The Formyl Peptide N‐Formyl‐methionyl‐leucyl‐phenylalanine Downregulates the Expression of FcγRs in Interferon‐γ‐Activated Monocytes/Macrophages In Vitro and In Vivo

N‐Formyl peptides are cleavage products of bacterial and mitochondrial proteins that have pro‐inflammatory activities and play an important role in antibacterial host defence. FcγRI is a receptor for the Fc portion of immunoglobulin G expressed in monocytes that mediates cytotoxicity and is upregulated by interferon‐γ (IFN‐γ) and interleukin‐10 (IL‐10). In this report, we demonstrate that N‐formyl‐methionyl‐leucyl‐phenylalanine (FMLP) downregulates the expression of FcγRI in IFN‐γ‐treated monocytes, but not in IL‐10‐treated monocytes. We determine that supernatants obtained from monocytes treated with IFN‐γ and then exposed to FMLP induce the downregulation of FcγRI in naïve monocytes. This effect is abrogated by the protease inhibitors phenylmethylsulphonyl fluoride and phosphoramidon, which inhibit serine and metalloproteases, respectively. Supernatants from FMLP‐treated neutrophils also induce the downregulation of FcγRI, when added to naïve monocytes. Similar observations were obtained in vivo in a mouse model of chronic inflammation. In vivo, FMLP also downregulates the expression of FcγRs in IFN‐γ‐activated macrophages. Our results support the existence of a new mechanism through which FMLP could modulate the activity of monocytes/macrophages during bacterial infections.


Introduction
The N-formyl peptides are cleavage products of bacterial and mitochondrial proteins that attract phagocytes to sites of infection, and therefore play an important role in antibacterial host defence in vivo [1,2]. In fact, it has been demonstrated that the interaction between formyl peptides and their receptors enhances the killing of Leishmania donovani [3], while the lack of these receptors renders mice more susceptible to Listeria monocytogenes infection [4] or impairs the recruitment of monocytes in a mouse model of pneumococcal pneumonia [5].
The receptors for the Fc portion of immunoglobulin G (IgG) (FcgRs) belong to a series of receptors involved in different effector and inflammatory functions, such as phagocytosis, respiratory burst, antibody-dependent cell-mediated cytotoxicity (ADCC) and secretion of inflammatory cytokines [6,7]. Although formyl peptides are considered potent pro-inflammatory agents, capable of triggering respiratory burst and degranulation of neutrophils and monocytes [8,9], we have recently demonstrated that the prototypic chemotactic peptide N-formyl-methionyl-leucylphenylalanine (FMLP), identical in sequence to the major chemotactic peptide produced by Escherichia coli [10], exerted, paradoxically, a severe anti-inflammatory effect manifested by the downregulation of receptors for the FcgRs on human leucocytes [11±13]. Thus, in neutrophils, we have observed that the exposure of the cells to FMLP downregulated both FcgRII and FcgRIIIB [11,12], while the pretreatment of monocytes with FMLP only impaired the upregulation of FcgRI induced by interferon-g (IFN-g) or interleukin-10 (IL-10) [13].
Inflammation involves several factors, and hence during the development of the inflammatory phenomenon, monocytes could be exposed first to cytokines and then to formyl peptides. In this context, we evaluated the effect of the chemotactic peptide on the expression of FcgRI in monocytes previously activated with IFN-g or IL-10. We also studied the effect of FMLP on the expression of FcgRII/III in IFN-g-activated macrophages, using a mouse model of chronic inflammation.
Preparation of human monocytes. Fresh human blood was obtained by venipuncture from healthy adult volunteers and collected on citrate/dextrose/adenine. Blood was diluted 1 : 2 with saline, layered on a Ficoll±Hypaque cushion and centrifuged at 850 Â g for 25 min, as previously described [14]. Peripheral blood mononuclear cells (PBMCs) were collected, centrifuged on a Percoll gradient and resuspended in RPMI 10% heat-inactivated (56 C, 30 min) fetal calf serum (FCS) and 50 mg/ml of gentamicin. Viability of monocytes was always more than 95% as measured by the Trypan blue exclusion test.
