| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* World Health Organization (WHO) Immunology Research and Training Center, Department of Biochemistry, Epalinges, Switzerland; and
Department of Pathology, University of Bern, Bern, Switzerland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
) in regulating neutrophil number. Inflammatory cells were recruited by i.p. injection of either 2% starch or L. major promastigotes. Neutrophils were isolated and cultured in the presence of increasing numbers of M. Extent of neutrophil apoptosis positively correlated with the number of M added. This process was strictly dependent on TNF because M from TNF-deficient mice failed to induce neutrophil apoptosis. Assays using M derived from membrane TNF knock-in mice or cultures in Transwell chambers revealed that contact with M was necessary to induce neutrophil apoptosis, a process requiring expression of membrane TNF. L. major was shown to exacerbate M-induced apoptosis of neutrophils, but BALB/c M were not as potent as C57BL/6 M in this induction. Our results emphasize the importance of M-induced neutrophil apoptosis, and membrane TNF in the early control of inflammation. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Infection with the intracellular protozoan L. major leads to self-healing cutaneous lesions in mice of most inbred strains such as C57BL/6. A small lesion develops that heals slowly until it becomes fully resolved within 5–6 wk after L. major inoculation. These mice are subsequently protected from developing any lesions upon reinfection and are thus referred to as "resistant" to infection with L. major. In contrast, in a few susceptible strains of mice such as BALB/c mice, inoculation of L. major leads to the development of a lesion that does not heal. These mice will develop a lesion even upon reinfection and are referred to as "susceptible" to infection with L. major. Resistance and susceptibility to infection have been correlated with the development of CD4+ Th1 and Th2 responses, respectively (reviewed in Ref. 7).
Following s.c. infection with L. major, the composition of the early cellular infiltrate differs significantly between resistant C57BL/6 and susceptible BALB/c mice. As early as 1 day after infection, in resistant C57BL/6 mice, the macrophage (M) to neutrophil ratio is two times higher than in BALB/c mice, and the number of PMN was shown to decrease significantly 72 h after parasite inoculation with M, becoming more dominant in the cellular infiltrates thereafter. In susceptible BALB/c mice, elevated numbers of PMN were reported to persist for >10 days after infection, reflecting sustained inflammation (8, 9, 10).
We have previously shown that removal of PMN in susceptible BALB/c mice before infection with L. major induces a partial healing of the otherwise nonhealing lesions and modifies the profile of cytokines secreted by CD4+ Th cells. On the contrary, transient removal of neutrophils in resistant strains of mice does not affect the final outcome of the disease (10).
Because the early persistence of neutrophils in susceptible mice was shown to have a significant impact on the course of infection with L. major, we investigated in this study the mechanism(s) determining the control of neutrophil number in L. major-induced lesions in resistant vs susceptible mouse strains.
In this study, we report that one of the mechanisms regulating the number of neutrophils at the site of infection includes M-induced neutrophil apoptosis, a process which is further amplified in the presence of L. major. Expression of membrane TNF (mTNF) on M is necessary to induce PMN apoptosis, indicating that induction of neutrophil apoptosis by M is tightly controlled by TNF. This mechanism may constitute an essential process in the control of the early inflammatory lesions and crucially determine the control and outcome of the disease.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Female BALB/c and C57BL/6 mice were purchased from Harlan Olac. Female memTNF
/ mice (11) were bred in the animal facility of the Department of Pathology (University of Bern, Bern, Switzerland). B6.TNF–/– (12) were obtained from Dr. Marino (Ludwig Institute for Cancer Research, New York, NY) and subsequently bred in the pathogen-free facility at the BIL Epalinges Center. All mice were used at 6–8 wk of age.
Parasites and recruitment of M and PMN cells
L. major LV 39 (MRHO/Sv/59/P strain) were maintained in vivo and grown in vitro as described previously (13). Groups of 2–3 mice were infected i.p. with 5 x 107 stationary phase L. major promastigotes in a final volume of 1 ml, or mice were injected with 0.5 ml of 2% starch in PBS. Peritoneal neutrophils were obtained 4 h after injection, whereas M were isolated 16 h later.
PMN and M isolation
Peritoneal neutrophils isolated 4 h after inoculation of 5 x 107 L. major, or bone marrow (BM) neutrophils, were isolated by positive MACS selection (Miltenyi Biotec) with PE-conjugated anti-Ly6G (1A8), purchased from BD Pharmingen. Flow cytometry analysis was performed on a FACScan using CellQuest software (BD Biosciences). Neutrophils of 95–98% purity, as assessed both by FACS analysis, and cytospins of isolated cells were stained using Diff-Quick (Dade Behring) and then counted and incubated in presence of M.
