M-CSF induces the expression of a membrane-bound form of IL-18 in a subset of human monocytes differentiating in vitro toward macrophages

Authors

  • Francesca Bellora,

    1. Dipartimento di Medicina Sperimentale, Università degli Studi di Genova, Genova, Italy
    Search for more papers by this author
  • Roberta Castriconi,

    1. Dipartimento di Medicina Sperimentale, Università degli Studi di Genova, Genova, Italy
    2. Centro di Eccellenza per le Ricerche Biomediche, Università degli Studi di Genova, Genova, Italy
    Search for more papers by this author
  • Andrea Doni,

    1. Istituto Clinico Humanitas IRCCS, Rozzano, Milano, Italy
    2. Dipartimento di Medicina Traslazionale, Università degli Studi di Milano, Milano, Italy
    Search for more papers by this author
  • Claudia Cantoni,

    1. Dipartimento di Medicina Sperimentale, Università degli Studi di Genova, Genova, Italy
    2. Centro di Eccellenza per le Ricerche Biomediche, Università degli Studi di Genova, Genova, Italy
    3. Istituto Giannina Gaslini, Genova, Italy
    Search for more papers by this author
  • Lorenzo Moretta,

    1. Istituto Giannina Gaslini, Genova, Italy
    Search for more papers by this author
  • Alberto Mantovani,

    1. Istituto Clinico Humanitas IRCCS, Rozzano, Milano, Italy
    2. Dipartimento di Medicina Traslazionale, Università degli Studi di Milano, Milano, Italy
    Search for more papers by this author
  • Alessandro Moretta,

    1. Dipartimento di Medicina Sperimentale, Università degli Studi di Genova, Genova, Italy
    2. Centro di Eccellenza per le Ricerche Biomediche, Università degli Studi di Genova, Genova, Italy
    Search for more papers by this author
    • These authors contributed equally to this study.

  • Cristina Bottino

    Corresponding author
    1. Istituto Giannina Gaslini, Genova, Italy
    • Dipartimento di Medicina Sperimentale, Università degli Studi di Genova, Genova, Italy
    Search for more papers by this author
    • These authors contributed equally to this study.


Full Correspondence: Dr. Cristina Bottino, Instituto Giannina Gaslini, L. go G. Gaslini 5, 16147 Genova, Italy

Fax: +39-010-3537576

e-mail: Cristina.Bottino@unige.it

See accompanying Commentary:

http://dx.doi.org/10.1002/eji.201142635

Abstract

IL-18 is a proinflammatory cytokine belonging to the “IL-1 family” that has been shown to play a prominent role in the induction of type 1 immune responses. Here, we show that M-CSFinduces the expression of a membrane-bound form of IL-18 (mIL-18) in a subset of human blood monocytes differentiating toward macrophages. While monocytes, DC, and GM-CSF-treated monocytes did not express mIL-18, its expression was detected in approximately 30–40% of M-CSF-primed macrophages differentiating from both CD16 and CD16+ monocytes. Treatment with the caspase-1 inhibitor significantly reduced mIL-18 expression suggesting the requirement of an assembled inflammasome for IL-18 surface expression. Polarization toward M2 did not modify mIL-18 expression. On the contrary, LPS stimulation of both M0 and M2 (mIL-18+) macrophages induced shedding of mIL-18, which was likely mediated by the activation of cellular protease(s). Importantly, the soluble form IL-18 (sIL-18) induced in autologous resting NK cells both the expression of CCR7 and the production of high amounts of IFN-γ, which was virtually abrogated by Ab-mediated neutralization of sIL-18. Overall our data shed new light on the cells and mechanisms leading to the release of sIL-18, the major IFN-γ-inducing factor in both physiological and pathological immune responses.

