* Immunobiology and
Molecular Immunology Units, The Institute of Child Health, and
Department of Immunology, Great Ormond Street Hospital, University College London, United Kingdom
Correspondence: Joanne R. Brown, Immunobiology Unit, The Institute of Child Health, Great Ormond Street Hospital, University College London, 30 Guilford Street, London, WC1N 1EH, UK. E-mail: J.Brown{at}ich.ucl.ac.uk
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: inflammation • reactive oxygen species • prostaglandins
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(TNF-)], interleukin-1 (IL-1; IL–6), eicosanoids [prostaglandin E2 (PGE2)]; leukotriene B4, CXC chemokines (macrophage inflammatory protein-2); and keratinocyte-derived chemokine, whose actions orchestrate vasodilatation and increased vascular permeability and amplify further recruitment of leukocytes to aid in microbial clearance. For many events, such as those surrounding the apoptosis of polymorphonuclear neutrophils (PMN) or phagocytic clearance of apoptotic PMN, the overall balance is tilted more toward suppression of inflammation. For example, the presence and removal of intact apoptotic cells are associated with the release of anti-inflammatory mediators such as IL-10 and transforming growth factor-ß (TGF-ß) production from activated monocytes [1
, 2
] and macrophages [3
, 4
], respectively. The clearance of apoptotic neutrophils is mediated via their surface changes and the engagement of recognition receptors on macrophages, ensuring safe and efficient clearance [5
]. Phosphatidylserine (PS) externalization from the inner leaflet of the apoptotic neutrophil membranes has been shown to be specific for a PS macrophage receptor as an ingestion signal [6
]. Other factors augmenting inflammatory resolution include the recently identified cyclopentenone PGs [cyPGs; PGD2, 
12-PGJ2, and 15-deoxy-12,14-PGJ2 (15d-PGJ2)], thought to derive from the inducible isoform of cyclo-oxygenase (COX)-2 [7
]. Anti-inflammatory properties of these mediators have been displayed in the carrageenin-induced pleurisy model of acute inflammation [7
, 8
] and in chronic inflammatory models of arthritis [9
, 10
], colitis [11
], and multiple sclerosis [12
].
The nicotinamide adenine dinucleotide phosphate (NADPH) oxidase is a multicomponent enzyme system responsible for the production of reactive oxygen species (ROS; reviewed in ref. [13
]). The functional significance of this membrane-associated enzyme system has been revealed in studies of patients with chronic granulomatous disease (CGD), a rare, X-linked or autosomal-inherited disorder of diminished or completely absent ROS production as a result of a genetic defect in any one of the four components of the NADPH oxidase enzyme system [14
]. Failure to produce an oxidative burst results in characteristic susceptibility to severe and recurrent infections by catalase-positive bacteria, such as Staphylcoccus aureus, Burkholderia cepacia, and Serratia marcescens, and fungi such as Aspergillus sp. However, CGD is not only characterized by susceptibility to severe infection; we have previously shown many patients develop inflammatory granulomas in hollow organs, such as gastric and urinary tract obstructions and skin, without clinical evidence of infections (sterile granulomas; see ref. [15
]). Similarly, mouse models of X-linked (gp91phox-/-; see ref. [16
]) and autosomal-recessive (p47phox-/-) forms of CGD [17
] show not only susceptibility to infections but also exaggerated inflammatory reactions to heat-killed fungus, with increased neutrophil and mononuclear cell accumulation compared with wild-type animals and increased early production of the proinflammatory cytokines IL-1ß and TNF- [18
]. Histologically, these inflammatory reactions comprise numerous areas of focal neutrophilic infiltrations containing centers of eosinophilic clusters surrounded by mononuclear cells, lesions that can persist for 21 days after the administration of sterile Aspergillus fumigatus hyphae.
