Inhibition of Histone Deacetylases Permits Lipopolysaccharide-Mediated Secretion of Bioactive IL-1β via a Caspase-1−Independent Mechanism This information is current as of September 28, 2021. Dominik Stammler, Tatjana Eigenbrod, Sarah Menz, Julia S. Frick, Matthew J. Sweet, Melanie R. Shakespear, Jonathan Jantsch, Isabel Siegert, Sabine Wölfle, Julian D. Langer, Ina Oehme, Liliana Schaefer, Andre Fischer, Judith Knievel, Klaus Heeg, Alexander H. Dalpke and Konrad A. Bode Downloaded from J Immunol published online 30 October 2015 http://www.jimmunol.org/content/early/2015/10/30/jimmun ol.1501195 http://www.jimmunol.org/

Supplementary http://www.jimmunol.org/content/suppl/2015/10/30/jimmunol.150119 Material 5.DCSupplemental

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 28, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

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 © 2015 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published October 30, 2015, doi:10.4049/jimmunol.1501195 The Journal of Immunology

Inhibition of Histone Deacetylases Permits Lipopolysaccharide-Mediated Secretion of Bioactive IL-1b via a Caspase-1–Independent Mechanism

Dominik Stammler,* Tatjana Eigenbrod,* Sarah Menz,† Julia S. Frick,† Matthew J. Sweet,‡ Melanie R. Shakespear,‡ Jonathan Jantsch,x,{ Isabel Siegert,{ Sabine Wo¨lfle,* Julian D. Langer,‖ Ina Oehme,# Liliana Schaefer,** Andre Fischer,†† Judith Knievel,‡‡ Klaus Heeg,* Alexander H. Dalpke,* and Konrad A. Bode*

Histone deacetylase (HDAC) inhibitors (HDACi) are clinically approved anticancer drugs that have important immune- modulatory properties. We report the surprising finding that HDACi promote LPS-induced IL-1b processing and secretion in Downloaded from human and murine dendritic cells and murine macrophages. HDACi/LPS-induced IL-1b maturation and secretion kinetics differed completely from those observed upon inflammasome activation. Moreover, this pathway of IL-1b secretion was depen- dent on caspase-8 but was independent of the inflammasome components NACHT, LRR, and PYD domains-containing protein 3, -associated speck-like protein containing a carboxyl-terminal caspase-recruitment domain, and caspase-1. Genetic stud- ies excluded HDAC6 and HDAC10 as relevant HDAC targets in this pathway, whereas pharmacological inhibitor studies impli- cated the involvement of HDAC11. Treatment of mice with HDACi in a dextran sodium sulfate–induced colitis model resulted in a http://www.jimmunol.org/ strong increase in intestinal IL-1b, confirming that this pathway is also operative in vivo. Thus, in addition to the conventional inflammasome-dependent IL-1b cleavage pathway, dendritic cells and macrophages are capable of generating, secreting, and processing bioactive IL-1b by a novel, caspase-8–dependent mechanism. Given the widespread interest in the therapeutic target- ing of IL-1b, as well as the use of HDACi for anti-inflammatory applications, these findings have substantial clinical implications. The Journal of Immunology, 2015, 195: 000–000.

ppropriate regulation of local immunity is a prerequisite APC, such as macrophages and dendritic cells (DC) respond, for the maintenance of local tissue homeostasis, and depending on the stimulus, with the secretion of several in- A dysregulated innate immunity can result in acute or flammatory cytokines, including TNF-a, IL-12p40, IL-6, and by guest on September 28, 2021 chronic inflammation. Both hematopoietic and nonhematopoietic IL-1b (1). cells of the innate recognize pathogen- and The IL-1b cytokine exhibits diverse biological activities and damage-associated molecular patterns via a variety of pattern- plays a central role in acute and chronic inflammation. Because of recognition receptors, including the TLR, nucleotide-binding oligo- its potent proinflammatory effects, IL-1b production is tightly merization domain protein-like receptors (NLR), such as nucleotide- regulated. At least two stimuli are needed for the secretion of binding oligomerization domain-containing protein (NOD)1 and bioactive IL-1b: a first stimulus (e.g., a TLR ligand) is required for NOD2 and the inflammasome-associated NLRP/NLRC sensor inducible expression of pro–IL-1b and priming of the inflamma- proteins, C-type lectin receptors, and Rig-I–like receptors (1). some, after which a second signaling event is necessary for pro–IL- Stimulation of these receptors results in cellular activation and 1b processing and the secretion of active IL-1b (2, 3). Pro–IL-1b the tailored release of a panel of cytokines and chemokines that itself is biologically inactive and requires proteolytic cleavage to is appropriate to the class of pathogen encountered. Professional enable secretion and biological responses. Classically, pro–IL-1b is

*Department of Infectious Diseases, Medical Microbiology and Hygiene, Heidelberg Address correspondence and reprint requests to Dr. Konrad A. Bode, Department of University Hospital, Heidelberg 69120, Germany; †Institute of Medical Microbiology Infectious Diseases, Medical Microbiology and Hygiene, Heidelberg University Hos- and Hygiene, University of Tubingen,€ Tubingen€ 70276, Germany; ‡Institute for Mo- pital, Im Neuenheimer Feld 324, Heidelberg 69120, Germany. E-mail address: lecular Bioscience, University of Queensland, Brisbane, St. Lucia, Queensland 4072, [email protected] Australia; xInstitute of Clinical Microbiology and Hygiene, University Hospital Regens- { The online version of this article contains supplemental material. burg, Regensburg 93053, Germany; Institute of Microbiology, University of Erlangen- Nuremberg, Erlangen 91054, Germany; ‖Department of Molecular Membrane Biology, Abbreviations used in this article: ASC, apoptosis-associated speck-like protein con- Max-Planck-Institute for Biophysics, Frankfurt 60438, Germany; #Clinical Cooperation taining a carboxyl-terminal caspase-recruitment domain; BMDC, bone marrow–de- Unit Pediatric Oncology, German Cancer Research Center, Heidelberg 69120, Ger- rived DC; BMDM, BM-derived macrophage; DC, dendritic cell; DSS, dextran many; **Institute of Pharmacology, Goethe University Frankfurt, Frankfurt 60590, sodium sulfate; HATi, histone acetyltransferase inhibitor; HDAC, histone deacety- Germany; ††German Center for Neurodegenerative Diseases, Department of Psychiatry lase; HDACi, HDAC inhibitor; IQ, imiquimod; KO, knockout; LTA, lipoteichoic and Psychotherapy, University Medical Center Go¨ttingen, Go¨ttingen 37077, Germany; acid; MDP, muramyl dipeptide; NLR, nucleotide-binding oligomerization domain and ‡‡Department of Gastroenterology, Hepatology, and Infectious Disease, University protein-like receptor; NLRP3, NACHT, LRR, and PYD domains-containing protein Hospital, Heinrich Heine University of Dusseldorf,€ Dusseldorf€ 40225, Germany 3; NOD, nucleotide-binding oligomerization domain-containing protein; PD106, pimelic diphenylamide 106; p(I:C), poly(deoxyinosinic-deoxycytidylic) acid; poly ORCIDs: 0000-0003-0974-7402 (J.K.); 0000-0001-9085-965X (K.A.B.). (I:C), polyinosinic-polycytidylic acid; SAHA, suberanilohydroxamic acid; siRNA, Received for publication June 3, 2015. Accepted for publication October 5, 2015. small interfering RNA; TRIF, TIR domain-containing adapter-inducing IFN-b; TSA, ; WT, wild-type. This work was supported by German Research Foundation Grants BO 3673/1-1 (to K.A.B.) and INST 114089/4-1 FUGG (to K.H.). M.J.S. is supported by a National Ó Health and Medical Research Council of Australia Senior Research Fellowship Copyright 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00 (APP1003470).