Introduction

Macrophages produce IL-1 (IL-1α and IL-1β) and IL-18, proinflammatory cytokines that have been generally assigned to a larger group of cytokines termed “IL-1 family” since they present related receptor structures and utilize similar signal transduction pathways [[1, 2]]. IL-1α is synthesized as pro-IL-1α, a biologically active precursor that is constitutively present in cells and can be cleaved into a mature cytokine by calpain, a membrane-bound cysteine protease. IL-1α is not readily secreted by cells and remains either bound to plasma membrane [[3]] or inside the cells where it acts as an autocrine growth factor. IL-1β, the prototype of primary proinflammatory cytokines is not present in cells under steady state conditions. Its production is highly regulated and requires a first signal-inducing de novo transcription of the gene and accumulation of the inactive precursor cytokine (pro-IL-1β) and a second signal-driving inflammasome assembly [[4]] and activation of caspase-1 (also known as IL-1β-converting enzyme, ICE), which converts pro-IL-1β into the IL-1β bioactive mature cytokine. These signals are provided by pathogen-associated molecular patterns (PAMPs) that act through specific receptors such as Toll-like receptors (TLRs), cytokines such as IL-33, IL-1β itself, and IL-1α, which is released passively upon cell injury, danger-associated molecular patterns (DAMPs) such as urate crystals or pathogenic dusts and extracellular ATP, which after binding to P2 receptors (P2R) induces a rapid exit of potassium from the cell. Interestingly, also a caspase-1 independent type of IL-1β activation has been described. Thus, in different models of sterile inflammation, macrophages and monocytes would be able to release pro-IL-1β that is cleaved by proteinase-3 produced by neutrophils [[1, 5]].

IL-18 originally identified as an IFN-γ-inducing factor (IGIF), has been shown to play a prominent role in the induction of type 1 immune responses [[1, 6, 7]]. Although belonging to the same cytokine family, IL-18 and IL-1β display both similarities and important differences [[1, 2]]. Similar to IL-1β, IL-18 is synthesized as a biologically inactive precursor peptide (pro-IL-18) that is subsequently cleaved by caspase-1. Unlike IL-1β, however, PAMPs such as LPS while inducing IL-18 release have little impact on IL-18 gene transcription [[2]].

Monocytes are a considerable reservoir of precursors for both macrophages and dendritic cells (DCs) [[8]]. During inflammatory processes, monocytes can be recruited into tissues and skewed toward macrophages by the macrophage colony-stimulating factor (M-CSF) [[9]]. Subsequently, the tissue microenvironment polarizes macrophages toward either M1 or M2, which display different surface phenotype, cytokine/chemokine gene expression profiles, and functional properties [[10]]. In general, M1-polarized macrophages present immunostimulatory Th1-oriented properties and antitumoral activity while M2 cells display immunomodulatory capacity, suppress Th1 adaptive immunity, and support tumor growth [[11]].

Recently, we described that unpolarized (M0) and M2-polarized macrophages following capture of microbial products polarize toward M1 and induce strong activation of autologous resting NK cells [[7]]. In particular, we showed that M0 and M2 cells expressed a membrane-bound form of IL-18 (mIL-18) that upon TLR engagement by LPS or BCG was released as soluble IL-18 (sIL-18). sIL-18 by acting in close cell-to-cell contacts induced in resting NK cells the expression of CCR7 and the release of large amounts of IFN-γ.

The present study was designed to characterize the cells expressing mIL-18 and to analyze the mechanisms allowing IL-18 expression at the cell surface as well as its release upon stimulation.

Results

Membrane-bound IL-18 is selectively expressed by M0 and M2 macrophages

We analyzed by flow cytometry the expression of mIL-18 in monocytes either freshly purified from peripheral blood mononuclear cells (PBMC) or after differentiation toward macrophages (with M-CSF)(M0) or DC (with GM-CSF and IL-4) (Fig. 1A and B). Monocytes, immature DCs (iDCs) as well as mature DCs (mDCs) did not express mIL-18 or release sIL-18. On the contrary, mIL-18 was present in approximately 30–40% of M0 cells. During M2 polarization, macrophages maintained the expression of mIL-18, while LPS stimulation of both M0 and M2 cells resulted in disappearance of mIL-18, which was paralleled by the release of sIL-18 in the supernatant. M1 cells further challenged with LPS, that is, “endotoxin tolerant” (ET) macrophages [[12, 13]], did not present mIL-18 or released detectable amounts of sIL-18 [[7]].

