Published online before print July 15, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Laboratoire d’Immunodifférenciation, EA 1556, Université Denis Diderot, Paris, France; and
Laboratório de Imunogenética, Instituto Butantan, São Paulo, Brazil
1Correspondence: Laboratoire d’Immunodifférenciation, EA 1556, Université Denis Diderot-Paris 7, Tour 54, CP7124, 2 Place Jussieu 75 251, Paris Cedex 05, France. E-mail: seman{at}paris7.jussieu.fr
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: bone marrow • polymorphonuclear neutrophils • genetics • polyacrylamide • apoptosis
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Phenotype-based genetic selection represents an alternative to models of knockout or transgenic mice for further exploration of PMN homeostasis. Indeed, this approach allows the evaluation of allelic gene combinations instead of gene mutations and might thus be more relevant to humans. Accordingly, two mouse lines have been selected on the basis of the AIR induced by subcutaneous (s.c.) injection of polyacrylamide beads [16 ]. The F0 founding population was generated from the intercross of eight inbred strains. The two mouse lines, AIRmax and AIRmin, for maximum and minimum AIR, respectively, were then generated by bi-directional, selective breeding based on cell counts and protein concentration in inflammatory exudates, 48 h and then 24 h upon injection of acrylamide [16 ]. The main phenotypic interline difference was characterized by a strong differential PMN count in inflammatory exudates. As could be anticipated, AIRmax mice were also shown remarkably resistant in various models of infection [17 ].
Genetics studies indicated that the interline difference involves seven to 11 loci, which are currently under investigation [16 ]. Gene combinations responsible for AIRmax and AIRmin phenotypes can possibly affect the production of chemotactic factors at the inflammatory site, PMN circulation, and homing or neutrophil production and differentiation in the BM. Experiments reported herein were designed to explore these possibilities at the biological level. AIRmax and AIRmin inflammatory parameters were compared with those in an unselected third party represented by BALB/c mice. In AIRmax, we demonstrate that a combination of convergent factors affecting granulopoiesis, chemotactic activity, and neutrophil apoptosis has accumulated during selective genetic breeding. Several candidate genes can be proposed, which may act synergistically and account for the unusual high responder phenotype of these mice. Similar combinations of individual alleles might exist in the human population with important consequence on the compartment of innate immunity.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Inflammation
Inflammation was induced as described previously [16
]. Briefly, animals were shaved, and 750 µl sterile 67% suspension (53 mg dry weight/ml) Biogel P100 (Biorad, France) in phosphate-buffered saline (PBS) was injected s.c. in the back. At given times, pouches were rinsed twice with 1 ml PBS–20 U/ml heparin on killed animals. Biogel beads were allowed to sediment. Cells were then recovered by centrifugation, and cell-free exudate was immediately frozen and kept at -20°C. Total cells in exudates were counted on a hemocytometer, and cytospins of the cells were stained with Wright-Giemsa. Cell subpopulations were differentially counted on cytospin preparations or by flow cytometry. Peripheral blood was obtained by retro-orbital sampling using heparinized capillary tubes and collected on PBS–heparin. Cells were analyzed after erythrocyte lysis using lysis buffer [4.15 g ammonium chloride, 0.84 g sodium bicarbonate, and 1 ml 0.5 M EDTA (pH 8) for 500 ml distilled water].
Assessment of cell phenotypes by flow cytometry
All antibodies used were from PharMingen (San Diego, CA) and used as fluorescein isothiocyanate (FITC) or phycoerythrin (PE) conjugates in two-color analyses by flow cytometry on a FACScalibur flow cytometer using the Cellquest software (Becton Dickinson, San Jose, CA). Ten thousand cells were analyzed.
Protein analysis
Twenty-four-hour inflammatory exudates from AIRmax and AIRmin mice adjusted to the same protein concentration were submitted to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 4–12% NuPAGETM gel gradients (Invitrogen, Carlsbad, CA). For two-dimensional (2-D) gel analysis, 100 µg exudate proteins were applied at the cathodic end of the immobilized pH gradient (IPG) gel strips (pH 3–10 L; Amersham Pharmacia, Little Chalfont, UK), and isoelectrofocusing was conducted for 18–20 h (32,000 vh) using IPGphor unit (Amersham Pharmacia). The electrophoresis in the second dimension was performed by 14% SDS-PAGE in a Hoefer DALT system (Amersham Pharmacia), run at a 40-mA constant current for 5 h [18
]. Silver- staining was performed according to standard protocol [19
]. Other 2-D gel preparations were blotted to polyvinylidene difluoride (PVDF) protein transfer and sequencing membranes, pore size 0.45 µm (Westran, Schleicher & Schuell BioScience, Keene, NH), and were stained with Coomassie brilliant blue R-250. Protein identification was performed by N-terminal sequence analysis according to Edman degradation. Spots on PVDF membranes were excised and applied to automated microsequencer PPSQ/23 (Shimadzu Corp., Kyoto, Japan) at the biophysics department, UNIFESP (Brazil). Six to 10 peptides were used for the database searches against the SWISS-PROT (http://www.expasy.ch) site.
