We recently reported that phagocytosis of apoptotic cells inhibits the release of inflammatory cytokines by human macrophages. In this paper we show that apoptotic cell uptake by mouse J774 macrophages also inhibits the synthesis and secretion of the chemokines, macrophage inflammatory protein-2 (Mip-2), KC, and Mip-1α (but not that of monocyte chemoattractant protein-1 (MCP-1)/JE), and increases TGF-β formation. Anti-TGF-β neutralizing Abs largely reversed the inhibitory effect of apoptotic cell uptake, and accordingly, exogenous TGF-β down-regulated the synthesis of the same mediators. Apoptotic cell ingestion or TGF-β also inhibited Mip-2 and Mip-1α gene expression in LPS-treated J774 cells, whereas TNF-α mRNA levels were unaffected. Importantly, TGF-β pretreatment of J774 cells did not significantly alter chemokine and TNF mRNA stability. Finally, we found that apoptotic cell uptake and TGF-β did not modulate NF-κB or AP-1 DNA binding in J774 cells. We conclude that the decreased production of chemokines and TNF resulting from apoptotic cell ingestion is largely mediated by a common event, i.e., feedback inhibition by endogenous TGF-β, but involves different mechanisms. Whereas TNF-α production appears to be translationally down-regulated, the suppression of most chemokines investigated appears to reflect transcriptional inhibition. In a broader context, the impairment of chemokine and TNF generation by apoptotic cell uptake might represent an important mechanism contributing to the resolution of inflammation. An additional consequence could be the selective recruitment of monocytes into inflammatory sites, as MCP-1/JE production by mouse macrophages was unaffected by apoptotic cell uptake, in contrast to other chemokines.

Most cell types have a limited life span, which ends through the process of apoptosis, or programmed cell death. In vivo, apoptotic cells are usually engulfed by neighboring cells or professional phagocytes, such as macrophages (reviewed in Ref. 1). As they become apoptotic, cells undergo changes in the composition of their surface, which allows their recognition by phagocytes, and subsequent removal. Because removal occurs before cell lysis, the release of immunogenic and potentially toxic intracellular contents in the surrounding tissue is avoided. As a result, apoptotic cell clearance is believed to represent a critical process in tissue remodeling, maintenance of immune homeostasis, and resolution of inflammation. In the latter instance, data derived from in vivo experimental models and from clinical studies have led to the suggestion that impaired recognition and/or removal of apoptotic cells from inflammatory foci contributes to a persistent state of inflammation in various organs and tissues (2, 3, 4).

Whereas the identification of surface molecules involved in the recognition of apoptotic cells by macrophages has been the subject of sustained research efforts in recent years, much remains to be learned about the effect of apoptotic cell phagocytosis on macrophage function. This represents an intriguing issue, as phagocytosis of a variety of targets normally triggers a battery of pro-inflammatory responses in macrophages, including the generation of reactive oxygen-derived intermediates, the release of proteolytic enzymes, the synthesis of lipid mediators (such as leukotrienes and thromboxanes), and the production of numerous inflammatory cytokines and growth factors. In sharp contrast, ingestion of apoptotic cells by macrophages induces an anti-inflammatory phenotype. For instance, it has been reported that thromboxane B2, GM-CSF, or lysosomal enzymes are no longer released under these conditions (5, 6). A recent study from our laboratory significantly extended these findings by showing that apoptotic cell uptake by human macrophages strongly inhibits their ability to generate IL-1β, IL-8, TNF-α, and leukotriene C4, in addition to thromboxane B2 and GM-CSF (7). Conversely, human macrophages secreted increased amounts of TGF-β and PGE2 under the same conditions, which repressed inflammatory cytokine release in an autocrine fashion (7). Importantly, apoptotic cells opsonized with Abs behaved as inflammatory phagocytic stimuli (7), thereby emphasizing the unique effect of apoptotic cell uptake on macrophage function. Collectively, these studies suggest that macrophages not only contribute to the resolution of inflammation through apoptotic cell removal, but also by actively suppressing inflammatory mediator production. A similar process may conceivably prevent the onset of inflammatory responses in situations such as tissue remodeling.

Despite recent advances (outlined above) in our understanding of how apoptotic cell uptake modifies macrophage function, many essential questions remain to be elucidated. In particular, the prominent role played by endogenous TGF-β in this context prompted us to investigate the molecular mechanisms involved in its action. For this purpose, we sought a cellular model in which TGF-β would account for most of the effect of apoptotic cell uptake. In addition, our previous finding that IL-8 production was down-regulated following apoptotic cell ingestion led us to determine whether other chemokines might be similarly affected. We now report that in mouse J774 macrophages, apoptotic cell ingestion results in the early release of TGF-β, and that the latter mainly accounts for the decreased production of TNF and several chemokines. More importantly, TGF-β was found to exert this action by acting at the level of gene transcription or protein translation, depending on the inflammatory mediator. Finally, the sparing of the chemokine monocyte chemoattractant protein-1 (MCP-1)3 from this suppressive effect may selectively recruit monocytes into inflamed lesions, presumably to participate in the reparative process.

Neutralizing anti-TGF-β (pan TGF) and anti-mouse IL-10 Abs, as well as recombinant mouse TNF-α, recombinant human TGF-β1, and recombinant human IL-10, were from R&D Systems (Minneapolis, MN). A commercial kit for preparing F(ab′)2 fragments was from Pierce (Rockford, IL). LPS (LPS, from Escherichia coli 0111:B4) was from List Biological Laboratories (Campbell, CA); and PGE2 and indomethacin were from the Cayman Chemical (Ann Arbor, MI). For EMSA analyses, an oligonucleotide containing tandemly repeated NF-κB sites identical to those of the HIV promoter (5′-gatcaGGGACTTTCCgctgGGGACTTTCC-3′) was synthesized, whereas an oligonucleotide containing a consensus AP-1 sequence (5′-cgcttgaTGAGTCAgccggaa-3′) was from Promega (Madison, WI). Poly(dI-dC) and T4 polynucleotide kinase were from Pharmacia (Uppsala, Sweden); [γ-32P]ATP and [α-32P]UTP were from ICN (Cleveland, OH). Acetylated BSA, diisopropyl fluorophosphate (DFP), and phenylethanesulfonyl fluoride (PMSF) were from Sigma-Aldrich (St-Louis, MO). Aprotinin, leupeptin, pepstatin, 4-(2-aminomethyl)benzenesulfonyl fluoride (AEBSF) and Nonidet P-40 (NP-40) were from Boehringer Mannheim (Mannheim, Germany). Polystyrene flasks and plates for cell culture were from Becton Dickinson (Lincoln Park, NJ). DMEM and RPMI 1640 were from Life Technologies (Gaithersburg, MD), X-Vivo 10 medium was from BioWhittaker (Walkersville, MD), and endotoxin-free FCS was from HyClone (Logan, UT). All other reagents were molecular biology grade.

Mouse J774A.1 macrophages (obtained from the American Type Culture Collection (ATCC), Manassas, VA) were cultured in DMEM supplemented with 10% heat-inactivated FCS, 2 mM l-glutamine, 100 μg/ml streptomycin, and 100 U/ml penicillin under a humidified 5% CO2 atmosphere at 37°C. Human Jurkat cells (also from the ATCC) were cultured (37°C, 5% CO2) in RPMI 1640 supplemented with 10% FCS, l-glutamine, and antibiotics and were used as a source of apoptotic cells. Apoptosis was induced by irradiating Jurkat cells under UV light (254 nm) for 10 min; cells were then further cultured for 3.5 h. This typically resulted in >70% apoptotic cells (as assessed morphologically), with ≤5% necrotic cells (as assessed by trypan blue staining). Apoptotic cells were used at a 5:1 ratio (Jurkat:J774 cells), as this optimally inhibited inflammatory mediator production, as previously reported (7). Under these conditions, a substantial proportion of J774 macrophages engulfed apoptotic Jurkat cells or human neutrophils, whereas nonapoptotic targets were not ingested (Fig. 1). The percentages of ingestion depicted in Fig. 1 are consistent with previously published observations made in J774 cells (8 and do not reflect their general ability to ingest particles, as latex beads are engulfed by virtually all macrophages (our unpublished data).

FIGURE 1.