Peritoneal M were isolated 16 h after injection with L. major, or 2% starch, by peritoneal washings and were further purified by Ficoll gradient (Amersham Biosciences), counted, and left to adhere for 3 h in 24-well plates at 37°C. After removal of nonadherent cells and several washings, M were fixed with 500 µl of 1% paraformaldehyde in PBS for 45 min, washed three times with PBS, and incubated with neutrophils. In selected experiments, M were detached with 5 mM EDTA and incubated for 1 h with LPS (1 µg/ml). Cells were preincubated with 24G2 mAb (anti-FcR
III Ab) and/or unconjugated rat anti-mouse TNF mAb (XT22) (both from BD Biosciences) and stained with FITC-conjugated F4/80 mAb (Caltag Laboratories) and PE-conjugated rat anti-mouse TNF mAb (XT22), or appropriate control mAb.
BM M were obtained by in vitro differentiation in RPMI 1640 supplemented with 20% horse serum and 30% supernatant from L929 cells as a source of M-CSF. After 7 days of culture, nonadherent cells were removed, and the remaining adherent M were detached and cultured in RPMI 1640 with 10% heat-inactivated FCS, L-glutamine, and 10 mM HEPES, as described previously (14).
Bioassay for TGF-
activity in PMN cell culture
TGF- was measured using a bioassay described previously (15, 16). The (PAI-L) cells used in this assay are mink lung epithelial cells transfected with a luciferase reporter driven by the plasminogen activator inhibitor-1 construct, a gift from Dr. D. Rifkin (New York University Medical Center, New York, NY). This quantitative bioassay is based on the ability of TGF- to up-regulate plasminogen activator inhibitor-1 expression. The induction of luciferase in these cells is very sensitive (picograms) and specific for TGF- (as verified by including mAb against TGF- (Genzyme) in the assays). Luminescence was detected with a Packard counter. Relative light unit values were converted to TGF- activity (picograms per milliliter) using a rTGF- (rHuman TGF-1; R&D Systems) standard curve. Mature PMN (1A8+) were purified by MACS and cultured in DMEM 5% FCS, in the presence or absence of L. major (5:1 parasite:cell ratio), or LPS (40 ng/ml; Sigma-Aldrich) and IFN- (50 U; BD Pharmingen) during 24 h. Culture supernatants were harvested, filtered, and frozen at –80°C. To measure total TGF- levels (active + latent), samples were diluted (2.5 times) in DMEM 0.1% BSA and acid-activated by adding 1 N HCl, then incubated for 1 min at room temperature, and neutralized using adding 1 M NaOH before addition to the assay plate. The treated samples were then used in the PAI-L assay. The luciferase assay was performed as described previously (16), and luciferase activity was determined using the luciferase assay substrate (Promega).
Assessment of PMN apoptosis
PMN apoptosis was quantified by morphology on May-Grünwald-stained cytospins of PMN, by counting at least 600 cells on three different fields per cytospin in a minimum of three different experiments, using bright field microscopy. Cells with chromatin aggregation, cytoplasmic vacuolation, and cell shrinkage were scored as apoptotic.
Annexin V binding
PMN apoptosis was evaluated by a combination of several methods. Early apoptosis was quantified by the expression of membrane phosphatidylserine (PS) detected by annexin V binding using annexin V-FITC as recommended by the manufacturer (BD Pharmingen) and identified by flow cytometry. Neutrophils were stained for surface markers such as Ly6G (1A8-PE mAb) and analyzed using a FACSCalibur and the CellQuest software (BD Biosciences). Because translocation of PS to the external cell surface also occurs during necrosis, it was used in conjunction with the 7-aminoactinomycin D (7AAD) nucleic dye, which stains nonviable cells. This allowed the distinction of early apoptotic cells (annexin V-FITC positive, 7AAD negative) from late apoptotic cells (annexin V-FITC positive, 7AAD-bright) (15).
TUNEL assay of chromatin fragmentation
Chromatin fragmentation was detected by the TUNEL assay (In Situ Cell Detection kit; Roche Diagnostics) to visualize late apoptotic cell death by enzymatic labeling of DNA strand breaks with dUTP-FITC and TdT. TUNEL was performed either on cytospins of neutrophils and detected by fluorescent microscopy or in cell suspension in 96-well plates, and detected by FACS analysis.
M:PMN coculture experiments
In the coculture experiments, the initial number of neutrophils was constant (5 x 105 PMN/ml), whereas the number of M varied from 1.3, 2.5, 5, 10, or 25 x 105 M/well. A M:PMN ratio of 2:1 (106 M:5 x 105 PMN) was used in most experiments.