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1501195 2 HDACi PERMITS LPS-MEDIATED SECRETION OF BIOACTIVE IL-1b cleaved by caspase-1. Stimulation of primed cells with damage- Cell Signaling Technology (Danvers, MA). Acetylated histone H3 was associated molecular patterns, such as ATP and urate crystals, or detected by an Ab purchased from Active Motif (Carlsbad, CA). The Ab pore-forming toxins, like nigericin, results in the assembly of a against b-actin was from Upstate Biotechnologies (Lake Placid, NY). MyD88-, apoptosis-associated speck-like protein containing a carboxy- large protein complex, the so-called “inflammasome,” which is terminal caspase-recruitment domain (ASC)-, NACHT, LRR, and PYD responsible for activation of caspase-1. Despite a predominant role domains-containing protein 3 (NLRP3)-, HDAC6-, cathepsin B–, cathep- for caspase-1 in pro–IL-1b cleavage, other , such as sin S–, granzyme A–, granzyme B–, IL-1b–, and caspase-1/11–knockout neutrophil elastase (4), proteinase 3 (5), chymase, cathepsin G (6), (KO) mice were described (18–27); lps2 mutant strains carry a mutation in TIR domain-containing adapter-inducing IFN-b (TRIF) (28). HDAC10- granzyme A (7), the matrix metalloproteinases stromelysin 1, KO mice were purchased from The Jackson Laboratory (Bar Harbor, plasmin, gelatinase A, and gelatinase B (8), and Staphylococcus ME). BALB/c mice and C57BL/6 mice were purchased from Charles River aureus protease (9), were proposed to promote IL-1b maturation. Laboratories (Sulzfeld, Germany). All KO animals were on a C57BL/6 Hence, mechanisms for inflammasome-independent IL-1b process- background with the exception of the caspase-1/11 mice used for the ing exist, but the biological relevance remains unclear. in vivo dextran sodium sulfate (DSS) colitis experiment, which were on an SV129X1 background. All animal experiments were performed in Lysine acetylation is a posttranslational protein modification that specific pathogen–free facilities following national, state, and institu- regulates many cellular processes. Although initially identified as a tional guidelines. histone modification involved in the regulation of chromatin Generation of primary mouse bone marrow–derived DC, structure, thousands of nuclear and cytosolic proteins can be re- macrophages, and human DC versibly acetylated on lysine residues (10). Histone acetyl- transferases and histone deacetylases (HDAC), the enzymes re- Mouse bone marrow–derived DC (BMDC) and BM-derived macrophages sponsible for these modifications, are found in the nucleus and in (BMDM), as well as human DC, were generated as previously described Downloaded from (15, 29) (also see supplemental material). the cytoplasm, thereby extending the mechanistic relevance and research interest in HDAC well beyond the field of chromatin bi- Determination of cytokine secretion ology. HDAC are classified according to their distinct activation Cells were stimulated in RPMI 1640 medium containing 10% FCS. Cell- mechanism into two groups: the classical HDAC and the sirtuins. free supernatants were harvested and analyzed for cytokines using com- Each group is subdivided based on sequence identity and domain mercially available ELISA kits (OptEIA; Becton Dickinson, Heidelberg, Germany). organization, with up to seven isoenzymes/class, associated with http://www.jimmunol.org/ varying substrate specificities and biological functions. HDAC in- Western blotting hibitors (HDACi) promote cellular apoptosis and/or differentiation Cell lysates. For the determination of intracellular IL-1b, cells were and originally were developed as anticancer drugs (11). However, scraped, pelleted (4000 3 g, 4˚C), washed twice in PBS, and lysed for more recent evidence demonstrates that HDACi also have anti- 30 min on ice in 100 ml lysis buffer (50 mM Tris-HCl [pH 7.4], 1% Ige- inflammatory effects in several animal models of inflammatory Pal, 0.25% sodium deoxycholate, complete protease inhibitor mixture diseases (12–14). [Roche Diagnostics, Mannheim, Germany]). Our group (15) and other investigators (16, 17) showed that Total lysates. For the determination of total processed IL-1b, extracts were prepared from cells and culture supernatants by adding Nonidet P-40 and HDACi inhibit the TLR-inducible expression of a diverse array of PMSF to a final concentration of 0.5% and 1 mM, respectively. Lysates proinflammatory mediators. In this article, we report the surprising were cleared by centrifugation at 4˚C for 10 min at 11,000 3 g. Equal by guest on September 28, 2021 finding that the specific combination of HDACi and LPS promotes amounts of protein were fractionated by 12% SDS-PAGE and electro- the secretion of processed, bioactive IL-1b from DC and macro- transferred to nitrocellulose membranes. The membranes were blocked phages. In DC and macrophages, the pathway is inflammasome with TBS (pH 7.8)/3% fraction V BSA/0.05% Tween 20 and blotted with the indicated Abs. Detection was done using ECL substrate (Amersham, independent, requires HDAC inhibition, and is associated with Freiburg, Germany). Abs were used according to the manufacturer’s pro- apoptosis. The existence of a new link between HDAC inhibition tocol. Equivalent protein loading was verified using an Ab against b-actin. and caspase-8–dependent generation of bioactive IL-1b has major Quantitative RT-PCR implications for targeting this pathway in therapeutic applications and for the development of HDAC-selective inhibitors for anti- Total RNA was isolated using RNeasy Kits (QIAGEN) or HighPure RNA inflammatory applications. Kits (Roche Diagnostics), which included DNase I digestion. A total of 1 mg RNA was reverse transcribed with a cDNA synthesis kit (MBI Fermentas, St. Leon-Rot, Germany) using oligo(desoxythymidin) primer. cDNA was Materials and Methods diluted 1:4, and 2.5 ml was used as template in 25 ml TaqMan-PCR mix, Reagents, Abs, and mice according to the manufacturer’s protocol (reagents from AbGene, Ham- burg, Germany; platform: ABI Prism 7700; Applied Biosystems, Darm- Phosphothioate-modified CpG-oligonucleotide 1668 (59-TCC ATG ACG stadt, Germany). The specificity of RT-PCR was controlled by no-template TTC CTG ATG CT-39) was custom synthesized by TIB Molbiol (Berlin, and no-RT controls. Quantitative PCR results were expressed relative to Germany). LPS from Salmonella minnesota was provided by U. Seydel the housekeeping gene b-actin (1/2[Ct (target) 2 Ct (actin)]). (Research Center Borstel, Borstel, Germany). R848, polyinosinic- polycytidylic acid [poly(I:C)], lipoteichoic acid (LTA), imiquimod (IQ), IL-1b bioassay muramyl dipeptide (MDP), Pam3CSK4, and flagellin were from InvivoGen The murine NOB-1 cell line responds to both human and mouse IL-1b by (San Diego, CA). Histone acetyltransferase inhibitor II, pimelic diphenyl- the production of IL-2. Furthermore, these cells are unresponsive to several amide 106 (PD106), 3MA, MeOSuc-AAPV-CMK, trichostatin A (TSA), other cytokines, including TNF-a, CSF, IL-3, IL-5, IL-6, and IFN-g (30). and BX-795 were from Calbiochem (Darmstadt, Germany). CAY10603, NOB-1 cells were plated in 12-well flat-bottom plates at a final density of M344, MS-275, and were from Cayman Chemical (Ann Arbor, 6 10 cells/ml and stimulated for 16 h using recombinant mouse IL-1b and MI). Staurosporine and were from Sigma-Aldrich different cell culture supernatants. IL-2 mRNA levels were determined as a (Taufkirchen, Germany). RGFP966, PCI-34051, and MC1568 were from read-out for active IL-1b levels. SelleckChem (Houston, TX). Nigericin, suberanilohydroxamic acid (SAHA), and necrostatin were from Enzo Lifesciences (Farmingdale, NY). Caspase DSS colitis model inhibitors (Z-VAD-FMK, Z-DEVD-FMK, Z-VEID-FMK, Z-IETD-FMK), droxinostat, and mocetinostat (MGCD0103) were from Santa Cruz Bio- Induction of