The IL-1−Dependent Sterile Inflammatory Response Has a Substantial -1− Independent Component That Requires C This information is current as of September 23, 2021. Hajime Kono, Gregory M. Orlowski, Zubin Patel and Kenneth L. Rock J Immunol published online 22 August 2012 http://www.jimmunol.org/content/early/2012/08/22/jimmun ol.1200136 Downloaded from

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 23, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published August 22, 2012, doi:10.4049/jimmunol.1200136 The Journal of Immunology

The IL-1–Dependent Sterile Inflammatory Response Has a Substantial Caspase-1–Independent Component That Requires Cathepsin C

Hajime Kono,* Gregory M. Orlowski,† Zubin Patel,† and Kenneth L. Rock†

The sterile inflammatory response to cell death and irritant crystals is medically important because it causes disease. Although these stimuli are structurally distinct, they cause inflammation through a common pathway that requires the cytokine IL-1. In vitro, the inflammasome, and in particular its generation of active caspase-1, is absolutely required to produce bioactive IL-1b. However, in this study, we report that caspase-1 is not required in vivo for much of the IL-1b–dependent sterile inflammatory response. Furthermore, we find that cathepsin C, which controls the activity of a number of leukocyte serine capable of processing IL-1b, plays a major role in this caspase-1–independent pathway. Mice that are deficient in cathepsin C have reduced inflam- Downloaded from matory responses to dying cells and silica crystals. In the absence of cathepsin C, caspase-1 becomes rate limiting such that mice doubly deficient in both of these proteases make little IL-1b in vivo and have markedly attenuated inflammatory responses to the sterile stimuli. In contrast, these mutant mice generate normal inflammation in response to exogenous IL-1b, indicating that cathepsin C and caspase-1 function upstream of IL-1b, and, in their absence, all components of the pathway downstream of mature IL-1b are intact. The Journal of Immunology, 2012, 189: 000–000. http://www.jimmunol.org/ ell death and a number of irritant particles, such as sili- a key role for IL-1b in responses to these particles has not been cates, calcium pyrophosphate, and urate, stimulate robust formally shown in vivo (6). C inflammatory responses in vivo and, as a result, can cause IL-1b is initially transcribed and translated into a proform that or exacerbate tissue damage and disease (1). Recently, it has been is biologically inactive (5). The generation of mature and active recognized that these diverse stimuli all elicit inflammation IL-1b requires proteolytic removal of a propeptide sequence from through a common pathway. These sterile inflammatory responses the cytokine precursor. This proteolytic processing of pro–IL-1b all require the IL-1R and its signaling adaptor molecules (2). in cells is controlled by a macromolecular complex, called the Moreover, where examined, IL-1 blocking reagents also inhibit inflammasome (7–11). This structure forms upon the oligomeri- these responses (3, 4). Therefore, it is now established that the zation of three distinct components. One subunit is a nucleotide- by guest on September 23, 2021 cytokine IL-1 plays an essential role in these sterile inflammatory binding domain, leucine-rich–containing (NLR) family protein responses. There are two distinct forms of IL-1, IL-1a and IL-1b, that is thought to control the activity of the complex; for many and both of these cytokines work through the same receptor, IL- sterile particulates, the NLR protein is NLRP3. A second com- 1R1 (5). It has generally been assumed that proinflammatory ponent is the apoptosis-associated speck-like protein containing particles cause inflammation by stimulating the production of IL- a caspase-recruitment domain subunit, which is thought to serve 1b, and, indeed, in vitro they stimulate macrophages to produce as a scaffold that allows the complex to oligomerize. The third this cytokine; however, IL-1a is also produced, and, in most cases, component is procaspase-1, which is the inactive precursor of the caspase-1. When the NLRP3 inflammasome is stimu- lated, procaspase is cleaved to its active form, and this in turn cleaves pro–IL-1b to mature IL-1b. The mature cytokine is then *Department of Internal Medicine, Teikyo University School of Medicine, Tokyo 173-8605, Japan; and †Department of Pathology, University of Massachusetts Med- released from cells via a nonclassical secretion pathway, and this ical School, Worcester, MA 01655 process may also be controlled by the inflammasome. Received for publication January 13, 2012. Accepted for publication July 25, 2012. The inflammasome has been shown to be absolutely essential for This work was supported by grants from the National Institutes of Health (to K.L.R.), the production of IL-1b in vitro (12, 13). Macrophages that ge- and core resources supported by Diabetes Endocrinology Research Center Grant netically lack NLRP3, apoptosis-associated speck-like protein DK32520 were also used. This work was also supported by a Grant for Research on Intractable Diseases from the Ministry of Health, Labour, and Welfare and a containing a caspase-recruitment domain, or caspase-1 fail to pro- Grant-in-Aid for Scientific Research (C) and Grant-in-Aid for Scientific Research duce any mature IL-1b in response to a number of sterile stimuli, on Innovative Areas from the Ministry of Education, Culture, Sports, Science, and including crystals of silica, urate, or other particles (14, 15). Based Technology of Japan. H.K. is supported by Mochida Memorial Foundation for Med- ical and Pharmaceutical Research, Public Trust Cardiovascular Research Fund, on these results, it has generally been assumed that the inflamma- Senshin Medical Research Foundation, Kowa Life Science Foundation, Naito Foun- some similarly controls the production of IL-1b in vivo, and, in- dation, Takeda Science Foundation, and NOVARTIS Foundation (Japan) for the deed, genetic knockouts of inflammasome components do reduce Promotion of Science. G.M.O. is supported by Medical Scientist Training Program Training Grant T32 AI095213-01 from the National Institutes of Health. responses in animals in which IL-1 participates (12, 16). However, Address correspondence and reprint requests to Dr. Kenneth L. Rock, University of where examined, at least some of these responses are not com- Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655. pletely eliminated in inflammasome-deficient mice (14, 16, 17), E-mail address: [email protected] although it has not been clear why. The inflammasome-independent Abbreviations used in this article: NLR, nucleotide-binding domain, leucine-rich– portion of these responses could be due to IL-1a (that doesn’t re- containing; WT, wild-type. quire caspase-1 for activity), other cytokines, or other ways of Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 producing IL-1b.

