Interleukin-33 Is Biologically Active Independently of Caspase-1 Cleavage^*

Materials

Cell culture reagents were obtained from Invitrogen. rmIL-33 and rhIL-33 produced in Escherichia coli were obtained from Alexis Corp. (Lausen, Switzerland). Soluble (s) ST2-Fc fusion protein was obtained from Amgen (Thousand Oaks, CA). T1/ST2 knock-out BALB/c mice were generously provided by Dr. A. McKenzie (Medical Research Council, Cambridge, UK).

Cloning of Mouse and Human Pro-IL-33 and Pro-IL-1β Expression Vectors

An IMAGE expressed sequence tag clone containing the cDNA encoding the full-length mouse pro-IL-33 (GenBank^TM accession number BC003847) in the pCMV-Sport6 vector was used as expression vector for full-length mouse pro-IL-33 in 293T human embryonic kidney cells. Mutagenesis of Asp to Ala in mouse D88A and D106A pro-IL-33 mutants was performed using the QuikChange kit (Stratagene, La Jolla, CA) and appropriate primer pairs, according to the manufacturer's instructions. The cDNA corresponding to the full-length human pro-IL-33 (GenBank^TM accession number AY905581) was amplified by reverse transcription-PCR using total RNA extracted from cultured rheumatoid arthritis synovial fibroblasts (forward primer, 5′-cgggatccatgaagcctaaaatgaagtattc-3′; and reverse primer, 5′-gctctagactaagtttcagagagcttaaac-3′) and cloned into the pCAGGSβ vector (10) for expression in 293T cells. For expression in THP-1 cells, full-length human pro-IL-33 cDNA was cloned into the pWPI lentiviral vector (11). Viral particles were produced in 293T cells using a standard protocol as described (12). The cDNA corresponding to full-length human pro-IL-1β (GenBank^TM accession number NM_000576) was amplified by reverse transcription-PCR using RNA extracted from phorbol 12-myristate 13-acetate (PMA)-stimulated THP-1 cells (forward primer, 5′-cgggatccggctgctctgggattctcttcag-3′; and reverse primer, 5′-gcctcgaggggtacagctctctttaggaag-3′) and cloned into the pcDNA3.1⁺ vector (Invitrogen) for expression in 293T cells. The sequence of the entire coding region was verified before use for all constructs.

Cell Culture, Transfection, and Lentiviral Infection

Human embryonic kidney 293T cells were cultured in Dulbecco's modified Eagle's medium (4.5 g/liter glucose) supplemented with l-glutamine, streptomycin, penicillin, and 10% fetal calf serum at 37 °C in a humidified atmosphere containing 5% CO₂. For transfection, 293T cells were plated at a density of 100,000 cells/ml and transfected 24 h after seeding with 1 μg/ml pro-IL-33 or pro-IL-1β expression vector, or empty vectors as a control, by calcium phosphate precipitation (13). THP-1 cells were cultured in RPMI 1640 medium (4.5 g/liter glucose) supplemented with 10 mm Hepes, 1 mm sodium pyruvate, 5 × 10⁻⁵ m β-mercaptoethanol, streptomycin, penicillin, and 10% fetal calf serum. For viral infection, THP-1 cells were plated at a density of 250,000 cells/well and infected with freshly produced lentiviruses encoding hIL-33 and green fluorescent protein (GFP), or GFP alone as a control, using a double infection protocol (12). P815 cells were cultured in Dulbecco's modified Eagle's medium (4.5 g/liter glucose) supplemented with 6% fetal calf serum, 10 mm Hepes, 10 mm minimum Eagle's medium nonessential amino acids, penicillin, and streptomycin. Primary mouse bone marrow-derived mast cells (BMMC) were generated as described (14) from wild-type and T1/ST2 knock-out BALB/c mice (15) and used after 10 weeks of culture.