DSS colitis was performed according to a published protocol (31). technology (Heidelberg, Germany) and Alexis Biochemicals (Lausanne, Female C57BL/6 mice, 6–8 wk old with an initial weight of 17.6 6 3.2 g, Switzerland). Recombinant murine and human cytokines were purchased were housed in groups of five/cage. Colonic inflammation was induced by from R&D Systems (Wiesbaden, Germany). adding 3.5% DSS to their drinking water for 7 d. SAHA (50 mg/kg body Immunoblot Ab against IL-1b was purchased from Abcam (ab9722; weight) was administered via oral gavage in 100 ml, once daily, starting at day Cambridge, U.K.). Ab for phosphorylated IRF3 (#4947) was bought from 1 of colitis induction. The first administration of SAHA was performed 1 h The Journal of Immunology 3 before the first exposure to DSS. Control animals were treated for 30 s with an induce the processing of LPS-induced pro–IL-1b to IL-1b and the anesthesia of 80% CO2 and 20% O2. During the experimental period, body secretion of pro–IL-1b and IL-1b into the medium. weight was determined every day (n = 9 mice/group [caspase-1/11 and wild- type (WT)]; n = 6 mice/group [IL-1b]). One WT mouse was removed from The effect of HDACi on IL-1b secretion is TRIF dependent but thedatasetduetosevereillness. MyD88 independent Disease activity index A particularly striking finding from our initial analysis was that Macroscopic symptoms of inflammation (stool consistency, presence of HDACi/TLR-induced IL-1b secretion was exclusively observed blood in stool, presence of blood at the anus, presence of relieving posture, after stimulation with the TLR4 ligand LPS but not with other and appearance of the fur) were assessed daily during the course of the TLR ligands. TLR initiate signaling by recruiting one or more of experiment. The following scores were assigned to stool consistency: 0, formed stool; 1, formed and soft stool; or 3, diarrhea. Blood in stool was four adaptor proteins (MAL, MyD88, TRAM, and TRIF). TLR4 scored as 0 (no presence of fecal blood) or 3 (presence of fecal blood). Anal signals via the MAL/MyD88 and TRAM/TRIF pathways (33, 34). blood was scored as 0, no blood; 1, inflamed anus; 2, slight bleeding; or 3, BMDC of MyD88-KO mice showed a strong impairment in TLR- bleeding and edema. Relieving posture was scored as 0–3 points, and inducible pro–IL-1b mRNA expression (Fig. 2A). Of the TLR unkempt fur was given 1 point. The scores were added, and the disease agonists examined, only LPS still induced some residual pro–IL- activity index was calculated. Maximum possible score was 13. 1b mRNA and pro–IL-1b protein expression in MyD88-KO cells. Statistical analyses Despite the low levels of LPS- and LPS/SAHA-induced pro–IL- Unless otherwise indicated, all experiments were performed at least three 1b protein expression in MyD88-deficient cells (Fig. 2C), HDACi/ times. ELISA and mRNA experiments were performed in at least three LPS-induced processing remained intact in BMDC of MyD88-KO independent experiments with duplicates; mean 6 SD are plotted. Sig- mice, as visualized by a decreased, yet still detectable, IL-1b p17 Downloaded from nificant differences were evaluated by the unpaired Student t test with two- fragment (Fig. 2C). In contrast, TRIF-defective lps2 BMDC tailed distributions using Prism 6 software (GraphPad, La Jolla, CA). The p values ,0.05 were considered significant. showed no deficiency in HDACi/LPS-inducible pro–IL-1b mRNA expression, yet they had a pronounced reduction in pro–IL-1b processing and secretion (Fig. 2D–F). The slight increase in pro– Results IL-1b protein in TRIF-defective cells was not reproducible in HDACi and LPS cotreatment induces IL-1b secretion from DC other experiments. Thus, these data indicate that HDACi/LPS- http://www.jimmunol.org/ and macrophages induced IL-1b activation depends on TRIF-dependent signal Treatment of DC and macrophages with the HDACi TSA or SAHA transduction. The absence of IL-1b secretion in BMDC stimulated (also known as ) strongly inhibits the TLR-induced ex- with p(I:C)/SAHA can be explained by the marginal expression pression of IL-12p40, IFN-b, type I IFN–dependent genes, IL-6, level of p(I:C)-induced pro–IL-1b in this cell type. and TNF-a (15). In marked contrast, we found that cotreatment of TRIF activation results in the activation of two distinct signaling murine DCs with the TLR4 ligand LPS and the broad-spectrum arms: one involving IRF3 and IKKε and the other involving RIP1. HDACi SAHA promoted IL-1b secretion (Fig. 1A), whereas IL- Western blot analyses of IRF3 phosphorylation showed that 12p40 secretion was concomitantly reduced, as previously observed HDACi had no influence on LPS-induced IRF3 phosphorylation (Fig. 1B). As expected, given the requirement for a trigger signal to (Fig. 2G). Inhibition of IRF3 and IKKε signaling with the phar- by guest on September 28, 2021 promote IL-1b processing, LPS alone did not trigger IL-1b secre- macological inhibitor BX-795, used in a concentration that com- tion. These results were reproducible when LPS derived from pletely abolished LPS-mediated IFN-b induction (Fig. 2H), had Escherichia coli was used instead of LPS derived from Salmonella no effect on SAHA/LPS-induced IL-1b secretion (Fig. 2I). An (Supplemental Fig. 1A). Strikingly, HDACi-dependent IL-1b se- involvement of RIP1 was excluded using the inhibitor necrostatin cretion was only observed after stimulation with LPS, whereas no (Fig. 2J, Supplemental Fig. 1B). Collectively, these results dem- comparable release of IL-1b was observed after costimulation of onstrate that the combination of HDACi and a TLR4 agonist trig- BMDC with SAHA plus Pam3CSK4 (TLR1/2), LTA (TLR2), poly gers TRIF-dependent, but TBK1-, IKK-, and RIP1-independent, (I:C) [poly(deoxyinosinic-deoxycytidylic) acid; [p(I:C)]) (TLR3), processing and secretion of IL-1b. flagellin (TLR5), IQ (TLR7), R848 (TLR7/8), CpG-DNA (TLR9), b or MDP (NOD2) (Fig. 1A). Treatment of BMDC with HDACi did HDACi-induced IL-1 secretion likely depends on inhibition of not affect LPS-stimulated pro–IL-1b mRNA synthesis (Fig. 1C), HDAC11 indicating that the above effects were independent of gene ex- To examine the role of protein acetylation in SAHA/LPS-induced pression. Rather, treatment of BMDC with HDACi/LPS resulted IL-1b secretion, the effect of the histone acetyltransferase inhib- in processing of pro–IL-1b to IL-1b p17. This maturation event itor (HATi) II was analyzed. HDACi/LPS-induced IL-1b secretion was not detectable in HDACi/CpG DNA–stimulated cells (Fig. was substantially reduced by preincubation with a HATi (Fig. 3A). 1D), again indicating that the effect of HDACi on IL-1b matu- Interestingly, the HATi had no effect on LPS-induced pro–IL-1b ration was selective for LPS-stimulated BMDC. HDACi/LPS mRNA (Fig. 3B). This is in line with the notion that the IL-1b stimulated similar levels of IL-1b secretion from BMDC as was gene possesses an open promoter configuration, and histone observed with inflammasome triggers (LPS/ATP, LPS/nigericin), acetylation is not required for pro–IL-1b transcription, whereas and the effects were not additive to these stimuli (Fig. 1E, 1F). IL-12p40 is a cytokine with a promoter that is strictly dependent Release of IL-18, another caspase-1–dependent cytokine, was on histone acetylation (Fig. 3C). These data suggest that the effect only modestly affected by HDACi+LPS (at the limits of detec- of SAHA is due to an alteration of the acetylation status of a target tion), in contrast to the effect observed with LPS/ATP stimulation protein involved in IL-1b processing. (data not shown). Cotreatment with HDACi and LPS also pro- The broad-spectrum HDACi, like SAHA and TSA, which target moted IL-1b secretion in BMDM (Fig. 1G) and in human primary class I, II, and IV HDAC, but not the sirtuins (class III