Figure 1.

mIL-18 is selectively expressed and released by macrophages. (A) Representative flow cytometric analysis of mIL-18 expression in monocytes, unpolarized (M0), M1- or M2-polarized macrophages, immature DCs (iDCs), and mature DCs (mDCs). Mean fluorescence intensity (MFI) and percentage of positive cells are indicated. (B) The supernatants of the indicated cell types were analyzed by ELISA for the presence of sIL-18. Data shown are pooled from nine independent experiments. Mean, 95% confidence intervals, and significance (Wilcoxon–Mann–Whiteny p-value test) are shown. ***p < 0.001 (C) During the differentiation from monocytes to macrophages, cells were analyzed daily for the expression of mIL-18 by flow cytometry. Squares and triangles indicate the MFI and the percentage (%) of positive cells, respectively. Data pooled from nine independent experiments. Mean, 95% confidence intervals, and significance (Wilcoxon–Mann–Whiteny p-value test) are shown. *p < 0.05, **p < 0.01.

To determine the kinetic of expression of mIL-18, flow cyto-metry was performed daily during the in vitro differentiation of monocytes to macrophages (Fig. 1C and Supporting Information Fig. 1A). Although a certain degree of variability existed among the various donors analyzed, mIL-18 appeared at the cell surface on average at day 2 and the surface density reached the highest level at the end of the differentiation process (day 7).

Similar results were obtained when cells were analyzed by confocal microscopy. As shown in Fig. 2A and Supporting Information Fig. 1B, IL-18 expression was detected at the cell surface in approximately 20% of nonpermeabilized macrophages in the course of differentiation process. Monocytes and iDCs as well as mDCs resulted negative. The mIL-18 expression was not significantly modified when macrophage membranes were permeabilized with Triton-X100, indicating that its expression was predominantly restricted to cell membrane (Fig. 2B).

Figure 2.

Analysis by confocal microscopy of mIL-18 expression. Representative fluorescence and differential interference contrast (DIC) images of mIL-18 expression in macrophages were taken on the indicated days. (A) IL-18 expression (green) was determined on nonpermeabilized monocytes and differentiating macrophages. Nuclei are white. Scale bar, 40 μm. (B)IL-18 expression (green) was determined on nonpermeabilized (left) and permeabilized (right) M0 macrophages. Scale bar, 80 μm. Data shown are representative of five independent experiments.

In search of markers identifying the mIL-18+ cell subset, macrophages were analyzed for the expression of different molecules present in M0 and/or M2 cells. As detected by confocal microscopy, M0 cells expressed the Folate receptor β (FRβ), stress-responsive heme-detoxification enzyme Heme Oxygenase-1 (HO-1), and CD209 (DC-SIGN), whose densities were strongly upregulated upon M2 polarization [[14-16]]. As shown in Supporting Information Fig. 1C, the expression of these markers did not coincide with that of mIL-18. Moreover, flow cytometry demonstrated that also the expression of surface molecules such as CD14, CD204, NTB-A, CD163 (Fig. 3A), CD80, CD206, and CD68 [[7]] was not restricted to the mIL-18+ cell subset.

Figure 3.

M-CSF but not GM-CSF induces mIL-18 expression. (A) Monocytes were cultured for 7 days in the presence of GM-CSF. Cells were analyzed by flow cytometry for the expression of the indicated surface molecules. M0s or iDCs obtained from the same donor were used for comparison. Values inside each histogram indicate the MFI and percentage of positive cells. Data shown are representative of three independent experiments. (B) Monocytes were treated with M-CSF or GM-CSF for 7 days followed, or not, by 18-h stimulation with LPS. The supernatants were analyzed by ELISA for the presence of sIL-18. Data pooled from three independent experiments. Mean, 95% confidence intervals, and significance (Wilcoxon–Mann–Whiteny p-value test) are shown. **p < 0.01.

M-CSF but not GM-CSF induces the expression of mIL-18 in monocytes

It has been suggested that macrophages can arise from monocytes cultured in the presence of either M-CSF or GM-CSF [[9]]. To clarify the role played by these factors in the expression of mIL-18, purified monocytes were cultured for 7 days either in the presence of M-CSF (M0) or GM-CSF and analyzed by flow cytometry (Fig. 3A). After GM-CSF treatment, differentiating monocytes displayed a surface phenotype different from that observed in M0 macrophages. In particular, GM-CSF-treated cells lacked CD163, expressed low levels of CD204 and NK-T-B-antigen (NTB-A, CD352), and were characterized by substantial expression of CD1a, a marker typical of DCs. Importantly, unlike M0 macrophages, GM-CSF-treated monocytes lacked mIL-18 (Fig. 3A) and did not acquire its expression after further treatment with M-CSF (Supporting Information Fig. 2A). In this context, M-CSF-mediated induction of mIL-18 correlated with the expression of the specific receptor. Indeed, M-CSF acted on monocytes, which expressed the M-CSF receptor but not on M-CSFR negative cells such as DCs and GM-CSF-treated monocytes (Supporting Information Fig. 2A).