Preparation of human neutrophils. Fresh human blood was obtained by venipuncture from healthy adult volunteers and collected on citrate/dextrose/adenine. Polymorphonuclear leucocytes (PMNs) were isolated from the bulk of red cells by sedimentation with 3% dextran and Ficoll± Hypaque cushion, as previously described [11]. The pellet containing PMNs was collected, and the contaminating erythrocytes were removed by hypotonic lysis. The cell suspension contained 95±98% PMNs. Viability of PMN was always more than 95% as measured by the Trypan blue exclusion test.
FMLP treatment. FMLP was dissolved in dimethyl sulphoxide at a concentration of 0.1 M. The subsequent dilutions of FMLP were made in saline. Monocytes were incubated in polypropylene tubes with medium, IFN-g (240 U/ml) or IL-10 (100 U/ml) at 37 C in a 5% CO 2 atmosphere for 24 h. Then, medium or FMLP 1 mM was added to the cells for two additional hours. After that period, the cells were used as effector cells in ADCC or for flow cytometry studies.
Supernatant collection. Monocytes (0.5 Â 10 6 cells/ml) were treated with medium, IFN-g (240 U/ml) or IL-10 (100 U/ml) at 37 C in a 5% CO 2 atmosphere for 24 h. After that period, the cells were incubated with medium or FMLP 1 mM for two additional hours. Then, supernatants were collected, spun down to remove cellular debris (10 min at 700 g), stored at À70 C and dialysed exhaustively against RPMI to eliminate residual FMLP before their usage. PMNs (1 Â 10 6 cells/ml) were treated with medium or FMLP 1 mM at 37 C in a 5% CO 2 atmosphere for 3 h, in the presence or absence of different protease inhibitors. Then, supernatants were collected, spun down to remove cellular debris (10 min at 700 g), stored at À70 C and dialysed exhaustively against RPMI to eliminate residual FMLP before their usage.
ADCC assay. When PBMCs at 4 Â 10 6 cells/ml (100 ml) were used as effector cells, ADCC was performed in 96-well polystyrene plates. The adherent PBMC population (monocytes) was obtained from 4 Â 10 6 PBMCs/ml that were left to adhere in 96-well, round-bottom plates for 1 h at 37 C. After that, the nonadherent cells were removed, and ADCC was performed. The cells were incubated with 1 Â 10 5 51 Cr-labelled chicken red blood cells ( 51 CRBC) and a suboptimal concentration of rabbit IgG anti-CRBC, as previously described [15]. After 18 h of incubation at 37 C and 5% CO 2 , the culture plate was centrifuged, and the percentage of cytotoxicity was calculated as follows: % ADCC 51 Cr released to the supernatant Â 100/total radioactivity. This value was corrected by subtracting the percentage of 51 Cr released in the absence of antibody (spontaneous release). Quadruplicates were set up for each sample.
In vivo studies. Glass cylinders of 2 cm length and 8 mm width were introduced subcutaneously into 8-to 10-week old BALB/c female mice. After 20 days, the cylinders caused a chronic inflammatory process with the open ends of the cylinders`closed' by fibrotic tissue and predominantly infiltrated by macrophages. At that moment, 50 ml of pyrogen-free saline or 50 ml of IFN-g (3000 U/ml) was injected into the cylinder. After 24 h, 40 ml of pyrogenfree saline or 40 ml of FMLP at a concentration of 50 mM was added for two additional hours. Considering that the volume of each cylinder is approximately 200 ml, the final concentration of FMLP should be approximately 10 mM. Then, the cylinders were extracted from the mice, and the cells that had been recruited were stained with an antibody against murine FcgRII/III (clone 2.4G2) and then with a goat FITC-labelled antirat IgG (H L). Flow cytometry was performed, and the macrophage population was analysed using macrophage-specific forward light scatter (FSC) and side light scatter (SSC) gates. Flow cytometry. After different treatments, 0.5 Â 10 6 monocytes/ml or 0.35 Â 10 6 cells obtained from the glass cylinders introduced into BALB/c mice, were washed and incubated with the monoclonal antibodies indicated above. The cells were washed and resuspended in ISO-FLOW, and flow cytometry was performed on FACS Analyser (Becton-Dickinson Immunocytometry System, San Jose, CA, USA). The results corresponding to human monocytes were expressed as the percentage of median of fluorescence intensity (% MFI) of untreated cells (controls). Purity of monocytes, determined as CD14-positive cells, was >90% in all donors. Nonspecific binding was determined using the appropriate mouse IgG isotype-matched control antibody.