In Vitro Protein Synthesis

Wild-type and D88A and D106A mutant mouse pro-IL-33 as well as human pro-IL-33 and pro-IL-1β were synthesized in vitro using the PURExpress in vitro protein synthesis kit (New England BioLabs, Ipswich, MA). This kit contains a cell-free coupled transcription/translation system reconstituted from purified components and is thus essentially free of any contaminant proteases. To generate templates for the in vitro protein synthesis, the various cDNAs were amplified from the above described expression vectors using appropriate primers to insert a T7 promoter and an E. coli ribosome entry site upstream of the gene-specific sequence, according to the manufacturer's instructions. The sequence of all PCR products was verified before use. Protein synthesis was carried out using 250 ng of DNA template in a final volume of 25 μl. Reactions without cDNA were set up as negative controls.

In Vitro Caspase-1 Assay

Transfected 293T cells were used 72 h after transfection. Cells were lysed in caspase assay buffer (10 mm Hepes, pH 7.5, 0.1% CHAPS, 2 mm EDTA, 10% glycerol, 10 mm dithiothreitol) by brief sonication on ice and stored at −80° C until used. Cell lysates (10 μl) were incubated for 6 h at 37° C without or with 2 units of recombinant active human caspase-1 (ALX-201-056, Alexis Corp.). In vitro synthesized proteins (0.7 μl) were incubated in 20 μl of caspase buffer for 6 h at 37 °C without or with 2 units of recombinant active human caspase-1. Reactions were stopped by the addition of Laemmli buffer, and samples were resolved on SDS-PAGE and analyzed by Western blotting.

Caspase-1 Activation in THP-1 Cells

Lentivirally transduced THP-1 cells were stimulated for 3 h with 0.5 μm PMA in complete RPMI 1640 medium and left to adhere overnight. Cells were then switched to Opti-MEM (Invitrogen) and stimulated or not with 50 μg/ml lipopolysaccharide (LPS) for 6 h, as described previously (16). Culture supernatants were removed, and cell lysates were prepared by freeze-thawing after addition of fresh Opti-MEM. Supernatant and cell lysate proteins were concentrated by methanol precipitation, resuspended in Laemmli buffer, resolved by SDS-PAGE, and analyzed by Western blotting.

Western Blotting

Samples were fractionated by SDS-PAGE and transferred to a Porablot membrane (Macherey-Nagel, Dueren, Germany). The membrane was blocked in 0.05% Tris-buffered saline, Tween 20, containing 5% bovine serum albumin or 5% Blotto (Santa Cruz Biotechnology, Santa Cruz, CA), and immunoblotting was performed with one of the following antibodies, as indicated: biotinylated monoclonal anti-IL-33 antibody (Nessy-1, 1:1,000, Alexis Corp.); biotinylated polyclonal anti-mIL-33 antibody (BAF3626, 1:2,000), polyclonal anti-hIL-33 antibody (AF3625, 1:1,000), and monoclonal anti-hIL-1β (MAB201, 1:1,000, R&D Systems); or the D116 antibody specific for mature 17-kDa hIL-1β (1:1,000, Cell Signaling Technology, Danvers, MA). Immunoreactive bands were visualized by ECL (Amersham Biosciences) using appropriate horseradish peroxidase-coupled secondary reagents.

Immunofluorescence

Lentivirally transduced THP-1 cells were seeded in chamber slides, primed with 0.5 μm PMA, and stimulated or not with 50 μg/ml LPS as described above. The cells were fixed (phosphate-buffered saline, 4% paraformaldehyde; 10 min at room temperature), permeabilized (phosphate-buffered saline, 0.1% Triton X-100; 10 min), and blocked (phosphate-buffered saline, 1% bovine serum albumin, 10% normal goat serum; 1 h) before staining with a monoclonal anti-hIL-33 antibody (ALX-804-726, 1:100, Alexis Corp.) and detection with a Texas Red-labeled anti-mouse IgG secondary antibody. To assess staining specificity, negative controls were performed in the absence of the primary antibody. Slides were mounted using a 4′,6-diamidino-2-phenylindole-containing mounting medium (Vector Laboratories), and images were acquired on an LSM510 Meta confocal microscope (Carl Zeiss AG, Feldbach, Switzerland).