Uptake of apoptotic cells by J774 macrophages. Apoptotic (Apo) or nonapoptotic Jurkat cells or human neutrophils (pmn) were added to J774A.1 cells at a 5:1 ratio and incubated for 1 h. Phagocytosis was evaluated by light microscopy, as described previously (7 ); values are mean ± SEM of at least seven independent experiments.

FIGURE 1.

Uptake of apoptotic cells by J774 macrophages. Apoptotic (Apo) or nonapoptotic Jurkat cells or human neutrophils (pmn) were added to J774A.1 cells at a 5:1 ratio and incubated for 1 h. Phagocytosis was evaluated by light microscopy, as described previously (7 ); values are mean ± SEM of at least seven independent experiments.

Close modal

Mouse J774 cells were cultured in 12-well plates until nearly confluent, washed twice with PBS, and further cultured in X-Vivo 10 medium (without serum) in the presence or absence of the stimuli or apoptotic cells. The necessity to stimulate the cells in serum-free medium stems from the presence of high levels of latent TGF-β in serum, which interferes with the detection of cell-derived TGF-β by ELISA. Following stimulation, culture supernatants were collected at the indicated times, centrifuged (1500 × g for 5 min at 4°C) to pellet intact cells, and snap-frozen in liquid nitrogen before storage at −70°C. When cell-associated cytokines were measured, 0.5 ml of ice-cold PBS was added to the wells, and macrophages were gently scraped and combined with the small cellular pellet resulting from the centrifugation of culture supernatants. Pooled cells were then spun (2000 × g for 5 min at 4°C); the resulting pellets were snap-frozen in liquid nitrogen and stored at −70°C. Immediately before ELISA analysis, the cell pellets were resuspended in 1 ml of cold lysis buffer (PBS supplemented with 0.5% NP-40, 5 mM EDTA, 0.5 mM AEBSF, 1 mM PMSF, and 10 μg/ml each of aprotinin, leupeptin, and pepstatin A), vigorously mixed for 30 s, and centrifuged (15000 × g for 10 min at 4°C) to remove insoluble material. Cytokine concentrations in culture supernatants or in the corresponding cell lysates were analyzed by ELISA using commercially available kits, according to the manufacturer’s (R&D Systems) instructions. For TGF-β1, samples were acid treated (to activate latent TGF-β) and later neutralized before ELISA analysis, as recommended by the kit manufacturer (Genzyme, Cambridge, MA).

Mouse J774 cells were cultured and stimulated as described above for the ELISA experiments. After the desired incubation times, culture medium was carefully removed, and cell lysates (prepared from ∼3 × 106 cells) were directly analyzed in RPA using the DirectProtect Lysate RPA kit (Ambion, Austin, TX) and multiprobe templates mCK3b or mCK5 (PharMingen, San Diego, CA) that had been transcribed with T7 polymerase and labeled with [α-32P]UTP using a Riboquant kit (PharMingen). Samples were electrophoresed on 5% acrylamide sequencing gels containing 8 M urea in 0.5× TBE; dried gels were exposed to PhosphorScreens (Molecular Dynamics, Sunnyvale, CA) before quantitation of RNA bands using a STORM 840 PhosphorImager and ImageQuant software (Molecular Dynamics). For mRNA stability experiments, cells were pretreated with or without 5 ng/ml TGF-β for 60 min, stimulated with 1 ng/ml LPS, and actinomycin D was added at a final concentration of 5 μg/ml. Cells were then further cultured for the indicated times, and processed for RPA analysis as described above.

Mouse J774 cells were cultured in 6-well plates until nearly confluent; cells were washed twice with PBS and further cultured in X-Vivo 10 medium (without serum) in the presence or absence of the stimuli or apoptotic cells. Incubations were stopped by adding an equal volume of ice-cold X-Vivo 10 medium to the wells, and nuclear extracts were then prepared by a modified Dignam procedure (9), as follows. Cells were collected by gentle scraping and centrifuged at 1000 × g for 3 min at 4°C. The resulting cell pellets were resuspended in ice-cold lysis buffer (10 mM HEPES (pH 7.90), 10 mM NaCl, 1.5 mM MgCl2, 1 mM EDTA, 0.5 mM EGTA, and 0.5 mM DTT) containing an antiprotease mixture (0.5 mM DFP, 0.5 mM AEBSF, 1 mM PMSF, and 10 μg/ml each of aprotinin, leupeptin, and pepstatin A, final concentrations). After a 10-min incubation on ice, an equal volume of lysis buffer containing the antiprotease mixture as well as 0.2% NP-40 was added (to yield a final concentration of 0.1% NP-40). Samples were immediately vortex mixed for 15 s before centrifugation at 1200 × g (5 min at 4°C). The resulting nuclear pellets were washed once with lysis buffer containing the antiprotease mixture before being resuspended in ice-cold nuclear extraction buffer (20 mM HEPES (pH 7.90), 400 mM NaCl, 1.5 mM MgCl2, 1 mM EDTA, 0.5 mM DTT, and 10% (v/v) glycerol) containing the antiprotease mixture. After a 20-min incubation on ice (with frequent mixing), samples were spun (15,000 × g for 15 min at 4°C), and supernatants (the nuclear extracts) were snap-frozen in liquid nitrogen and stored at −70°C. Aliquots of the extracts were routinely processed for protein content determination. Nuclear extracts (amounts used are specified in the figure legends) were analyzed in EMSA as follows. For NF-κB binding, the extracts were processed exactly as described previously (10) using 30,000 cpm of oligonucleotide probe (end-labeled with T4 kinase). For the analysis of AP-1 DNA binding, nuclear extracts were incubated in a modified binding buffer (20 mM HEPES (pH 7.50), 50 mM KCl, 1 mM EDTA, 5 mM DTT, 0.1% NP-40, and 6% glycerol) supplemented with 0.4 μg poly(dI-dC) and 8 μg acetylated BSA before the addition of 30,000 cpm of labeled oligonucleotide probe. For supershift experiments, binding reactions were conducted in the presence of specific antisera to individual NF-κB/Rel, Jun, or Fos proteins (30 min at 4°C), before the addition of 32P-labeled probes. Samples were electrophoresed on 6% acrylamide gels at 4°C in 0.5× TBE; dried gels were then exposed to PhosphorScreens (Molecular Dynamics).

Where mentioned, statistical significance was assessed using the Student’s t test for paired data (one-tailed).

Ingestion of apoptotic cells by human monocyte-derived macrophages (HMDM) profoundly alters their ability to generate pro-inflammatory cytokines and IL-8 (6, 7). We first examined whether apoptotic cell uptake by murine macrophage-like J774 cells similarly affects the production of a cytokine (TNF-α), and of C-X-C chemokines (Mip-2, KC) and C-C chemokines (Mip-1α, MCP-1). As shown in Fig. 2, apoptotic cell uptake markedly decreased the ability of LPS-treated J774 cells to secrete TNF-α, Mip-2, KC, and Mip-1α, whereas the release of MCP-1/JE was not significantly affected. By comparison, the LPS-elicited release of IL-10 was only slightly suppressed following apoptotic cell uptake (Fig. 2). Finally, the secretion of TGF-β1 was substantially increased in J774 cells that had ingested apoptotic cells, both in the presence and absence of LPS (Fig. 2). Similar results were obtained when J774 cells were stimulated with 10 ng/ml LPS instead of 1 ng/ml, or when apoptotic cell uptake took place 3 h before LPS stimulation (data not shown). Table I summarizes the effect of apoptotic cell uptake on chemokine and TNF-α release in J774 cells.

FIGURE 2.

Apoptotic cell uptake modulates chemokine and TNF-α release by J774 macrophages. Mouse J774A.1 cells were cultured for 90 min in the presence (open symbols) or absence (filled symbols) of apoptotic Jurkat cells (at a 1:5 ratio); uningested Jurkat cells were removed by mild washings before further culture in the presence or absence of 1 ng/ml LPS for varying lengths of time. Culture supernatants were then analyzed by ELISA; values are mean ± SD of duplicate samples. This experiment is representative of at least four.

FIGURE 2.