In order to investigate the possible effect of M-CSF at the level of IL-18 gene transcription, we performed real-time PCR on monocytes treated with M-CSF at different time points. As shown in Supporting Information Fig. 2B, M-CSF treatment did not induce IL-18 mRNA increase. Monocytes incubated with M-CSF from 6 to 48 h expressed even lower levels of IL-18 mRNA as compared with those of untreated monocytes.

These data show that M-CSF induced the expression of the membrane-bound form of IL-18 in a subset of monocytes differentiating toward macrophages. On the other hands, GM-CSF was unable to promote IL-18 surface expression as demonstrated by the lack of mIL-18 in both classical DCs (GM-CSF + IL-4) and GM-CSF-treated monocytes. It is of note however, that unlike DCs, GM-CSF-treated monocytes released small but detectable amounts of sIL-18 upon LPS stimulation (Fig. 3B).

mIL-18 identifies a cell subset in both CD16 and CD16+ monocytes-derived macrophages

Two major monocyte subsets have been identified in blood according to differences in CD16 (FcγRIII) surface expression, that is CD14+/CD16 and CD14+/CD16+ monocytes [[8, 17, 18]]. In the attempt to characterize the subset of monocytes that acquired the expression of mIL-18 during differentiation toward macrophages, the two monocyte subsets were purified and cultured in the presence of M-CSF for 7 days. A mIL-18+ cell subset could be detected in macrophages differentiating from both CD16 and CD16+ monocytes (Fig. 4A).

Figure 4.

mIL-18 identifies a subset of macrophages differentiating from both CD16 and CD16+ blood monocytes. (A) CD16+ or CD16 monocyte subsets were isolated from total blood monocytes, differentiated toward macrophages with M-CSF (7 days), and analyzed by immunofluorescence and flow cytometry for the expression of the indicated surface markers. Data are representative of three independent experiments. (B) Resting NK cells were cocultured overnight with autologous macrophages (from either CD16 or CD16+ monocytes) in the presence of LPS and analyzed by flow cytometry for the expression of the indicated markers. Data are representative of three independent experiments. (C) Supernatants from LPS-treated macrophages (left and middle) and from NK cell–macrophage cocultures (right) were analyzed by ELISA for the presence of the indicated cytokines. White and grey bars indicate macrophages from CD16 or CD16+ monocytes, respectively. Data pooled from three independent experiments. Mean, 95% confidence intervals, and significance (Wilcoxon–Mann–Whiteny p-value test) are shown. *p < 0.05.

To verify whether macrophages derived from either CD16 or CD16+ monocytes might promote NK cell activation, macrophages were cultured overnight with autologous resting NK cells in the presence of LPS. After coculture, NK cells were recovered (>99% purity) and analyzed for their surface phenotype. Both CD16- and CD16+-derived macrophages induced in NK cells the expression of CD69 and CD25 activation markers as well as that of CCR7 (Fig. 4B). Moreover, both types of macrophages released detectable amounts of sIL-18 and promoted the production by NK cells of high amounts of IFN-γ that was virtually abrogated by Ab-mediated neutralization of IL-18 (Fig. 4C).

It is of note that the percentage of mIL-18+ cells as well as the mIL-18 surface density was higher in macrophages from CD16 monocytes as compared with that derived from CD16+ monocytes (Fig. 4A). Accordingly, while both types of macrophages released comparable amount of TNF-α, macrophages from CD16 monocytes secreted larger amounts of sIL-18 and induced the release by NK cells of higher amounts of IFN-γ (Fig. 4C).

Caspase-1-mediated expression of mIL-18 and protease-dependent release of sIL-18

To understand the role played by caspase-1 in the expression of IL-18 at the cell surface, monocytes were differentiated toward M0 in the presence of the caspase-1 inhibitor Z-YVAD-FMK [[19]] and analyzed daily by flow cytometry for the expression of mIL-18 (Fig. 5 and Supporting Information Fig. 3A). Caspase-1 inhibition strongly reduced IL-1β release while did not modify that of IL-12 (Supporting Information Fig. 3B). Notably, treatment with the caspase-1 inhibitor significantly inhibited mIL-18 expression during macrophage differentiation suggesting the requirement of caspase-1 processing for the expression of IL-18 at the cell surface (Fig. 5 and Supporting Information Fig. 3A).

Figure 5.

Caspase-1 is involved in the expression of mIL-18. Monocytes were treated for different time intervals with M-CSF either in the absence (open symbols) or in the presence (black symbols) of the caspase-1 inhibitor (Z-YVAD-FMK). Cells were analyzed daily by flow cytometry for the expression of mIL-18 by MFI (squares) and percentage of positive cells (triangles). Data pooled from four independent experiments. Mean, 95% confidence intervals, and significance (Wilcoxon–Mann–Whiteny p-value test) are shown. *p < 0.05.

To investigate the mechanism underlying cell surface retention of IL-18 and its shedding after LPS treatment, M0 macrophages were analyzed for mIL-18 expression and sIL-18 secretion after treatment with a glycine buffer, pH 3 or Phorbol Myristate Acetate (PMA), a broad spectrum activator of cellular proteases [[20]] (Supporting Information Fig. 4A). Both types of treatment did not affect the expression of various cell surface molecules including CD48 and Polio Virus Receptor (PVR, CD155). The acidic treatment did not modify the expression of mIL-18 by macrophages, while dissociating HLA class I molecules [[21]]. Conversely, activation of cellular proteases had no effect on HLA class I expression while inducing the disappearance of mIL-18 and the release of sIL-18 in the supernatant (Supporting Information Fig. 4A). These results suggest that IL-18 is firmly attached to the plasma membrane and that secretion of sIL-18 depends on the activity of a LPS-inducible protease(s). The protease responsible for mIL-18 cleavage would not be represented by calpain, a membrane-bound cysteine protease involved in the conversion of pro-IL-1α to IL-1α [[1, 2]]. Thus, treatment with the calpain inhibitor I [[22]] did not affect the release of sIL-18, while strongly affecting that of IL-1α (Supporting Information Fig. 4B and C).

Discussion

In the present study, we analyzed the expression of the membrane form of IL-18 in macrophages and DCs differentiating in vitro from human blood monocytes. We show that mIL-18 is expressed in unpolarized (M0) and M2-polarized macrophages but not in monocytes and DCs. Importantly, the ability to release sIL-18 upon bacterial stimuli strictly correlates with the presence of mIL-18. In fact, upon LPS treatment, sIL-18 was produced in significant amounts by mIL-18+ M0 and M2 cells but not by mIL-18 negative DCs. Accordingly, Gardella et al. did not detect sIL-18 in culture fluids of DCs [[23]]. We also show that sIL-18 secretion by macrophages could depend on the protease-mediated shedding of its membrane-bound form. These data are in line with our previous observation that blocking of intracellular protein trafficking does not affect LPS-mediated release of sIL-18 [[7]]. The molecular events allowing cell surface retention of IL-18 and the protease(s) involved in its shedding upon stimulation are currently under investigation.

mIL-18 expression was detected on average at day 2 in M-CSF-primed monocytes but not in GM-CSF-treated cells. M-CSF is a stimulating factor that preferentially generates macrophages populations both in vitro and in vivo [[7, 9, 24]]. In mice, inactivating mutations of M-CSF resulted in a deficiency of macrophage development in many tissues [[25]]. Although GM-CSF is generally thought to preferentially generate classical monocyte-derived DCs [[26]], other studies suggested that it could be involved in both differentiation and M1 polarization of macrophages [[16, 27]]. In particular, GM-CSF gave rise to macrophages that in response to TLR stimulation produced proinflammatory cytokines such as TNF-α and IL-6. On the contrary, treatment of monocytes with M-CSF was found to generate M2 macrophages that are considered as anti-inflammatory because they produce IL-10 upon TLR stimulation.

In our view, M-CSF, which is constitutively produced by several cell types including fibroblasts, endothelial cells, and smooth muscle cells, would play a homeostatic role by promoting macrophage differentiation of monocytes and preventing their polarization in order to avoid inappropriate inflammatory signals. In fact, the exposure of purified blood monocytes to M-CSF generates steady-state macrophages (unpolarized M0) that only following exposure to appropriate stimuli such as LPS or IL-4 polarize toward either proinflammatory (M1) or immunomodulatory/tissue-repairing (M2) macrophages [[7, 24]]. Importantly, our present data also show that M-CSF can induce the expression in M0 cells of mIL-18 that is conserved during M2 polarization. Thus, M-CSF (but not GM-CSF) would render both M0 and M2 macrophages prompt to release sIL-18 in response to M1 polarizing stimuli. It is of note however that while in DCs, the lack of mIL-18 correlated with the inability to release the soluble form of the cytokine, in GM-CSF-treated monocytes, a small but measurable amount of sIL-18 could be detected (approximately eight times less that in M-CSF macrophages). Thus, in these cells, the amount of mIL-18 might be below the limit of detection in flow cytometry. Alternatively, a different mechanism of sIL-18 secretion may exist that under certain circumstances could bypass cell membrane docking.

Monocytes have been found to display phenotypic and functional heterogeneity in both mouse and human [[8, 18]]. In humans, CD16 is expressed by all macrophages while it is present only in a minor subset of circulating monocytes. Both CD16 and CD16+ monocytes are equipped with chemokine receptors (CCR2 or CCR5) that are necessary for their migration into peripheral tissues and can differentiate into macrophages upon M-CSF-mediated stimulation. It has been suggested that the two monocyte subsets could have different functional capabilities including phagocytosis, Ag presentation, cytokine production, and response to given stimuli [[8, 18]]. Here, we show that although all blood monocytes homogeneously expressed the M-CSF receptor, the M-CSF-induced expression of mIL-18 is confined to a subset of monocytes differentiating toward macrophages. This would suggest that only a fraction of circulating monocytes is prone to give rise to macrophages expressing mIL-18 and releasing sIL-18 upon appropriate stimulation. This novel monocyte subset is not restricted to either the CD16 or CD16+ fraction as indicated by the finding that mIL18+ cells were detected in macrophages derived from both monocyte subpopulations.

The identification of mIL-18+ and mIL-18 macrophages could be relevant for a better understanding of inflammatory responses in physiological and pathological conditions. Previous studies indicated that this cytokine by acting on effectors such as T and NK cells promotes type 1 immune responses. In particular, our previous and present data show that during NK/macrophage interactions, microbial products (LPS or BCG) induce the release by M0 and M2 macrophages of sIL-18. This in turn, by acting at the synaptic cleft, induces in autologous resting NK cells the release of high amounts of IFN-γ [[7]], a cytokine that activates neutrophils and macrophages for intracellular killing of pathogens and promotes M1 polarization of macrophages [[6, 7, 24]]. Moreover, sIL18 induces the acquisition by NK cells of CCR7, a chemokine receptor that allows the recruitment of NK cells to secondary lymphoid organs in response to CCL19 and CCL21 [[28]]. Thus, it is conceivable that NK cells activated in peripheral tissues by sIL18, which is released by a subset of M1-polarizing mIL-18+ macrophages, could participate also in Th1 polarization within lymph nodes [[29]].

Il-18 belongs to the IL-1 family, a group of molecules that either activate or suppress inflammation [[1]]. IL-1β is the most-studied member and, although measurable levels in supernatants as well as serum are low and, in disease, rarely correlate with severity, it is involved in the pathogenesis of different autoinflammatory diseases. IL-1β neutralization by IL-1Ra, a natural-occurring IL-1β antagonist known as anakinra or by anti-IL-1β mAbs successfully entered the clinical practice and in some cases was life saving [[5, 30, 31]]. Monocytes and macrophages do not constitutively express mRNA for pro-IL-1β and IL-1β secretion requires two signals. For example, the engagement of TLR4 by LPS induces mRNA transcription and synthesis of the IL-1β precursor while cleavage of pro-IL-1β and the secretion of the mature/active IL-1β requires ATP-mediated assembling of inflammasome and caspase-1 activation. It is of note that caspase-1 is constitutively active in blood monocytes that release significant amounts of ATP in the extracellular space. On the contrary, caspase-1 is inactive in macrophages that are unable to produce ATP even after stimulation [[1]]. Overall available data suggest that blood monocytes are well equipped to quickly release IL-1β in response to danger signals. On the contrary, IL-1β secretion by macrophages would require accessory cells such as platelets, neutrophils, and endothelial cells as exogenous sources of second signals such as ATP [[6]].