Purification of Mouse Pro-IL-33

293T cells were seeded in 10-cm Petri dishes and transiently transfected with the mouse pro-IL-33 expression vector. Cells were harvested 72 h after transfection in TNT (50 mm Tris, pH 7.4, 150 mm NaCl, 0.5% Triton X-100) supplemented with a protease inhibitor mixture (Roche Diagnostics) and lysed by brief sonication on ice. Alternatively, transfected cells were used to prepare nuclear extracts following standard procedures (17). Briefly, cells were lysed in sucrose buffer on ice, and nuclei were pelleted, resuspended in low salt buffer, and lysed by the addition of high salt buffer. Mouse pro-IL-33 was affinity-purified from TNT total cell extracts or from nuclear extracts using mouse sST2-Fc and the Seize X protein G immunoprecipitation kit (Pierce) and eluted in 100–200-μl fractions using the acid elution buffer (pH 2.8) provided in the kit. Eluted fractions were neutralized using 5–10 μl of Tris (1 m, pH 9.5) and used for Western blotting (10 μl/lane) or, after the addition of 0.1% bovine serum albumin, stored at −80° C until they were used to stimulate P815 cells or BMMC.

Bioactivity of Purified and in Vitro Translated Mouse Pro-IL-33

Bioactivity of purified mouse pro-IL-33 or of in vitro synthesized wild-type, D88A, and D106A mouse pro-IL-33 was assessed using P815 mastocytoma cells and BMMC (14). P815 cells were plated in 96-well plates in 200 μl of complete culture medium at a density of 20,000 cells/well and treated with the indicated concentrations of rmIL-33 or with the indicated dilutions of in vitro synthesis products for 24 h before culture supernatants were harvested. To assess specificity of the observed effects, rIL-33 or in vitro synthesized proteins were preincubated with sST2-Fc (50 μg/ml) for 15 min at 37 °C before addition to the cells. Alternatively, BMMC generated from wild-type or T1/ST2 knock-out mice were incubated without or with purified pro-IL-33 (10 μl of 200 μl of total eluate), rmIL-33 (1 ng/ml), or LPS (1 μg/ml) for 24 h before culture supernatants were harvested. Levels of IL-6 in P815 and BMMC culture supernatants were assessed using an enzyme-linked immunosorbent assay DuoSet kit (R&D Systems).

Statistical Analysis

The significance of differences was assessed by analysis of variance. A difference between experimental groups was considered significant when the p value was <0.05.

Generation of a 20-kDa Fragment of IL-33

Human pro-IL-33, like pro-IL-1β, was reported to be cleaved by caspase-1 in vitro to generate an 18-kDa C-terminal fragment (1). However, the proposed caspase-1 cleavage site is poorly conserved between species (5). Therefore, we further investigated caspase-1 cleavage of mouse and human pro-IL-33. Mouse pro-IL-33 was overexpressed in 293T cells, and total cell lysates were incubated with recombinant human active caspase-1 in vitro. Western blot analysis revealed the presence of a major band corresponding to the 30-kDa full-length form of pro-IL-33, even after incubation with caspase-1 (Fig. 1A). In addition, we detected a less intense band at 20 kDa, the abundance of which was increased in the presence of active caspase-1. No band corresponding to the size of rmIL-33 was detected either in the absence or presence of caspase-1. Similar results were obtained also upon incubation of the cell lysates with recombinant active mouse caspase-1 and using two different anti-mIL-33 antibodies (BAF3626 and Nessy-1) for detection (Fig. 1A and data not shown).