The mechanisms behind these aberrant, inflammatory responses remain unknown, although neutrophils from CGD patients have previously been shown to be relatively resistant to apoptosis in vitro [19 , 20 ]. In this study, we have investigated whether the persistence of inflammation could be linked to an endogenous defect or imbalance in the release of pro- and anti-inflammatory mediators in phagocytes from CGD patients. We show that PMN from CGD patients are more resistant to spontaneous apoptosis and produce significantly less of the anti-inflammatory mediator PGD2. Furthermore, macrophages from CGD patients are markedly inhibited in their ability to produce PGD2 and TGF-ß during phagocytosis of nonopsonized and opsonized apoptotic targets.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Spontaneous PMN apoptosis
The study population included four X-linked CGD patients, median age 23.5 years, ranging from ages 6 to 30 years, following ethical approval and written consent as described above. Eight healthy individuals were included from local laboratory staff and were used as controls (median age 24.5 years; range 22–31 years). PMN from healthy, control individuals (control PMN) and CGD patients (CGD PMN) were isolated from heparinized blood by dextran sedimentation and density gradient centrifugation. Residual erythrocytes were removed by hypotonic lysis. The density of PMN was adjusted to 2 x 106 cells/ml in RPMI 1640 supplemented with 25 mM HEPES and 300 mg/L L-glutamine (Gibco-Life Technologies, Paisley, UK) medium containing 10% lipopolysaccharide (LPS) low, heat-inactivated fetal calf serum (Myclone Superplus, Gibco-Life Technologies) and 10,000 IU/ml penicillin and streptomycin (Gibco-Life Technologies; complete medium), and 1 ml was seeded into 24-well plates. LPS (1 ng/ml) was added to corresponding wells, and the plates were incubated at 37°C to initiate a time course of spontaneous apoptosis over 48 h. At each time point, the cells were separated by centrifugation (350 g), and the supernatants were stored in aliquots at -80°C for the measurement of PGD2. The cells were washed with phosphate-buffered saline (PBS) and stained with annexin V–fluorescein isothiocyanate (FITC; Becton Dickinson, Oxford, UK) antibody and propidium iodide (PI) following the manufacturer’s instructions (Becton Dickinson). Using a Beckman Coulter XL flow cytometer with Expo2 software (Beckman Coulter, High Wycombe, Bucks, UK), apoptotic PMN were read at 525 nm (FITC) and 575 nm (PI), where a total of 10,000 events was counted—gating on PMN using forward-scatter/side-scatter (FS/SS) plots—and values were stated as the percentage of positive events. After fluorescein-activated cell sorter (FACS) analysis, the remaining cells were fixed in 70% ethanol and stored at 4°C to measure cell-cycle apoptosis measuring a sub-G1 population of PI-stained, fragmented DNA. At the end of each time course, fixed apoptotic PMN were washed with ice-cold PBS and were then lysed and labeled with PI using PI lysis buffer (H2O containing 1 mg/ml sodium citrate and 1 µg/ml Triton X-100). Apoptotic PMN were read at 575 nm (PI), where a total of 20,000 events was counted—gating on PMN using FS/SS plots—and values were stated as the percentage of positive events. In separate experiments, the effects of exogenous PGs were examined by incubating the PMN with a concentration range of 0.1–10 µM PGE2, PGD2, 12-PGJ2, and 15d-PGJ2 diluted in dimethyl sulfoxide (DMSO) or methyl acetate vehicle control to a final concentration of 0.01% over a time course of 24 h. All PGs were purchased from Alexis Corporation (Nottingham, UK). All treatments were repeated twice per donor from one healthy individual and one CGD patient and were expressed as a percentage of 0.01% DMSO or methyl acetate-treated controls.
Macrophage isolation
The study population included eight X-linked CGD patients, with a median age of 13 years, ranging from ages 4 to 21 years. Ten healthy individuals were included from local laboratory staff and were used as controls (median age 26 years, ranging from ages 22 to 30 years). Human monocytes were isolated from peripheral blood using LymphoprepTM (Gibco-Life Technologies), a density gradient with a centrifugation method described previously [21
]. Briefly, peripheral blood mononuclear cells were plated at 1 x 106 cells/ml into six-well plates and incubated in complete medium for 2 h at 37°C. Nonadherent cells were removed, and the remaining adherent mononuclear cells were cultured in complete medium containing 10 ng/ml recombinant human macrophage-colony stimulating factor (M-CSF; R&D Systems, Abingdon, UK), added every 2 days. Phenotyping was performed on day 6 to assess monocyte differentiation into macrophages.