BM cell (BMC) cultures
Liquid myeloid cultures were performed by seeding 5 x 105 BMC in 1 ml RPMI-1640 (Gibco-BRL, Grand Island, NY) medium supplemented with 2 mM glutamine, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 µg/ml streptomycin, 50 µM 2-mercaptoethanol, and 10% heat-inactivated fetal calf serum (FCS). Cells were plated in 24-well dishes together with 5 pg/ml granulocyte-colony stimulating factor (G-CSF) or 50 ng/ml granulocyte macrophage (GM)-CSF, 20 ng/ml stem cell factor (SCF), 0.1 ng/ml interleukin-3 (IL-3; PreproTech, Rocky Hill, NJ), 0.2 ng/ml IL-5 (PharMingen), or combinations of these cytokines. Pools of inflammatory exudates were used alone at variable dilutions or in combination with cytokines at a 10% final dilution. All-trans retinoic acid (ATRA; Sigma Chemical Co., St. Louis, MO) was used at doses ranging from 1 x 10-9 to 1 x 10-6 M. After 5 days of incubation at 37°C, 5% CO2, cells were carefully harvested, counted with a hemocytometer, and stained with indicated antibodies for phenotype analysis by flow cytometry.
Oxidative burst
Oxidative burst was performed on exudate cells, total BMC, or white blood cells as described previously [20
]. Cells (1x106) were suspended in 400 µl RPMI 1640 without serum. Catalase (3 µl) at 140 U/µl (Sigma Chemical Co.) and 1.8 µl 29 mM dihydrorhodamine 123 (DHR 123; Molecular Probes, Junction City, OR) were added to samples. After 5 min incubation at 37°C, 100 µl phorbol myristate acetate (PMA; Sigma Chemical Co.) at 2 µg/ml was added except in control tubes, which received RPMI. After additional 15 min incubation at 37°C, cells were analyzed in flow cytometry, and DHR 123 fluorescence was measured in FL2.
In vitro chemotaxis assays
Chemotaxis assays were performed using 24-well Transwell plates with a 3-µm pore size polycarbonate filter (Costar, Cambridge, MA) on unpurified BMC or white blood cells to avoid neutrophil preactivation by antibodies. Cells (1x106) in 200 µl RPMI-1640 medium without serum were loaded in the upper chamber. RPMI 1640 (600 µl) was placed in the lower chamber with 1 µM formyl-Met-Leu-Phe (fMLP; Sigma Chemical Co.) or crude exudate proteins. Plates were incubated for 1 h at 37°C. Transmigrated cells were then collected from the lower chamber and counted on the flow cytometer as described [13
]. Neutrophils were defined on light scatters.