In an attempt to identify which of the two N-terminal Asp residues might represent a potential caspase-1 cleavage site in mouse pro-IL-33, we generated two mutant forms of mouse pro-IL-33, D88A and D106A, in which either Asp⁸⁸ or Asp¹⁰⁶ was replaced by Ala. Surprisingly, neither of these two mutations interfered with the generation of the 20 kDa band, which was detectable already in untreated cell lysates and slightly enhanced in the presence of caspase-1 (Fig. 1B).

Similarly, lysates of 293T cells overexpressing human pro-IL-33 contained essentially the 30-kDa full-length form of pro-IL-33, even after incubation with caspase-1, as well as a minor band at 20 kDa (Fig. 1C). Again, similar results were obtained using two different anti-hIL-33 antibodies (AF3625 and ALX-804-726) for detection (Fig. 1C and data not shown).

For both mIL-33 and hIL-33, the observed 20-kDa fragment was of higher molecular mass than 18-kDa rIL-33. This 20 kDa band was generally detectable already before caspase-1 treatment, although its abundance was variable in different cell lysate preparations (Fig. 1 and data not shown). However, the 30-kDa full-length pro-IL-33 remained the most abundant form detected in all preparations, even in the presence of caspase-1. Finally, lysates of 293T cells overexpressing pro-IL-1β were used as a positive control for efficient caspase-1 activity in our experimental conditions, and indeed, pro-IL-1β was cleaved efficiently into its 17-kDa mature form (Fig. 1D).

IL-33 Is Not a Direct Substrate for Caspase-1

The presence of the 20 kDa band in untreated cell lysates as well as in cells transfected with mutant pro-IL-33 suggested that generation of this IL-33 fragment is not mediated directly by caspase-1. In fact, 293T cell lysates contain a number of proteases that might be involved in IL-33 processing. We thus sought to assay caspase-1 cleavage of pro-IL-33 in a protease-free system. To this end, we synthesized pro-IL-33 using an in vitro transcription/translation system reconstituted from purified and recombinant components, which is essentially protease-free. Interestingly, when incubating in vitro synthesized mouse (Fig. 2A) or human (Fig. 2B) pro-IL-33 with recombinant human active caspase-1, we could not observe any caspase-1-dependent cleavage of pro-IL-33, which was recovered exclusively in its 30-kDa full-length form, even after 6 h of incubation. In contrast, using the same experimental conditions, human pro-IL-1β was totally cleaved to generate 17-kDa mature IL-1β (Fig. 2C). Taken together, these observations demonstrate that, in contrast to pro-IL-1β, pro-IL-33 is not a direct substrate for caspase-1.

IL-33 Is Released in Its Precursor Form after Activation of Endogenous Caspase-1 in THP-1 Cells

We next investigated the effect of caspase-1 activation on pro-IL-33 processing and secretion in living cells. For these experiments, we used THP-1 cells because caspase-1 activation and IL-1β secretion have been described extensively in these cells, in response to PMA and LPS, in well established experimental conditions (16). In our hands, THP-1 cells did not express endogenous pro-IL-33 as assessed by reverse transcription-PCR and Western blotting (Fig. 3B and data not shown). We thus used a lentiviral vector to express human pro-IL-33 in these cells. Following lentiviral transduction, expression of pro-IL-33 was observed essentially in the nuclei of THP-1 cells, with and without LPS stimulation (Fig. 3A and data not shown). We monitored potential pro-IL-33 cleavage and secretion upon activation of endogenous caspase-1 in response to LPS in PMA-primed THP-1 cells (16). Cleavage and secretion of endogenous IL-1β were monitored in parallel in the same samples as a positive control. Western blot analysis revealed the presence of 30-kDa full-length pro-IL-33 as well as the 20-kDa fragment in cell lysates of THP-1 cells transduced with the IL-33 expression vector (Fig. 3B). The abundance of the 20-kDa fragment was not enhanced in LPS-treated cells. In addition, no band corresponding to the size of rhIL-33 was detected either in the absence or presence of LPS. Interestingly, we observed release of full-length pro-IL-33 into culture supernatants specifically in LPS-stimulated THP-1 cells. LPS-induced secretion of mature 17-kDa IL-1β, as an indication of effective caspase-1 activation, was easily observed in both IL-33-expressing and control THP-1 cells (Fig. 3C).