Phenotyping
All antibodies were purchased from Becton Dickinson. Approximately 5 x 104 macrophages/100 µl were transferred into each FACS tube and were stained with 5 µl of a panel of antibodies, including CD11c [phycoerythrin (PE)], CD45 (FITC), human leukocyte antigen (HLA)-DR (PE), CD14 (FITC), CD86 (PE), CD83 (PE), CD3 (FITC), or CD19 (PE), along with immunoglobulin G (IgG)1 and IgG2 FITC/PE isotype controls. Macrophages were left incubating with the antibodies in the dark for 30 min on ice. To each tube, 2 ml PBS was added, and the tubes were centrifuged at 350 g for 5 min. Cell pellets were then resuspended in 1% paraformaldehyde, and marker expression was analyzed at 525 nm (FITC) and 575 nm (PE). A total of 10,000 events was counted—gating on macrophages using FS/SS plots—and values were stated as the percentage of positive events.
PMN isolation, induction of apoptosis, and opsonization
Isolated PMN were rendered apoptotic essentially as described previously [3
]. Briefly, isolated PMN were irradiated with 20 min UV exposure (254 nm) and were then maintained in RMPI 1640 at room temperature for 2 h while rotating end-over-end in polypropylene tubes. These conditions induced 50–60% apoptosis as assessed by annexin V/PI labeling. Opsonized, apoptotic PMN were labeled with mouse anti-human CD45 (Becton Dickinson) for 30 min at 4°C rotating end-over-end, followed by a secondary rabbit anti-mouse (Stratech Scientific, Luton, UK) for a further 30 min. After washing in RPMI, cells were maintained in RPMI 1640 without serum until phagocytosis was initiated.
Phagocytosis
Target apoptotic PMN were labeled with 50 µg/ml FITC and incubated for 10 min at room temperature. After two PBS washes, apoptotic PMN were incubated with differentiated macrophages for 1 h at 37°C. Undigested cells were washed, the macrophages scraped, and any surface-associated cells quenched using 5 µg/ml crystal violet with 10% trypan blue [22
] for 5 min at room temperature. Excess quench solution was washed with PBS, and pellets were resupended in 400 µl PBS and kept on ice to be read within 30 min for FACS analysis. A total of 10,000 events was counted, gating on macrophages using FS/SS plots to exclude debris and unphagocytosed material, and was read at 525 nm (FITC). A phagocytosis time course was established for the measurement of liberated cytokines into the surrounding medium, where supernatants were aspirated at each time point and stored in aliquots at -80°C. The control for phagocytosis of nonopsonized, apoptotic PMN was macrophages incubated with no target cell (nonphagocytosing, resting-state macrophages) for the measurement of basal cytokine levels. The control for phagocytosis of opsonized, apoptotic PMN was PMN labeled with CD45 antibody alone. In a series of experiments, 1 ng/ml LPS was added at the same time as the target cell to activate the macrophages.
Analysis of inflammatory mediators
TNF-, IL-10 (both from eBioscience, Wembley, UK), and TGF-ß (BD, Abingdon, UK) were measured using enzyme-linked immunosorbent assay. PGE2 was measured by enzyme immunoassay, and PGD2 was measured using a radioimmunoassay (both from Amersham Pharmacia Biotech, Bucks, UK).
Statistics
Data are expressed as mean ± SEM. Differences between data sets were evaluated by performing an unpaired Student’s t-test using the GraphPad PrismTM statistical package (GraphPad Software, San Diego, CA). A level of P < 0.05 was considered significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (36K): [in a new window] |
Figure 1. Control (Con; open bars) and CGD (solid bars) neutrophil apoptosis occurring spontaneously over time. Apoptosis was assessed labeling neutrophil PS with annexin V–FITC antibody and measuring membrane integrity with PI (a). Also, cell-cycle apoptosis was assessed measuring a sub-G1 population of PI-stained, fragmented DNA (b). Corresponding PGD2 release (pg/ml; c) was also measured. Data are displayed as mean ± SEM within the study population of four CGD patients ranging from ages 6 to 30 years and eight healthy individuals of a similar age range. *, P < 0.05; **, P < 0.01; ***, P < 0.001 compared with control neutrophils.
|
and PGE2 between control and CGD PMN (data not shown). However, after 6 h, apoptosing CGD PMN produced significantly less PGD2 compared with control PMN (P=0.012; Fig. 1c
), a reduction that persisted throughout. The addition of LPS did not restore PGD2 production (data not shown). To determine whether deficiency of PG production contributed to the retardation of CGD–PMN apoptosis, exogenous cyPGs were added to control and CGD PMN. PGE2, PGD2, 12-PGJ2, and 15d-PGJ2 had no effect on the accumulation of the sub-G1 population of PI-stained, fragmented DNA after 2, 6, or 12 h incubation. At 18 and 24 h, PGE2 had some protective effect on control and CGD PMN from cell death (Fig. 2a
), although this failed to reach significance. In addition, the cyPGs appeared to induce a trend toward increased apoptosis of control and CGD PMN in a concentration-dependent manner after 18 and 24 h (Fig. 2b
and 2d
; nonsignificant), although the CGD PMN were consistently less sensitive, and 15d-PGJ2 perhaps had the most potent effect (Fig. 2d)
.