Statistical analysis. Statistical significance of the results was calculated using the nonparametric Mann±Whitney U-test or Wilcoxon tests (two-tailed). For the in vivo experiment, the unpaired t-test was used.

Effect of FMLP on the expression of FcgRI in monocytes pretreated with IFN-g or IL-10
We evaluated the effect of FMLP on the expression of FcgRI in monocytes that had been previously treated with either IFN-g or IL-10. As shown in Fig. 1A, FMLP 1 mM did not alter the basal level of the expression of FcgRI. However, when monocytes were treated with IFN-g for 24 h and then exposed to FMLP for two additional hours, a significant downregulation of the expression of FcgRI was observed (Fig. 1B, Table 1). Lower concentrations of FMLP ranging from 1 to 100 nM were devoid of effect (data not shown). In contrast to the results obtained with IFN-g-treated monocytes, FMLP was completely unable to downregulate the expression of FcgRI in monocytes previously exposed to IL-10 ( Fig. 1C). The levels of expression of other surface molecules, such as HLA-DR, were not altered by FMLP in monocytes previously treated with IFN-g (% MFI of untreated cells: IFN-g, 150%; IFN-g FMLP, 150%; FMLP, 99%; representative experiment of n 3).
Effect of FMLP on the enhancement of ADCC by IFN-g As the upregulation of FcgRI induced by IFN-g could correlate with FcgR-dependent functions, such as ADCC [16], this cytotoxic activity was evaluated. As shown in Fig. 2, the enhancement of ADCC induced by IFN-g in human monocytes was completely inhibited when the cells were treated with FMLP.

Effect of supernatants obtained from monocytes treated with IFN-g plus FMLP on the expression of FcgRI
In order to investigate whether the factors released by monocytes were responsible for the downregulation of FcgRI triggered by FMLP, experiments were carried out in which supernatants from IFN-g-treated monocytes cultured with FMLP were collected, and their ability to modulate the expression of FcgRI in naõ Ève monocytes was evaluated. Supernatants were dialysed exhaustively to eliminate FMLP before their usage. The results summarized in Table 2 indicate that these supernatants significantly reduce the expression of FcgRI in naõ Ève Effect of N-formyl-methionyl-leucyl-phenylalanine (FMLP) on the expression of FcgRI in human monocytes treated with interferon-g (IFN-g) or interleukin-10 (IL-10). Monocytes (0.5 Â 10 6 cells/ml) were incubated for 24 h with: (A) medium (B) IFN-g (240 U/ml) or (C) IL-10 (100 U/ml). After that period, the cells were incubated with medium or FMLP 1 mM for two additional hours. Then, monocytes were stained with an anti-FcgRI antibody. The histograms correspond to a representative experiment of n 26. Nonspecific binding (filled peak) was determined using control isotype antibodies. X-axis: fluorescence intensity (arbitrary units); y-axis: cell number. monocytes, confirming the role of released product(s) in this process. In contrast, supernatants obtained from FMLP-or IL-10 plus FMLP-treated monocytes were not capable of inducing the downregulation of FcgRI ( Table 2).

Effect of protease inhibitors on the downregulation of FcgRI
With the purpose of investigating whether proteases released by monocytes were responsible for the downregulation of FcgRI, monocytes were incubated with IFN-g for 24 h and then treated with FMLP for 2 h in the presence of different protease inhibitors. These inhibitors included PMSF 2 mM, which inhibits all serine proteases with a trypsin-like specificity [17]; pepstatin 1 nM, an inhibitor of aspartic protease [18]; phosphoramidon 10 mM, a metalloprotease inhibitor [19] and the cysteine protease inhibitor, leupeptin 50 mM [20]. The results shown in Fig. 3 indicate that the downregulation of FcgRI is completely abrogated by PMSF and phosphoramidon, supporting the role of a serine protease(s) and a metalloprotease(s) in this process. The lack of effect of the other inhibitors led us to discard the participation of a wide range of proteases. Similar results were obtained when supernatants from monocytes treated with IFN-g plus FMLP in the presence of the protease inhibitors were added to naõ Ève monocytes (data not shown).

