbach-Mansky, N. J. Dailey, S. W. Canna, A. Gelabert, J. Jones, B. I. Rubin, H. J. Kim, C. Brewer, C. Zalewski, E. Wiggs, et al. Neonatal-Onset Multisystem Inflammatory Disease Responsive to Interleukin-1{beta} Inhibition N. Engl. J. Med., August 10, 2006; 355(6): 581 - 592. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Chida, T. Osaka, O. Hashimoto, and Y. Iwakura Combined interleukin-6 and interleukin-1 deficiency causes obesity in young mice. Diabetes, April 1, 2006; 55(4): 971 - 977. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Chida, T. Imaki, T. Suda, and Y. Iwakura Involvement of Corticotropin-Releasing Hormone- and Interleukin (IL)-6-Dependent Proopiomelanocortin Induction in the Anterior Pituitary during Hypothalamic-Pituitary-Adrenal Axis Activation by IL-1{alpha} Endocrinology, December 1, 2005; 146(12): 5496 - 5502. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |