METHODS AND AGENTS FOR CONTROLLING INTESTINAL INFLAMMATION AND MUCOSAL
IMMUNITY USING AGENTS INTERACTING WITH TIR8/SIGIRR
The present invention relates to methods for controlling inflammation in the gastrointestinal tract by modulating the activity of TIR8. TIR8, also known as single immunoglobulin IL-1 -related receptor (SIGIRR), is a member of the IL-1 receptor (IL-lR)/Toll-like receptor (TLR) superfamily which acts as an intracellular decoy for components of the signaling pathway. Here we report that TirS has a unique pattern of expression, which includes mucosal tissues and dendritic cells (DC). 7ϊr8-deficient DC showed increased cytokine production in response to TLR agonists (lipopolysaccharide (LPS), CpG oligodeoxynucleotides). 77r<_?-deficient mice had normal susceptibility to systemic LPS toxicity and to intraperitoneal or subcutaneous inflammation. However, 77r#-deficient mice were more susceptible to intestinal inflammation. Thus, TIR8 represents a negative pathway of regulation of the IL- 1R/TLR system, expressed in epithelial cells and DC, crucial for tuning inflammation in the gastrointestinal tract. Introduction Members of the Toll-like receptor (TLR)-IL-l receptor (IL- 1R) superfamily play a key role in innate immunity and inflammation (1-4). TLRs act as sensors for the presence of microorganisms and activate a complex, multifaceted cellular response. Agonists which interact with signaling IL-1R and IL- 18 receptor complexes initiate an amplification cascade of innate resistance, contribute to the activation and orientation of adaptive immunity and play a key role in inflammatory conditions (5). The activity of members of the IL-1R/TLR superfamily is tightly regulated at multiple levels (5, 6). Released and intracellular isoforms of the IL-1 receptor antagonist (IL-lra) block agonistic interactions with formation
of a signaling IL- 1 receptor complex. The IL- 1 RII and IL- 18 binding protein act as decoys for IL- 1 and IL-18 and as dominant negatives at the level of signaling receptor complexes (6-8). IRAK-M is a negative regulator of TLR signaling in macrophages (9) and a splice variant of the adaptor MyD88 (MyD88s) inhibits recruitment of IRAK4 and IL- l/TLR signaling (10). SOCS-1 is a further negative regulator of the IL- 1R/TLR system (1 1). TIR8, also known as single immunoglobulin IL-1 -related receptor (SIGIRR) (12), is a member of the IL-1R family with unique properties. Structurally it is characterized by a single extracellular Ig domain, an intracellular Toll/IL- IR (TIR) domain and 95 aa cytoplasmic tail (12, 13). Ligands for TIR8/SIGIRR have not been identified and searches for accessory functions in signaling complexes have yielded negative results (12). Recently, TIR8 was shown to inhibit NFkB activation by members of the IL-l/TLR family (13, 14). The inhibitory activity of TIR8 was associated with trapping of TRAF6 and IRAKI (14). Wald et al. have recently generated SIGIRR/Tir8 deficient mice and found that these animals are more susceptible to the systemic toxicity of bacterial LPS (14). Here we report that Tir8 has a distinct pattern of expression which includes epithelial tissues and dendritic cells (DC). We generated TϊrS-deficient mice and these mice showed a selective increase in susceptibility to intestinal inflammation. Thus, TIR8, expressed in epithelial cells and sentinel DC, represents a negative pathway of regulation of the IL-1R TLR system and plays a crucial role in tuning inflammation in the gastrointestinal tract. Disclosure of the invention In a first aspect, the invention provides a method for controlling inflammation in the gastrointestinal tract which comprises administering to a subject in need thereof, preferably a human subject, an agent interacting with
TIR8/SIGIRR. In an alternative embodiment, the invention provides a method for modulating mucosal or epithelial immunity, which comprises contacting dendritic or epithelial cells with an agent interacting with TIR8/SIGIRR. As a result of its interaction with TIR8/SIGIRR, the agent may modulate, e.g. inhibit or stimulate, the intracellular signaling of TLR/IL-1 superfamily. Preferably, the agent is selected from small molecules, antibodies and natural or modified interleukins. In a further embodiment, the invention provides a method of screening compounds useful for the treatment of gastrointestinal inflammation, which comprises in vitro assaying a candidate compound for its ability to modulate TIR8/SIGIRR. Preferably, the candidate compound is assayed on cells expressing TIR8/SIGIRR and members of the IL-1R/TLR superfamily, including dendritic and epithelial cells. The candidate compound may be tested and selected for its ability to modulate IL- 1R/TLR signaling, preferably for its effects on TIR8/SIGIRR-mediated TRAF6 and/or IRAK trapping by IL- 1-TLR complex. In a preferred embodiment, the effect of the candidate compound on NFkB and JNK activation is tested. Materials and methods Generation of .Ttrtf-deficient mice. The genomic library consisted of mouse 129/Sv liver DNA cloned in the Xhol site of the Lambda FIX II vector (Stratagene, La Jolla, CA). A total of 30 x 104 plaques were screened with the 32P-labeled Tir8 cDNA. Ten μg of a monophage containing a 15.3 kb insert were analyzed by Southern blot with P-labeled oligonucleotides. Three Bam HI/Bam HI and two Notl/Bam HI fragments were subcloned in pBluscript and partially sequenced. To generate the targeting vector, a Bam Hl/Stul and a Stul/Stul fragments (6kb and 2.3kb) from a genomic clone containing exon 1 to 8 were used. An IRES (Internal Ribosomal Entry Site)-LacZ cassette followed by the
PGK-neomycin resistance gene was inserted into exon 2, 51 bp downstream the first ATG, in the Stul site. After electroporation of Rl embryonic stem cells with the construct, neomycin resistant clones were isolated and analyzed by Southern blot analysis after Bgl II or Hind III digestion using an external probe (A) and an internal Neo probe (B), respectively (Fig. lA and IB). Chimeric mice from one out of 28 targeted embryonic stem cell clone were established by injection of C57BL/6J blastocysts and mated with C57BL/6J females (Charles River, Calco, Italy) to obtain an outbred line carrying the mutated Tir8 alleles. Mice were routinely genotyped by PCR with two primers sets that detected the wild type and targeted allele, respectively (Fig. IB). Phenotypic analysis was performed in a 129/Sv and C57BL/6J mixed genetic background. Tir8 +1+ mice were littermates of Tir8 -I- mice. Procedures involving animals and their care were conformed with institutional guidelines in compliance with national (4D.L. N. l 16, G.U., suppl. 40, 18-2-1992) and international law and policies (EEC Council Directive 86/609, OJ L 358,1 ,12-12-1987; NIH Guide for the Care and Use of Laboratory Animals, US National Research Council 1996). All efforts were made to minimize the number of animals used and their suffering. Cells Mouse DC were generated from pooled CD34+ bone marrow cells from
3-4 mice using GM-CSF (40 ng/ml) and Flt3 ligand (100 ng/ml) as described (15). On day 7 DC were plated at 1 x 106/ml in 0.5 ml and cultured for 24 hr in the presence of different stimuli (LPS 1-lOOng/ml, bacterial CpG oligodeoxynucleotide motif GACGTT (CpG 1826) 0.02-2 μg/ml (Invitrogen Life Technologies, Rockville, MD), heat-killed Candida albicans 10- 100 μg/ml, poly (I):(C) 5-50 μg/ml). Murine macrophages were generated from bone marrow cells or collected from the peritoneal cavity 5 days after the injection of 1.5 ml 3%
sterile thioglycollate (Difco, Detroit, MI). Human monocytes, monocyte- derived macrophages and DC were obtained as described (16). mRNA levels and cytokine production mRNA expression was assessed by Northern blot analysis on total RNA as described (13). IL-6, IL-10, IL- 12, CCL2, CXCL10, were measured in DC supernatants by commercial sandwich ELISA using three replicates per experimental group. LPS-induced inflammation Mortality was evaluated twice a day after injection of 60 mg/kg i.p. of
LPS (E.coli O55:B5, Sigma, St. Luis, MO). Neutrophyl recruitment in the lung was assessed by measuring τnyeloperoxidase activity as reported (17) 2 hr after LPS treatment (200 μg/kg i.p.). Leukocyte recruitment in the air pouch model was assessed by local injection of 20 or 200 ng LPS as reported (18). Intestinal inflammation For acute colitis, mice were fed 3.5% of dextran sulfate sodium (DSS) (molecular mass 40 kDa; ICN) dissolved in sterile, distilled water ad libitum for 5 days followed by 5 days of regular drinking water. For chronic colitis, mice were fed 2% DSS dissolved in sterile, distilled water ad libitum for 5 days followed by 5 days of normal drinking water; this cycle was repeated three times, resulting in a 30-day experimental period. Clinical and histological scores were assessed as described (19). Briefly, for the determination of clinical scores, the body weight and the presence of occult or gross blood per rectum were determined daily during the acute course and every other day during the chronic colitis induction. For bleeding, we used hemoccult (Beckman Coulter) in coded samples, giving a score from 0 to 4 depending on the colour intensity.
Statistical analysis 5-12 animals per experimental group were used throughout. Experiments were repeated 1 to 6 times as detailed. Fisher's exact test and the two-tailed Student's t-test were used as appropriate. Results Expression of Tir8 in murine mucosal tissues and DC As shown in Fig. 2A, Tir8 mRNA was highly expressed in the gut (small intestine, colon, cecum), the lung, the kidney and the liver. Accordingly, the murine colon epithelial cell line colon 26, the mammary epithelial line TSA and the kidney RENCA line expressed high amounts of the transcript. By contrast, the fibroblast cell line L929 did not express Tir8 (Fig. 2B). As shown in Fig. 2C, immature bone marrow-derived DC expressed appreciable levels of Tir8, whereas mature DC obtained by exposure to LPS expressed lower levels of the transcript. By contrast, Tir8 expression in macrophages was almost undetectable. TIR8 expression was also analysed in human monocytes, monocyte-derved DC and macrophages. As shown in Fig. 2D, immature DC expressed TIR8 and LPS treatment downmodulated TIR8 expression in this cell type (mature DC). Monocytes expressed low levels of the transcript that were further reduced upon maturation to macrophages. Generation of Tir8-deficient mice The murine Tir8 gene consists of 9 exons spanning about 4 kb (Fig. 1A). To assess the in vivo role of TIR8 we generated Tir8 deficient mice by homologous recombination. The targeting vector consisted of a genomic DNA fragment of 8.3 kb encompassing exon 1 through 8 of the mouse Tir8, with an IRES-lacZ cassette followed by the PGK-neomycin resistance gene integrated
in exon 2, 51 bp downstream of the first coding ATG (Fig. 1A). 28 independently targeted Rl embryonic stem cell clones of 235 tested were identified by Southern blot hybridization using probe A (Fig. 1A and IB) after digestion with Bgl II. No evidence for random integration was detected with probe B (from the neomycin gene) after digestion with Hind III. One recombinant clone was injected into C57BL/6J blastocysts to generate germ line chimeras. Tir8 deficiency at the mRNA level was assessed by Northern blot analysis of tissues using a 450 bp cDNA probe encompassing exon 1 through 4 (Fig. 1C). Heterozygous females and males were normal and fertile and breeding yielded the predicted number of homozygous null mice at a mendelian frequency. Tir8 -I- mice were viable, fertile and displayed a normal life span in a conventional mouse facility. Microscopic examination of organs and tissues of young or adult mice (1 and 4 months old) did not reveal gross morphological abnormalities. Complete peripheral blood cell counts and microscopic examination of blood smears of Tir8 -I- mice (n=4) did not show statistically significant differences compared to Tir8 +1+ animals. Increased cytokine production by Tir8 -I- DC As shown in Fig. 2, DC, unlike mature macrophages, express Tir8. Given the key role of DC as sentinel cells it was important to assess the significance of this observation. As shown in Fig. 3A, where one experiment representative of 6 performed is shown, Tir8 -I- DC showed increased responsiveness to LPS in terms of IL-6 production. Similar results were obtained when a CpG oligodeoxynucleotide engaging TLR9 was used, whereas no consistent difference was observed when Candida albicans or Poli(I):(C) (3 experiments) were used (Fig. 3B). The better responsiveness of Tir8 -I- DC compared to wild type was also observed when CXCL10/IP 10 (Fig. 3C), IL- 10 or IL-12 (not shown, 1 experiment performed) were
measured. Bone marrow-derived (not shown) or peritoneal macrophages (Fig. 3D) from Tir8 -I- mice showed normal responsiveness to LPS, as expected on the basis of low expression of Tir8. Selective increase in susceptibility to intestinal inflammation in Tirδ -/- mice The susceptibility of Tir8 deficient mice to a variety of inflammatory conditions was then investigated. LPS-induced mortality was evaluated by injecting 60 mg/kg of LPS i.p., a dose corresponding to about the LD50 in
Tir8 +/+ mice. In four experiments performed (two with males and two with females) with 8-12 mice per group, we did not observe significant differences in susceptibility to LPS between Tir8 -I- and +/+ mice. Fig. 4A shows the cumulative result of the 4 experiments performed with a total of 43 Tir8 +/+ and 45 -/- mice. Systemic LPS-induced inflammation was also assessed by measuring serum levels of the proinflammatory cytokine IL-6 at 2 hr after i.p. injection of 200 μg/kg LPS. As shown in Fig. 4B no significant difference was found between Tir8 +1+ and -/- mice (n=5, 1 experiment performed). We then analysed LPS-induced polymorphonuclear cell recruitment in the lungs by measuring lung myeloperoxidase activity at 2 hr after systemic LPS injection (200 μg/kg, i.p.). In these experimental conditions we observed a 8-fold increase in myeloperoxidase activity between untreated or LPS-treated mice, with no significant differences between Tir8 +/+ and -/- mice (Fig. 4C) (n=5, 1 experiment performed). We studied local inflammation by analysing leukocyte recruitment in two different models, the air pouch and thioglycollate-induced peritonitis. As shown in Fig. 4D, at 4 hr after the injection of 20 or 200 ng LPS in the air pouch, we did not observe a significant difference in leukocyte recruitment between Tir8 +/+ and -/- mice (n=6). Finally, Tir8 deficiency did not modify
leukocyte recruitment in the peritoneal cavity at the time points analysed (5 and 24 hr and 5 days) after thioglycollate injection (not shown). Given the expression of Tir8 in the intestinal tract and in DC, it was important to investigate the role played in vivo by Tir8 in intestinal inflammation. DSS-induced colitis is due to a toxic effect of DSS on colon epithelial cells followed by phagocytosis by lamina propria cells and production of pro-inflammatory cytokines such as TNFα, IL-6, IL- 1 (20, 21). In the chronic phase, the slow regeneration after DSS damaged the colon epithelial barrier causes further perpetuation of intestinal inflammation by bacterial products from the lumen, activation of DC and of a T-cell mediated colitis (22). In a first series of 3 experiments, DSS-induced acute colitis was studied. Although only a non-significant trend in grater body weight loss in Tir8 -I- mice was observed, TϊrS-deficient mice showed increased blood loss with a score of 4 + 0 compared to 2.3 + 0.7 of control mice (data not shown). Chronic bowel inflammation was then investigated. As shown in Fig. 5 A, in chronic colitis, Tir8 -I- mice showed an increased weight loss compared to Tir8 +/+ mice, with, for instance, on day 10 a body weight of 19 + 0.8 g and 21.9 + 0.4 g for Tir8 -I- and +/+, respectively (R<0.01, Student's t-test). The difference in body weight was paralleled by the degree of intestinal bleeding (Fig. 5B). By histology, Tir8 -I- mice showed more severe damage of intestinal mucosa with erosion and inflammatory cell recruitment (Fig. 5C). The number of mice with gross focal ulcerations was 5/8 in Tir8 -I- mice and 2/9 in Tir8 +/+. Discussion The results presented here show that TIR8 has a unique pattern of expression which includes epithelial cells and DC. Previous studies have reported that TIR8 is expressed in epithelial cells and tissues but not in myelomonocytic cells (13, 14). Therefore, the finding that DC derived from
monocytes or bone marrow precursors, which have a definite relationship to the myelomonocytic differentiation pathway (23), express TIR8, is unexpected. DC are heterogeneous. It remains to be established whether expression of TIR8 is shared by diverse DC populations, including, for instance, the recently characterized plasmacytoid DC (23). Given the sentinel function of DC and their localization at epithelial surfaces, the expression of TIR8 in this cell type is consistent with the view that this molecule has a regulatory role in epithelial tissues and at mucosal sites. Genetically deficient mice were generated in an effort to define non- redundant functions of Tir8. rrS-deficient DC, but not macrophages, showed increased responsiveness to LPS and CpG oligodeoxynucleotides in terms of production of cytokines and chemokines (IL-6, CXCL10, IL-12, IL- 10). LPS and CpG interact with signaling receptor complexes which include TLR4 and TLR9, respectively (4). Therefore, the finding that 7ϊr5-deficient DC show increased responsiveness to TLR agonists is consistent with its pattern of expression and its proposed function as a negative regulator of IL-1R/TLR signaling. Tir8 -I- mice showed increased severity of colitis induced by DSS. Recognition of microbial moieties of the enteric flora and production of inflammatory cytokines play a key role in intestinal inflammation in experimental systems and in humans (24-27). For instance, evidence points to a key role of IL- 18 and IL-l α, which is the predominant IL-1 form expressed in epithelial cells, in intestinal inflammation (1, 5, 19, 24, 25). Therefore the observation of increased severity of colitis in Tir8 -I- mice is consistent with a non-redundant regulatory role of this molecule in the gastrointestinal mucosa. It is conceivable that increased production of inflammatory cytokines in response to tissue damage and exposure to microbial molecules by DC in the lamina propria and possibly by epithelial cells is responsible of a more severe
colon inflammation in Tir8 -/- mice. TϊrS-deficient mice showed normal inflammatory reactions at sites other than the gastrointestinal tract, including normal peritoneal inflammation to thiogly collate and normal systemic or local inflammation in response to LPS. Wald et al. (14) reported that SIGIRR/Tir8 -I- mice showed increased susceptibility to systemic administration of LPS in terms of mortality. In the present study, in 4 separate experiments we observed no significant difference in LPS toxicity between Tir8 -I- and +/+ mice. Mononuclear phagocytes and endothelial cells, which generally do not express TIR8 (13, 14), are credited to play a central role in endotoxic shock (28, 29). The apparent discrepancy between results reported here and those of Wald et al. is likely due to the different genetic background and may reflect a differential involvement in the systemic toxicity of LPS of cellular components other than myelomonocytic cells and endothelial cells. The restricted pattern of expression of TIR8 and the selectivity of the inflammatory phenotype of deficient mice are consistent with a selective regulatory role of TIR8 in epithelial tissues and mucosal surfaces. TIR8 has been shown to dampen signaling in response to activation of members of IL-1R/TLR superfamily (13, 14). Evidence suggests that TIR8 recruited at signaling receptor complexes may act as an intracellular decoy trapping key components of the transduction cascade (TRAF6 and IRAK) (14). The increased responsiveness of DC to TLR engagement, in terms of cytokine production, and the increased severity of intestinal inflammation in Tir8 -I- mice is consistent with the view of this molecule as a molecular trap for components of the signaling cascade. Thus, TIR8 represents a negative pathway of regulation of the IL-1R/TLR system, with a unique pattern of expression in epithelial cells and DC, crucial for tuning inflammation in the gastrointestinal tract.