Effect of FMLP on the expression of FcgRs in a mouse model of chronic inflammation
In order to investigate the in vivo effect of FMLP on the expression of FcgRs, a glass cylinder was introduced subcutaneously into female BALB/c mice [21]. After 20 days, the cylinders caused a chronic inflammatory process. At that moment, the open ends of the cylinders had beeǹ closed' by fibrotic tissue, and were predominantly infil-trated by macrophages. Then, murine IFN-g was injected into the cylinder, and 24 h later FMLP was added. Two hours after the injection of FMLP, the cylinders were extracted from mice, and the cells that had been recruited into the cylinder were stained with an antibody against murine FcgRII/III. Flow cytometry was performed, and  After that period, supernatants were collected, dialysed to eliminate residual FMLP and added to naõ Ève monocytes (0.5 Â 10 6 cells/ml). These cells were incubated with the supernatants for 3 h, and then were stained with an anti-FcgRI antibody. Data are expressed as percentage of median of fluorescence intensity (% MFI) of monocytes treated with supernatants from untreated PMNs AE standard error of the mean, n 7. *P < 0.0007 significantly different from monocytes treated with supernatants from untreated PMNs. Statistical significance was calculated using the Mann± Whitney U-test, two-tailed. SN, supernatants. PMNs (1 Â 10 6 cells/ml) were incubated with medium or FMLP 1 mM for 3 h in the presence or absence of different protease inhibitors. After that period, supernatants were collected, dialysed to eliminate residual FMLP and added to naõ Ève monocytes (0.5 Â 10 6 cells/ml). These cells were incubated with the supernatants for 3 h, and then were stained with an anti-FcgRI antibody. The table summarizes a representative experiment of the macrophage population was studied. Figure 5 shows that FMLP did not modify the basal level of expression of FcgRII/III, but induced a drastic downregulation of these receptors in macrophages that had been previously treated with IFN-g. This result is in accordance with our in vitro experiments with human monocytes although, in this case, we have not observed any upregulatory effect of IFN-g in the murine FcgRs studied.

Discussion
The lysis of bacteria at the site of infection provokes the release of formyl peptides [8,22] and generates a concentration gradient that induces the unidirectional migration of neutrophils and monocytes to the inflammatory focus [23]. The aim of this report was to study the effect of FMLP, a prototype of N-formyl peptides, on the expression of FcgRs in human monocytes as well as in murine macrophages.
In this report, we demonstrate that when monocytes are previously activated by IFN-g, FMLP 1 mM induces a significant downregulation of FcgRI that have been already expressed. However, this effect of FMLP is not observed in naõ Ève monocytes, indicating that the activity of FMLP depends on the pretreatment of the cells with IFN-g. This result could be owing to the well-known effect of priming induced by IFN-g, which enhances the activity of different agonist on monocytes [24]. This is different from our previous observations, showing that preincubation of monocytes with lower concentrations of FMLP, ranging from 10 to 100 nM, inhibited the ability of IFN-g to upregulate FcgRI, without affecting the already-expressed FcgRI. Taken together, these results suggest that FMLP could downregulate FcgRI in monocytes through their way to the inflammatory site in two different ways: at a relatively low concentration, it could inhibit the overexpression of FcgRI induced by IFN-g or IL-10, while at higher concentrations, it could induce the secretion of enzymes which downregulate the FcgRI induced only by IFN-g. The relevance of these mechanisms in the context of an ongoing bacterial induced inflammation is that, independently of the sequence of events, the expression of FcgRI will be downregulated, either before or after its expression has been induced by cytokines.
The downregulation of FcgRI correlates with the complete inhibition of ADCC (Fig. 2), a well-known FcgRdependent cytotoxic mechanism. The mechanism of downregulation of FcgRI by FMLP has not been completely elucidated. However, it seems to be mediated by the action of serine and metalloproteases released by monocytes (Fig. 3). The role of secretory products is supported by the fact that the downregulation of FcgRI is also observed when using supernatants from IFN-g plus FMLP-treated monocytes ( Table 2).