 View larger version (43K): [in a new window] |
Figure 2. Exogenous PGs and neutrophil apoptosis. PMN isolated from control (Con; open squares) and CGD (solid squares) individuals were incubated in the presence of PGE2 (a), PGD2 (b), 12-PGJ2 (c), and 15d-PGJ2 (d) in a concentration range of 0.1–10 µM, diluted in DMSO or methyl acetate vehicle control to a final concentration of 0.01% over a time course of 24 h. Apoptosis was assessed measuring a sub-G1 population of PI-stained, fragmented DNA. The concentration of vehicle was shown not to significantly alter PMN cell death at each time point. All treatments were repeated twice per donor from one healthy individual and one CGD patient and were expressed as a percentage compared with 0.01% DMSO or methyl acetate-treated controls.
|
Release of inflammatory mediators from macrophages phagocytosing normal apoptotic PMN
Resolution and suppression of potentially damaging inflammation are tightly linked to the clearance of microbial pathogens and apoptotic, inflammatory cells and the controlled release of pro- and anti-inflammatory mediators. These processes were therefore investigated in more detail. Resting control and CGD macrophages were compared with those cocultured with nonopsonized, apoptotic PMN (to mimic uptake of apoptotic cells and debris), following which the release of inflammatory mediators was determined. Control macrophages cocultured with nonopsonized, apoptotic PMN produced similar levels of the proinflammatory mediators TNF- (Fig. 3a
) and PGE2 (Fig. 3b)
when compared with macrophages in their nonphagocytosing, resting state. However, increased levels of all the anti-inflammatory mediators were measured and compared with macrophages in their nonphagocytosing, resting state. Specifically, they produced 69%, 99% (P=0.002), and 86% higher levels of IL-10, TGF-ß, and PGD2, respectively, in the presence of nonopsonized PMN (Fig. 3c
3d
3e)
. Similar observations have been made in other studies [1
2
3
]. When CGD macrophages were compared with control macrophages, resting CGD macrophages produced significantly lower basal levels of TNF- (a 95% reduction; Fig. 3a
; P=0.009) and reduced levels of PGE2, IL-10, and PGD2 (nonsignificant; Fig. 3b
3c
and 3e , respectively). In the presence of nonopsonized, apoptotic targets, CGD macrophages produced 92% less TGF-ß (P=0.01; Fig. 3d
) and 99% less PGD2 (P=0.05; Fig. 3e
) than control macrophages. As before, the levels of TNF-, PGE2, and IL-10 were also lower, although these differences did not reach statistical significance.
 View larger version (29K): [in a new window] |
Figure 3. Inflammatory mediator release from control macrophages (Con; open bars) and CGD (solid bars) macrophages (M ), incubated alone or cocultured with nonopsonized, apoptotic PMN for 18 h. Macrophages were incubated alone or with UV-irradiated, apoptotic PMN for 18 h, and inflammatory mediators were measured from the supernatants. Data are displayed as mean ± SEM within the study population of eight X-linked, CGD patients, ranging from ages 4 to 30 years and 10 healthy individuals of an age range of 22–30 years. *, P < 0.05; **, P < 0.01 compared with control macrophages alone.