Assessment of neutrophil apoptosis
Neutrophils isolated from 24-h Biogel-induced inflammatory pouches were suspended at 5 x 105 cells/ml in complete RPMI-1640–10% FCS medium and plated in 24-well dishes previously coated with Poly-2-hydroxyethylmethacrylate (Sigma Chemical Co.) to prevent adhesion. Cells were incubated for 0, 2, 4, or 6 h at 37°C, 5% CO2. For propidium iodide (PI) staining, pelleted cells were resuspended in 0.3 ml PBS. Ice-cold 100% ethanol (0.7 ml) was added drop-wise, and samples were stored at 4°C before analysis. After washing, cell pellets were resuspended in 0.25 ml PBS. RNase A (0.25 ml, 1 mg/ml; Sigma Chemical Co.) and 0.25 ml PI (50 µg/ml; Sigma Chemical Co.) were added. After 15 min of incubation at room temperature, cells were analyzed by flow cytometry. In parallel, staining with Annexin V–FITC (PharMingen) and PI was performed as described previously [21
]. Cytospins of cells were also done to detect apoptotic nuclei after Wright-Giemsa coloration.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
20 times higher than in AIRmin after 24 h. The average cell number was 75 ± 3.8 x 106 per pouch in AIRmax, 4.6 ± 1.2 x 106 in AIRmin, and 10.6 ± 0.7 x 106 in BALB/c, which thus displayed an intermediate–low phenotype compared with the selected mouse lines. Cell phenotypes were analyzed using Gr-1/CD11b double-staining in flow cytometry. At 24 h in all mice (Fig. 1B) , over 85% of exudate cells were Gr-1bright/CD11b+ and corresponded to neutrophils as confirmed by cytology. The other cells were Gr-1low/CD11b+ and represented macrophages. Despite a low number of PMN in exudates of AIRmin mice, there was no evidence for abnormal expression of MEL14 (CD62L), CD11b (Fig. 1C)
, intercellular adhesion molecule-1 (CD54), IL-1 receptor (IL-1R), or IL-6R among cells of either strain (data not shown). As also illustrated in Figure 1C
, CD11b integrin expression increased in a similar way on Gr-1+ cells from BM, blood, and exudates, whereas CD62L selectin conversely decreased on Gr-1+ cells from the same tissues. Despite important quantitative differences in numbers, mature neutrophils from AIRmax, AIRmin, or BALB/c had the same phenotype and gave similar responses to PMA-induced oxidative burst (Fig. 1D) or in phagocytosis assays using FITC-labeled Escherichia coli (not shown). This indicated that PMN from the three mouse lines had no major functional difference.
 View larger version (26K): [in a new window] |
Figure 1. Inflammatory response in AIRmax and AIRmin mice. (A) Kinetics of cell migration to the Biogel-induced inflammatory site in AIRmax ( ), AIRmin (•), and BALB/c mice ( ). (B) Gr-1/CD11b profile of extravasated cells at the inflammatory site after 24 h in BALB/c, AIRmax, and AIRmin. (C) Relative expression of CD62L and CD11b on Gr-1+ cells in BM, blood, and 24-h inflammatory exudates (Ex) from BALB/c (dotted line), AIRmax (bold line), and AIRmin mice (thin line). (D) Oxidative burst in neutrophils from AIRmax, AIRmin, and BALB/c mice. Neutrophils from 24-h inflammatory exudates were stimulated with 2 µg/ml PMA for 5 min. DHR 123 fluorescence was measured after additional 15 min incubation at 37°C. Results are typical from three different experiments.
|
 View larger version (67K): [in a new window] |
Figure 2. Biochemical analysis and chemotactic activity of inflammatory exudates from AIRmax and AIRmin mice. (A) Kinetic evolution of protein concentration in exudates from AIRmax (), AIRmin ( ), and BALB/c ( ) mice ± SE (n=5). (B) SDS-PAGE analysis of 24-h inflammatory exudates from four individual AIRmax and AIRmin mice. Exudates were adjusted to the same protein concentration. (C) Two-dimensional electrophoresis of pools from six AIRmax and six AIRmin inflammatory exudates harvested 24 h after s.c. Biogel injection. Several indicated spots were identified by microsequencing and are: 48, C5a anaphylatoxin, P06684*; 37, C3a anaphylatoxin, P01027*; 41, macrophage inflammatory protein 2 (MIP-2) chemokine CxC, P10889*; 39, tumor necrosis factor receptor (TNFR), fibroblast growth factor-inducible, immediate-early response protein 14 (FGF-inducible 14), Q9CR75*; 42, lymphotactin precursor (XCL1; cytokine single C motif-1; lymphotaxin; small inducible cytokine C1), P47993*. *, SwissProt accession number. (D) Chemotactic activity on BALB/c BM neutrophils of 24-h exudates from AIRmax and AIRmin mice (results are the mean±SE of three different experiments). (E) Chemotactic response of BM neutrophils from AIRmax, AIRmin, and BALB/c to 10% 24-h AIRmax exudate. The number of transmigrated neutrophils ± SE is adjusted for 1 x 106 neutrophils in the upper chamber defined by Gr-1/CD11b staining and light scatters on total BMC (n=3). Nb, Number.