Pro-IL-33 Exerts T1/ST2-dependent Biological Effects on Mast Cells

Because we observed secretion of IL-33 in its pro form, we next investigated whether full-length pro-IL-33, like pro-IL-1α, might be able to activate its cognate cell surface receptor. We thus purified mouse pro-IL-33 from transfected 293T total cell lysates or nuclear extracts using ST2-Fc/protein G beads. Although pro-IL-33 readily bound to these beads, it proved very difficult to elute efficiently using elution conditions compatible with subsequent bioactivity tests, such as elution using a low pH (2.8) buffer (Fig. 4A, left panel) or, in contrast, a high pH (11.5) triethanolamine buffer (data not shown). Efficient recovery of pro-IL-33 could only be achieved using stringent denaturing conditions, as illustrated by elution in Laemmli buffer (Fig. 4A, right panel). In fact, control experiments indicated that pro-IL-33 exhibited considerable background binding to protein G beads even in the absence of ST2-Fc (data not shown), and this “stickiness” may thus account for our difficulties in recovering bound pro-IL-33. In addition, this observation indicates that binding of pro-IL-33 to ST2-Fc/protein G beads cannot be inferred to demonstrate specific binding of pro-IL-33 to ST2.

Despite this lack of specificity, we used binding of pro-IL-33 to ST2-Fc/protein G beads followed by acid elution as a tool to prepare protein fractions enriched in pro-IL-33. Full-length 30-kDa pro-IL-33 was the only form of IL-33 detectable in ST2-Fc/protein G beads eluates (Fig. 4A, right panel). Addition of eluted fractions to P815 mastocytoma cells induced IL-6 secretion, and this effect was prevented in the presence of sST2-Fc (data not shown). Consistently, eluted pro-IL-33 induced IL-6 secretion in wild-type BMMC but not in ST2 knock-out BMMC (Fig. 4B), further suggesting that pro-IL-33 exerts T1/ST2-dependent biological effects in mast cells. The responsiveness to LPS was similar in wild-type and ST2 knock-out BMMC, indicating that there is no intrinsic defect in the induction of IL-6 secretion by proinflammatory stimuli in ST2 knock-out cells.

In view of the difficulties encountered with the purification of pro-IL-33 from cell lysates, we used in vitro synthesized proteins as an alternative source of pro-IL-33 to characterize its biological effects further. In vitro synthesized mouse pro-IL-33 dose-dependently induced IL-6 production in P815 cells, and this effect was inhibited in the presence of sST2-Fc (Fig. 5A). The two mutant forms of mouse pro-IL-33, D88A and D106A, similarly induced IL-6 production in P815 cells, and their effect was also inhibited by sST2-Fc (Fig. 5B).

Finally, we investigated potential in situ processing of pro-IL-33 to shorter forms in the culture medium of P815 cells. These cells do not express caspase-1, as assessed by reverse transcription-PCR (data not shown), but we hypothesized that, after addition to the cultures, pro-IL-33 might be processed by other secreted or cell surface proteases. Indeed, we could detect processing of pro-IL-33 to shorter forms, in particular 28-, 20-, and 18-kDa fragments (Fig. 5C), although, after a 24-h incubation with P815 cells, the relative abundance of full-length pro-IL-33 and these shorter fragments was variable between experiments (Fig. 5C and data not shown). The presence of shorter forms was also observed when P815 cells were incubated with the D88A and D106A mutant forms of pro-IL-33 (data not shown). It thus remains to be determined which forms of IL-33 indeed correspond to biologically active molecules.