HDAC), monocyte-derived DCs (Fig. 1H), thus indicating that the effect promoted similar levels of secreted IL-1b from LPS-stimulated was not restricted to a particular cell type or host species. Sur- BMDC (Fig. 3D). To gain further insights into which HDAC’s prisingly, but in accordance with a publication by Carta et al. (32), inhibition is likely responsible for the observed effects, selective SAHA inhibited LPS- and LPS/ATP-induced IL-1b secretion by HDACi were tested for their ability to induce IL-1b secretion PBMCs (Fig. 1I). Collectively, these data demonstrate that HDACi from LPS-stimulated BMDC. Costimulation of BMDC with LPS 4 HDACi PERMITS LPS-MEDIATED SECRETION OF BIOACTIVE IL-1b Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 1. HDACi induces LPS-dependent IL-1b secretion in BMDC. (A and B) BMDC were stimulated in the presence or absence of SAHA (4 mM), together with Pam3CSK4 (10 mg/ml), LTA (10 mg/ml), poly(I:C) (50 mg/ml), LPS (100 ng/ml), flagellin (flag) (10 mg/ml), IQ (10 mg/ml), R848 (0.5 mg/ml), CpG-DNA (1 mM), MDP (100 mg/ml), or TNF-a (50 ng/ml) for 14 h. IL-1b (A) and IL-12p40 (B) were determined in supernatants by ELISA. (C) BMDC were stimulated with CpG-DNA (1 mM) or LPS (100 ng/ml), with or without SAHA (4 mM). After 4 h, total RNA was isolated, and expression of IL-1b was quantified by quantitative RT-PCR. (D) BMDC were stimulated for 7 h with CpG-DNA (1 mM) or LPS (100 ng/ml), with or without SAHA (4 mM). Equal amounts of protein were immunoblotted with an Ab against IL-1b.(E and F) BMDC were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM) for 7 h. For the last 30 min, cells were exposed to ATP (5 mM) (E and F) or nigericin (nig; 20 mM) (E). IL-1b was determined in the supernatants by ELISA (E), and equal amounts of total lysates (cells plus cell culture supernatant) were immunoblotted with an Ab against IL-1b (F). BMDM (G), human DC [hDC (H)], and human PBMC [hPBMC (I)] were stimulated for 7 h with LPS (100 ng/ml), with or without 4 mM SAHA. (I) Cells were exposed to ATP (5 mM) for the last 30 min. (G–I) IL-1b was determined in supernatants by ELISA. All experiments were performed at least three times. Data are mean 6 SD. ***p , 0.001. C, CpG-DNA; L, LPS. and HDACi MS-275 (HDAC1, HDAC2, and HDAC3), PD106 HDACi (i.e., spontaneous secretion of active IL-1b upon LPS (HDAC1, HDAC2, HDAC3, and HDAC8), RGFP966 (HDAC3), stimulation would be expected). However, BMDC and BMDM or PCI-34051 (HDAC8) at concentrations that promoted histone derived from HDAC6- or HDAC10-deficient animals did not show hyperacetylation (Supplemental Fig. 2A, data not shown) induced increased IL-1b secretion in this setting, although HDACi/LPS- no or only modest levels of secreted IL-1b (Fig. 3D, 3E). These triggered IL-1b release was very modestly reduced in HDAC10- data indicate that involvement of the class I HDAC (HDAC1, deficient BMDC compared with those from WT mice (Fig. 3F, HDAC2, HDAC3, and HDAC8) was unlikely. The HDAC class I 3G). Similar results were observed in BMDC after knockdown of and IIa inhibitor valproate did not induce IL-1b, even when ap- HDAC6 or HDAC10 (data not shown), and even cells deficient in plied at high concentrations. Considering that SAHA, a strong both HDAC6 and HDAC10 (HDAC6 knockdown in HDAC10- inducer of IL-1b secretion, is a poor HDAC class IIa inhibitor deficient cells, Supplemental Fig. 2B, 2C) showed no pronounced (35), HDAC4, HDAC5, HDAC7, and/or HDAC9 inhibition was IL-1b secretion upon stimulation with LPS alone (Fig. 3H). also unlikely to be responsible for the observed effect. The These data suggest that neither inhibition of HDAC6 or HDAC10 class IIb HDACi CAY10603 and M344 were also strong in- alone nor inhibition of both HDAC6 and HDAC10 is sufficient to ducers of IL-1b secretion (Fig. 3D, 3E). Therefore, the class promote IL-1b maturation from LPS-stimulated BMDC. To in- IIb HDAC HDAC6 and HDAC10, as well as the class IV HDAC vestigate the potential involvement of HDAC11, cells were treated HDAC11, remained as the most likely candidates, and avail- with LPS plus mocetinostat, an inhibitor of HDAC1, HDAC2, able KO mice were analyzed. Deficiency in the relevant HDAC HDAC3, and HDAC11. This inhibitor induced comparable was predicted to mimic the effect induced by the respective amounts of IL-1b to TSA and SAHA in LPS-activated BMDC The Journal of Immunology 5 Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 2. HDACi/LPS-induced IL-1b secretion is TRIF dependent. BMDC from WT, MyD88-KO (A–C), and TRIF-defective (D–F) mice were stimulated with CpG-DNA (1 mM) or LPS (100 ng/ml), with or without SAHA (4 mM). (A and D) After 4 h, total RNA was isolated, and expression of IL-1b was quantified by quantitative RT-PCR. (B and E) After 8 h, IL-1b concentrations were determined in the cell culture supernatant by ELISA. (C and F) After 6 h, cells were lysed, and equal amounts of protein were immunoblotted with Abs against IL-1b.(G) Immunoblot against p-IRF3 in lysates of WT BMDC stimulated with CpG-DNA (1 mM) or LPS (100 ng/ml), with or without SAHA (4 mM). WT BMDC were preincubated with the IRF3/IKKε inhibitor BX795 (BX; 100 nM (H and I) or the RIP 1 inhibitor necrostatin-1 (nec; 10 mM) (J) and then stimulated with CpG-DNA (1 mM) or LPS (100 ng/ml), with or without SAHA (4 mM). IL-1b secretion was measured in cell culture supernatants by ELISA (I and J), and IFN-b mRNA expression was determined by RT-qPCR (H). All experiments were performed at least three times. Data are mean 6 SD. **p , 0.01, ***p , 0.001. C, CpG-DNA; L, LPS; ns, not significant.

(Fig. 3I). Because the HDAC1 and HDAC2 inhibitors MS-275 adaptor molecule involved in the formation of NLRP3 inflam- and PD106, as well as the HDAC3 inhibitor droxinostat, did not masome, produced similar levels (NLRP32/2, Fig. 4A) or only promote IL-1b secretion in LPS-stimulated cells (Fig. 3D), slightly reduced levels (ASC2/2, Fig. 4B) of IL-1b compared with these results suggest that HDAC11 inhibition is most likely WT BMDC upon HDACi/LPS stimulation. Although caspase-1 responsible for LPS-induced IL-1b secretion from DC and activation upon HDACi/LPS treatment was faint (Fig. 4C), macrophages. This is in line with the finding that sodium bu- caspase-12/2 BMDC produced only slightly reduced levels of tyrate, an HDAC class I and class IV inhibitor (36), was also a secreted IL-1b (Fig. 4D, 4E). The peculiarity of this IL-1b mat- strong inducer of IL-1b (Fig. 3D). Unfortunately, only limited uration was also evident in a kinetic analysis of the response. data are available on HDAC11, and no specific inhibitor has HDACi/LPS-induced IL-1b was substantially slower than that been described. Induction by CAY1063 does not exclude in- induced by LPS/ATP (Fig. 4F). In classical inflammasome- volvement of HDAC11, because the effect of this inhibitor on dependent IL-1b processing, cells require priming with a TLR HDAC11 has not been reported. stimulus to induce pro–IL-1b synthesis before the second stimulus (ATP, nigericin) activates IL-1b processing. Preincubation of cells b HDACi-induced IL-1 processing is inflammasome and with the “second signal” is not accompanied by IL-1b secretion. caspase-1 independent In contrast, HDACi needed to be added before or simultaneously Because prototypical activation of pro–IL-1b usually depends on with LPS to induce noteworthy amounts of IL-1b. If the HDACi caspase-1–mediated processing by the inflammasome, BMDC was given 120 min after the TLR stimulus, the secreted amount of derived from KO mice of various inflammasome proteins were IL-1b declined to ,25% of the maximum (Fig. 4G). Further, investigated. Strikingly, BMDC deficient in NLRP3 or ASC, an when BMDC were extensively washed 3 h after HDACi/LPS 6 HDACi PERMITS LPS-MEDIATED SECRETION OF BIOACTIVE