Apoptotic cell uptake modulates chemokine and TNF-α release by J774 macrophages. Mouse J774A.1 cells were cultured for 90 min in the presence (open symbols) or absence (filled symbols) of apoptotic Jurkat cells (at a 1:5 ratio); uningested Jurkat cells were removed by mild washings before further culture in the presence or absence of 1 ng/ml LPS for varying lengths of time. Culture supernatants were then analyzed by ELISA; values are mean ± SD of duplicate samples. This experiment is representative of at least four.

Close modal
Table I.

Effect of apoptotic cell phagocytosis on chemokine and TNF release by macrophagesa

CytokineEffect of Apoptotic Cell Uptake (% control)
In unstimulated J774 cellsIn LPS-stimulated J774 cells
TNF-α 63 ± 3 (n = 4)* 41 ± 5 (n = 11)** 
Mip-2 61 ± 11 (n = 4)* 60 ± 3 (n = 11)** 
KC 35 ± 14 (n = 4)* 50 ± 5 (n = 11)** 
Mip-1α 51 ± 9 (n = 4)* 61 ± 4 (n = 10)** 
MCP-1/JE 102 ± 4 (n = 4) 95 ± 7 (n = 9) 
IL-10 80 ± 9 (n = 6)* 96 ± 4 (n = 11) 
CytokineEffect of Apoptotic Cell Uptake (% control)
In unstimulated J774 cellsIn LPS-stimulated J774 cells
TNF-α 63 ± 3 (n = 4)* 41 ± 5 (n = 11)** 
Mip-2 61 ± 11 (n = 4)* 60 ± 3 (n = 11)** 
KC 35 ± 14 (n = 4)* 50 ± 5 (n = 11)** 
Mip-1α 51 ± 9 (n = 4)* 61 ± 4 (n = 10)** 
MCP-1/JE 102 ± 4 (n = 4) 95 ± 7 (n = 9) 
IL-10 80 ± 9 (n = 6)* 96 ± 4 (n = 11) 
a

Mouse J774 cells were cultured for 90 min in the absence or presence of apoptotic Jurkat cells (at a 1:5 ratio). Uningested Jurkat cells were removed by mild washings prior to further culture for 12 h with or without 1 ng/ml LPS. Culture supernatants were analyzed by ELISA. Values are mean ± SEM of the indicated number of experiments.

b

, Significantly different than unstimulated J774 cells at p ≤ 0.04.

c

∗, Significantly different than LPS-treated J774 cells at p ≤ 0.0001.

Our previous work showed that the effect of apoptotic cell uptake on cytokine production by human macrophages involved several endogenous mediators, foremost among which was TGF-β (7). For this reason, and because TGF-β1 is released early following apoptotic cell uptake by J774 cells, we first examined whether neutralizing anti-TGF-β F(ab′)2 Abs would affect inflammatory cytokine release under these conditions. The need to use F(ab′)2 fragments stems from the fact that whole Abs (as well as irrelevant rabbit IgG) potently induced the release of TNF and chemokines by themselves. As shown in Fig. 3,A, the apoptotic cell-mediated inhibition of chemokine and TNF-α release was substantially reversed by the anti-TGF-β F(ab′)2 Abs both in unstimulated and LPS-treated J774 cells. Neither the constitutive nor the LPS-induced secretion of these mediators was affected by the anti-TGF-β F(ab′)2 Abs alone (Fig. 3,A), consistent with the fact that little or no TGF-β is detected in J774 macrophages that did not ingest apoptotic cells (Fig. 2).

FIGURE 3.

The inhibition of chemokine and TNF-α release resulting from apoptotic cell uptake by J774 macrophages is mainly mediated by endogenous TGF-β. A, Mouse J774A.1 cells were exposed to apoptotic Jurkat lymphocytes (at a 1:5 ratio) for 90 min; uningested Jurkat cells were removed by mild washings, before further culture in the absence (−) or presence of 1 ng/ml LPS for 12 h. The entire experiment was conducted in the continued presence or absence of neutralizing anti-TGF-β F(ab′)2 Abs (7 μg/ml, i.e., the equivalent of 10 μg/ml of whole Ab). Culture supernatants were analyzed by ELISA; values are mean ± SEM of three independent experiments. ∗, Significantly different from matching controls (i.e., without Abs) at p ≤ 0.027. B, Mouse J774 cells were pretreated with either 10 μM indomethacin (indo) or 100 μM WEB 2086 (WEB) before the addition of apoptotic Jurkat cells, as described in A; macrophages were then stimulated with 1 ng/ml LPS for 12 h. The entire experiment was conducted in the continued presence or absence of the drugs. Values are mean ± SEM of at least three independent experiments. C, Mouse J774 cells were cultured in the presence or absence of apoptotic Jurkat cells as described in A and stimulated with 1 ng/ml LPS for 12 h. The entire experiment was conducted in the continued presence or absence of neutralizing anti-IL-10 Abs (3 μg/ml). Values are mean ± SEM of three independent experiments. ∗, Significantly different from matched controls (i.e., without Abs) at p < 0.045.

FIGURE 3.

The inhibition of chemokine and TNF-α release resulting from apoptotic cell uptake by J774 macrophages is mainly mediated by endogenous TGF-β. A, Mouse J774A.1 cells were exposed to apoptotic Jurkat lymphocytes (at a 1:5 ratio) for 90 min; uningested Jurkat cells were removed by mild washings, before further culture in the absence (−) or presence of 1 ng/ml LPS for 12 h. The entire experiment was conducted in the continued presence or absence of neutralizing anti-TGF-β F(ab′)2 Abs (7 μg/ml, i.e., the equivalent of 10 μg/ml of whole Ab). Culture supernatants were analyzed by ELISA; values are mean ± SEM of three independent experiments. ∗, Significantly different from matching controls (i.e., without Abs) at p ≤ 0.027. B, Mouse J774 cells were pretreated with either 10 μM indomethacin (indo) or 100 μM WEB 2086 (WEB) before the addition of apoptotic Jurkat cells, as described in A; macrophages were then stimulated with 1 ng/ml LPS for 12 h. The entire experiment was conducted in the continued presence or absence of the drugs. Values are mean ± SEM of at least three independent experiments. C, Mouse J774 cells were cultured in the presence or absence of apoptotic Jurkat cells as described in A and stimulated with 1 ng/ml LPS for 12 h. The entire experiment was conducted in the continued presence or absence of neutralizing anti-IL-10 Abs (3 μg/ml). Values are mean ± SEM of three independent experiments. ∗, Significantly different from matched controls (i.e., without Abs) at p < 0.045.

Close modal

By comparison, neither indomethacin (a cyclooxygenase inhibitor) nor WEB 2086 (a PAF receptor antagonist) significantly altered the secretion of chemokines and TNF-α in J774 cells that had ingested apoptotic cells (Fig. 3,B). Finally, neutralizing anti-IL-10 Abs only marginally affected the release of these mediators by LPS-treated J774 cells, whether or not the macrophages had ingested apoptotic targets (Fig. 3,C). This is consistent with the fact that J774 cells release very little IL-10 under these conditions (Fig. 2). Taken together, these results clearly show that in J774 cells, the repression of TNF-α and chemokine release resulting from apoptotic cell uptake is mainly mediated by endogenous TGF-β (as opposed to IL-10, PAF, or prostanoids). Thus, the interaction between TGF-β and J774 cells is likely to represent a useful model to investigate the mechanisms whereby apoptotic cell ingestion regulates inflammatory mediator production.

The above findings led us to investigate the effect of exogenous TGF-β on the secretion of TNF-α and chemokines. As shown in Fig. 4, preincubation of J774 cells with TGF-β1 markedly inhibited the LPS-induced release of TNF-α, Mip-2, KC, and Mip-1α, whereas MCP-1/JE release was only marginally decreased. This TGF-β-mediated inhibition was already evident at early time points (Fig. 4) and was found to be dose-dependent, a threshold effect being observed using as little as 100 pg/ml TGF-β1 (data not shown). Table II summarizes the effect of exogenous TGF-β on chemokine and TNF-α release in J774 cells. In addition, we found that pretreatment of J774 cells with TGF-β followed by washing and subsequent LPS stimulation in the absence of TGF-β was just as effective in inhibiting inflammatory mediator release as when J774 cells were stimulated with LPS in the continued presence of TGF-β (Fig. 5,A). Thus, mere exposure of the macrophages to TGF-β is sufficient to down-regulate TNF-α and chemokine secretion. Finally, pretreatment of J774 macrophages with TGF-β for up to 20 h still resulted in a marked inhibition of chemokine and TNF secretion (Fig. 5,B); addition of TGF-β 1 h after LPS stimulation also led to the same result (Fig. 5 B).