Cocultures were incubated for 18 h (or 24 h when indicated) at 37°C in a CO2 incubator in RPMI 1640 + 5% FCS (LPS-free as determined by the Limulus amebocyte lysate assay; Invitrogen Life Technologies). The neutrophils were then collected and processed either for cytospins or FACS analysis. In selected experiments, 2.5 x 106 metacyclic parasites were added to the culture.
Transwell experiments
Experiments were performed in 24-well Transwell plates (Costar) in a final volume of 700 µl. PMN (5 x 105) were in the upper compartment of the wells, and (1x or 2.5 x 106) M were seeded in the lower compartment. Cells were cultured for 18 h as described above, and neutrophils were collected, stained, and analyzed by FACS as described above.
Statistics
Statistical analysis was done using the two-tailed t test for unpaired data.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
after 18 h of culture of C57BL/6 cells
It was previously reported that following s.c. infection with L. major, significant differences exist in the composition of the inflammatory cellular infiltrates in mice resistant (C57BL/6) or susceptible (BALB/c) to infection. A few hours after infection, a higher M:PMN ratio occurs in the cellular infiltrates of C57BL/6 mice than in BALB/c mice (8, 9). To investigate whether a higher number of M present at the site of parasite inoculation might reduce the numbers of neutrophils in L. major lesions, we first explored whether M could directly affect PMN apoptosis in vitro. Because recovery of neutrophils in footpads of mice is technically not feasible and reliable, we recovered neutrophils from the peritoneal cavity 4 h after the injection i.p. of 2% starch. A total of 5 x 105 PMN was cultured for 18 h without M and at increasing M:PMN ratios. Early apoptosis was detected by the presence of PS on the PMN surface detected by annexin V staining followed by FACS analysis. Twenty percent of C57BL/6 PMN cultured alone were annexin V+, and culture in the presence of M increased this percentage to 32 and 40% for the 2:1 and 5:1 M:PMN ratios, respectively (Fig. 1A).
|
. A significant increase in the number of apoptotic PMN was measured (18.7 ± 2.9) when the M:PMN ratio was increased to 2:1 (Fig. 1B). M-induced PMN apoptosis was confirmed by TUNEL staining of C57BL/6 cells 18 h after coculture of M with PMN at a 2:1 ratio. The percentage of TUNEL-positive neutrophils was 44% for PMN cultured alone, and increased to 69% in M:PMN cocultures (Fig. 1C). Thus, an increase in M numbers in the cocultures further increases apoptosis of PMN from C57BL/6 mice.
BALB/c-derived M are less potent inducers of PMN apoptosis in vitro than C57BL/6-derived M
To investigate whether M could also induce apoptosis of PMN from BALB/c mice, cocultures of M and neutrophils recruited i.p. with 2% starch were initiated. Morphological analysis and differential counting of cytospins of cocultures of M and PMN revealed no increased PMN apoptosis when PMN were cultured in presence of equal (1:1) or twice (2:1) the number of M (Fig. 2A). To investigate whether a higher number of M could induce PMN apoptosis, the M:PMN ratio was increased to 5:1. No significant difference in the number of apoptotic PMN was counted in four independent experiments (Fig. 2A). FACS analysis of early apoptosis (annexin V+) confirmed that M at a 2:1 PMN ratio did not affect PMN apoptosis, but an increase of early apoptosis was measured at a 5:1 M:PMN ratio (Fig. 2B). Hence, BALB/c M are able to induce PMN apoptosis but are less potent than C57BL/6-derived M in this respect.
|
recruited with L. major to induce PMN apoptosis
To study the effect of the mode of cell recruitment on PMN apoptosis, inflammatory cells from C57BL/6 mice were recruited following the injection i.p. with 2% starch or 5 x 107 stationary L. major promastigotes and cultured in the absence, or presence, of M. Late apoptosis was assessed by differential counting of May-Grünwald-stained cytospins. The level of PMN apoptosis in C57BL/6 PMN cultured alone was significantly increased when cells were recruited with L. major, compared with that of cells recruited with starch (Fig. 3A). This increase was further enhanced in presence of M (Fig. 3A). A similar increase in PMN apoptosis was also observed when BALB/c cells were recruited following injection of L. major promastigotes (data not shown).