IL-1b Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021 FIGURE 3. HDACi-induced IL-1b secretion likely depends on inhibition of HDAC11. (A–C) After preincubation (1 h) with the HATi II (2,6-bis(BrHOB) C6one) (5 mM), BMDC were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM). After 14 h, IL-1b (A) and IL-12p40 (C) concentrations were determined in the cell culture supernatants by ELISA. (B) After 4 h, total RNA was isolated, and expression of IL-1b was quantified by quantitative RT- PCR. (D and E) BMDC were stimulated for 7 h with LPS plus SAHA (4 mM), M344 (0.5 mM), TSA (100 nM), PCI-34051 (10 mM), valproate (Val; 5 mM), PD106 (10 mM), MS-275 (5 mM), sirtuin inhibitor (10 mM), RGFP966 (10 mM), MC1568 (10 mM), droxinostat (drox; 10 mM), CAY10603 (500 nM), or sodium butyrate (butyrate) (5 mM). (D) IL-1b concentrations were determined in the supernatants by ELISA. (E) Where indicated, the cells were exposed to ATP (5 mM) for the last 30 min. Equal amounts of total lysates (cells plus cell culture supernatant) were immunoblotted with an Ab against IL-1b. BMDC of HDAC10-KO mice (F) or HDAC6-KO mice (G) or BMDC of HDAC10-KO mice in which HDAC6 had been knocked down (H) were stimulated for 14 h with LPS (100 ng/ml) together with SAHA (4 mM). IL-1b concentrations were determined in the supernatants by ELISA. (I) BMDC were stimulated for 7 h with LPS plus either SAHA (4 mM) or mocetinostat (mo; 20 mM). IL-1b concentrations were determined in the supernatants by ELISA. All experiments were performed at least three times. Data are mean 6 SD. **p , 0.01, ***p , 0.001. L, LPS; ns, not significant. stimulation and fresh medium was readded, the final IL-1b con- of HDACi/LPS-stimulated WT BMDC induced similar amounts centration in culture supernatants was comparable to that of cells of IL-2 mRNA compared with rIL-1b (Fig. 5A), whereas those of incubated without exchange of medium (Fig. 4H). Hence, once IL-1b–deficient cells showed significantly less IL-2 mRNA pro- activated by HDACi/LPS stimulation, BMDC continue to release duction. Thus, these data confirm the biological activity of IL-1b, with detectable IL-1b in cells, which contrasts with the HDACi/LPS-induced IL-1b. To determine whether the unknown rapid release that occurs upon inflammasome activation (Fig. 4I). pro-IL-1b–processing was located extracellularly or in- Collectively, these data demonstrate that the mechanism of tracellularly, cells were stimulated with HDACi/LPS for 6 h and HDACi/LPS-induced IL-1b secretion is mechanistically distinct washed intensively before they were lysed. Western blot analysis from the canonical pathway of inflammasome-dependent IL-1b of these lysates showed that pro–IL-1b processing occurred in- maturation. tracellularly (Fig. 5B). Based on these results, we set out to identify the intracellular b IL-1 induced by an alternative pathway upon HDACi/LPS protease involved in HDACi-induced IL-1b activation. Different treatment is biologically active proteases, like granzyme A, granzyme B, cathepsin B, and ca- In addition to caspase-1, other enzymes were reported to process thepsin S, could be ruled out in BMDC using the respective KO pro–IL-1b to IL-1b, and the biological activity of the processed mice, whereas the cytosolic serine protease proteinase 3 could be IL-1b can differ depending on the cleavage site (37). To test the excluded using pharmacological inhibitors (Supplemental Fig. activity of IL-1b formed by HDACi/LPS stimulation, supernatants 3A–D). of HDACi/LPS-stimulated WT and IL-1b–deficient BMDC were Strikingly, pretreatment with the broad-spectrum caspase in- compared with rIL-1b in an IL-1b bioassay (30). The supernatants hibitor Z-VAD-FMK reduced HDACi/LPS-induced IL-1b matu- The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/

FIGURE 4. HDACi-induced IL-1b processing is inflammasome- and caspase-1–independent. BMDC of WT and NLRP3-KO mice (A), ASC-KO mice (B), or caspase-1–KO mice (D) were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM). IL-1b concentrations were determined in the 14-h cell culture supernatant by ELISA. (C) BMDC were stimulated with LPS (100 ng/ml) for 7 h. IL-1b and caspase-1 processing was determined in the cell E lysates by immunoblotting with the IL-1b and caspase-1 Abs. ( ) BMDC of WT and caspase-1–KO mice were stimulated with LPS (100 ng/ml) and/or by guest on September 28, 2021 SAHA (4 mM) for 7 h. For the last 30 min, cells were exposed to 5 mM ATP, as indicated. IL-1b processing was determined in the total lysates by immunoblotting with the IL-1b Ab. (F) BMDC were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM), as indicated. For the last 30 min, cells were exposed to 5 mM ATP. (G) BMDC were incubated with SAHA (4 mM), as indicated. At 0 h, cells were stimulated with LPS (100 ng/ml) or CpG (1 mM) for 8 h. IL-1b levels in culture supernatants were determined by ELISA. (H) BMDC were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM). After 3 h, cells were washed three times, and the medium was replaced with fresh medium and incubated for an additional 10 h. IL-1b levels in the culture supernatants were determined by ELISA. (I) BMDC were stimulated for 4 h. For the last 30 min, the cells were exposed to 5 mM ATP. IL-1b processing was determined in the supernatant and the lysates by immunoblotting with the IL-1b Ab. All experiments were performed at least three times. Data are mean 6 SD. **p , 0.01. L, LPS; n.s., not significant. ration completely (Fig. 5C, 5D) but had no influence on the LPS- antagonized IL-1b release from cells costimulated with HDACi/ induced synthesis of pro–IL-1b mRNA (Supplemental Fig. 3E) or LPS (Fig. 5K). In conclusion, these results demonstrate that pro–IL-1b protein (Fig. 5D). In accordance with data using HDAC inhibition is able to induce pro–IL-1b cleavage in a caspase- BMDC from caspase-1–KO mice (Fig. 4D, 4E), a caspase-1 8–dependent, but inflammasome-independent, manner. inhibitor (Z-YVAD-FMK) had only a minor effect on HDACi/ LPS-induced IL-1b secretion (Fig. 5E), yet it blocked LPS/ATP- HDACi treatment of mice with acute colitis increases intestinal b induced IL-1b, as expected. Caspase-3 and caspase-6 inhibitors IL-1 (Z-DEVD-FMK, Z-VEID-FMK) did not affect HDACi/LPS- , a metabolic end product of a variety of intestinal triggered IL-1b maturation (Fig. 5E). Recently, doxorubicin-, , including Clostridia and Eubacteria, possesses HDACi staurosporine-, and FAS-meditated caspase-8–dependent matura- activity with an extended spectrum of target enzymes comparable tion of pro–IL-1b was reported (38, 39). Indeed, pretreatment with that of TSA (33). Butyric acid concentrations of up to 20 mM can the caspase-8 inhibitor Z-IETD-FMK in higher concentrations be found in the mammalian intestine (40). To examine the hy- reduced HDACi/LPS-triggered IL-1b maturation, although at the pothesis that HDACi regulate IL-1b production in the intestine, applied concentrations, inflammasome-dependent (ATP-triggered) we monitored intestinal IL-1b production after HDACi treatment IL-1b processing was inhibited concomitantly (Fig. 5F–H). of mice during experimental colitis. WT mice were treated with HDACi/LPS stimulation of BMDC induced activation of caspase- SAHA or vehicle during a course of mild DSS-induced colitis. As 8 (Fig. 5H). Moreover, the dependency of HDACi/LPS-induced hypothesized, HDACi treatment of WT animals resulted in a IL-1b activation on caspase-8 was confirmed by small interfering strong upregulation of IL-1b in the colon (Fig. 6A), whereas IL-6 RNA (siRNA)-knockdown experiments targeting capsase-8 (Fig. was significantly reduced (Fig. 6B). To confirm the caspase-1 5I, 5J, Supplemental Fig. 3F). In accordance, stimulation of independency that was observed in the in vitro experiments, IL- BMDC with staurosporine and doxorubicin elicited only modest 1b production in caspase-1/11–KO mice treated with SAHA or levels of IL-1b in LPS-stimulated cells, and both stimuli actually vehicle was investigated. In line with the in vitro data, HDACi 8 HDACi PERMITS LPS-MEDIATED SECRETION OF BIOACTIVE IL-1b Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021