|
|
View this table: [in a new window] |
Table 1. Relative Percentage of Neutrophils in BM and Blood from AIRmax, AIRmin, and BALB/c Mice
|
 View larger version (34K): [in a new window] |
Figure 3. High in vitro granulopoiesis of AIRmax BMC in response to cytokines and exudates. (A) Stimulation of BALB/c–BMC proliferation and PMN differentiation by inflammatory exudates (Ex) from AIRmax and AIRmin mice. Cells (2x105) were seeded in 0.2 ml cultures. Proliferation and differentiation were measured on day 5. Values ± SE are the mean of five experiments (comparison with control cultures; *, P<0.001). (B) AIRmax and AIRmin BMC response to 10% inflammatory exudates. Experiments were performed as in A. (C and D) BMC (5x105) from AIRmax, AIRmin, and BALB/c mice were cultured with different cytokines. Total cell number (Nb; C) or neutrophils defined by Gr-1/CD11b staining (PMN; D) were measured on day 5. Values shown are mean ± SE from eight different experiments. Comparisons between AIRmax and AIRmin (*), AIRmax and BALB/c (), and BALB/c and AIRmin (•) by the Student’s t-test; ***, P  0.001.
|
60% in AIRmax and BALB/c but only 38% in AIRmin. An additive effect on total number and percentage of GR-1bright/CD11b+ cells was detected in cultures containing GM-CSF and exudate or even GM-CSF, exudates, and SCF. It is important that a significant difference was observed between AIRmin– and AIRmax–BMC stimulation standing out the response to GM-CSF, IL-3, and exudate. When compared with BALB/c, AIRmax average response was also constantly higher. Yet, BALB/c response was also higher than that observed in AIRmin cultures. Hence, AIRmax–BMC appeared more reactive to granulopoietic cytokines than those from AIRmin but also from genetically unselected BALB/c mice.
High expression of CD131 on AIRmax BM cells
GM-CSF, IL-3, and IL-5 have specific receptors, which are heterodimers composed of 
and ß chains [22
]. Specificity for each cytokine is supported by the subunit, whereas GM-CSFR, IL-3R, and IL-5R share a common ßc chain (CD131), which plays an essential role in signal transduction [14
]. Variation in CD131 structure or expression could thus contribute to the differential BMC response to GM-CSF and IL-3 among AIRmax, AIRmin, and BALB/c mice. As illustrated in Figure 4A
, CD131 expression was detected ex vivo on BMC from the three mouse lines among Gr-1high and Gr-1low but not on Gr-1- cells. Among Gr-1low cells, CD131 expression was slightly higher in AIRmax, with a mean fluorescence value of 100.4 versus 53.4 and 48 in AIRmin and BALB/c, respectively. More importantly, in the Gr-1high population, which corresponds to cells engaged into the neutrophil differentiation pathway, CD131 expression was broad in AIRmin or BALB/c. In contrast, most Gr-1high cells from AIRmax mice expressed CD131 at a very high level (R1 gate in Fig. 4A
). To further control the specificity of CD131 induction during neutrophil differentiation, we then tested CD131 expression on unstimulated BMC or on BMC stimulated by different cytokines for 5 days in vitro. A high level of CD131 expression was induced on AIRmax cells cultured with GM-CSF, IL-3, IL-5, or exudate, although at a lower level in the last two conditions (Fig. 4B)
. No CD131 expression was observed in unstimulated cultures or in cultures stimulated with SCF or G-CSF. This was consistent with the existence of preformed receptors on myelocytic precursors up-regulated by cytokines requiring their expression for signal transduction [23
]. We then compared CD131 induction in BMC cultures from AIRmax and AIRmin mice stimulated with GM-CSF or GM-CSF plus inflammatory exudate. Under these conditions, CD131 was induced on cells from both lines on a population of large, granular Gr-1low cells, as seen ex vivo. Again, the mean CD131 fluorescence was twice as high in AIRmax compared with AIRmin. In the Gr-1high population, CD131 level was clearly higher in AIRmax than in AIRmin (Fig. 4C) . This was in agreement with our ex vivo observations (Fig. 4A) . In an attempt to identify the cytokine receptor expressed at a high level on AIRmax Gr-1high cells, IL-3R/GR-1 double-stainings were performed on BMC. IL-3R was not found on the brightest GR-1+ cells but only on GR-1low immature precursors in all strains (Fig. 4D
and data not shown).