|
to induce PMN apoptosis was due to the M, PMN, or both, criss-cross experiments were performed with L. major-recruited cells. C57BL/6 M were incubated for 18 h in presence of C57BL/6- or BALB/c-derived PMN at a M:PMN ratio of 5:1. Similarly, coculture experiments with C57BL/6- or BALB/c- derived PMN were performed using BALB/c M (Fig. 3B). C57BL/6 M could induce equally well apoptosis of BALB/c and C57BL/6 PMN (70 vs 68%). BALB/c-derived M also induced apoptosis of PMN from both strains but to a slightly lower percentage, as shown in Fig. 3B. This small difference in the capacity of BALB/c- and C57BL/6-derived M to induce PMN apoptosis was consistently measured in three independent experiments. A similar difference (10%) in M-induced PMN apoptosis of BALB/c vs C57BL/6 was also observed when cells were recruited with 2% starch (data not shown). Thus, BALB/c and C57BL/6 PMN are equally susceptible to M-induced apoptosis, and M from both strains induce PMN apoptosis, although BALB/c M are less efficient inducers of PMN apoptosis.
Recruitment of cells and stimulation with L. majorincreases the M-induced apoptosis of PMN in cocultures
To further investigate the effect of L. major, and/or a combined effect of both L. major and M on PMN apoptosis, L. major-recruited neutrophils were cultured for 18 h with or without L. major metacyclic promastigotes, in the presence or absence of M (2:1 M:PMN ratio). Early apoptosis was measured by FACS, using annexin V staining as described in Materials and Methods. A representative experiment of three is shown in Fig. 4. C57BL/6 and BALB/c M induced PMN apoptosis, with a lower induction for BALB/c M (Fig. 4, left panels), in line with the results presented above. Incubation of PMN with L. major in the absence of M had no effect on induction of C57BL/6 PMN apoptosis (Fig. 4, middle panel). In four independent experiments, incubation of L. major-recruited C57BL/6 PMN with L. major did not result in a statistically significant induction of apoptosis (p = 0.3). In contrast, addition of L. major to cocultures of M and PMN significantly induced >80% of apoptosis in C57BL/6 PMN (Fig. 4, right panel). Thus, L. major synergizes with M in enhancing C57BL/6 PMN apoptosis. Addition of L. major promastigotes to BALB/c PMN induced a tiny increase in PMN apoptosis by 
1.3 times (Fig. 4, middle panel). This induction of early PMN apoptosis by L. major was small but statistically significant (p = 0.016), as determined in four independent experiments. L. major also synergized with BALB/c M in the induction of PMN apoptosis, but only 50% of the PMN became apoptotic (Fig. 4, right panel). Thus, L. major alone mediates an incremental increase in the percentage of apoptotic PMN and synergizes with M in inducing BALB/c PMN apoptosis, but approximately half of the cultured BALB/c PMN remained nonapoptotic. Altogether, these results show that 1) L. major synergizes with M in the induction of PMN apoptosis, and 2) that there exist strain-specific differences between the induction of PMN apoptosis by M and L. major, with BALB/c-derived M being less potent inducers of PMN apoptosis than C57BL/6-derived M.
|
in cocultures
Following infection with L. major, TNF has been reported to be essential in the resolution of the inflammatory lesion (14, 17, 18, 19, 20, 21). Indeed, TNF–/– mice, on a resistant background, infected with L. major developed unhealing lesions comparable to those developed by BALB/c mice (14). Therefore, we investigated whether TNF, a cytokine secreted by both activated M and neutrophils, was involved in the M-induced PMN apoptosis. To this end, coculture of PMN and M from B6.TNF–/– mice were set up. Cells were recruited i.p either with starch or stationary phase L. major and cultured in the absence, or presence, of M at a M:PMN ratio of 2:1. Apoptotic PMNs were scored on cytospins as described in Materials and Methods. M from B6.TNF–/– mice were not able to induce PMN apoptosis following both modes of recruitment (Fig. 5A). Spontaneous apoptosis of TNF–/– PMN was higher than that measured in C57BL/6 PMN, thus revealing either a direct and/or indirect role for soluble TNF in attenuating spontaneous PMN apoptosis. To assess whether PMN from TNF–/– mice had any intrinsic defect and could be susceptible to M-induced apoptosis, M from C57BL/6 mice were cultured in presence of TNF–/– PMN. The presence of C57BL/6 M induced TNF–/– PMN apoptosis by 10% (Fig. 5B, middle panel); however, in contrast to the PMN cultured alone in which most apoptotic PMN (37%) were annexin V+ 7AAD– (early apoptosis), in the presence of C57BL/6 M, a substantial part (31%) of the apoptotic PMN became annexin V+ 7AAD+ (late apoptosis/necrosis; data not shown). As a positive control, C57BL/6 M were able to induce C57BL/6 PMN apoptosis. In contrast, M derived from TNF–/– mice did not induce apoptosis of either TNF–/– PMN or C57BL/6 PMN (Fig. 5B, right panel). These results demonstrate that induction of TNF expression by M is needed for PMN apoptosis induction, although TNF production by PMN may also contribute to some extent to apoptosis induction.