Description of the Figures Fig. 1. Targeting of the Tir8 gene. (A) The targeting vector (TV), the Tir8 wild type allele (WT) and the homologous recombinant allele (HR) are shown. Exons (white blocks), introns (thick line), probes (lettered black blocks), restriction fragments, primers (numbered arrows) are shown. The targeting vector contained a 6 kb BamHI-StuI fragment, an IRES-lacZ and a PGK-neomycin resistance cassette (gray block) cloned in the second coding exon and a 2.3 kb Stul/Stul fragment. (B) Left: Southern blot of Bgl Il-digested genomic DNA hybridized with probe A, generating a 6.5 kb wild type restriction fragment and a 5 kb homologous recombinant restriction fragment. Right: PCR analysis of genomic DNA from Tir8 +/+, +1- and -/- mice. (C) Northern blot of kidney (K) and colon (C) from Tir8 +1+ and -/- mice. 10 μg of total RNA was used in each lane. Ethidium bromide stain of the gel is shown in the lower panel. Fig. 2. Tir8 gene expression. (A) Northern blot analysis of murine tissue total RNA (10 μg/lane). Specific Tir8 transcripts are indicated by the arrow. The lower part of the panel shows the ethidium bromide staining after RNA transfer to the membrane. (B) murine cell lines, (C) freshly isolated or cultured murine or (D) human immunocompetent cells (Mo, monocytes ; Mφ, macrophages; iDC, immature DC; mDC, mature DC). Fig. 3. Cytokine and chemokine production by bone marrow-derived DC and macrophages from Tir8 -I- mice. Cytokine and chemokine levels were undetectable in the supernatants of unstimulated cells from Tir8 +/+ and -/- mice and these data are not shown. (A) IL-6 production in response to different LPS concentrations. Results are mean + SE of values from 1 representative experiment of 6 performed. (B) IL-6 production in response to different stimuli. Results are mean + SE of values from 1 experiment for CpG 1826 and 3 experiments for Candida albicans and Poli (I):(C). (C) CXCL10
production (1 experiment). (D) Responsiveness to 100 ng/ml LPS of DC and thioglycollate-elicited peritoneal macrophages. Results are mean + SE from 1 experiment with 4 individual mice for macrophages and from 6 experiments for DC; * P <0.05 and ** P <0.01 by Student's t-test. Fig. 4. Responsiveness of Tir8 -I- mice to local and systemic inoculation of LPS. (A) Mortality following injection of LPS i.p. (60 mg/kg). Results presented are a compound of four experiments with a total of 43 Tir8 +/+ and 45 -/- mice. No significant difference was observed when each of the four experiments was analysed individually. (B) Serum IL-6 levels 2 hr after injection of LPS i.p. (200 μg/kg). (C) Myeloperoxidase (MPO) activity in the lungs 2 hr after i.p. injection of 200 μg/kg LPS. (D) Leukocyte recruitment in the air pouch following local injection of LPS. Fig. 5. Intestinal inflammation in Tir8 -I- mice. Ten mice per group fed with DSS at the indicated time points (bottom bars). Body weight (A) and bleeding (B) were determined at different time points. Results are mean + se from 10 mice (A) or 3 experiments (B). (C) Histology of control untreated, Tir8 +/+ DSS-treated and -/- DSS-treated mice on day 30. Magnification bar: 100 μm.
REFERENCES
1. Dinarello, C. A. (1996) Blood SI, 2095-2147.
2. Medzhitov, R. (2001) Nat. Rev. Immunol. 1, 135-45.
3. O'Neill, L. A. (2002) Trends Immunol. 23, 296-300. 4. Takeda, K., Kaisho, T. & Akira, S. (2003) Annu. Rev. Immunol. 21, 335-76.
5. O'Neill, L. A. & Dinarello, C. A. (2000) Immunol. Today 21, 206-9.
6. Mantovani, A., Locati, M., Vecchi, A., Sozzani, S. & Allavena, P. (2001) Trends Immunol. 22, 328-36. 7. Colotta, F., Dower, S.K., Sims, J.E. & Mantovani, A. (1994) Immunol. Today 15, 562-566.
8. Novick, D., Kim, S. H., Fantuzzi, G., Reznikov, L.L., Dinarello, CA. & Rubinstein, M. (1999) Immunity 10, 127-136.
9. Kobayashi, K., Hernandez, L.D., Galan, J.E., Janeway, C.A., Jr., Medzhitov, R. 8c Flavell, R A. (2002) Cell 110, 191-202.