In contrast to these observations, FMLP does not induce any effect on the expression of FcgRI in IL-10treated monocytes, despite the fact that this cytokine also induces the overexpression of FcgRI (Fig. 1C). In addition, supernatants from monocytes that have been stimulated with IL-10 are devoid of effects (Table 2). These results could be explained by the fact that IL-10 is a powerful inhibitor of monocyte secretion of pro-inflammatory products, including proteases [25], and suggests that the state of activation of the cells is a crucial event to allow FMLP to mediate the downregulation of FcgRI.
In previous reports, we had demonstrated that FMLP was capable of inducing the downregulation of both FcgRII and FcgRIIIB in human neutrophils [11,12], and that this effect was owing to the secretion of serine protease(s) by these cells. In this report, we observe that serine proteases secreted not only by PMNs but also by monocytes are involved in the downregulation of FcgRI in human monocytes, showing the existence of autocrine and paracrine mechanisms, leading to the downregulation of FcgRs. This mechanism is not unique to FcgRs, as it has been demonstrated that CD14 expressed in human monocytes is downregulated by serine proteases released by FMLP-stimulated PMNs [26]. Moreover, proteases released by activated human monocytes downregulate the expression of major histocompatibility complex class I (MHC-I) [27] as well as that of MHC-II molecules [28].
When we assayed FMLP in a mouse model of chronic inflammation [21], the formyl peptide was capable of inducing the downregulation of FcgRII/III in IFN-gactivated macrophages present at the inflammatory focus, Effect of N-formyl-methionyl-leucyl-phenylalanine (FMLP) on the expression of FcgRII/III in macrophages treated with interferon-g (IFN-g )in a mouse model of chronic inflammation. A glass cylinder was introduced subcutaneously into BALB/c female mice. After 20 days, the cylinders caused a chronic inflammatory process and pyrogen-free saline or 50 ml of IFN-g (3000 U/ml) was injected into the cylinder. After 24 h, pyrogen-free saline or 40 ml of FMLP at a concentration of 50 mM was added to the cylinder for two additional hours. The concentration of FMLP inside the cylinder was 10 mM approximately. Then, the cylinders were extracted from the mice, and the cells that had been recruited were stained with an antibody against murine FcgRII/III. Flow cytometry was performed, and the macrophage population was analysed using macrophage-specific forward light scatter (FSC) and side light scatter (SSC) gates. Data are expressed as median of fluorescence intensity (MFI) and each symbol (&) corresponds to the data of one mouse. *P < 0.0001 significantly different from untreated cells. Statistical significance was calculated using the unpaired t-test, two-tailed.
suggesting that our observations performed in vitro could be relevant to the development of the inflammatory phenomenon in vivo. This is supported by the fact that previous reports have shown that formyl peptides, through their receptors, are important in antibacterial host defence in vivo, because mice lacking these receptors are more susceptible to challenge with L. monocytogenes [4] or L. donovani [3]. Also, it has been demonstrated that FMLP mediates the recruitment of monocytes and macrophages in a mouse model of pneumococcal pneumonia [5]. We can speculate that when cells are gradually entering into the infectious focus, they come in contact with different products such as IFN-g, which is produced early by activated natural killer and T cells in response to pathogens like L. monocytogenes and Leishmania sp. [29,30]. At the site of infection, the cells are exposed to the highest concentration of formyl peptides (i.e. FMLP 1 mM), which induces both the release of products such as proteases and oxygen species, enabling inflammatory cells to kill micro-organisms, and the downregulation of receptors (FcgRs). The regulation of the expression of FcgRI is a key event in the control of bacterial infections, as it has been demonstrated that FcgRI knockout mice show impaired antibacterial host defence [31]. According to our results, formyl peptides are able to downregulate the expression of FcgRI even in IFN-g-activated monocytes, and this event could help control an exacerbated host response to infection as well as to operate as a bacterial mechanism of evasion of the immune response.
Further studies about the nonconventional (antiinflammatory) effects of formyl peptides will be necessary in order to have a comprehensive view of inflammation.