|
To demonstrate that both M and neutrophils need to be activated to respond to M-induced apoptosis, we isolated BM-derived neutrophils and L. major-recruited neutrophils for subsequent coculture with L. major-recruited M or BM-derived M. Neutrophil apoptosis was quantified by morphology on May-Grünwald-stained cytospins of PMN, as described in Material and Methods. Only a few apoptotic PMN were visually scored on cytospins, and no M-induced PMN apoptosis was detected in these cocultures (Fig. 6A). These results were confirmed with 7AAD staining and FACS analysis (data not shown). L. major-derived M also failed to induce early apoptosis of BM-derived neutrophils, as detected by annexin V staining (Fig. 6B). Thus, to induce PMN apoptosis, both PMN and M need to be activated following recruitment to the site of the inflammatory stimulus.
|
Morphological examination of cytospins obtained from cocultures of M with neutrophils revealed that neutrophils that were often found in close contact with M, presented the typical features of apoptotic cells including chromatin aggregation, cytoplasmic, and cell shrinkage (Fig. 7A, right panel). PMN cultured alone, however, predominantly represented healthy neutrophils with their characteristic multilobular nucleus (Fig. 7A, left panel). In the experiments shown in Fig. 7A, the M used in the cocultures were fixed first for 45 min in PBS 1% PFA, as described in Materials and Methods, to prevent secretion of TNF while still retaining membrane TNF on the cell surface. These results thus suggest that it is the mTNF expressed on M which is important in the induction of PMN apoptosis.
|
, cells from memTNF mice, in which an uncleavable mutant form of TNF was knocked in (11), were used in coculture experiments (Fig. 7B). The presence of mTNF was sufficient to restore the induction of PMN apoptosis by M, as revealed by FACS analysis of cells stained with 1A8, annexin V, and 7AAD. The percentage of nonapoptotic PMN decreased from 59 to 25% when the PMN were cultured in presence of unfixed mTNF+ M, and the percentage of late apoptotic cells (annexin V+ 7AAD+) increased from 13 to 48% of the PMN population (Fig. 7B).
To confirm these results and establish that cell-to-cell contact is required for the induction of PMN apoptosis, L. major-recruited C57BL/6 M and neutrophils were cultured in different compartments of Transwell plates, separated by membranes that prevented direct cell-to-cell contact while allowing the diffusion of soluble mediators. For these experiments, unfixed M were used. Eighteen hours later, neutrophils were collected, stained with annexin V and 7AAD, and analyzed by FACS. A representative experiment performed with unfixed M is presented in Fig. 7C. No induction of PMN apoptosis was observed using unfixed M when these M were physically separated from neutrophils. Such an absence of M-induced PMN apoptosis was observed at M:PMN ratios of 2:1 and 5:1 (Fig. 7C, middle and right panels) using cells isolated from both C57BL/6 and BALB/c mice (Fig. 7). Thus, the absence of induction of PMN apoptosis by unfixed M, physically separated from PMN, strongly argues against the involvement of a soluble factor in M-induced PMN apoptosis. Fixed M also failed to induce PMN apoptosis when cultured under the same Transwell culture (data not shown).
To investigate whether the higher ratio of M to PMN required to induce apoptosis of BALB/c PMN was due to lower expression of mTNF in BALB/c M, we measured by FACS the level of mTNF expression in peritoneally recruited BALB/c and C57BL/6 PMN M, 16 h after injection i.p. of L. major. Detection of mTNF with anti-TNF mAbs, however, turned out to be difficult, possibly resulting from a conformational change of the membrane bound TNF. For example in BY L6 hybridomaT cells that are transfected with an uncleavable form of TNF, >80% of the cells stained positive for intracellular TNF, whereas surface expression of mTNF was only shifted by <1.5% compared with mock-transfected cells (data not shown); nevertheless, mTNF on these cells was previously shown to be biologically functional (24). Similarly, only low levels of mTNF were detectable in both BALB/c and C57BL/6 M recruited i.p. following injection with L. major. However, with the available reagents, no substantial differences in the cell surface expression of mTNF on M from BALB/c and C57BL/6 mice were observed. Increased levels of cell surface mTNF were reported on M following LPS stimulation in vitro (11); therefore, L. major-recruited M were stimulated in vitro during 1 h with LPS and analyzed for mTNF expression by FACS. C57BL/6 M expressed slightly higher levels of mTNF than BALB/c M that were treated identically (data not shown). Thus, it was not possible with the available tools to detect a significant change in mTNF surface expression in BALB/c vs C57BL/6 L. major-recruited M. Significant expression of mTNF was detectable only subsequent to further stimulation with LPS. Therefore, it is not possible to rule out the existence of small differences in mTNF expression levels between both strains of M during in vivo stimulation.