FIGURE 5. LPS/HDACi-induced IL-1b is bioactive and this pathway is blocked by the broad-spectrum caspase inhibitor Z-VAD-FMK. (A) Deter- mination of biological activity of IL-1b in cell culture supernatants of LPS/HDACi-stimulated WT and IL-1b–KO BMDC using a cell-based bioassay. BMDC were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM) for 14 h. Cell culture supernatants of the BMDC were taken, and Nob-1 cells were incubated in the cell culture supernatant, with or without rIL-1b, for 14 h. After incubation, total RNA was isolated, and expression of IL-2 mRNA was quantified by quantitative RT-PCR. (B) BMDC were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM) for 7 h. After incubation, cells were washed three times. Cells and lysates were immunoblotted with the anti–IL-1b Ab. BMDC were incubated with inhibitors targeting the caspase family (Z-VAD-FMK; 10 mM) (C), (Z-YVAD-FMK; 20 mM), caspase 3 (Z-DEVD-FMK; 20 mM), caspase 6 (Z-VEID-FMK; 20 mM) (E), or caspase 8 (Z-IETD-FMK; 50 and 100 mM) (F and G) and were stimulated with LPS (100 ng/ml), with or without SAHA (4 mM) for 7 h. Where indicated, cells were exposed to 5 mM ATP for the last 30 min. IL-1b was determined in supernatants by ELISA (C and E–G). (D) Cells and lysates were immunoblotted with the anti–IL-1b Ab. (H) BMDC were incubated with SAHA (4 mM), with or without LPS (100 ng/ml), as indicated. After 6 h, washed lysates (Lys) and cell culture supernatants (SN) were immunoblotted with Abs against caspase-8 and IL-1b.(I and J) BMDC were transfected with caspase-8 siRNA or control siRNA and stimulated for 6 h with SAHA (4 mM), with or without LPS (100 ng/ml), as indicated. Secretion of IL-1b was determined in supernatants by ELISA (I), and expression of caspase-8 mRNA was assessed by quantitative RT-PCR (J). (K) BMDC were primed with LPS (100 ng/ml). After 4 h, cells were treated with doxorubicin (doxo; 5 mM) or staurosporin (stau; 200 nM), with or without simultaneously added SAHA (4 mM) and/or LPS (100 ng/ml) as indicated. IL-1b concentration was determined in the supernatants by ELISA. All experiments were performed at least three times. Data are mean 6 SD. *p , 0.05, ***p , 0.001. L, LPS; S, SAHA. treatment of caspase-1/11–KO animals resulted in a significant provide in vitro and in vivo evidence for a novel HDAC-dependent increase in IL-1b levels in the colon (Fig. 6C). However, treatment pathway for IL-1b maturation that is independent of caspase-1/11 of WT or caspase-1/11–KO mice with SAHA did not cause a and that contributes to pathological inflammatory processes. significant alteration in the severity of DSS colitis, possibly as a result of the antagonistic effects of IL-1b upregulation and Discussion downregulation of proinflammatory cytokines, such as IL-6 In recent years, growing evidence demonstrated that HDACi have (Fig. 6B, 6F). Indeed, SAHA treatment reduced disease severity immunosuppressive and anti-inflammatory effects and may have (Fig. 6H) only in IL-1b–KO mice (Fig. 6D). As was observed with potential as new drugs for the treatment of acute or chronic in- WT mice (Fig. 6B), SAHA again reduced IL-6 levels on this flammatory diseases (1, 12, 41, 42). This is consistent with the genetic background (Fig. 6E). In conclusion, the data presented suppressed production of many proinflammatory mediators, such The Journal of Immunology 9 Downloaded from

FIGURE 6. HDACi treatment of mice with acute colitis increases intestinal IL-1b C57BL/6 WT mice (A, B, and F), caspase-1/11–KO mice (C and G), and IL-1b–KO mice (D, E, and H) were exposed to 3.5% DSS for 7 d. Mice were treated with SAHA (50 mg/kg body weight) or with vehicle once daily, http://www.jimmunol.org/ starting on the first day of the experiment. Effect of SAHA treatment on the mucosal IL-1b (A, C, and D) and IL-6 (B and E) cytokine concentration in colon in mice with acute colitis. (F–H) Disease activity index (DAI) of the mice during the course of the colitis. **p , 0.01. ns, not significant. as IL-12 and IL-6 (Figs. 1, 6) (15). In contrast, our finding that Recent evidence from other investigators also supports the HDAC inhibition, in combination with LPS, triggers pronounced existence of a caspase-1–independent, but caspase-8–dependent, IL-1b secretion was very surprising, given the potent proin- pro-IL-1b–maturation pathway. Maelfait et al. (48) described flammatory properties of this cytokine (3). Strikingly, this IL-1b caspase-8–dependent pro–IL-1b processing after prolonged secretion was caspase-8 dependent. The data provided indicate treatment of the cells with agonists of TLR3 or TLR4. They that HDAC11 is the most likely HDAC responsible for the ob- showed that recombinant caspase-8 cleaves pro–IL-1b at the same by guest on September 28, 2021 served effect, although a functional redundancy of different site as caspase-1, suggesting that caspase-8 may directly cleave HDAC cannot definitely be excluded. pro–IL-1b. Antonopoulos et al. (39) reported that treatment of Increased IL-1b production has been linked to a wide spectrum LPS-primed BMDC with the chemotherapeutic drugs doxorubicin of diseases (3, 4). Much attention has been paid to inflammasome- and staurosporine induced the secretion of bioactive IL-1b via and caspase-1–dependent IL-1b production. However, several caspase-8. In line with these data, we found that stimulation of a recent reports support the assumption that caspase-1–independent TLR induces pro–IL-1b expression, but activation of the TRIF IL-1b maturation is also of physiological relevance. Caspase-1– pathway is necessary for caspase-8–dependent IL-1b maturation. KO mice can produce considerable amounts of IL-1b in different Butyric acid, a physiologically occurring broad-range HDACi, disease models (arthritis, peritonitis) (5, 43); in the DSS colitis exhibits several additional functions in the intestine. It is the model, NLRP6-KO mice showed impaired caspase-1–dependent primary energy source for intestinal epithelial cells and, further, it is IL-18 cleavage, whereas IL-1b production was actually en- recognized by important receptors of the GPCR family (GPR41 and hanced (44). In addition to caspase-1, many proteases are known GPR43) (49). To overcome the limitation of pleiotropic effects of to be capable of cleaving pro–IL-1b. Most publications on inflam- butyric acid, SAHA was used to investigate the influence of HDACi masome-independent IL-1b activation focus on nonmacrophages/ on intestinal immune responses in a DSS colitis model. As proof of DC, and the majority of reports described extracellular activation principle for the in vivo relevance, treatment of mice with SAHA of pro–IL-1b (4, 7–9, 45). The results of the current study clearly during DSS-induced colitis caused a significant increase in IL-1b demonstrate caspase-1–independent, but caspase-8–dependent, IL- production in WT mice, as well as in caspase-1/11–KO mice 1b secretion by both murine and human DC and macrophages, (Fig. 6A, 6C). As expected, HDACi treatment of the animals de- underscoring the fact that these cells are able to activate IL-1b by creased other proinflammatory cytokines, such as IL-6 and IL- more than one mechanism. Interestingly, in PBMC, HDACi did not 12p35 (Fig. 6B, 6E, data not shown). This reverse regulation of promote IL-1b secretion, and it even reduced this response (Fig. 1H) proinflammatory cytokines likely explains the lack of a net effect of (32). However, this is in line with reports of constitutively activated SAHA in regulating pathology