|
-mediated uptake of microbial pathogens, apoptotic PMN were preopsonized and cocultured as before with control and CGD macrophages. Coculture with opsonized, apoptotic PMN induced high levels of mediator production, irrespective of their pro- or anti-inflammatory nature. When compared with control macrophages, cocultured with control targets, coculture with opsonized cells increased TNF- levels by 89% (P<0.0001; Fig. 4a
), PGE2 levels by 85% (P<0.0001; Fig. 4b
), IL-10 levels by 99% (P<0.0001; Fig. 4c
), TGF-ß levels by 95% (P<0.0001; Fig. 4d
), and PGD2 levels by 76% (P=0.03; Fig. 4e ). As before, differences in the profiles of mediator release were noted when CGD macrophages were substituted for control macrophages. Opsonized, apoptotic PMN induced CGD macrophages to produce higher levels of TNF- (92% increase, P=0.02; Fig. 4f ), PGE2 (76% increase; Fig. 4g
), IL-10 (99% increase, P=0.03; Fig. 4h
), and TGF-ß (99% increase; Fig. 4i
) compared with CGD macrophages cocultured with nonopsonized targets. Moreover, CGD macrophages were unable to produce detectable amounts of PGD2 under these conditions (Fig. 4j)
. However, the magnitude of these increases was somewhat less than those of control macrophages under the same conditions. More specifically, compared with control macrophages, opsonized, apoptotic PMN induced CGD macrophages to produce 54% less TNF- (not significant), 74% less PGE2 (P=0.02), 51% less IL-10 (not significant), 52% less TGF-ß (not significant), and 99% less PGD2 (P=0.008).
 View larger version (31K): [in a new window] |
Figure 4. Inflammatory mediator release from control and CGD macrophages (M) cocultured with nonopsonized and opsonized, apoptotic PMN for 18 h. Macrophages were incubated with UV-irradiated, apoptotic PMN, nonopsonized (open bars) or opsonized (hatched bars) with human anti-CD45 followed by a secondary rabbit anti-mouse and inflammatory mediators measured from the supernatants. Data are displayed as mean ± SEM within the study population of 10 healthy individuals of an age range of 22–30 years and eight X-linked, CGD patients, ranging from ages 4 to 30 years. *, P < 0.05; **, P < 0.01; ***, P < 0.001 compared with control or CGD macrophages cocultured with apoptotic PMN labeled with a mouse anti-human CD-45 antibody alone.
|
, peaking at 12 h, but control macrophages produced 77% higher levels after 8 h (P=0.04; Fig. 5a
) and 46% higher levels after 18 h (P=0.05). Similarly, the profile of PGE2 release peaked at 8 h for control and CGD macrophages, but the former produced 88% higher levels after 1 h (P=0.006; Fig. 5b
), 88% higher levels after 12 h (P=0.012), and 74% higher levels after 18 h (P=0.021). IL-10 levels were similar for both types of macrophages throughout the time course (Fig. 5c)
. In contrast to control cells, CGD macrophages produced undetectable amounts of TGF-ß after 1, 8, and 12 h (P=0.004, P=0.002, and P=0.03, respectively; Fig. 5d
), although this increased to control levels at 18 h. Control macrophages produced a biphasic PGD2 profile, peaking early at 1 h and increasing again at 12 and 18 h (Fig. 5e)
. Strikingly, CGD macrophages were unable to produce PGD2 throughout the entire time course. Therefore, in this experimental system, the overall balance of mediator release from CGD macrophages during phagocytosis of nonopsonized and opsonized targets appears to be shifted away from the normal mechanisms that suppress unwanted inflammation and encourage resolution.
 View larger version (36K): [in a new window] |
Figure 5. Inflammatory mediator release from control (Con; open bars) and CGD (closed bars) macrophages cocultured with nonopsonized and opsonized, apoptotic PMN for over 18 h. Macrophages were incubated with UV-irradiated PMN labeled with mouse anti-human CD-45 antibody alone (PMNs) or with opsonized PMN labeled with anti-CD45 followed by a rabbit anti-mouse secondary antibody during an 18-h time course where inflammatory mediators were measured from the supernatants. Data are displayed as mean ± SEM within the study population of eight X-linked, CGD patients, ranging from ages 4 to 30 years and 10 healthy individuals of an age range of 22–30 years. *, P < 0.05; **, P < 0.01 compared with control macrophages cocultured with opsonized, apoptotic PMN at each time point.
|
and PGE2 for both normal and CGD cells, although the response of CGD cells appeared to be slightly attenuated (Fig. 6 a
1b
1c
1d
1e)
. Similarly, the addition of LPS restored the ability of CGD macrophages to produce PGD2. This illustrates that the CGD macrophages have the capability to produce inflammatory mediators when phagocytosing opsonized, apoptotic PMN but only when the macrophages have received an additional inflammatory signal such as LPS.