The autocrine production of cytokines may differ between BALB/c and C57BL/6 PMN and thus could contribute to their distinct susceptibility to M-induced PMN apoptosis. Differential production of cytokines blocking PMN apoptosis such as GM-CSF, or TGF-, may explain the distinct sensitivity of C57BL/6 PMNs compared with BALB/c PMNs. Therefore, we have assessed the production of TGF- in BM-derived mature BALB/c and C57BL/6 PMN by bioassay (15) (16). PMN were stimulated or not with L. major stationary phase promastigotes (5:1 ratio) for 24 h, and the supernatant was collected for analysis. As a positive control, PMN were activated with LPS and IFN-. Significantly higher levels of biologically active TGF- were secreted by BALB/c PMN compared with the levels measured in supernatants of C57BL/6 PMN cultured in the presence or absence of L. major (Fig. 7D). Similar differences in TGF- secretion were also measured in L. major-recruited peritoneal PMN cultured in vitro; however, lower levels of biologically active TGF- were measured in the supernatant, possibly due to the rapid release of this cytokine in the peritoneum (data not shown). Thus, the significantly higher secretion of bioactive TGF- by BALB/c PMN could contribute to their lower sensitivity to M-induced apoptosis.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
from C57BL/6 mice induce a cell contact-dependent apoptosis of PMN, a process that is further amplified following infection with L. major. We show that increasing the number of M in cocultures of PMN and M results in a significant increase of PMN apoptosis. Moreover, we demonstrate that TNF is required in this process. Indeed, M recruited from TNF–/– mice failed to induce PMN apoptosis. We further show, using inflammatory cells derived from mice expressing only mTNF (memTNF/), that expression of mTNF on the M membrane, without secretion of soluble TNF, is necessary to induce PMN apoptosis in vitro.
The murine model of infection with L. major has been studied extensively to analyze the factors leading to resistance or susceptibility to this protozoan parasite. Following infection with L. major, most inbred mouse strains such as C57BL/6 develop only a small lesion that is self-healing, and they are able to control parasite replication efficiently. On the contrary, BALB/c mice fail to control lesion size, develop nonhealing lesions, and do not control parasite replication (25). It was previously reported that the ratio of M to PMN is lower at the site of parasite inoculation in BALB/c mice with equal numbers of PMN and M present from day 2 after infection, persisting for at least 10 days; whereas in C57BL/6 mice, the M:PMN ratio measured 2 days after infection (3.5:1) further increases to 30:1 on day 3 days after infection (8). We thus investigated a potential role of M in PMN-induced apoptosis. To this end, we isolated i.p. inflammatory cells and set up in vitro cocultures of M and PMN. It is important to consider that other factors found at the site of Leishmania inoculation by the phlebotome (the skin) may also contribute to neutrophil apoptosis/survival, which may require further analysis. Indeed, 4 h after infection of C57BL/6 with L. major i.p, the ratio of M to PMN in the cellular infiltrate was significantly higher than in L. major-infected BALB/c mice, as observed following infection s.c. of a high inoculum of L. major. However, the kinetics of neutrophils turnover following injection of L. major in the peritoneal cavity was more rapid than following infection s.c. of 3 x 106 parasites in the footpads (C. Allenbach and F. Tacchini-Cottier, unpublished observation), suggesting that additional mediators secreted at the site of s.c. infection, such as chemokines, may also contribute to the observed differential persistence and/or removal of PMN at the site of parasite inoculation.
The observed M-mediated induction of neutrophil apoptosis followed by the ingestion of apoptotic cells by M during the early stages of infection very likely contribute to the subsequent rapid resolution of the inflammation in C57BL/6 mice infected s.c. with L. major. This may occur by switching off the production of proinflammatory mediators, preventing the release of disease promoting mediators from necrotic cells, and stimulating the production of anti-inflammatory cytokines such as TGF- (23, 26), thus substantially modifying the microenvironment at the site of parasite inoculation. Interestingly, it was recently reported that coculture of L. major-infected M from C57BL/6 mice with dead neutrophils promoted the killing of L. major and secretion of TNF (27). Thus, induction of neutrophil apoptosis by M could contribute efficiently to the control of infection with L. major. The same authors also reported that on the contrary, i.e., incubation of dead neutrophils with L. major-infected M from BALB/c mice, exacerbated replication of L. major induced M production of TGF- but did not induce production of TNF (27). Hence, our present observations of an attenuated TNF-dependent neutrophil apoptosis induction by BALB/c M might be explained by such a reduced TNF induction in inflammatory cells of BALB/c mice. Unfortunately, the available reagents did not allow for a conclusive quantitation of cell surface-expressed mTNF in BALB/c- vs C57BL/6-derived M. Furthermore, additional mechanisms that may influence PMN apoptosis cannot be excluded, for instance, involving differences in the expression pattern of cell surface molecules including the presence of integrins such as 2 molecules on M (28, 29), that may lead to a differential recognition and/or attachment of M to neutrophils in the two different strains.