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1200136 2 CATHEPSIN C IN IL-1–DEPENDENT STERILE INFLAMMATION

In addition to the inflammasome pathway, there are other po- then heat shocked at 45˚C for 10 min, followed by 37˚C incubation for 5 h; tential mechanisms by which pro–IL-1b could be cleaved into this resulted in necrosis (7-aminoactinomycin D/propidium iodide-positive mature IL-1b. In vitro, a number of cellular proteases, including cells). neutrophil serine proteases (elastase, , and protease 3), Neutrophil and monocyte recruitment to peritoneal cavity matrix metalloproteinases, and mast cell chymase, can cleave b b Quantification of recruited neutrophils and monocytes to the peritoneal purified pro–IL-1 to bioactive IL-1 (18–21). Whether and to cavity was described before (6). Mice were injected i.p. with indicated what extent these mechanisms contribute to IL-1 production in amount of silica crystal or necrotic EL4 cells in 150 ml PBS. After 4 or various settings is unclear. There are limited data that chemical 16 h of injection, the peritoneum was lavaged with 6 ml PBS with 2% FCS, 3 mM EDTA, and 10 U/ml heparin. The absolute number of neutrophils (Ly- inhibitors of elastase and chymase could reduce the processing of + + 2 + m b 6G 7/4 ) and monocytes (Ly-6G 7/4 )in100 l lavage was counted using pro–IL-1 by neutrophils or mast cells, respectively, in vitro and a flow cytometer equipped with a high throughput sampler (BD Biosciences). reduce IL-1b levels and inflammation in vivo (22). Another neu- b trophil , protease 3, has been implicated in IL-1b– Measurement of the mature IL-1 of the peritoneal cavity dependent chronic arthritis in a streptococcal cell wall model (23). Mature IL-1b was measured using a previously described MRC-5 fibro- A limitation of studies using protease inhibitors is the potential blast bioassay (30). The peritoneal cavity of the silica-stimulated mouse that they may have pleotropic effects that can influence responses. was lavaged using 1 ml PBS. The supernatant of the lavage was treated b Overall, the role of these alternate IL-1–processing mechanisms in with anti–IL-1 or control Ab (24). These specimens were added to the MRC-5 fibroblast, and the IL-8 level of the supernatant was determined by most situations in vivo is poorly understood. ELISA. The IL-1b bioactivity was calculated by subtracting the result of We previously found that the sterile inflammatory response to the anti–IL-1b Ab-treated samples from the control Ab-treated ones. cell death required the IL-1R, but was substantially caspase-1 Downloaded from Production and measurement of IL-1b and TNF-a from in vitro independent (3). We initially attributed this caspase-1 indepen- cultures dence to a dominant role of IL-1a in these responses. However, subsequent experiments with mutant mice found that IL-1b was Peritoneal exudate cells were elicited by i.p. injection of 3 ml 1% thio- also critical for the inflammation (6). This led us in this report glycolate and collected after 72 h by peritoneal lavage. Bone marrow neutrophils were isolated from whole bone marrow, following RBC lysis, to examine the contribution of caspase-1 and other proteolytic using the anti–Ly-6G Microbead Kit from Miltenyi Biotec (Cambridge, mechanisms in the generation of IL-1–dependent inflammation to MA). Purity was assessed to be .95–98% by flow cytometry. Murine bone http://www.jimmunol.org/ cell death and other sterile particulate stimuli. marrow-derived mast cells were derived from whole bone marrow using murine rIL-3 (PeproTech, Rocky Hill, NJ), and purity was assessed to be .95% by toluidine blue (31). In all cases, cells were primed in 96-well Materials and Methods plates in RPMI 1640 [or MC/9 medium for mast cells (31)] for 3 h with Reagent and Abs LPS (200 ng/ml) prior to stimulation for an additional 6 h with either silica, nigericin, or poly(dA-dT). Supernatants were collected, and cyto- Abs against Ly-6G (clone 1A8) were obtained from BD Biosciences. Anti- kine levels were analyzed by ELISA. 7/4 Ab was purchased from Serotec. Ab against IL-1b (24) was obtained from Leinco Technologies (St. Louis, MO). The 7-aminoactinomycin D Statistical analyses was obtained from Molecular Probes. Recombinant MIP-2 and human IL-8 6 ELISA kit were purchased from PeproTech (Rocky Hill, NJ). IL-1b and Data are reported as means SEs. Statistical analyses in each independent by guest on September 23, 2021 TNF-a ELISA kits were purchased from BD Biosciences and eBioscience, experiment were performed with an unpaired, two-tailed Student t test. respectively (San Diego, CA). Ultrapure LPS from Salmonella minnesota One-way ANOVA and Dunnett’s multiple comparison posttest were used was purchased from Invivogen (San Diego, CA). Poly(dA-dT) and nigericin to compare the means of multiple groups to the control group. A p value , were purchased from Sigma-Aldrich (St. Louis, MO). Silica crystal (MIN- 0.05 was considered statistically significant. U-SIL 15) was obtained from U.S. Silica (Frederick, MD). Animal and cell lines Results Caspase-1–independent component of IL-1–driven sterile Wild-type (WT) C57BL/6 mice were purchased from The Jackson Labo- inflammation ratory or Japan SLC. Caspase-1–deficient mice (25), cathepsin C-deficient mice (26), IL-1–deficient mice (27), and IL-1R–deficient mice (28) were We have previously reported that the sterile inflammatory response previously described. Caspase-1–deficient mice also lack caspase-11 (29). to dying cells was IL-1b dependent but caspase-1 independent, at All animal protocols were approved by the University of Massachusetts least at the peak of responses (3). In subsequent investigation, we and Teikyo University animal care and use committee. EL4 cells were maintained in RPMI 1640 with 10% FCS and antibiotics. have confirmed this result, although in some experiments we have observed a partial reduction (0–50%) of these responses in cas- Preparation of necrotic cells pase-1–deficient mice. The reason for this variable contribution of Necrotic EL4 cells were prepared as described (6). EL4 cells were washed caspase-1 is not entirely clear; presumably at some times it is five times with PBS, resuspended in PBS at 10 million cells/50 ml, and a rate-limiting component for maximal responses, and at others it

FIGURE 1. Inflammation to necrotic cells after 4 h of injection to peritoneal cavity. The total neutrophil (A) and monocyte (B) numbers in peritoneal cavity of WT C57BL/6, caspase-1–deficient, cathepsin C-defi- cient, caspase-1 and cathepsin C double-deficient, IL- 1b–deficient, and IL-1R–deficient mice after 4-h i.p. injection of 10 million heat-shocked necrotic EL4 cells. The data are combined results of four experiments and represented as means 6 SEM (n = total number of mice from the multiple experiments for each group). PBS groups, WT mice received i.p. PBS. *p , 0.05, **p , 0.01, NS versus WT groups using ANOVA with Dun- nett’s multiple comparison tests The Journal of Immunology 3

FIGURE 2. Inflammation to silica crystal after 4 h of injection to peritoneal cavity. The total neutrophil (A) and monocyte (B) numbers in peritoneal cavity of WT C57BL/6, caspase-1–deficient, cathepsin C-defi- cient, caspase-1 and cathepsin C double-deficient, IL- 1b–deficient, and IL-1R–deficient mice after 4-h i.p. injection of 0.125 mg silica crystal. The data are combined results of three experiments and represented as means 6 SEM (n = total number of mice from the multiple experiments for each group). PBS groups, WT mice received i.p. PBS. *p , 0.05, **p , 0.01, NS versus WT groups using ANOVA with Dunnett’s multiple comparison tests. is not. This led us to examine the contribution of caspase-1 at mice that lacked caspase-1 (Fig. 2). These results again demonstrated earlier time points in the response. a contribution of IL-1b that could be partially or completely inde- Cell death-induced inflammation can be detected within 4 h of pendent of caspase-1.

the injection of dying cells. To examine the role of caspase-1 in the Downloaded from initial phase of the inflammatory response, we injected sterile Role of cathepsin C in sterile inflammatory responses necrotic EL4 cells i.p. into mice, and 4 h later the cellular influx into In vitro, a number of serine proteases have been shown to process the peritoneum was evaluated by immunofluorescent staining and pro–IL-1b into mature IL-1, and very limited data suggest that flow cytometry (3, 6). We found that sterile dead cells elicited an these may function similarly in vivo (19–21, 33). We equivalent influx of neutrophils and monocytes in caspase-1– were interested in taking a genetic approach to evaluate the po- deficient and WT mice (Fig. 1). In contrast, these sterile inflamma- tential role of these proteases in the caspase-1–independent http://www.jimmunol.org/ tory responses were markedly attenuated in mice that lacked IL-1b, pathway of IL-1b production in vivo, but mutant animals lacking and the magnitude of this reduction was similar to that observed in all of these proteases are not available. However, these proteases IL-1R–deficient animals. Therefore, in the total absence of caspase-1, have a common mechanism that controls their activity and that IL-1b was contributing to this sterile inflammatory response. was amenable to a genetic disruption. Given our findings of a caspase-1–independent pathway in cell The neutrophil serine proteases (cathepsin G, elastase, and death-induced inflammation, we sought to analyze another stim- proteinase 3) contain a leader sequence that targets them into the ulus that also elicits an IL-1–dependent sterile inflammatory re- exocytic pathway (34). After transport into the endoplasmic re- sponse. Sterile silica crystals, like many other particles, stimulate ticulum, the leader is removed by the signal peptidase, but two inflammatory responses in vivo that require the IL-1R (32). The extra amino acids remain on the N terminus of the proteases that by guest on September 23, 2021 species of IL-1 that drives these responses in animals has not been keep them in an inactive state (35). These zymogens are trans- elucidated. However, in vitro, these particles stimulate macro- ported into the endocytic compartments of cells. In this location, phages to produce IL-1b (32). Moreover, we have previously the two N-terminal residues are removed by cathepsin C (dipep- shown (32) and confirmed again in this study (described below) tidyl peptidase I), resulting in the proteases becoming catalytically that the production of IL-1b in response to silica stimulation active (36). Because of this mechanism, genetic inactivation of in vitro is absolutely dependent on the activation of caspase-1. cathepsin C will result in the loss of activity of all of these serine When silica is injected i.p. into mice, it stimulates an inflam- proteases (26). Cathepsin C-deficient mice are viable, fail to ac- matory response. This response is markedly reduced in mice that tivate their neutrophil serine proteases, and were shown to have lack the IL-1b (Figs. 2, 3). The loss of inflammatory reduced inflammatory responses in some settings (in arthritis responses in the IL-1b–deficient mice was similar in magnitude to models, to microbial Ags and immune complexes), but not others what was observed with animals that lacked either the IL-1R or (to thioglycolate injection) (37). We therefore examined the cas- IL-1a plus b (Figs. 2, 3). In contrast, at the peak of the response pase-1–independent pathway of IL-1b–dependent inflammation in (16 h), there was no reduction in inflammation in caspase-1–de- a cathepsin C-deficient mouse model. ficient mice (Fig. 3). Similarly, at very early time points (4 h), there The inflammatory response to silica was reduced in cathepsin C- was only a modest reduction (∼25%) in inflammatory responses in deficient mice at both initial and peak time points (Figs. 2, 3).

FIGURE 3. Inflammation to silica crystal after 16 h of injection to peritoneal cavity. The total neutrophil (A) and monocyte (B) numbers in peritoneal cavity of WT C57BL/6, caspase-1–deficient, cathepsin C-defi- cient, caspase-1 and cathepsin C double-deficient, IL- 1b–deficient, and IL-1a and b double-deficient mice after 16-h i.p. injection of 0.5 mg silica crystal. The data are combined results of three experiments and represented as means 6 SEM (n = total number of mice from the multiple experiments for each group). PBS groups, WT mice received i.p. PBS. *p , 0.05, **p , 0.01, NS versus WT groups using ANOVAwith Dunnett’s multiple comparison tests. 4 CATHEPSIN C IN IL-1–DEPENDENT STERILE INFLAMMATION

FIGURE 4. Inflammation to necrotic cells after 16 h of injection to peritoneal cavity. The total neutrophil (A) and monocyte (B) numbers in peritoneal cavity of WT C57BL/6, caspase-1–deficient, cathepsin C-defi- cient, caspase-1 and cathepsin C double-deficient, IL- 1b–deficient, and IL-1R–deficient mice after 16-h i.p. injection of 30 million heat-shocked necrotic EL4 cells. The data are combined results of four experi- ments and represented as means 6 SEM (n = total number of mice from the multiple experiments for each group). PBS groups, WT mice received i.p. PBS. *p , 0.05, **p , 0.01, NS versus WT groups using ANOVA with Dunnett’s multiple comparison tests.

Similarly, inflammation to dead cells was also decreased in these be other proteases contributing to the processing of pro–IL-1b mutant mice at early and later time points (Figs. 1, 4). These in vivo. Because these animals still had caspase-1, we evaluated reductions in inflammation were partial and of lesser magnitude whether this protease might become rate limiting in the absence of than was observed in mice lacking IL-1b (Figs. 1–4). cathepsin C and be responsible for the cathepsin C-independent Downloaded from mechanism of IL-1 generation. Loss of cathepsin C reveals a contribution of caspase-1 to the Indeed, the combined loss of cathepsin C and caspase-1 reduced sterile inflammatory response the initial cell death-induced inflammatory response down to the Our finding that the loss of cathepsin C only partially inhibited the level of inflammation observed in the IL-1b–deficient mice (Fig. IL-1b–dependent inflammatory responses indicated that there must 1). This increased impairment of inflammation in the double- http://www.jimmunol.org/

FIGURE 5. Caspase-1–dependent, but not ca- by guest on September 23, 2021 thepsin C-dependent secretion of IL-1b from macrophages (A, B), mast cells (C, D), and neu- trophils (E–H) in vitro. IL-1b (A) and TNF-a (B) levels in the supernatants of murine peritoneal exudate macrophages from WT, cathepsin C-de- ficient, or caspase-1–deficient C57BL/6 mice. Cells were primed for 3 h with LPS (200 ng/ml) and treated with silica (35 mg/ml), nigericin (1.5 mM), or poly(dA-dT) (0.5 mg/ml) for an addi- tional 6 h. IL-1b (C) and TNF-a (D) levels in the supernatants of murine bone marrow-derived mast cells from WT, cathepsin C-deficient, or caspase- 1–deficient C57BL/6 mice. Cells were primed for 3 h with LPS (200 ng/ml) and treated with silica (40 mg/ml) or poly(dA-dT) (0.3 mg/ml) for an additional 6 h. IL-1b (E, G) and TNF-a (F, H) levels in the supernatants of murine bone marrow neutrophils from WT, cathepsin C-deficient, or caspase-1–deficient C57BL/6 mice. Cells were primed for 3 h with LPS (200 ng/ml) and treated with silica (80–100 mg/ml; E, F) or nigericin (1 mM; G, H) for an additional 6 h. In all experi- ments, cytokine levels in the supernatants were analyzed by ELISA and represented as means 6 SEM. Data are representative of more than three experiments. The Journal of Immunology 5 deficient mice was not evident at the peak of responses (Fig. 4; NS by one-way ANOVA with a Bonferroni test, Casp12/2 versus Casp12/2CatC2/2). The inflammatory response to silica was also further decreased in the double-deficient mice compared with single knockouts, and this greater reduction was observed both at 4 and 16 h (Figs. 2, 3). The responses to silica in the double-deficient mice, although substantially attenuated, were not reduced down to the level seen in IL-1b–deficient mice, especially at the peak of the response (p , 0.01 by one-way ANOVA with a Bonferroni test, Casp12/2 CatC2/2 versus IL-1b2/2; Fig. 3). These results demonstrate enzymes other than caspase-1, and those activated by cathepsin C, can also participate in the maturation of IL-1b in certain settings.

The production of IL-1b by macrophages, neutrophils, and mast cells in vitro is dependent on caspase-1 but not cathepsin C Given the contribution of cathepsin C to IL-1b production in vivo, we sought to determine whether it was involved in the generation FIGURE 6. Caspase-1 and cathepsin C are not required to induce in- of this cytokine by various leukocytes in in vitro assays. Silica flammation to IL-1b. The total neutrophil numbers in peritoneal cavity of Downloaded from stimulates LPS-primed cathepsin C-deficient macrophages, mast WT C57BL/6, caspase-1, and cathepsin C double-deficient mice after 6-h cells, and neutrophils to produce IL-1b, and the magnitude of i.p. injection of 2 mg rIL-1b. The data are combined results of two these responses is similar to that of their WT counterparts (Fig. 5). experiments and represented as means 6 SEM (n = total number of mice Similar results were obtained with other inflammasome stim- from the multiple experiments for each group). PBS groups, WT mice ulators, poly(dA-dT) or nigericin. In contrast, we had previously received i.p. PBS. NS versus WT groups using t test. reported that the production of IL-1b by macrophages stimulated with silica or nigericin was absolutely dependent on caspase-1, Discussion http://www.jimmunol.org/ and we again confirmed this finding (32). Similarly, we found that In this study, we demonstrate that there is a caspase-1–independent caspase-1–deficient mast cells and neutrophils also fail to produce pathway that in vivo contributes to IL-1b–dependent sterile in- IL-1b to inflammasome stimulators. Neutrophils secreted rela- flammatory responses to cell death and silica crystals. The extent to tively low amounts of IL-1b in response to silica compared with which this pathway contributed varied somewhat over the course macrophages and mast cells (Fig. 5E), but produced an equivalent of responses, but it was always quite substantial. This finding is amount of IL-1b in response to nigericin (Fig. 5G). surprising because there is an appreciable literature that the inflammasome complex, which controls caspase-1 activity, and b The requirement for cathepsin C is upstream of IL-1b caspase-1 itself, is required for the generation of IL-1 , including by guest on September 23, 2021 in response to stimuli like silica (25, 32). However, the studies Whereas our data showed a role for cathepsin C and caspase-1 in showing that the inflammasome is required for IL-1b production the IL-1b–dependent sterile inflammatory response in vivo, the are primarily ones performed in vitro (7–11, 32, 38). Our findings results do not define how these proteases are participating in this show that the situation in vivo is clearly more complex. response. To investigate whether cathepsin C and caspase-1 were required for aspects of inflammation aside from the production of IL-1b, we injected rIL-1b into WT versus the double-deficient mice and compared the resulting inflammatory responses (Fig. 6). Injection of IL-1b is sufficient to stimulate a neutrophilic in- flammatory response in WT mice (monocyte recruitment was not seen in these experiments, at least at the time point analyzed). Importantly, this cytokine stimulated an equivalent inflammatory response when injected into cathepsin C and caspase-1 mutant mice. These results indicated that the double-deficient mice are fully capable of mounting IL-1b–dependent inflammatory respon- ses. Moreover, they strongly suggest that the locus of action of cathepsin C and caspase-1 in vivo is upstream of IL-1b. To further investigate this issue, we investigated whether mature IL-1b levels were altered in the sterile inflammatory responses. We focused on responses to silica at 4 h because, technically, we were able to consistently measure mature IL-1b in the peritoneal lavage fluid from mice injected i.p. with these crystals using a bioassay. To eliminate the participation of IL-1a or other ac- FIGURE 7. The production of bioactive IL-1b in vivo. The level of b tivities in the bioassay, the level of IL-1b was determined by mature IL-1 in peritoneal cavity of WT C57BL/6, caspase-1–deficient, subtracting the anti–IL-1b Ab-treated sample from the control cathepsin C-deficient, and caspase-1 and cathepsin C double-deficient mice after 4-h i.p. injection of 0.5 mg silica crystal. Functional IL-1b activities IgG-treated sample. IL-1b levels were significantly reduced in the were measured using MRC-5 fibroblast bioassay, in which the MRC-5 peritoneal fluid of cathepsin C and caspase-1 double-deficient fibroblasts produce IL-8 in response to IL-1. The data are combined results mice injected with silica i.p. (Fig. 7). Levels of this cytokine of three experiments and represented as means 6 SEM (n = total number were also decreased in caspase-1 or cathepsin C-deficient mice of mice from the multiple experiments for each group). *p , 0.05, NS compared with WT mice, although these differences were not versus WT groups using ANOVA with Dunnett’s multiple comparison statistically significant. tests. Bkg, Background level in IL-1b–deficient mice. 6 CATHEPSIN C IN IL-1–DEPENDENT STERILE INFLAMMATION

We show that the inflammasome-independent pathway of the cosis, gout, pseudogout, asbestosis, and possibly also even ath- inflammatory responses in vivo is dependent on cathepsin C. Our erosclerosis (14, 17, 32). IL-1b has also been implicated in met- data strongly suggest that cathepsin C is acting in ways that control abolic syndrome and type II diabetes, among other diseases (40). the generation of bioactive IL-1b. Specifically, bioactive IL-1b Our findings have implications for potential therapeutic targets to levels are reduced in sterile inflammation in cathepsin C-deficient block IL-1 production and may help to explain why caspase-1 mice, and inflammatory responses to exogenous IL-1b are intact inhibitors were not effective in clinical trials (41). in these animals (Fig. 5). Presumably, cathepsin C is contributing to the caspase-1–independent pathway by controlling the activa- Acknowledgments tion of neutrophil serine proteases that can cleave pro–IL-1b into We thank Sharlene Hubbard, Janice BelleIsle, and Tamiko Yanagida for its mature and bioactive form (37). This is consistent with earlier technical assistance. work that showed that these proteases can process pro–IL-1b and that inhibiting of these enzymes reduced inflammatory responses Disclosures 3 in arthritis models (an acute one induced by K/B N serum and The authors have no financial conflicts of interest. a chronic one induced by streptococcal cell wall) and to mono- sodium urate (22, 23). It should be noted that, whereas these enzymes are often collectively referred to as neutrophil serine References proteases, this is a bit of a misnomer because they are expressed in 1. Kono, H., and K. L. Rock. 2008. How dying cells alert the immune system to danger. Nat. Rev. Immunol. 8: 279–289. a number of cell types, including macrophages. 2. Shito, M., G. Wakabayashi, M. Ueda, M. Shimazu, N. Shirasugi, M. Endo, The difference in the contribution of caspase-1 to responses M. Mukai, and M. Kitajima. 1997. Interleukin 1 receptor blockade reduces tumor Downloaded from in vitro versus in vivo was surprising. It is possible that this di- necrosis factor production, tissue injury, and mortality after hepatic ischemia- reperfusion in the rat. Transplantation 63: 143–148. chotomy is due to some cell type whose production of IL-1b in vivo 3. Chen, C. J., H. Kono, D. Golenbock, G. Reed, S. Akira, and K. L. Rock. 2007. is inflammasome independent. However, we examined macro- Identification of a key pathway required for the sterile inflammatory response triggered by dying cells. Nat. Med. 13: 851–856. phages, neutrophils, and mast cells, and all three cell types require 4. So, A., T. De Smedt, S. Revaz, and J. Tschopp. 2007. A pilot study of IL-1 caspase-1 to make IL-1b in vitro. Moreover, we have previously inhibition by anakinra in acute gout. Arthritis Res. Ther. 9: R28. 5. Dinarello, C. A. 2009. Immunological and inflammatory functions of the shown that dendritic cells are dispensable for cell death-induced, http://www.jimmunol.org/ interleukin-1 family. Annu. Rev. Immunol. 27: 519–550. IL-1–dependent responses in vivo (6). Perhaps there is another cell 6. Kono, H., D. Karmarkar, Y. Iwakura, and K. L. Rock. 2010. Identification of the involved or that macrophages or other leukocytes manifest a cas- cellular sensor that stimulates the inflammatory response to sterile cell death. J. pase-1–independent pathway in vivo but not in vitro. There is the Immunol. 184: 4470–4478. 7. Martinon, F., K. Burns, and J. Tschopp. 2002. The inflammasome: a molecular opposite dichotomy with cathepsin C, wherein this protease is platform triggering activation of inflammatory and processing of proIL- required for the inflammasome-independent pathway of IL-1 beta. Mol. Cell 10: 417–426. b 8. Ogura, Y., F. S. Sutterwala, and R. A. Flavell. 2006. The inflammasome: first line production in vivo, but not for the processing of IL-1 in vitro. of the immune response to cell stress. Cell 126: 659–662. These dichotomies may actually make sense when the likely 9. Franchi, L., T. Eigenbrod, R. Mun˜oz-Planillo, and G. Nun˜ez. 2009. The mechanisms of action are considered. Cathepsin C-dependent IL- inflammasome: a caspase-1-activation platform that regulates immune responses and disease pathogenesis. Nat. Immunol. 10: 241–247. b by guest on September 23, 2021 1 processing is presumably mediated through cathepsin C- 10. Martinon, F., A. Mayor, and J. Tschopp. 2009. The inflammasomes: guardians of dependent proteases, like elastase, cathepsin G, and protease 3, the body. Annu. Rev. Immunol. 27: 229–265. and in living cells this hydrolytic machinery is localized in vesi- 11. Schroder, K., and J. Tschopp. 2010. The inflammasomes. Cell 140: 821–832. 12. Mariathasan, S., K. Newton, D. M. Monack, D. Vucic, D. M. French, W. P. Lee, cles and granules. In contrast, procaspase-1 is a cytosolic protein, M. Roose-Girma, S. Erickson, and V. M. Dixit. 2004. Differential activation of as is pro–IL-1b. Consequently, pro–IL-1b would be accessible to the inflammasome by caspase-1 adaptors ASC and Ipaf. Nature 430: 213–218. 13.Agostini,L.,F.Martinon,K.Burns,M.F.McDermott,P.N.Hawkins,and processing by activated caspase-1, but not the vacuolar proteases. J. Tschopp. 2004. NALP3 forms an IL-1beta-processing inflammasome with in- Therefore, processing of IL-1b by cells in vitro would be caspase- creased activity in Muckle-Wells autoinflammatory disorder. Immunity 20: 319–325. 1 dependent and cathepsin C independent. In contrast, pro–IL-1b 14. Martinon, F., V. Pe´trilli, A. Mayor, A. Tardivel, and J. Tschopp. 2006. Gout- associated uric acid crystals activate the NALP3 inflammasome. Nature 440: can be released from dying cells undergoing primary or secondary 237–241. necrosis, and many leukocytes containing this cytokine precursor 15. Eisenbarth, S. C., O. R. Colegio, W. O’Connor, F. S. Sutterwala, and die at sites of inflammation in vivo. The released pro–IL-1b could R. A. Flavell. 2008. Crucial role for the Nalp3 inflammasome in the immu- nostimulatory properties of aluminum adjuvants. Nature 453: 1122–1126. then be cleaved into mature IL-1b by the cathepsin C-dependent 16.Dostert,C.,V.Pe´trilli, R. Van Bruggen, C. Steele, B. T. Mossman, and proteases when these enzymes are released into the extracellular J. Tschopp. 2008. Innate immune activation through Nalp3 inflammasome sensing of asbestos and silica. Science 320: 674–677. fluids or possibly in vesicular compartments, if cells internalize 17. Duewell, P., H. Kono, K. J. Rayner, C. M. Sirois, G. Vladimer, the cytokine precursor by endocytic mechanisms. F. G. Bauernfeind, G. S. Abela, L. Franchi, G. Nun˜ez, M. Schnurr, et al. 2010. In some of our experiments, we observed that whereas in- NLRP3 inflammasomes are required for atherogenesis and activated by cho- lesterol crystals. Nature 464: 1357–1361. flammation was markedly reduced in cathepsin C and caspase-1 18. Hazuda, D. J., J. Strickler, F. Kueppers, P. L. Simon, and P. R. Young. 1990. double-deficient mice, it was still higher than in animals lacking Processing of precursor interleukin 1 beta and inflammatory disease. J. Biol. IL-1b (Figs. 2–4). This finding suggests that there are yet other Chem. 265: 6318–6322. 19. Mizutani, H., N. Schechter, G. Lazarus, R. A. Black, and T. S. Kupper. 1991. mechanisms in vivo that can generate mature IL-1, albeit more Rapid and specific conversion of precursor interleukin 1 beta (IL-1 beta) to an minor ones. Presumably, the IL-1 processing is mediated by other active IL-1 species by human mast cell chymase. J. Exp. Med. 174: 821–825. 20. Scho¨nbeck, U., F. Mach, and P. Libby. 1998. Generation of biologically active proteases whose activity is not dependent on cathepsin C. One IL-1 beta by matrix metalloproteinases: a novel caspase-1-independent pathway such protease could be matrix metalloproteinase 9 (20). This of IL-1 beta processing. J. Immunol. 161: 3340–3346. protease can process pro–IL-1b into its mature form in vitro, and 21. Coeshott, C., C. Ohnemus, A. Pilyavskaya, S. Ross, M. Wieczorek, H. Kroona, A. H. Leimer, and J. Cheronis. 1999. Converting -independent release of metalloproteinase 9-deficient mice have been reported to reduce tumor necrosis factor alpha and IL-1beta from a stimulated human monocytic IL-1b–dependent neuropathic pain (39). cell line in the presence of activated neutrophils or purified proteinase 3. Proc. We believe that understanding how IL-1b is produced during Natl. Acad. Sci. USA 96: 6261–6266. 22. Guma, M., L. Ronacher, R. Liu-Bryan, S. Takai, M. Karin, and M. Corr. 2009. sterile inflammation is important because this process underlies -independent activation of interleukin-1beta in neutrophil-predominant a number of diseases. The inflammation that occurs in response to inflammation. Arthritis Rheum. 60: 3642–3650. 23. Joosten, L. A., M. G. Netea, G. Fantuzzi, M. I. Koenders, M. M. Helsen, cell death can cause tissue damage and disease. Sterile inflam- H. Sparrer, C. T. Pham, J. W. van der Meer, C. A. Dinarello, and W. B. van den mation to crystals underlies a number of diseases, including sili- Berg. 2009. Inflammatory arthritis in caspase 1 gene-deficient mice: contribution The Journal of Immunology 7

of proteinase 3 to caspase 1-independent production of bioactive interleukin- the NALP3 inflammasome through phagosomal destabilization. Nat. Immunol. 9: 1beta. Arthritis Rheum. 60: 3651–3662. 847–856. 24. Rogers, H. W., K. C. Sheehan, L. M. Brunt, S. K. Dower, E. R. Unanue, and 33. Irmler, M., S. Hertig, H. R. MacDonald, R. Sadoul, J. D. Becherer, A. Proudfoot, R. D. Schreiber. 1992. Interleukin 1 participates in the development of anti-Listeria R. Solari, and J. Tschopp. 1995. Granzyme A is an interleukin 1 beta-converting responses in normal and SCID mice. Proc. Natl. Acad. Sci. USA 89: 1011–1015. enzyme. J. Exp. Med. 181: 1917–1922. 25. Kuida, K., J. A. Lippke, G. Ku, M. W. Harding, D. J. Livingston, M. S. Su, and 34. Jenne, D. E., and J. Tschopp. 1988. Granzymes, a family of serine proteases R. A. Flavell. 1995. Altered cytokine export and apoptosis in mice deficient in released from granules of cytolytic T lymphocytes upon T cell receptor stimu- interleukin-1 beta converting enzyme. Science 267: 2000–2003. lation. Immunol. Rev. 103: 53–71. 26. Pham, C. T., and T. J. Ley. 1999. I is required for the 35. Caputo, A., R. S. Garner, U. Winkler, D. Hudig, and R. C. Bleackley. 1993. processing and activation of granzymes A and B in vivo. Proc. Natl. Acad. Sci. Activation of recombinant murine cytotoxic cell proteinase-1 requires deletion of USA 96: 8627–8632. an amino-terminal dipeptide. J. Biol. Chem. 268: 17672–17675. 27. Horai, R., M. Asano, K. Sudo, H. Kanuka, M. Suzuki, M. Nishihara, 36. McGuire, M. J., P. E. Lipsky, and D. L. Thiele. 1993. Generation of active M. Takahashi, and Y. Iwakura. 1998. Production of mice deficient in for myeloid and lymphoid granule serine proteases requires processing by the interleukin (IL)-1alpha, IL-1beta, IL-1alpha/beta, and IL-1 receptor antagonist granule thiol protease dipeptidyl peptidase I. J. Biol. Chem. 268: 2458–2467. shows that IL-1beta is crucial in turpentine-induced fever development and 37. Adkison, A. M., S. Z. Raptis, D. G. Kelley, and C. T. Pham. 2002. Dipeptidyl glucocorticoid secretion. J. Exp. Med. 187: 1463–1475. peptidase I activates neutrophil-derived serine proteases and regulates the de- 28. Glaccum, M. B., K. L. Stocking, K. Charrier, J. L. Smith, C. R. Willis, velopment of acute experimental arthritis. J. Clin. Invest. 109: 363–371. C. Maliszewski, D. J. Livingston, J. J. Peschon, and P. J. Morrissey. 1997. 38. Cassel, S. L., S. C. Eisenbarth, S. S. Iyer, J. J. Sadler, O. R. Colegio, Phenotypic and functional characterization of mice that lack the type I receptor L. A. Tephly, A. B. Carter, P. B. Rothman, R. A. Flavell, and F. S. Sutterwala. for IL-1. J. Immunol. 159: 3364–3371. 2008. The Nalp3 inflammasome is essential for the development of silicosis. 29. Kayagaki, N., S. Warming, M. Lamkanfi, L. Vande Walle, S. Louie, J. Dong, Proc. Natl. Acad. Sci. USA 105: 9035–9040. K. Newton, Y. Qu, J. Liu, S. Heldens, et al. 2011. Non-canonical inflammasome 39. Kawasaki, Y., Z. Z. Xu, X. Wang, J. Y. Park, Z. Y. Zhuang, P. H. Tan, Y. J. Gao, activation targets caspase-11. Nature 479: 117–121. K. Roy, G. Corfas, E. H. Lo, and R. R. Ji. 2008. Distinct roles of matrix 30. Dinarello, C. A., K. Muegge, and S. K. Durum. 2001. Measurement of soluble metalloproteases in the early- and late-phase development of neuropathic pain. and membrane-bound interleukin 1 using a fibroblast bioassay. Curr. Protoc. Nat. Med. 14: 331–336. Immunol. Chapter 6: Unit 6.2. 40. Larsen, C. M., M. Faulenbach, A. Vaag, A. Vølund, J. A. Ehses, B. Seifert, Downloaded from 31. Jensen, B. M., E. J. Swindle, S. Iwaki, and A. M. Gilfillan. 2006. Generation, T. Mandrup-Poulsen, and M. Y. Donath. 2007. Interleukin-1-receptor antagonist isolation, and maintenance of rodent mast cells and mast cell lines. Curr. Protoc. in type 2 diabetes mellitus. N. Engl. J. Med. 356: 1517–1526. Immunol. Chapter 3: Unit 3.23. 41. Cornelis, S., K. Kersse, N. Festjens, M. Lamkanfi, and P. Vandenabeele. 2007. 32. Hornung, V., F. Bauernfeind, A. Halle, E. O. Samstad, H. Kono, K. L. Rock, Inflammatory caspases: targets for novel therapies. Curr. Pharm. Des. 13: K. A. Fitzgerald, and E. Latz. 2008. Silica crystals and aluminum salts activate 367–385. http://www.jimmunol.org/ by guest on September 23, 2021