in the DSS colitis model. Consistent caspase-1 in PBMC due to increased levels of autosecreted ATP with this hypothesis, SAHA exerted therapeutic effects only in the (46), as well as with the results obtained by pretreatment of BMDC absence of IL-1b (Fig. 6H). In line with previous publications, IL- with ATP, which inhibited SAHA/LPS-induced IL-1b secretion 1b deficiency alone did not influence the time course of the DSS (data not shown). Differential effects of HDACi on systemic versus colitis (Supplemental Fig. 4). A comparison with the caspase-1/11– peripheral IL-1b production may enable the involvement of HDACi- KO experiments is not possible because different mouse strains induced IL-1b in the maintenance of local homeostasis, as hypoth- were used (Sv129 versus C57BL/6). The discrepancy between our esized for intestinal butyric acid (47), without inducing systemic data and the findings of Glauben et al. (50, 51), who described a inflammation. milder phenotype in SAHA-treated mice, probably can be explained 10 HDACi PERMITS LPS-MEDIATED SECRETION OF BIOACTIVE IL-1b by the milder experimental settings used in the current study. Be- 8. Scho¨nbeck, U., F. Mach, and P. Libby. 1998. Generation of biologically active b IL-1 beta by matrix metalloproteinases: a novel caspase-1-independent pathway cause IL-1 was implicated as a major regulator of intestinal ho- of IL-1 beta processing. J. Immunol. 161: 3340–3346. meostasis under steady-state conditions, it will be interesting to 9. Black, R. A., S. R. Kronheim, M. Cantrell, M. C. Deeley, C. J. March, investigate the role of butyrate in this setting. K. S. Prickett, J. Wignall, P. J. Conlon, D. Cosman, T. P. Hopp, et al. 1988. Generation of biologically active interleukin-1 beta by proteolytic cleavage of HDACi are considered promising candidates for the treatment of the inactive precursor. J. Biol. Chem. 263: 9437–9442. a variety of diseases. Beyond their established usage for the 10. Choudhary, C., C. Kumar, F. Gnad, M. L. Nielsen, M. Rehman, T. C. Walther, treatment of some malignancies, various HDACi have been ex- J. V. Olsen, and M. Mann. 2009. Lysine acetylation targets protein complexes and co-regulates major cellular functions. Science 325: 834–840. amined in a range of animal models of inflammatory diseases, 11. Barneda-Zahonero, B., and M. Parra. 2012. Histone deacetylases and cancer. including arthritis. Indeed, in an initial clinical trial of the HDACi Mol. Oncol. 6: 579–589. , therapeutic benefit was observed in a small cohort of 12. Roger, T., J. Lugrin, D. Le Roy, G. Goy, M. Mombelli, T. Koessler, X. C. Ding, A.-L. Chanson, M. K. Reymond, I. Miconnet, et al. 2011. patients with systemic-onset juvenile idiopathic arthritis (52). In inhibitors impair innate immune responses to Toll-like receptor agonists and to animal models of rheumatoid arthritis, MS-275, which was not infection. Blood 117: 1205–1217. b 13. Shakespear, M. R., M. A. Halili, K. M. Irvine, D. P. Fairlie, and M. J. Sweet. capable of inducing significant amounts of IL-1 (Fig. 3D, 3E), 2011. Histone deacetylases as regulators of inflammation and immunity. Trends displayed strong antirheumatic activities (53), whereas SAHA had Immunol. 32: 335–343. only a weak or no effect. A possible explanation for this obser- 14. Sweet, M. J., M. R. Shakespear, N. A. Kamal, and D. P. Fairlie. 2012. HDAC inhibitors: modulating leukocyte differentiation, survival, proliferation and in- vation is that the additional effect of anti-HDAC11 activity is flammation. Immunol. Cell Biol. 90: 14–22. responsible for local IL-1b release that can counteract the other- 15. Bode, K. A., K. Schroder, D. A. Hume, T. Ravasi, K. Heeg, M. J. Sweet, and wise anti-inflammatory effects of SAHA. Thus, future studies A. H. Dalpke. 2007. Histone deacetylase inhibitors decrease Toll-like receptor- mediated activation of proinflammatory gene expression by impairing tran- should address the roles of enhanced IL-1b secretion during scription factor recruitment. Immunology 122: 596–606. Downloaded from HDACi therapy in different inflammatory disease models. It is 16. Halili, M. A., M. R. Andrews, L. I. Labzin, K. Schroder, G. Matthias, C. Cao, likely that HDAC11 inhibition can have both desirable and un- E. Lovelace, R. C. Reid, G. T. Le, D. A. Hume, et al. 2010. Differential effects of selective HDAC inhibitors on macrophage inflammatory responses to the Toll- desirable effects on inflammatory responses, depending on the like receptor 4 agonist LPS. J. Leukoc. Biol. 87: 1103–1114. specific inflammatory disease model. Moreover, the potency of 17. Brogdon, J. L., Y. Xu, S. J. Szabo, S. An, F. Buxton, D. Cohen, and Q. Huang. HDACi in cancer therapy was attributed to their apoptosis- 2007. Histone deacetylase activities are required for innate immune cell control of Th1 but not Th2 effector cell function. Blood 109: 1123–1130. inducing properties, as well as their immune-modulating effects. 18. Adachi, O., T. Kawai, K. Takeda, M. Matsumoto, H. Tsutsui, M. Sakagami, http://www.jimmunol.org/ Thus, the development of HDAC11-specific inhibitors may pro- K. Nakanishi, and S. Akira. 1998. Targeted disruption of the MyD88 gene results in loss of IL-1- and IL-18-mediated function. Immunity 9: 143–150. vide opportunities for more efficacious agents for anticancer ap- 19. Mariathasan, S. 2007. ASC, Ipaf and Cryopyrin/Nalp3: bona fide intracellular plications. Conversely, HDACi that do not target HDAC11 may adapters of the caspase-1 inflammasome. Microbes Infect. 9: 664–671. provide an avenue for the development of anti-inflammatory 20. Kanneganti, T. D., N. Ozo¨ren, M. Body-Malapel, A. Amer, J. H. Park, L. Franchi, J. Whitfield, W. Barchet, M. Colonna, P. Vandenabeele, et al. 2006. agents with fewer undesirable side effects. Bacterial RNA and small antiviral compounds activate caspase-1 through In summary, our study identified a new and unexpected role for cryopyrin/Nalp3. Nature 440: 233–236. HDAC inhibition in promoting noncanonical caspase-8–dependent 21. Ebnet, K., M. Hausmann, F. Lehmann-Grube, A. Mullbacher,€ M. Kopf, M. Lamers, and M. M. Simon. 1995. Granzyme A-deficient mice retain potent IL-1b maturation. Human and mouse DC and macrophages are cell-mediated cytotoxicity. EMBO J. 14: 4230–4239. capable of producing bioactive IL-1b by a novel, inflammasome- 22. Heusel, J. W., R. L. Wesselschmidt, S. Shresta, J. H. Russell, and T. J. Ley. 1994. by guest on September 28, 2021 independent mechanism. Further elucidation of this pathway is Cytotoxic lymphocytes require granzyme B for the rapid induction of DNA fragmentation and apoptosis in allogeneic target cells. Cell 76: 977–987. likely to reveal key insights into inflammatory processes and has 23. Shornick, L. P., P. De Togni, S. Mariathasan, J. Goellner, J. Strauss-Schoenberger, major implications for the development of selective HDACi as R. W. Karr, T. A. Ferguson, and D. D. Chaplin. 1996. Mice deficient in IL-1beta manifest impaired contact hypersensitivity to trinitrochlorobenzone. J. Exp. Med. anti-inflammatory drugs. 183: 1427–1436. 24. Kuida, K., J. A. Lippke, G. Ku, M. W. Harding, D. J. Livingston, M. S. Su, and R. A. Flavell. 1995. Altered cytokine export and apoptosis in mice deficient in Acknowledgments interleukin-1 beta converting enzyme. Science 267: 2000–2003. We thank Selina Hassel and Rebecca Benz for excellent technical support. 25. Li, P., H. Allen, S. Banerjee, S. Franklin, L. Herzog, C. Johnston, J. McDowell, M. Paskind, L. Rodman, J. Salfeld, et al. 1995. Mice deficient in IL-1 beta- converting enzyme are defective in production of mature IL-1 beta and resis- Disclosures tant to endotoxic shock. Cell 80: 401–411. The authors have no financial conflicts of interest. 26. Halangk, W., M. M. Lerch, B. Brandt-Nedelev, W. Roth, M. Ruthenbuerger, T. Reinheckel, W. Domschke, H. Lippert, C. Peters, and J. Deussing. 