The effect of an autocrine production of cytokines by PMN may differ in L. major susceptible vs resistant mice and also contribute to the observed M-induced PMN apoptosis. Distinct secretion of cytokines by PMN was previously reported in PMN of mouse strains susceptible vs resistant to infections with pathogens such as Candida albicans (30), Staphylococcus aureus (6), and L. major (10, 31). Differential production of cytokines blocking PMN apoptosis such as GM-CSF or TGF- may explain the distinct sensitivity of C57BL/6 and BALB/c PMNs to M-induced apoptosis. In this line, we report that BALB/c PMN secrete significantly higher levels of biologically active TGF- compared with that released by C57BL/6 PMN. These higher levels of TGF- may contribute to the delay of M-induced PMN apoptosis observed in BALB/c PMN. This point will deserve further investigation both in vitro and in vivo.
Early after s.c. or i.p. infection with L. major, the proportion of M in the inflammatory cells recruited in BALB/c mice is significantly lower than in C57BL/6 mice. Thus, during the first days following infection with L. major, the number of M present in the cellular infiltrate of BALB/c mice may not be sufficient to induce PMN apoptosis. Following s.c. inoculation of the parasite, this delay may be further increased by a lower level of TNF secreted locally in BALB/c mice, and thus lower levels of mTNF expressed at the surface of inflammatory M. However, due to technical limitations, it was difficult to reliably measure in vitro differences in mTNF expression on M, unless the M were activated with LPS. In the latter case, levels of mTNF were slightly but not significantly higher in C57BL/6 than in BALB/c M. At this stage, it is difficult to extrapolate these results in vivo, but it has been reported that following s.c. infection with L. major, activation of BALB/c M differ from that of C57BL/6 M (32), suggesting that expression of M surface molecules may also be distinct.
The role of TNF has been extensively studied in the model of infection with L. major. TNF has been shown to synergize with IFN- in inducing parasite killing (33, 34). In vivo, mice resistant to infection with L. major were shown to produce significant amounts of TNF in their draining lymph nodes during the course of infection, whereas no TNF was detectable in susceptible animals (17). Following infection with L. major, transgenic mice on a C57BL/6 resistant genetic background expressing high levels of soluble hTNFRp55 fusion protein, or TNF–/– mice, ultimately developed severe ulcerating lesions but were able to clear their parasites (14, 18). Mice genetically deficient for TNF receptor 1, but not TNF receptor 2, developed nonhealing lesions with a transient increase in parasite burden, but eventually cleared the parasites from their lesions (19, 20, 21). We report in this study that M from TNF–/– mice cultured with neutrophils from either TNF–/– or C57BL/6 mice were unable to induce neutrophil apoptosis. We further showed, using M from memTNF/ mice and Transwell experiments, that the presence of membrane TNF on M was necessary to induce neutrophil apoptosis and required cell-to-cell contacts.
Thus, in C57BL/6 mice infected with L. major s.c., TNF could play a major role in the control of neutrophil apoptosis at the site of parasite inoculation. This control appears essential in the resolution of the inflammatory lesion elicited following infection with L. major. Interestingly, the roles of soluble vs membrane TNF in PMN apoptosis appear distinct, because soluble TNF may contribute to neutrophil survival while membrane TNF induces their apoptosis.