FIGURE 4.

Exogenous TGF-β modulates chemokine and TNF-α release in J774 macrophages. Mouse J774A.1 cells were cultured for 90 min with 5 ng/ml TGF-β1 (open symbols) or diluent control (filled symbols), before stimulation with 1 ng/ml LPS for the indicated times. Culture supernatants were analyzed by ELISA; values are mean ± SD of duplicate samples. This experiment is representative of at least three.

FIGURE 4.

Exogenous TGF-β modulates chemokine and TNF-α release in J774 macrophages. Mouse J774A.1 cells were cultured for 90 min with 5 ng/ml TGF-β1 (open symbols) or diluent control (filled symbols), before stimulation with 1 ng/ml LPS for the indicated times. Culture supernatants were analyzed by ELISA; values are mean ± SD of duplicate samples. This experiment is representative of at least three.

Close modal
Table II.

Effect of exogenous TGF-β on chemokine and TNF release by macrophagesa

CytokineEffect of exogenous TGF-β (% control)
In unstimulated J774 cellsIn LPS-stimulated J774 cells
TNF-α 78 ± 7 (n = 9)* 42 ± 4 (n = 12)** 
Mip-2 60 ± 7 (n = 9)* 38 ± 5 (n = 14)** 
KC 23 ± 9 (n = 6)* 41 ± 5 (n = 8)** 
Mip-1α 55 ± 11 (n = 4)* 49 ± 6 (n = 5)** 
MCP-1/JE 98 ± 2 (n = 5) 85 ± 7 (n = 9)** 
IL-10 95 ± 22 (n = 6) 46 ± 11 (n = 9)** 
CytokineEffect of exogenous TGF-β (% control)
In unstimulated J774 cellsIn LPS-stimulated J774 cells
TNF-α 78 ± 7 (n = 9)* 42 ± 4 (n = 12)** 
Mip-2 60 ± 7 (n = 9)* 38 ± 5 (n = 14)** 
KC 23 ± 9 (n = 6)* 41 ± 5 (n = 8)** 
Mip-1α 55 ± 11 (n = 4)* 49 ± 6 (n = 5)** 
MCP-1/JE 98 ± 2 (n = 5) 85 ± 7 (n = 9)** 
IL-10 95 ± 22 (n = 6) 46 ± 11 (n = 9)** 
a

Mouse J774 cells were incubated for 60–90 min with 5 ng/ml TGF-β1 (or diluent control) and further cultured with or without 1 ng/ml LPS for 12 h. Culture supernatants were analyzed by ELISA. Values are mean ± SEM of the indicated number of experiments.

b

, Significantly different than unstimulated J774 cells at p ≤ 0.014.

c

∗, Significantly different than LPS-treated J774 cells at p ≤ 0.031.

FIGURE 5.

Effect of the continuous presence of TGF-β, and of the pre-incubation time with TGF-β on chemokine release by J774 macrophages. A, Mouse J774A.1 cells were cultured for 90 min in the absence of TGF-β1 and stimulated for 12 h in the presence of 1 ng/ml LPS (open bar). Alternatively, macrophages were pretreated for 90 min in the presence of 5 ng/ml TGF-β1; cells were washed free of the culture medium and subsequently cultured for 12 h, either in the combined presence of both TGF-β1 and 1 ng/ml LPS (shaded bar) or in the presence of LPS only (filled bar). Culture supernatants were then analyzed by ELISA for Mip-2; similar results were obtained when the supernatants were analyzed for TNF-α, KC, or Mip-1α release (data not shown). Values are mean ± SEM of three independent experiments. Chemokine release was not significantly different (p = 0.43, n = 3) between cells stimulated in the continued presence of TGF-β (shaded bar) and cells stimulated without TGF-β (filled bar). B, Mouse J774 cells were cultured for varying lengths of time in the absence (open bar) or presence (shaded bars) of 5 ng/ml TGF-β1 and further stimulated for 12 h in the presence of 1 ng/ml LPS. Culture supernatants were then analyzed by ELISA for Mip-2; similar results were obtained when the supernatants were analyzed for TNF-α, KC, or Mip-1α release (data not shown). Values are mean ± SEM of three independent experiments. In all conditions tested, the inhibition by TGF-β of chemokine release was significant at p < 0.038.

FIGURE 5.

Effect of the continuous presence of TGF-β, and of the pre-incubation time with TGF-β on chemokine release by J774 macrophages. A, Mouse J774A.1 cells were cultured for 90 min in the absence of TGF-β1 and stimulated for 12 h in the presence of 1 ng/ml LPS (open bar). Alternatively, macrophages were pretreated for 90 min in the presence of 5 ng/ml TGF-β1; cells were washed free of the culture medium and subsequently cultured for 12 h, either in the combined presence of both TGF-β1 and 1 ng/ml LPS (shaded bar) or in the presence of LPS only (filled bar). Culture supernatants were then analyzed by ELISA for Mip-2; similar results were obtained when the supernatants were analyzed for TNF-α, KC, or Mip-1α release (data not shown). Values are mean ± SEM of three independent experiments. Chemokine release was not significantly different (p = 0.43, n = 3) between cells stimulated in the continued presence of TGF-β (shaded bar) and cells stimulated without TGF-β (filled bar). B, Mouse J774 cells were cultured for varying lengths of time in the absence (open bar) or presence (shaded bars) of 5 ng/ml TGF-β1 and further stimulated for 12 h in the presence of 1 ng/ml LPS. Culture supernatants were then analyzed by ELISA for Mip-2; similar results were obtained when the supernatants were analyzed for TNF-α, KC, or Mip-1α release (data not shown). Values are mean ± SEM of three independent experiments. In all conditions tested, the inhibition by TGF-β of chemokine release was significant at p < 0.038.

Close modal

By comparison, IL-10 potently inhibited the secretion of chemokines and TNF-α by J774 macrophages (Fig. 6); at a concentration of 100 U/ml, IL-10 was about as effective as 5 ng/ml TGF-β1. This is consistent with previous reports showing that IL-10 inhibits LPS-induced cytokine/chemokine secretion in macrophages (11, 12, 13, 14). A notable exception was MCP-1/JE, whose secretion was dramatically augmented by IL-10, even in LPS-stimulated cells (Fig. 6). That IL-10 can increase MCP-1 production has already been reported in monocytes/macrophages (15, 16, 17), as well as in fibroblasts and endothelial cells (18, 19). Thus, while endogenous IL-10 does not appear to mediate the inhibitory effect of apoptotic cells in our system, this cannot be attributed to a poor responsiveness of J774 cells to IL-10.

FIGURE 6.

Exogenous IL-10 modulates chemokine and TNF-α release by J774 macrophages. Mouse J774A.1 cells were incubated for 60–90 min in the presence or absence of 100 U/ml IL-10 and further cultured for 12 h with or without of 1 ng/ml LPS. Values are mean ± SEM of four independent experiments. In unstimulated cells (unstim), the modulation by IL-10 of inflammatory mediator release was significant at p ≤ 0.034, except in the case of Mip-2. In LPS-stimulated cells (LPS), the modulation by IL-10 of mediator release was significant at p ≤ 0.026.

FIGURE 6.

Exogenous IL-10 modulates chemokine and TNF-α release by J774 macrophages. Mouse J774A.1 cells were incubated for 60–90 min in the presence or absence of 100 U/ml IL-10 and further cultured for 12 h with or without of 1 ng/ml LPS. Values are mean ± SEM of four independent experiments. In unstimulated cells (unstim), the modulation by IL-10 of inflammatory mediator release was significant at p ≤ 0.034, except in the case of Mip-2. In LPS-stimulated cells (LPS), the modulation by IL-10 of mediator release was significant at p ≤ 0.026.