Although an in vivo confirmation is still missing, our data suggest a different scenario for IL-18. Monocytes constitutively express the pro-IL-18 mRNA while they do not express mIL-18. During monocyte differentiation into macrophages, M-CSF did not raise mRNA levels while inducing the expression of mIL-18. mIL-18 expression is restricted to a subpopulation of differentiating monocytes implying the existence in these cells of a second, endogenous signal that cooperates with M-CSF in inducing IL-18 surface expression. As suggested by the involvement of caspase-1, the membrane-bound form of IL-18 is likely represented by the mature form of the cytokine. Unlike other membrane-bound cyto-kines such as IL-15, mIL-18 is inactive and, to become active, needs protease-mediated shedding and release. Indeed, during NK/macrophage interactions, mIL-18+ M0 (and M2) macrophages require TLR stimulation and sIL-18 release to promote in NK cells IFN-γ secretion and CCR7 expression (Fig. 4) [[7]]. Finally, LPS induced the release of sIL-18 by mIL-18+ macrophages but not that of IL-1β that required exogenous ATP. Therefore, sIL-18 secretion in macrophages appears to be independent from that of IL-1β.

IL-18 plays an essential role in Th1 responses, primarily via its ability to induce IFN-γ production in T and NK cells [[7, 32, 33]]. IL-18-induced cellular activation is counteracted by the IL-18 binding protein (IL-18BP), a natural occurring antagonist that binds IL-18 with very high affinity [[34]]. Interestingly, a similar binding has been reported for viral IL-18BP homologs that are encoded by poxviruses such as Molluscum contagiosum virus (MCV) and represent a mechanism of escape from the immune response [[35, 36]]. Thus, similar to IL-1β, the balance between IL-18 and its antagonist dictates the role of the cytokine in both health and disease. In this context, a number of autoimmune or autoinflammatory diseases are thought to be mediated in part by IL-18. These include systemic lupus erythematosus, rheumatoid arthritis, Crohn's disease, psoriasis [[37]], and the macrophage activation syndromes (MAS) associated with the systemic-onset Juvenile idiopathic arthritis [[38]] and the adult-onset Still's disease [[39]]. These syndromes are systemic inflammatory disorders characterized by excessive activation of macrophages and elevated concentration of sIL-18 in serum. In light of our present results, it will be of interest to analyze in patients affected by inflammatory syndromes, the size and the phenotypic/functional properties of the mIL-18+ macrophage subpopulation. It will be also important to measure the endogenous levels of both IL-18 and its antagonist since IL-18BP could functionally keep IL-18 at bay in several diseases [[40]].

Materials and methods

Cells

Monocytes were purified from PBMC of healthy donors using the Human monocyte Cell Isolation kit II (Miltenyi Biotec, GmbH, Bergisch Gladbach, Germany). The CD16+ and CD16 monocyte subsets were isolated using the Human CD16+ Monocyte isolation kit (Miltenyi Biotec). To obtain macrophages (M0), monocytes were cultured (5 × 105/mL) for 7 days in RPMI, 10% FCS (Sigma-Aldrich, USA) with 100 ng/mL rM-CSF (PeproTech, London, United Kingdom) in 24 lumox multiwell TC-QUALITAET plates (Greiner bio-one GmbH, Frickenhausen, Germany). M1 and M2 polarization were obtained culturing M0 for 18 h with 100 ng/mL lipopolysaccharide (LPS) from Escherichia coli (Sigma-Aldrich) or 20 ng/mL rIL-4 (PeproTech), respectively. M2 macrophages were activated with LPS (100 ng/mL for 18 h)[[7]]. For IL-1β production, M0 were incubated overnight with 100 ng/mL of LPS plus 1 mM ATP (Sigma-Aldrich)[[19]].

To obtain iDCs, monocytes were cultured for 7 days with 50 ng/mL rGM-CSF (PeproTech) and 20 ng/mL rIL-4 (PeproTech). To generate mDC, iDCs were treated for 24 h with 1 μg/mL LPS [[41]]. To obtain GM-CSF-treated macrophages, monocytes were cultured 7 days with 100 ng/mL rGM-CSF [[9, 27]].