 View larger version (21K): [in a new window] |
Figure 6. Inflammatory mediator production over 18 h by activated macrophages (M) derived from controls (Con; open bars) and patients with CGD (closed bars) macrophages (M). Macrophages were incubated with UV-irradiated PMN labeled with mouse anti-human CD-45 antibody alone (PMNs) or with opsonized PMN (Op PMNs) labeled with anti-CD45 followed by a rabbit anti-mouse secondary antibody with or without activation by 1 ng/ml LPS and inflammatory mediators measured from the supernatants. Data are displayed as mean ± SEM within the study population of eight X-linked, CGD patients ranging from ages 4 to 30 years and 10 healthy individuals of an age range of 22–30 years. *, P < 0.05; **, P < 0.01 compared with control macrophages cocultured with opsonized PMN.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As has been shown previously, the apoptotic response of CGD PMN was clearly abrogated and kinetically delayed when compared with control cells, indicating that the NADPH oxidase contributes importantly to this process [19
, 20
]. Neutrophil apoptosis is highly sensitive to influential factors within the local microenvironment. For example, bacterial LPS, granulocyte M-CSF (GM-CSF), G-CSF, interferon- [23
], eicosanoids [24
], PMN transmigration [25
], and hypoxia [26
], particularly in an arthritic synovial joint [27
], can delay neutrophil apoptosis. Conversely, phagocytosis of Escherichia coli [28
] and IL-4, TGF-ß, glucocorticoids, and TNF- can accelerate the apoptotic process [29
30
31
]. Many agents that trigger apoptosis are oxidants or stimulators of cellular, oxidative metabolism, whereas inhibitors of apoptosis display largely antioxidant activity [32
]. For example, eosinophil apoptosis can be triggered by sodium arsenite, Fas, or aging, inducing a concomitant increase in intracellular hydrogen peroxide, which can be reversed by glutathione or N-acetyl-cysteine [33
]. Fas-induced apoptosis of activated neutrophils can also be inhibited by elevating intracellular glutathione levels [34
]. It therefore seems likely that deficient production of oxidants from CGD cells is contributory to the apoptotic defect and is supported by the observation that pharmacological inhibition of the NADPH oxidase results in a similar phenotype (unpublished observations). Persistence of neutrophils and other inflammatory cells, which do not undergo physiological apoptosis, leads to an increase in necrotic decay and release of toxic granule contents, which cause direct tissue damage. By this mechanism, chronic inflammation can be initiated and propagated.
Externalization of PS during physiological apoptosis is important for recognition of apoptotic cells by macrophages and for triggering their safe clearance [6 ]. The implications of abrogated apoptosis and therefore delayed PS externalization have been highlighted by several studies. Huynh et al. [35 ] showed that endotracheal administration of apoptotic Jurkat cells (human T cell line that expresses PS) to mice promotes TGF-ß secretion and the resolution of lung inflammation. However, PLB-985 (human monomyelocytes that do not express PS) did not induce TGF-ß, allowing lung inflammation to persist [35 ]. Furthermore, enzymatic cleavage of the macrophage PS receptor by elastase may further inhibit release of crucial anti-inflammatory signals from macrophages [36 ]. Clinically, delayed apoptotic cell clearance has been measured in sputa samples from patients with cystic fibrosis and bronchiectasis [37 ]. Here, airway fluid from these patients inhibited apoptotic cell removal by alveolar macrophages via the cleavage of the macrophage PS receptor in a neutrophil elastase-dependent way [37 ].
A link between production of ROS, apoptosis, and PS-mediated clearance of apoptotic cells has also been proposed. For example, Kagan et al. [38 ] showed that oxidized forms of PS were necessary for efficient phagocytosis by macrophages and that inhibition of PS oxidation abrogated this process. Similarly, diphenyleneiodonium chloride-treated (to inhibit the NADPH oxidase), healthy PMN or PMN from CGD patients were more resistant to macrophage uptake after apoptosis than control cells [39 ]. Therefore, abnormal PS externalization (present study) or PS oxidation [38 ] and subsequent uptake by macrophages may contribute to the underlying, aberrant, inflammatory response seen in CGD patients.