Close modal

We next investigated whether the inhibition of inflammatory mediator release by either apoptotic cell ingestion or exogenous TGF-β occurs at the level of secretion and/or protein synthesis. To this end, cell-associated levels of chemokines and TNF-α were analyzed by ELISA. Fig. 7 shows that the bulk of the chemokines being synthesized is secreted, and that chemokines (Mip-2, KC, Mip-1α) are more efficiently secreted than TNF-α. More importantly, both TGF-β and apoptotic cell uptake inhibited the cell-associated levels of chemokines and TNF-α to a similar extent as when secretion was examined (Fig. 7). Thus, apoptotic cells and TGF-β appear to inhibit inflammatory mediator production by acting at the level of, or upstream from, translation.

FIGURE 7.

Effect of exogenous TGF-β and of apoptotic cell uptake on the synthesis and release of chemokines and TNF-α by J774 macrophages. Mouse J774A.1 cells were cultured for 60 to 90 min in the absence or presence of either apoptotic Jurkat cells (at a 1:5 ratio) or 5 ng/ml TGF-β1; macrophages were then further cultured in the presence of 1 ng/ml LPS for 12 h. Culture supernatants and the corresponding cell pellets were then analyzed by ELISA for TNF-α, Mip-2, KC, and Mip-1α content; in the case of KC, only secretion is shown, as cell-associated levels were consistently undetectable. Values are mean ± SEM of at least five independent experiments.

FIGURE 7.

Effect of exogenous TGF-β and of apoptotic cell uptake on the synthesis and release of chemokines and TNF-α by J774 macrophages. Mouse J774A.1 cells were cultured for 60 to 90 min in the absence or presence of either apoptotic Jurkat cells (at a 1:5 ratio) or 5 ng/ml TGF-β1; macrophages were then further cultured in the presence of 1 ng/ml LPS for 12 h. Culture supernatants and the corresponding cell pellets were then analyzed by ELISA for TNF-α, Mip-2, KC, and Mip-1α content; in the case of KC, only secretion is shown, as cell-associated levels were consistently undetectable. Values are mean ± SEM of at least five independent experiments.

Close modal

To determine whether the inhibition of inflammatory mediator synthesis is paralleled by similar changes in mRNA levels, RPA analyses were performed using J774 cells that had either ingested apoptotic Jurkat cells or been pretreated with TGF-β. In the former instance, a human GAPDH band (originating from the ingested Jurkat cells) hybridized with our RNA probes, which precluded a reliable quantitation of the mouse cytokine mRNA bands in initial experiments. However, this difficulty was overcome by deliberately overrunning the gels, which effectively resolved the human and mouse GAPDH bands (Fig. 8,A). As shown in Fig. 8,B, the constitutive gene expression of the inflammatory mediators was generally unaltered by TGF-β or apoptotic cell uptake, with the exception of TNF-α and Mip-2 mRNA, which were modestly down-regulated. In LPS-stimulated cells, however, TGF-β pretreatment and apoptotic cell uptake markedly inhibited the accumulation of Mip-2 and Mip-1α mRNA, whereas that of TNF-α mRNA was not significantly affected (Fig. 8,B). In contrast, the LPS-elicited accumulation of MCP-1/JE mRNA was enhanced following TGF-β pretreatment (Fig. 8,B). Finally, TGF-β1 mRNA levels were unaffected by exogenous TGF-β, apoptotic cell uptake, and/or LPS stimulation (Fig. 8,B). Noteworthy is that only the β1 isoform of TGF was consistently detected in J774 cells, as mRNA encoding the β2 and β3 isoforms were either weak or undetectable (Fig. 8,A). In a final set of experiments, we also determined that chemokine and TNF-α mRNA stability remained essentially the same, whether or not J774 cells were pretreated with TGF-β before LPS stimulation (Fig. 9). Thus, it appears that apoptotic cell uptake and TGF-β inhibit stimulated TNF-α production via a translational effect, whereas the decreased production of Mip-1α and Mip-2 probably involves transcriptional events.

FIGURE 8.

Effect of exogenous TGF-β and of apoptotic cell uptake on chemokine and TNF-α gene expression in J774 macrophages. Mouse J774A.1 cells were cultured for 60–90 min in the absence (−) or presence of either apoptotic Jurkat cells (Apo) or 5 ng/ml TGF-β1, before further culture with or without 1 ng/ml LPS for 60–90 min. Cells were lysed directly in the culture plates, and lysates were analyzed in RPA as described in Materials andMethods. A, Representative gel is shown (MIF, macrophage inhibitory factor; mGAPDH, mouse GAPDH; hGAPDH, human GAPDH). B, After PhosphorImager quantitation of the mRNA bands, chemokine mRNA values were first expressed as a ratio to the corresponding GAPDH values (to normalize for loading); the resulting ratios were then compared with those obtained for the matching LPS samples. Values are mean ± SEM of at least four experiments (TGF-β vs unstimulated cells), at least eight experiments (TGF-β vs LPS-treated cells), and three experiments (apoptotic cell uptake). ∗, Significantly different from matched controls at p < 0.01; ∗∗, significantly different from matched control at p = 0.044.

FIGURE 8.

Effect of exogenous TGF-β and of apoptotic cell uptake on chemokine and TNF-α gene expression in J774 macrophages. Mouse J774A.1 cells were cultured for 60–90 min in the absence (−) or presence of either apoptotic Jurkat cells (Apo) or 5 ng/ml TGF-β1, before further culture with or without 1 ng/ml LPS for 60–90 min. Cells were lysed directly in the culture plates, and lysates were analyzed in RPA as described in Materials andMethods. A, Representative gel is shown (MIF, macrophage inhibitory factor; mGAPDH, mouse GAPDH; hGAPDH, human GAPDH). B, After PhosphorImager quantitation of the mRNA bands, chemokine mRNA values were first expressed as a ratio to the corresponding GAPDH values (to normalize for loading); the resulting ratios were then compared with those obtained for the matching LPS samples. Values are mean ± SEM of at least four experiments (TGF-β vs unstimulated cells), at least eight experiments (TGF-β vs LPS-treated cells), and three experiments (apoptotic cell uptake). ∗, Significantly different from matched controls at p < 0.01; ∗∗, significantly different from matched control at p = 0.044.

Close modal
FIGURE 9.

Effect of exogenous TGF-β on the stability of chemokine and TNF-α mRNA in LPS-stimulated J774 macrophages. Mouse J774A.1 cells were cultured for 60 min with 5 ng/ml TGF-β1 (▪) or diluent control (□), before stimulation with 1 ng/ml LPS for 60–90 min. Actinomycin D was then added, and cells were collected at the indicated times; cytokine mRNA levels were assessed by RPA, as described in the legend to Fig. 8 B. Values are mean ± SEM from at least three independent experiments.

FIGURE 9.

Effect of exogenous TGF-β on the stability of chemokine and TNF-α mRNA in LPS-stimulated J774 macrophages. Mouse J774A.1 cells were cultured for 60 min with 5 ng/ml TGF-β1 (▪) or diluent control (□), before stimulation with 1 ng/ml LPS for 60–90 min. Actinomycin D was then added, and cells were collected at the indicated times; cytokine mRNA levels were assessed by RPA, as described in the legend to Fig. 8 B. Values are mean ± SEM from at least three independent experiments.

Close modal

We finally investigated whether TGF-β or apoptotic cell uptake alter the LPS-induced activation of transcription factors NF-κB and AP-1 in J774 cells. The rationale for these experiments is that the LPS-inducible gene expression of the inflammatory mediators under investigation is largely dependent upon the activation of NF-κB or AP-1, or both (20, 21, 22, 23, 24, 25, 26). As shown in Fig. 10,A, TGF-β did not markedly affect the constitutive NF-κB DNA-binding activity detected in nuclear extracts from J774 cells, although a small increase (up to 2-fold) was observed in three of nine independent experiments. By contrast, stimulation of J774 cells with LPS consistently resulted in a dramatic induction of NF-κB activity (Fig. 10,A). Supershift analyses revealed that both the constitutive and LPS-inducible NF-κB complexes contain RelA and p50 NF-κB1 (data not shown). Pretreatment of J774 cells with TGF-β or PGE2 before LPS stimulation did not alter the intensity of the NF-κB signal (Fig. 10,A), regardless of the time of addition of TGF-β or PGE2 (from 90 min before to 30 min after LPS), and regardless of the time elapsed after LPS stimulation (from 15 to 180 min). Similar results were obtained when HMDM or mouse RAW 264.7 cells were used instead of J774 cells. When the same samples were analyzed in EMSA using an AP-1 probe, constitutive AP-1 binding was somewhat decreased by TGF-β and moderately enhanced by PGE2, whereas it was potently increased by LPS (Fig. 10,B); the AP-1 complex was found to contain c-Jun, JunD, and c-Fos (data not shown). Pretreatment of J774 cells with TGF-β moderately enhanced AP-1 activation by LPS, whereas PGE2 pretreatment had no effect (Fig. 10 B). Thus, it does not appear that TGF-β inhibits inflammatory mediator production at the level of NF-κB or AP-1 binding in macrophages.