10. Burns, K., Janssens, S., Brissoni, B., Olivos, N., Beyaert, R. & Tschopp, J. (2003) J. Exp. Med. 197, 263-8.
1 1. Kinjyo, I., Hanada, T., Inagaki-Ohara, K., Mori, H., Aki, D., Ohishi, M., Yoshida, H., Kubo, M. & Yoshimura, A. (2002) Immunity 17, 583-91. 12. Thomassen, E., Renshaw, B.R. & Sims, J.E. (1999) Cytokine 11,
389-399.
13. Polentarutti, N., Penton Rol, G., Muzio, M., Zoja, C, Benigni, A.,
Tomasoni, S., Vecchi, A., Garlanda, C. & Mantovani, A. (2003) Eur. Cytokine
Netw. 14, 1 -8. 14. Wald, D., Qin, J., Zhao, Z., Qian, Y., Naramura, M., Tian, L., Towne,
J., Sims, J.E., Stark, G.R. & Li, X. (2003) Nat. Immunol. 4, 920-927.
15. Vecchi, A., Massimiliano, L., Ramponi, S., Luini, W., Bernasconi, S.,
Bonecchi, R., Allavena, P., Parmentier, M., Mantovani, A. 8c Sozzani,
S. ( 1999) J. Leukoc. Biol. 66, 489-494.
16. Garlanda, C, Hirsch, E., Bozza, S., Salustri, A., De Acetis, M., Nota, R., Maccagno, A., Riva, F., Bottazzi, B., Peri, G., Doni, A., Vago, L., Botto, M., De Santis, R., Carminati, P., Siracusa, G., Altruda, F., Vecchi, A., Romani, L. & Mantovani, A. (2002) Nature 420, 182-186.
17. Goldblum, S.E., Wu, K.M. & Jay, M. (1985) J. Appl. Physiol. 59, 1978-85.
18. Romano, M., Sironi, M., Toniatti, C, Polentarutti, N., Fruscella, P., Ghezzi, P., Faggioni, R., Luini, W., van Hinsbergh, V., Sozzani, S., Bussolino, F., Poli, V., Ciliberto, G. & Mantovani, A. (1997) Immunity 6, 315-25.
19. Siegmund, B., Lehr, H.A., Fantuzzi, G. & Dinarello, CA. (2001) Proc. Natl. Acad. Sci. U S A 98, 13249-54.
20. Suzuki, A., Hanada, T., Mitsuyama, K., Yoshida, T., Kamizono, S., Hoshino, T., Kubo, M., Yamashita, A., Okabe, M., Takeda, K., Akira, S., Matsumoto, S., Toyonaga, A., Sata, M. & Yoshimura, A. (2001) J. Exp. Med. 193, 471-81.
21. Arai, Y., Takanashi, H., Kitagawa, H. & Okayasu, I. (1998) Cytokine 10, 890-6.
22. Dieleman, L.A., Pal en, M.J., Akol, H., Bloemena, E., Pena, A.S., Meuwissen, S.G. & Van Rees, E.P. (1998) Clin. Exp. Immunol. 114, 385-91.
23. Shortman, K. & Liu, Y.J. (2002) Nat. Rev. Immunol. 2, 151-61.
24. Dinarello, CA. (2000) N. Engl. J. Med. 343, 732-4.
25. Siegmund, B., Fantuzzi, G., Rieder, F., Gamboni-Robertson, F., Lehr, H.A., Hartmann, G., Dinarello, C.A., Endres, S. & Eigler, A. (2001) Am. J. Physiol. Regul. Integr. Comp. Physiol. 281, Rl 264-73.
26. Ardizzone, S. 8c Porro, G.B. (2002) J. Intern . Med. 252, 475-96.
27. Bouma, G. & Strober, W. (2003) Nat. Rev. Immunol. 3, 521-33.
28. Vallet, B. & Wiel, E. (2001) Crit. Care Med. 29, S36-41.
29. Froidevaux, C, Roger, T., Martin, C, Glauser, M.P. & Calandra, T. (2001) Crit. Care Med. 29, S I 3-5.