It is interesting that we found that apoptosing CGD PMN and phagocytosing CGD macrophages were markedly deficient in their production of PGD2. PGs have been shown to have important roles in the regulation of the inflammatory process. During resolution of acute inflammation in a carrageenin-induced pleurisy model, PGD2 levels are significantly elevated with a concomitant increase in COX-2 protein expression [7
]. When COX-2 activity is selectively inhibited, the inflammatory response is exacerbated; restoration of lost PGD2 returns inflammatory parameters back to controls levels [7
]. The relevance of decreased cyPG release from CGD macrophages and PMN to resolution of inflammation and to susceptibility of cells to apoptosis warrants further investigation, although our studies suggest that this may be an important effect. PGs of the J2 series are potent inducers of ROS in a human neuroblastoma cell line [40
], and exogenous treatment of these cells induces apoptosis [41
]. More recently, PGD2 and its metabolites have been shown to induce physiological caspase-dependent granulocyte apoptosis, suggesting that this may be an important mechanism of their anti-inflammatory actions [42
]. In fact, PGD2 and its metabolite PGJ2 selectively induce eosinophil apoptosis, whereas the sequential metabolites 12-PGJ2 and 15d-PGJ2 induce caspase-dependent apoptosis in neutrophils and eosinophils [42
]. It is interesting that pathological, inflammatory lesions in CGD usually involve an accumulation of inflammatory cells, including eosinophils, macrophages, and pigmented histiocytes [43
, 44
]. 15d-PGJ2 has been shown to inhibit multiple steps of the nuclear factor (NF)-
B pathway [45
] such that 15d-PGJ2 can induce granulocyte apoptosis and reverse the LPS-induced delay in apoptosis by inhibiting NF-B activation [42
]. In other systems, the entry of cyPGs into cells has been shown to activate a stress response [46
]. Therefore, cyPG release may be critical for normal induction of apoptosis, acting via NF-B activation or induction of the stress response.
In agreement with previous studies, we have confirmed that normal macrophages produce significant levels of TGF-ß when incubated with nonopsonized, apoptotic PMN, which supports the suggestion that in normal cells, this is important for the suppression of inflammation during this process [3 , 4 ]. Similar to CGD PMN, we have also shown that macrophages isolated from CGD patients are markedly compromised in their ability to produce the anti-inflammatory mediators PGD2 and TGF-ß during the phagocytosis of apoptotic cells, whether in the nonopsonized state or when opsonized. These findings indicate that there may be a disturbed balance of inflammatory mediators during the physiological clearance of apoptosing cells and during the resolution phase of inflammation following microbial infection; in CGD, this results in the initiation and persistence of chronic inflammation. It is interesting that in this model, the addition of LPS restored PGD2 production to near-control levels, indicating that CGD macrophages retain the capacity to produce near-normal levels of all inflammatory mediators if they receive sufficient, additional activating signals (in this case LPS), as may be sometimes encountered during acute clearance of infected material. Under many circumstances, the degree of activation may be insufficient to correct the defect of PGD2 production, under which conditions the balance would be shifted toward persistence of inflammation, even when the initiating microbial stimulus has been eliminated.
The signaling mechanisms that link failure of the NADPH oxidase to defects of inflammatory mediator production and apoptosis and therefore to dysfunctional resolution of inflammation are undetermined. One possibility lies in the recently identified anti-inflammatory action of the redox-sensitive transcription factor NF-B. Indeed, therapeutic inhibition of NF-B activity has been shown to exacerbate acute inflammation by preventing leukocyte apoptosis and TGF-ß production in carageenin-induced rat pleurisy and murine air pouch models [47
]. Furthermore, p47phox and gp91phox-/- mice challenged with various inflammatory stimuli are significantly deficient in NF-B activity compared with wild-type mice, resulting in elevated CXC chemokine-driven neutrophil accumulation [48
49
50
], suggesting a link (direct or indirect) among NADPH oxidase activity, NF-B activation, and persistence or resolution of inflammation.
Prolongation of the apoptotic process and failure to actively suppress or resolve inflammatory reactions may underlie the aberrant, inflammatory responses observed in CGD patients in response to microbial stimuli and clearance of apoptotic debris. The identification of defects in the production of mediators such as TGF-ß and cyPG PGD2 provides important clues to the pathogenetic mechanisms and offers significant opportunity to design and test novel therapies for these difficult complications.