FIGURE 10.

Effect of exogenous TGF-β and of apoptotic cell uptake toward transcription factor activation in J774 macrophages. A, Mouse J774A.1 cells were cultured for 60 min in the absence (−) or presence of 5 ng/ml TGF-β1 or 10 nM PGE2 before further culture with or without 1 ng/ml LPS for 45 min. Nuclear extracts (0.5 μg) were analyzed in EMSA using a NF-κB oligonucleotide probe. This experiment is representative of four PGE2-treated samples and of nine TGF-β-treated samples. B, The same samples as in A were analyzed in EMSA using an AP-1 oligonucleotide probe. C, Mouse J774 cells were cultured for 60 min in medium alone (−) or allowed to phagocytose apoptotic Jurkat cells, and further cultured with or without 1 ng/ml LPS for 60 min. Nuclear extracts were analyzed in EMSA (110,000 cell equivalents/well) using a NF-κB oligonucleotide probe. The need to load using cell equivalents stems from the fact that nuclear extracts derived from macrophages that ingested apoptotic cells contained six to eight times more protein than when macrophages were preincubated in medium alone. This experiment is representative of four.

FIGURE 10.

Effect of exogenous TGF-β and of apoptotic cell uptake toward transcription factor activation in J774 macrophages. A, Mouse J774A.1 cells were cultured for 60 min in the absence (−) or presence of 5 ng/ml TGF-β1 or 10 nM PGE2 before further culture with or without 1 ng/ml LPS for 45 min. Nuclear extracts (0.5 μg) were analyzed in EMSA using a NF-κB oligonucleotide probe. This experiment is representative of four PGE2-treated samples and of nine TGF-β-treated samples. B, The same samples as in A were analyzed in EMSA using an AP-1 oligonucleotide probe. C, Mouse J774 cells were cultured for 60 min in medium alone (−) or allowed to phagocytose apoptotic Jurkat cells, and further cultured with or without 1 ng/ml LPS for 60 min. Nuclear extracts were analyzed in EMSA (110,000 cell equivalents/well) using a NF-κB oligonucleotide probe. The need to load using cell equivalents stems from the fact that nuclear extracts derived from macrophages that ingested apoptotic cells contained six to eight times more protein than when macrophages were preincubated in medium alone. This experiment is representative of four.

Close modal

When similar experiments were performed using J774 or human macrophages that had ingested apoptotic cells, nuclear extracts consistently contained six to eight times more protein than extracts from macrophages that had not phagocytosed apoptotic targets, thereby preventing a reliable EMSA analysis. To circumvent this problem, gels were instead loaded using cell equivalents. As shown in Fig. 10 C, apoptotic cell uptake by itself moderately enhanced constitutive NF-κB binding, whereas it had little or no effect on NF-κB activation by LPS; similar results were obtained using an AP-1 probe (data not shown). Thus, the modulation of inflammatory mediator production resulting from apoptotic cell uptake is not likely to reflect altered binding of these transcription factor in macrophages.

Phagocytosis of apoptotic cells has been shown to inhibit the production of inflammatory cytokines and eicosanoids by HMDM (6, 7). In the current study, we extend these findings by showing that apoptotic cell uptake also suppresses the generation of several chemokines in mouse J774 macrophages, and by exploring the mechanism of this inhibition. Although we used human lymphocyte-like Jurkat cells as apoptotic targets in most of our experiments, ingestion of apoptotic human neutrophils by J774 cells had a similar outcome (our unpublished data). Likewise, inflammatory mediator production is down-regulated following the uptake of human eosinophils, neutrophils, or Jurkat cells, both in HMDM (6, 7) and in murine bone marrow-derived macrophages (our unpublished data). This indicates that the inhibition of macrophage function by apoptotic cells (even from a different species) is not specific to the apoptotic target. Another important feature of this phenomenon is that it occurs in unstimulated mouse and human macrophages, as well as in J774 cells stimulated with either LPS or TNF-α, and in LPS- or zymosan-stimulated HMDM (7). Thus, in macrophages that have ingested apoptotic cells, the suppression of inflammatory mediator production (and induction of TGF-β) appears to be a general phenomenon that does not depend on the nature of the apoptotic target, the macrophage activation state, or the type of stimulus used. The importance of this phenomenon is illustrated by the fact that chemokine production by J774 cells remained inhibited for 20 h after exposure to TGF-β (Fig. 5).

We also report that the inhibitory effect of apoptotic cell uptake is largely mediated by endogenous factors, among which TGF-β appears to play a paramount role. First, TGF-β is the only cytokine examined, whose release was up-regulated following ingestion of apoptotic cells, both in unstimulated and LPS-treated J774 cells. More compelling evidence is that anti-TGF-β neutralizing Abs largely reversed the inhibitory effects of apoptotic cell uptake. Finally, exogenously added TGF-β closely mimicked the suppressive action of apoptotic cell ingestion. These observations, together with the fact that endogenous TGF-β plays a similar role in primary human macrophages (7) and murine bone marrow-derived macrophages, suggest that TGF-β represents the main endogenous factor mediating the inhibitory effect of apoptotic cells. Although this phenomenon may also involve other endogenous factors (namely, PGE2 and PAF) in HMDM (7), both products proved to be potent inducers TGF-β release in these cells (7). In J774 macrophages, however, the suppressive effect of apoptotic cell uptake was unaffected by a PAF receptor antagonist and a cyclooxygenase inhibitor. Accordingly, J774 cells fail to synthesize detectable amounts of PAF in response to 1 ng/ml LPS (P. Henson and R. Murphy, unpublished data), and only release very small amounts of PGD2 and E2, amounting to <140 pM after 2–4 h (C. Amura and P. Henson, unpublished data). Such PGE2 levels are in keeping with results obtained in previous studies performed in J774 cells (27). Therefore, our data indicate that endogenous TGF-β can effectively mediate the action of apoptotic cells without an additional requirement for prostanoids or PAF. Such a central role for TGF-β emphasizes the necessity to elucidate the mechanisms regulating TGF-β1 induction in macrophages undergoing phagocytosis of apoptotic cells. In this regard, we showed herein that these mechanisms are probably posttranscriptional because TGF-β1 mRNA levels were unaffected by apoptotic cell uptake. Whether the regulation involves mRNA stabilization, increased translation or secretion, or activation of latent TGF-β currently represents an area of ongoing investigation in our laboratory.

The fact that neutralizing anti-TGF Abs did not completely block the effect of apoptotic cells raises the possibility that additional endogenous factors may also act as minor participants. Although the anti-inflammatory cytokine, IL-10, would theoretically represent an attractive candidate, its secretion was not increased following apoptotic cell uptake by J774 cells (this paper) or HMDM (7). Accordingly, we observed that although MCP-1/JE release was dramatically up-regulated by IL-10, no increase in MCP-1/JE secretion occurred in J774 cells that had ingested apoptotic targets. Finally, anti-IL-10-neutralizing Abs only marginally increased the production of TNF-α, KC, Mip-2, and Mip-1α in LPS-stimulated J774 cells, whether apoptotic cells had been engulfed or not. That similar observations were made in HMDM (7), further argues against a significant role for endogenous IL-10. In contrast to macrophages, human monocytes have been reported to release more IL-10 following coincubation with apoptotic lymphocytes and to secrete less TNF-α, IL-1β, and IL-12 under these conditions (28). Despite this, no evidence for an autocrine role of IL-10 was presented in that study. This being said, it remains likely that IL-10 would contribute to inhibit macrophage function in vivo, should the cytokine be produced by other cells in the microenvironment.