NK cells purified from PBMC of healthy donors (Human NK Cell Isolation kit, Miltenyi Biotec) were frozen and thawed 24 h before coculture experiments. Informed consent was obtained for all blood donors as approved by the Ethics Board.

Antibodies

The anti-hIL-18 mAb (IgG1) used for flow cytometry and IL-18 neutralization was generated by immunizing Balb/c mice with recombinant human IL-18 (clone 125-2H, MBL, Naka-ku Nagoya, Japan). MA 127 (IgG1, anti-NTB-A), 6A4 (IgG1, anti-HLA class I-A, -B, -C, and -E), M5A10 (IgG1, anti-PVR), CO202 (IgM, anti-CD48), FM184/703 (IgM, anti-CD1a), c227 (IgG1, anti-CD69), MAR93 (IgG1, anti-CD25) mAbs were produced in our labs [[7]]. Anti-CD14 (IgG2a, Immunotech, Marseille, France); anti-CD163 and anti-CD206 (IgG1, BD Bioscience, Franklin Lakes, NJ, USA); anti-CD204PE, anti-MCSF-R (IgG1), and anti-CCR7 (IgG2a) (R&D Systems, MN, USA). Rabbit polyclonals to DC-SIGN, Heme Oxygenase, and FRβ were purchased by Abcam (Cambridge, UK).

Flow cytometry and cytokine release

For immunofluorescence and flow cytometry (FACSCalibur Becton Dickinson & Co, Mountain View, CA, USA), cells were stained with PE-conjugated mAbs or with unconjugated mAbs followed by PE-conjugated isotype-specific goat anti-mouse second reagent (Southern Biotechnology Associated, Birmingham, AL, USA). Monocytes, macrophages, and DCs were preincubated for 30 min at 4°C with hIgG (1 mg/mL) (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) before specific mAb staining. The full gating strategy used for FACS analysis is shown in Supporting Information Fig. 5.

ELISA kits used were: IL-12p40/p70, TNF-α, IL-1α, and IL-1β (BIOSOURCE Int. Inc., Camarillo, CA, USA); IL-18 (MBL).

Acidic treatment was performed by treating macrophages for 15 min at 4°C with 25 mM glycine, 150 mM sodium chloride, pH 3 buffer. Cellular proteases activation was performed by culturing macrophages for 3 h at 37°C in FCS-free medium containing 200 ng/mL Phorbol 12-myristate 13-acetate (PMA) (Sigma) [[20]].

Inhibition of caspase-1 activity was performed by culturing monocytes for up to 72 h with 20 μM of the caspase-1 inhibitor (fluoromethylketone) Z-YVAD(oMe)-FMK (Enzo Life Sciences, NY, USA) [[19]]; the caspase-1 inhibitor was added to the culture every 24 h.

Inhibition of calpain activity was performed by treating macrophages for 18 h with LPS and 5 μM of calpain inhibitor I (kindly provided by E. Melloni and M. Averna, University of Genova, Italy) [[22]]

The methods used for confocal microscopy analysis, real-time PCR, and statistical analysis are provided as supplementary data.

Acknowledgments

This work was supported by Investigator Grants (10643, 10225, 9005) and special project 5×1000 (9962) from Associazione Italiana per la Ricerca sul Cancro (AIRC), Istituto Superiore di Sanità (ISS), Ministero del Lavoro, della Salute e delle Politiche Sociali and Ministero dell’Istruzione, dell'Università e della Ricerca (MIUR). F. B. is recipient of a fellowship awarded by AIRC (special project 5×1000, 9962). We thank B. Azzarone (INSERM, UMR 542,Villejuif, France), A. Rubartelli (Istituto nazionale per la Ricerca sul Cancro, Genova, Italy), E. Melloni, M. Averna (University of Genova, Italy), and G. Reggiardo (Medi Service, Genova, Italy) for helpful discussion.

Conflict of interest

A. M. is founder and shareholder of Innate Pharma (Marseille, France). The other authors declare no conflicts of interest.

Abbreviations
iDC

immature DC

mDC

mature dendritic cell

mIL-18

membrane-bound IL-18

sIL-18

soluble IL-18

Ancillary