2000. Role of cathepsin B in intracellular trypsinogen activation and the onset of acute pancreatitis. J. Clin. Invest. 106: 773–781. References 27. Govindarajan, N., P. Rao, S. Burkhardt, F. Sananbenesi, O. M. Schluter,€ 1. Newton, K., and V. M. Dixit. 2012. Signaling in innate immunity and inflam- F. Bradke, J. Lu, and A. Fischer. 2013. Reducing HDAC6 ameliorates cognitive mation. Cold Spring Harb. Perspect. Biol. 4: 1–19. deficits in a mouse model for Alzheimer’s disease. EMBO Mol. Med. 5: 52–63. 2. Klionsky, D. J., E. H. Baehrecke, J. H. Brumell, C. T. Chu, P. Codogno, A. M. 28. Hoebe, K., X. Du, J. Goode, N. Mann, and B. Beutler. 2003. Lps2: a new locus Cuervo, J. Debnath, V. Deretic, Z. Elazar, E.-L. Eskelinen, et al. 2011. A required for responses to lipopolysaccharide, revealed by germline mutagenesis comprehensive glossary of autophagy-related molecules and processes (2nd and phenotypic screening. J. Endotoxin Res. 9: 250–255. edition). Autophagy 7: 1273–1294. 29. Wo¨lfle, S. J., J. Strebovsky, H. Bartz, A. Sa¨hr, C. Arnold, C. Kaiser, 3. Dinarello, C. A. 2011. Interleukin-1 in the pathogenesis and treatment of in- A. H. Dalpke, and K. Heeg. 2011. PD-L1 expression on tolerogenic APCs is flammatory diseases. Blood 117: 3720–3732. controlled by STAT-3. Eur. J. Immunol. 41: 413–424. 4. Greten, F. R., M. C. Arkan, J. Bollrath, L.-C. Hsu, J. Goode, C. Miething, 30. Gearing, A. J., C. R. Bird, A. Bristow, S. Poole, and R. Thorpe. 1987. A simple S. I. Go¨ktuna, M. Neuenhahn, J. Fierer, S. Paxian, et al. 2007. NF-kappaB is a sensitive bioassay for interleukin-1 which is unresponsive to 10(3) U/ml of negative regulator of IL-1beta secretion as revealed by genetic and pharmaco- interleukin-2. J. Immunol. Methods 99: 7–11. logical inhibition of IKKbeta. Cell 130: 918–931. 31.Frick,J.S.,K.Fink,F.Kahl,M.J.Niemiec,M.Quitadamo,K.Schenk,and 5. Joosten, L. A., M. G. Netea, G. Fantuzzi, M. I. Koenders, M. M. Helsen, I. B. Autenrieth. 2007. Identification of commensal bacterial strains that modulate H. Sparrer, C. T. Pham, J. W. van der Meer, C. A. Dinarello, and W. B. van den Yersinia enterocolitica and dextran sodium sulfate-induced inflammatory responses: Berg. 2009. Inflammatory arthritis in caspase 1 gene-deficient mice: contribution implications for the development of probiotics. Infect. Immun. 75: 3490–3497. of proteinase 3 to caspase 1-independent production of bioactive interleukin-1b. 32. Carta, S., S. Tassi, C. Semino, G. Fossati, P. Mascagni, C. A. Dinarello, and Arthritis Rheum. 60: 3651–3662. A. Rubartelli. 2006. Histone deacetylase inhibitors prevent exocytosis of 6. Hazuda, D. J., J. Strickler, F. Kueppers, P. L. Simon, and P. R. Young. 1990. interleukin-1beta-containing secretory lysosomes: role of microtubules. Blood Processing of precursor interleukin 1 beta and inflammatory disease. J. Biol. 108: 1618–1626. Chem. 265: 6318–6322. 33. Beurel, E. 2011. HDAC6 regulates LPS-tolerance in astrocytes. PLoS One 6: 7. Irmler, M., S. Hertig, H. R. MacDonald, R. Sadoul, J. D. Becherer, A. Proudfoot, e25804. R. Solari, and J. Tschopp. 1995. Granzyme A is an interleukin 1 beta-converting 34. Akira, S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innate enzyme. J. Exp. Med. 181: 1917–1922. immunity. Cell 124: 783–801. The Journal of Immunology 11

35. Bradner, J. E., N. West, M. L. Grachan, E. F. Greenberg, S. J. Haggarty, (IL-1) activity and release of biologically active IL-1 from monocytes by T. Warnow, and R. Mazitschek. 2010. Chemical phylogenetics of histone trypsin and plasmin. J. Immunol. 136: 2883–2891. deacetylases. Nat. Chem. Biol. 6: 238–243. 46.Netea,M.G.,C.A.Nold-Petry,M.F.Nold,L.A.Joosten,B.Opitz,J.H.van 36. Cleophas, M. C., T. O. Cris¸an, H. Lemmers, H. Toenhake-Dijkstra, G. Fossati, T. der Meer, F. L. van de Veerdonk, G. Ferwerda, B. Heinhuis, I. Devesa, et al. L. Jansen, C. A. Dinarello, M. G. Netea, and L. A. Joosten. 2015. Suppression of 2009. Differential requirement for the activation of the inflammasome for monosodium urate crystal-induced cytokine production by butyrate is mediated processing and release of IL-1beta in monocytes and macrophages. Blood 113: by the inhibition of class I histone deacetylases. Ann. Rheum. Dis. annrheumdis– 2324–2335. 2014–206258. 47. Vinolo, M. A., H. G. Rodrigues, R. T. Nachbar, and R. Curi. 2011. Regulation of 37. Stehlik, C. 2009. Multiple interleukin-1beta-converting enzymes contribute to inflammation by short chain fatty acids. Nutrients 3: 858–876. inflammatory arthritis. Arthritis Rheum. 60: 3524–3530. 48. Maelfait, J., E. Vercammen, S. Janssens, P. Schotte, M. Haegman, S. Magez, and 38. Bossaller, L., P. I. Chiang, C. Schmidt-Lauber, S. Ganesan, W. J. Kaiser, R. Beyaert. 2008. Stimulation of Toll-like receptor 3 and 4 induces interleukin- V. A. Rathinam, E. S. Mocarski, D. Subramanian, D. R. Green, N. Silverman, 1beta maturation by caspase-8. J. Exp. Med. 205: 1967–1973. et al. 2012. Cutting edge: FAS (CD95) mediates noncanonical IL-1b and IL-18 49. Samuel, B. S., A. Shaito, T. Motoike, F. E. Rey, F. Backhed, J. K. Manchester, R. E. Hammer, S. C. Williams, J. Crowley, M. Yanagisawa, and J. I. Gordon. maturation via caspase-8 in an RIP3-independent manner. J. Immunol. 189: 2008. Effects of the gut microbiota on host adiposity are modulated by the short- 5508–5512. chain fatty-acid binding G protein-coupled receptor, Gpr41. Proc. Natl. Acad. 39. Antonopoulos, C., C. El Sanadi, W. J. Kaiser, E. S. Mocarski, and G. R. Dubyak. Sci. USA 105: 16767–16772. 2013. Proapoptotic chemotherapeutic drugs induce noncanonical processing and 50. Glauben, R., A. Batra, I. Fedke, M. Zeitz, H. A. Lehr, F. Leoni, P. Mascagni, release of IL-1b via caspase-8 in dendritic cells. J. Immunol. 191: 4789–4803. G. Fantuzzi, C. A. Dinarello, and B. Siegmund. 2006. Histone hyperacetylation 40. Cummings, J. H., E. W. Pomare, W. J. Branch, C. P. Naylor, and is associated with amelioration of experimental colitis in mice. J. Immunol. 176: G. T. Macfarlane. 1987. Short chain fatty acids in human large intestine, portal, 5015–5022. hepatic and venous blood. Gut 28: 1221–1227. 51. Glauben, R., A. Batra, T. Stroh, U. Erben, I. Fedke, H. A. Lehr, F. Leoni, 41. Bode, K. A., and A. H. Dalpke. 2011. HDAC inhibitors block innate immunity. P. Mascagni, C. A. Dinarello, M. Zeitz, and B. Siegmund. 2008. Histone Blood 117: 1102–1103. deacetylases: novel targets for prevention of colitis-associated cancer in mice. 42. Halili, M. A., M. R. Andrews, M. J. Sweet, and D. P. Fairlie. 2009. Histone Gut 57: 613–622. deacetylase inhibitors in inflammatory disease. Curr. Top. Med. Chem. 9: 309–319. 52. Vojinovic, J., N. Damjanov, C. D’Urzo, A. Furlan, G. Susic, S. Pasic, N. Iagaru, Downloaded from 43. Guma, M., L. Ronacher, R. Liu-Bryan, S. Takai, M. Karin, and M. Corr. 2009. M. Stefan, and C. A. Dinarello. 2011. Safety and efficacy of an oral histone Caspase 1-independent activation of interleukin-1b in neutrophil-predominant deacetylase inhibitor in systemic-onset juvenile idiopathic arthritis. Arthritis inflammation. Arthritis Rheum. 60: 3642–3650. Rheum. 63: 1452–1458. 44. Chen, G. Y., M. Liu, F. Wang, J. Bertin, and G. Nu´n˜ez. 2011. A functional role for 53. Lin, H.-S., C.-Y. Hu, H.-Y. Chan, Y.-Y. Liew, H.-P. Huang, L. Lepescheux, Nlrp6 in intestinal inflammation and tumorigenesis. J. Immunol. 186: 7187–7194. E. Bastianelli, R. Baron, G. Rawadi, and P. Cle´ment-Lacroix. 2007. Anti- 45. Matsushima, K., M. Taguchi, E. J. Kovacs, H. A. Young, and J. J. Oppenheim. rheumatic activities of histone deacetylase (HDAC) inhibitors in vivo in 1986. Intracellular localization of human monocyte associated interleukin 1 collagen-induced arthritis in rodents. Br. J. Pharmacol. 150: 862–872. http://www.jimmunol.org/ by guest on September 28, 2021