Our investigation of the mechanisms underlying the suppressive effect of apoptotic cell uptake and endogenous TGF-β on macrophage function revealed that the decreased TNF-α synthesis and release by LPS-stimulated J774 cells was not accompanied by changes in gene expression or mRNA stability, which points to a translational inhibition. This conclusion is consistent with studies performed in mouse peritoneal macrophages, in which TGF-β was found to reduce the LPS-elicited production of TNF-α by acting at a posttranscriptional step (29, 30). By contrast, we reported that apoptotic cell uptake down-regulated TNF-α mRNA levels in HMDM (7). This may be related to the concurrent increase PGE2 synthesis observed under these conditions, as PGE2 has long been known to inhibit TNF gene expression in LPS-stimulated mouse macrophages and HMDM (31, 32). In contrast to TNF-α, the secretion, total protein synthesis, and mRNA accumulation of Mip-2, Mip-1α and KC were all down-regulated to a similar extent in LPS-treated J774 cells following TGF-β exposure or apoptotic cell uptake, whereas no significant changes in mRNA stability were observed. This indicates that transcriptional inhibition is the mechanism whereby TGF-β represses inflammatory chemokine production. It must be stressed, however, that these results do not completely rule out an additional inhibitory mechanism at the level of translation, especially in view of the fact that addition of TGF-β 1 h after LPS inhibited chemokine production to a similar extent as when TGF-β was added before LPS (Fig. 5 B). This being said, the exact mechanism by which transcription is inhibited remains elusive, in view of the fact that TGF-β pretreatment and apoptotic cell uptake did not alter NF-κB or AP-1 DNA binding in LPS-stimulated J774 cells or HMDM. One possibility is that TGF-β and apoptotic cell uptake somehow impair the ability of NF-κB to transactivate chemokine promoters, despite normal binding. However, this implies that the expression of other κB-dependent genes (such as the one encoding TNF-α) would be similarly decreased, and we have shown that this is not the case. Moreover, alterations of the transactivating potential of NF-κB complexes would not explain the decreased activity of the Mip-1α gene, which does not contain κB or AP-1 motifs in its LPS response region (22). Alternatively, TGF-β could promote the binding of factors to repressor elements within chemokine gene promoters. Studies are in progress to further delineate the nature of the transcriptional events being affected by apoptotic cell uptake and TGF-β in activated macrophages.

In contrast to the other chemokines, MCP-1 release was only marginally inhibited by TGF-β, and this was accompanied by a moderate up-regulation at the mRNA level. Although the protein and mRNA data are in apparent contradiction in the latter instance, this may reflect a translational inhibition by TGF-β (as observed with TNF-α, and possibly with the other chemokines as well), especially because no difference in MCP-1/JE mRNA stability was noted between J774 cells cultured in the presence or absence of TGF-β. The latter observation also indicates that an increase in MCP-1 mRNA levels is likely to result from an enhanced transcription of the MCP-1/JE gene. In this regard, several studies have established that MCP-1/JE gene transcription is principally AP-1 driven (24, 25, 26) and accordingly, we found that TGF-β pretreatment moderately enhanced AP-1 activation in LPS-treated J774 cells. Thus, TGF-β appears to regulate the LPS-elicited production of MCP-1/JE by acting on at least two discrete steps. By comparison, MCP-1/JE production and gene expression were unaltered by apoptotic cell uptake. This may stem from the fact that exogenous TGF-β generally exerts more pronounced effects than apoptotic cell ingestion, presumably because J774 cells are exposed to less TGF-β in the latter instance. From a more general standpoint, our data are consistent with the fact that MCP-1 mRNA levels are also enhanced by TGF-β in other cell types, such as astrocytes, osteoblasts, and bone marrow stromal cells (33, 34, 35). However, our observations are in contrast with a recent report (36) in which similar concentrations of TGF-β inhibited the gene expression and release of MCP-1/JE in mouse primary macrophages and murine macrophage cell lines. Although cells were pretreated with TGF-β for 6 h in that study, we only observed a slight decrease in MCP-1/JE release under these conditions (our unpublished data). Macrophages were also stimulated with a LPS concentration 1000-fold higher than the one used herein, but we find it unlikely that this would explain our divergent results. Thus, unless the latter are related to the different J774 sublines used (i.e., J774.2 vs J774A.1 herein), we can offer no explanation for this discrepancy.

In a broader context, the cytokine production profile of macrophages that have ingested apoptotic cells could have profound implications in an inflammatory setting. In addition to producing lesser amounts of pro-inflammatory cytokines (such as TNF-α and IL-1β), macrophages are likely to exert an overall immunosuppressive action in the microenvironment through the production of TGF-β. Similarly, macrophages are believed to represent a major source of neutrophil chemoattractants; therefore, the decreased production of chemokines such as Mip-2 and KC by macrophages is likely to limit neutrophil recruitment into inflammatory sites, and the related tissue damage. Finally, the fact that MCP-1/JE production is not attenuated under the same conditions might result in the selective recruitment of monocytes into inflammatory sites. This could further contribute to the resolution of the inflammatory reaction, in view of the reported ability of monocytes to release anti-inflammatory cytokines such as IL-10 once they come in contact with apoptotic cells (28). Importantly, all of the effects reported herein were achieved despite the fact that only one quarter of the J774 macrophages actually ingested apoptotic targets, which further emphasizes the anti-inflammatory potential of this process. Collectively, the consequences of apoptotic cell uptake by macrophages outlined above lend support to the notion that apoptotic cell clearance from inflamed sites must constitute an important mechanism for the resolution of inflammation. Conversely, dysfunctions in the ability of macrophages to ingest apoptotic cells (and to release TGF-β) might represent an important component of several inflammatory pathologies, as illustrated by the fact that TGF-β1 knockout mice are afflicted by severe and generalized inflammation (37, 38).

We thank Dr. Claudia Amura for kindly providing apoptotic Jurkat cells on several occasions.

1

This work was supported by grants to P.M.H. from the National Institutes of Health (GM 48211, HL 60980, and HL 34303). P.P.M. is a Centennial Postdoctoral Fellow of the Medical Research Council of Canada.

3

Abbreviations used in this paper: MCP-1, monocyte chemoattractant protein-1; DFP, diisopropyl fluorophosphate; PMSF, phenylethanesulfonyl fluoride; AEBSF, 4-(2-aminomethyl)benzenesulfonyl fluoride; HMDM, human monocyte-derived macrophages; Mip, macrophage inflammatory protein; NP-40, Nonidet P-40; PAF, platelet-activating factor; RPA, ribonuclease protection assay.