Received December 9, 2002; revised January 10, 2003; accepted January 21, 2003.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
12,14-PGJ2 induces synoviocyte apoptosis and suppresses adjuvant-induced arthritis in rats J. Clin. Invest. 106,189-197[Medline]12,14-PGJ2 attenuates the development of acute and chronic inflammation Mol. Pharmacol. 61,997-1007 ligands to inhibit the epithelial inflammatory response J. Clin. Invest. 104,383-389[Medline] agonist 15-deoxy-12,14-prostaglandin J2 ameliorates experimental autoimmune encephalomyelitis J. Immunol. 168,2508-2515Related Article
This article has been cited by other articles:
 |
|
 |
|
  R. Rajakariar, J. Newson, E. K. Jackson, P. Sawmynaden, A. Smith, F. Rahman, M. M. Yaqoob, and D. W. Gilroy Nonresolving Inflammation in gp91phox-/- Mice, a Model of Human Chronic Granulomatous Disease, Has Lower Adenosine and Cyclic Adenosine 5'-Monophosphate J. Immunol., March 1, 2009; 182(5): 3262 - 3269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. F. Fernandez-Boyanapalli, S. C. Frasch, K. McPhillips, R. W. Vandivier, B. L. Harry, D. W. H. Riches, P. M. Henson, and D. L. Bratton Impaired apoptotic cell clearance in CGD due to altered macrophage programming is reversed by phosphatidylserine-dependent production of IL-4 Blood, February 26, 2009; 113(9): 2047 - 2055. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. M. Marriott, L. E. Jackson, T. S. Wilkinson, A. J. Simpson, T. J. Mitchell, D. J. Buttle, S. S. Cross, P. G. Ince, P. G. Hellewell, M. K. B. Whyte, et al. Reactive Oxygen Species Regulate Neutrophil Recruitment and Survival in Pneumococcal Pneumonia Am. J. Respir. Crit. Care Med., April 15, 2008; 177(8): 887 - 895. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kanayama and Y. Miyamoto Apoptosis triggered by phagocytosis-related oxidative stress through FLIPS down-regulation and JNK activation J. Leukoc. Biol., November 1, 2007; 82(5): 1344 - 1352. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Bedard and K.-H. Krause The NOX Family of ROS-Generating NADPH Oxidases: Physiology and Pathophysiology Physiol Rev, January 1, 2007; 87(1): 245 - 313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. N. Sanford, A. R. Suriano, D. Herche, K. Dietzmann, and K. E. Sullivan Abnormal apoptosis in chronic granulomatous disease and autoantibody production characteristic of lupus Rheumatology, February 1, 2006; 45(2): 178 - 181. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-L. N. Huynh, K. C. Malcolm, C. Kotaru, J. A. Tilstra, J. Y. Westcott, V. A. Fadok, and S. E. Wenzel Defective Apoptotic Cell Phagocytosis Attenuates Prostaglandin E2 and 15-Hydroxyeicosatetraenoic Acid in Severe Asthma Alveolar Macrophages Am. J. Respir. Crit. Care Med., October 15, 2005; 172(8): 972 - 979. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Dinauer Chronic Granulomatous Disease and Other Disorders of Phagocyte Function Hematology, January 1, 2005; 2005(1): 89 - 95. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Milla, S. Yang, D. N. Cornfield, M.-L. Brennan, S. L. Hazen, A. Panoskaltsis-Mortari, B. R. Blazar, and I. Y. Haddad Myeloperoxidase deficiency enhances inflammation after allogeneic marrow transplantation Am J Physiol Lung Cell Mol Physiol, October 1, 2004; 287(4): L706 - L714. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J P Buyon, A Rupel, and R M Clancy Neonatal lupus syndromes Lupus, September 1, 2004; 13(9): 705 - 712. [Abstract] [PDF]
|
|
|
|
 |
|
 B. E. Marciano, S. D. Rosenzweig, D. E. Kleiner, V. L. Anderson, D. N. Darnell, S. Anaya-O'Brien, D. M. Hilligoss, H. L. Malech, J. I. Gallin, and S. M. Holland Gastrointestinal Involvement in Chronic Granulomatous Disease Pediatrics, August 1, 2004; 114(2): 462 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Wahl, J. Swisher, N. McCartney-Francis, and W. Chen TGF-{beta}: the perpetrator of immune suppression by regulatory T cells and suicidal T cells J. Leukoc. Biol., July 1, 2004; 76(1): 15 - 24. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Clancy and J. P. Buyon Clearance of apoptotic cells: TGF-{beta} in the balance between inflammation and fibrosis J. Leukoc. Biol., December 1, 2003; 74(6): 959 - 960. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