M are known to ingest PMN that undergo spontaneous apoptosis. We show here, in experiments performed in vitro, that M can also actively induce neutrophil apoptosis. These results are in agreement with previous findings obtained in an experimental rat model of wound healing (29). M isolated from the wounds were found to induce neutrophil apoptosis in a cell contact-dependent manner. Circumstantial evidence suggested that in addition to CD36 and CD61, cell surface-associated TNF was involved in this process. Incubation of purified inflammatory mouse C57BL/6 PMN with L. major alone did not induce PMN apoptosis. This is in line with experiments performed with blood-derived human neutrophils (35, 36) in which L. major was reported to induce a delay in spontaneous neutrophil apoptosis. It would be interesting to investigate whether this effect would persist if human inflammatory neutrophils were used, and whether the neutrophils were cultured in presence of M, because it has been reported that following transmigration through endothelial cells, neutrophils differ functionally from blood-derived PMN also in their survival (1, 28). A delay in the induction of apoptotic PMN was also observed 3 days after inoculation of L. major s.c. in mice, using an air pouch technique (35). Interestingly, the mice used in those experiments were BALB/c mice that reportedly showed lower M:PMN ratios than in resistant C57BL/6 mice following s.c. infection with L. major (8, 10).
Taken together, our results demonstrate that M, through their mTNF, induce PMN apoptosis in a mTNF-dependent manner. The efficiency of M-induced apoptosis induction in PMN is further enhanced by the presence of the parasite. M-induced PMN apoptosis may also be observed in cells from BALB/c mice, however, at a substantially reduced efficiency. The molecular basis for this difference is currently under investigation. Spontaneous apoptosis of both BALB/c and C57BL/6 PMN is blocked by the caspase inhibitor z-VAD-fmk (inhibitor of pan-caspases). However, M-induced apoptosis is only partially blocked by this inhibitor (C. Allenbach and F. Tacchini-Cottier, unpublished results). Comparison of several genes that either promote or block apoptosis should reveal whether there exist strain differences at this level.
In L. major-infected BALB/c mice, the low apoptosis induction in PMN is likely to be further attenuated in vivo by the reported low M:PMN ratio in the cellular infiltrates in the lesions. The quantitative as well as qualitative differences in the inflammatory cells recruited within hours of infection could thus lead to a distinct resolution of the L. major-induced inflammatory processes. Further investigations on how to control the early neutrophil accumulation, but also mTNF expression on M, may reveal new strategies to prevent exacerbating tissue damage in infections that, as for experimental cutaneous leishmaniasis, require early control of granulocytic infiltration.
|
Acknowledgments |
|---|
bioassay, and Y. Hauyon and O. Froelicher for technical assistance. |
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by grants from the Swiss National Foundation for Scientific Research (3200-06744.01; to F.T.-C.). 
2 Address correspondence and reprint requests to Dr. Fabienne Tacchini-Cottier, WHO Immunology Research and Training Center, Department of Biochemistry, University of Lausanne, 155 Chemin des Boveresses, CH-1066 Epalinges, Switzerland. E-mail address: Fabienne.Tacchini-Cottier{at}unil.ch
3 Abbreviations used in this paper: PMN, polymorphonuclear neutrophil; mTNF, membrane TNF; M, macrophage; BM, bone marrow; PS, phosphatidylserine; 7AAD, 7-aminoactinomycin D.
Received for publication August 22, 2005. Accepted for publication March 10, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-mediated apoptosis of human neutrophils. Kidney Int. 53: 84-91. [Medline], and the chemokines macrophage-inflammatory protein-1 and -1 in response to Toxoplasma gondii antigens. J. Immunol. 162: 7369-7375. using cells transfected with a plasminogen activator inhibitor-1 promoter-luciferase construct. Anal. Biochem. 216: 276-284. [Medline] in biological material using cells transfected with a plasminogen activator inhibitor-1 promoter-luciferase construct. Anal. Biochem. 247: 45-51. [Medline] or cycloheximide. J. Immunol. 154: 2403-2412. [Abstract], PGE2, and PAF. J. Clin. Invest. 101: 890-898. [Medline] (TNF) mediates effects distinct from those of wild-type TNF in vitro and in vivo. J. Biol. Chem. 274: 38112-38118. 1 secretion and the resolution of inflammation. J. Clin. Invest. 109: 41-50. [Medline]This article has been cited by other articles:
|
|
 |
|
  E. McFarlane, C. Perez, M. Charmoy, C. Allenbach, K. C. Carter, J. Alexander, and F. Tacchini-Cottier Neutrophils Contribute to Development of a Protective Immune Response during Onset of Infection with Leishmania donovani Infect. Immun., February 1, 2008; 76(2): 532 - 541. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. L. Ribeiro-Gomes, M. C. A. Moniz-de-Souza, M. S. Alexandre-Moreira, W. B. Dias, M. F. Lopes, M. P. Nunes, G. Lungarella, and G. A. DosReis Neutrophils Activate Macrophages for Intracellular Killing of Leishmania major through Recruitment of TLR4 by Neutrophil Elastase J. Immunol., September 15, 2007; 179(6): 3988 - 3994. [Abstract] [Full Text] [PDF] |