1
Savill, J., V. Fadok, P. Henson, C. Haslett.
1993
. Phagocyte recognition of cells undergoing apoptosis.
Immunol. Today
14
:
131
2
Grigg, J. M., J. S. Savill, C. Sarraf, C. Haslett, M. Silverman.
1991
. Neutrophil apoptosis and clearance from neonatal lungs.
Lancet
338
:
720
3
Cox. G., J., J. Crossley, Z. Xing.
1995
. Macrophage engulfment of apoptotic neutrophils contributes to the resolution of acute pulmonary inflammation in vivo.
Am. J. Respir. Cell Mol. Biol.
12
:
232
4
Haslett, C., J. S. Savill, M. K. Whyte, M. Stern, I. Dransfield, L. C. Meagher.
1994
. Granulocyte apoptosis and the control of inflammation.
Philos. Trans. R. Soc. London
345
:
327
5
Meagher, L. C., J. S. Savill, A. Baker, R. W. Fuller, C. Haslett.
1992
. Phagocytosis of apoptotic neutrophils does not induce macrophage release of thromboxane B2.
J. Leukocyte Biol.
52
:
269
6
Stern, M., J. Savill, C. Haslett.
1996
. Human monocyte-derived macrophage phagocytosis of senescent eosinophils undergoing apoptosis: Mediation by αvβ3/CD36/thrombospondin recognition mechanism and lack of phlogistic response.
Am. J. Pathol.
149
:
911
7
Fadok, V. A., D. L. Bratton, A. Konowal, P. W. Freed, J. Y. Westcott, P. M. Henson.
1998
. Macrophages that have ingested apoptotic cells in vitro inhibit proinflammatory cytokine production through autocrine/paracrine mechanisms involving TGF-β, PGE2, and PAF.
J. Clin. Invest.
101
:
890
8
Pradhan, D., S. Krahling, P. Williamson, R. A. Schlegel.
1997
. Multiple systems for recognition of apoptotic lymphocytes by macrophages.
Mol. Biol. Cell
8
:
767
9
Dignam, J. D., R. M. Lebovitz, R. G. Roeder.
1983
. Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei.
Nucleic Acids Res.
11
:
1475
10
McDonald, P. P., A. Bald, M. A. Cassatella.
1997
. Activation of the NF-κB pathway by inflammatory stimuli in human neutrophils.
Blood
89
:
3421
11
Bogdan, C., Y. Vodovotz, C. Nathan.
1991
. Macrophage deactivation by interleukin-10.
J. Exp. Med.
174
:
1549
12
Fiorentino, D. F., A. Zlotnik, T. R. Mosmann, M. Howard, A. O’Garra.
1991
. IL-10 inhibits cytokine production by activated macrophages.
J. Immunol.
147
:
3815
13
Berkman, N., M. John, G. Roesems, P. J. Jose, P. J. Barnes, K. F. Chung.
1995
. Inhibition of macrophage inflammatory protein-1α expression by IL-10: differential sensitivities in human blood monocytes and alveolar macrophages.
J. Immunol.
155
:
4412
14
Flesch, I. E., J. Barsig, S. H. Kaufmann.
1998
. Differential chemokine response of murine macrophages stimulated with cytokines and infected with Listeria monocytogenes.
Int. Immunol.
10
:
757
15
Seitz, M., P. Loetscher, B. Dewald, H. Towbin, H. Gallati, M. Baggiolini.
1995
. Interleukin-10 differentially regulates cytokine inhibitor and chemokine release from blood mononuclear cells and fibroblasts.
Eur. J. Immunol.
25
:
1129
16
Sironi, M., C. Munoz, T. Pollicino, A. Siboni, F. L. Sciacca, S. Bernasconi, A. Vecchi, F. Colotta, A. Mantovani.
1993
. Divergent effects of interleukin-10 on cytokine production by mononuclear phagocytes and endothelial cells.
Eur. J. Immunol.
23
:
2692
17
Yano, S., H. Yanagawa, Y. Nishioka, N. Mukaida, K. Matsushima, S. Sone.
1996
. T helper 2 cytokines differently regulate monocyte chemoattractant protein-1 production by human peripheral blood monocytes and alveolar macrophages.
J. Immunol.
157
:
2660
18
Lukacs, N. W., S. W. Chensue, R. E. Smith, R. M. Strieter, K. Warmington, C. Wilke, S. L. Kunkel.
1994
. Production of monocyte chemoattractant protein-1 and macrophage inflammatory protein-1α by inflammatory granuloma fibroblasts.
Am. J. Pathol.
144
:
711
19
Mantovani, A., F. Bussolino, E. Dejana.
1992
. Cytokine regulation of endothelial cell function.
FASEB J.
6
:
2591
20
Collart, M. A., P. Baeuerle, P. Vassalli.
1990
. Regulation of tumor necrosis factor α transcription in macrophages: involvement of four κB-like motifs and of constitutive and inducible forms of NF-κB.
Mol. Cell. Biol.
10
:
1498
21
Shakhov, A. N., M. A. Collart, P. Vassalli, S. A. Nedospasov, C. V. Jongeneel.
1990
. κB-type enhancers are involved in lipopolysaccharide-mediated transcriptional activation of the tumor necrosis factor α gene in primary macrophages.
J. Exp. Med.
171
:
35
22
Widmer, U., K. R. Manogue, A. Cerami, B. Sherry.
1993
. Genomic cloning and promoter analysis of macrophage inflammatory protein (MIP)-2, MIP-1α, and MIP-1β, members of the chemokine superfamily of proinflammatory cytokines.
J. Immunol.
150
:
4996
23
Ohmori, Y., S. Fukumoto, T. A. Hamilton.
1995
. Two structurally distinct κB sequence motifs cooperatively control LPS-induced KC gene transcription in mouse macrophages.
J. Immunol.
155
:
3593
24
Timmers, H. T., G. J. Pronk, J. L. Bos, A. J. van der Eb.
1990
. Analysis of the rat JE gene promoter identifies an AP-1 binding site essential for basal expression but not for TPA induction.
Nucleic Acids Res.
18
:
23
25
Koike, M., T. Kuroki, K. Nose.
1993
. Common target for 12-O-tetradecanoylphorbol-13-acetate and ras in the transcriptional enhancer of the growth factor-inducible JE gene.
Mol. Carcinog.
8
:
105
26
Shyy, J. Y., M. C. Lin, J. Han, Y. Lu, M. Petrime, S. Chien.
1995
. The cis-acting phorbol ester 12-O-tetradecanoylphorbol 13-acetate-responsive element is involved in shear stress-induced monocyte chemotactic protein 1 gene expression.
Proc. Natl. Acad. Sci. USA
92
:
8069
27
Iwabuchi, K., S. Hatakeyama, A. Takahashi, M. Ato, M. Okada, Y. Kajino, K. Kajino, K. Ogasawara, K. Takami, H. Nakagawa, K. Onoe.
1997
. Csk overexpression reduces several monokines and nitric oxide productions but enhances prostaglandin E2 production in response to lipopolysaccharide in the macrophage cell line J774A.1.
Eur. J. Immunol.
27
:
742
28
Voll, R. E., M. Herrmann, E. A. Roth, C. Stach, J. R. Kalden, I. Girkontaite.
1997
. Immunosuppressive effects of apoptotic cells.
Nature
390
:
350
29
Bogdan, C., J. Paik, Y. Vodovotz, C. Nathan.
1992
. Contrasting mechanisms for suppression of macrophage cytokine release by transforming growth factor-β and interleukin-10.
J. Biol. Chem.
267
:
23301
30
Hausmann, E. H., S. Y. Hao, J. L. Pace, M. J. Parmely.
1994
. Transforming growth factor-β1 and γ-interferon provide opposing signals to lipopolysaccharide-activated mouse macrophages.
Infect. Immun.
62
:
3625
31
Kunkel, S. L., M. Spengler, M. A. May, R. Spengler, J. Larrick, D. Remick.
1988
. Prostaglandin E2 regulates macrophage-derived tumor necrosis factor gene expression.
J. Biol. Chem.
263
:
5380
32
Zhong, W. W., P. A. Burke, M. E. Drotar, S. R. Chavali, R. A. Forse.
1995
. Effects of prostaglandin E2, cholera toxin and 8-bromo-cyclic AMP on lipopolysaccharide-induced gene expression of cytokines in human macrophages.
Immunology
84
:
446
33
Hurwitz, A. A., W. D. Lyman, J. W. Berman.
1995
. Tumor necrosis factor α and transforming growth factor β upregulate astrocyte expression of monocyte chemoattractant protein-1.
J. Neuroimmunol.
57
:
193
34
Takeshita, A., Y. Chen, A. Watanabe, S. Kitano, S. Hanazawa.
1995
. TGF-β induces expression of monocyte chemoattractant JE/monocyte chemoattractant protein-1 via transcriptional factor AP-1 induced by protein kinase in osteoblastic cells.
J. Immunol.
155
:
419
35
Gautam, S. C., C. J. Noth, N. Janakiraman, K. R. Pindolia, R. A. Chapman.
1995
. Induction of chemokine mRNA in bone marrow stromal cells: modulation by TGF-β1 and IL-4.
Exp. Hematol.
23
:
482
36
Kitamura, M..
1997
. Identification of an inhibitor targeting macrophage production of monocyte chemoattractant protein-1 as TGF-β1.
J. Immunol.
159
:
1404
37
Shull, M. M., I. Ormsby, A. B. Kier, S. Pawlowski, R. J. Diebold, M. Yin, R. Allen, C. Sidman, G. Proetzel, D. Calvin, et al
1992
. Targeted disruption of the mouse transforming growth factor-β1 gene results in multifocal inflammatory disease.
Nature
359
:
693
38
Kulkarni, A. B., C. G. Huh, D. Becker, A. Geiser, M. Lyght, K. C. Flanders, A. B. Roberts, M. B. Sporn, J. M. Ward, S. Karlsson.
1993
. Transforming growth factor β1 null mutation in mice causes excessive inflammatory response and early death.
Proc. Natl. Acad. Sci. USA
90
:
770