 View larger version (30K): [in a new window] |
Figure 4. High expression of CD131 on BMC from AIRmax mice. (A) CD131 expression on Gr-1+ BMC ex vivo. (B) Induction of CD131 expression by GM-CSF, IL-3, IL-5, and inflammatory exudate (Ex) on AIRmax BMC (5x105/ml). Cells were analyzed on day 5. Results are representative of three different experiments. (C) Comparison of CD131 expression on Gr-1+ cells in AIRmax (bold) and AIRmin (gray area) after 5 days of BMC stimulation in vitro with GM-CSF or GM-CSF plus 10% 24-h inflammatory exudates (Ex). Results are typical from four different experiments. (D) IL-3R expression on Gr-1+ BMC ex vivo. (E) BMC (5x105/ml) response to IL-5 or IL-5 plus 10% 24-h inflammatory exudates (Ex). Total cell number and eosinophils, defined by light-scatter and Gr-1dull/CD11b+ phenotype in flow cytometry, were counted on day 5. Results shown are mean ± SE from three different experiments.
|
Altogether, these results suggested a fast or high propensity of AIRmax BM precursor cells to up-regulate the receptors to cytokines involved in neutrophil and eosinophil differentiation. The high CD131 level found in this line was likely associated with high IL-5R and more importantly, high GM-CSFR expression.
Differential response to RA
Neutrophil differentiation is under the control of several nuclear factors, some of them, such as CCAAT/enhancer-binding protein (C/EBP) or C/EBP
, playing a selective and critical role [25
]. These differentiation signals can alternatively or complementarily be provided through retinoic acid (RA) receptors (RAR) [26
]. To further explore the granulocyte-differentiation pathway, we compared the response to all-trans retenoic acid (ATRA) in BMC cultures of the three mouse lines, alone or in combination with GM-CSF. After 5 days in liquid cultures, no significant change in the number of cells was observed with ATRA concentrations ranging from 10-9 to 10-6 M or in the Gr-1/CD11b phenotype. When combined with GM-CSF, an approximate doubling in the number of cells per culture was observed in the three mouse strains with 10-6 M ATRA by comparison with GM-CSF alone. This was accompanied by an increase in the fraction of Gr-1bright/CD11b+ cells in all instances (data not shown). However, Gr-1/CD11b staining is not fully informative of the differentiation stage, as it depends on the fluorochrome to which the Gr-1 antibody is conjugated [26
]. FITC-conjugated anti-Gr-1 provides the best resolution, whereas Gr-1-PE gives a more compact staining. Cells were then further analyzed using Gr-1-FITC conjugate and CD38-PE. Indeed, CD38 is expressed at a high level on myelocytic cells but down-regulated on most mature neutrophils [27
]. Additionally, CD38 expression is induced by ATRA [28
]. This was confirmed in the three mouse lines (Fig. 5A
). With GM-CSF, two main populations of cells were found in vitro on day 5: one CD38+/Gr-1low corresponding to myelocytes and one CD38dull/Gr-1bright corresponding to neutrophils (Fig. 5A)
. With GM-CSF and 10-6 M ATRA in AIRmin and BALB/c mice, two populations could be distinguished on the basis of Gr-1 expression, which expressed CD38 at very a high level. Surprisingly, in AIRmax, a shift to a CD38bright phenotype was also observed in the Gr-1low population, but a large fraction of cells was Gr-1high/CD38low. Cytospin preparations (Fig. 5B)
revealed that most CD38bright neutrophils induced in the presence of ATRA had ring-shaped nuclei, whereas CD38low/Gr-1+ cells had segmented nuclei such as mature peripheral neutrophils (Fig. 5C) . This suggested an accelerated neutrophil differentiation in response to ATRA in AIRmax mice.
 View larger version (32K): [in a new window] |
Figure 5. Differential AIRmax BMC response to ATRA. (A) BMC (5x105/ml) were cultured for 5 days with GM-CSF or GM-CSF plus 1 µM ATRA and then analyzed by two-color in flow cytometry for Gr-1 and CD38 expression. Profiles are typical from three independent experiments. (B and C) Cytospin and Wright-Giemsa staining of neutrophils induced with GM-CSF and ATRA in BALB/c showing ring-shaped nuclei (B) and in AIRmax showing segmented nuclei (C).
|
20% in the two other strains (not shown). A more striking difference was observed after 6 h of incubation at 37°C for the three investigated criteria (Fig. 6A 6B
6C)
, indicating that neutrophils from AIRmax mice were less prone to spontaneous apoptosis than those from BALB/c or AIRmin mice.
 View larger version (66K): [in a new window] |
Figure 6. AIRmax neutrophils are resistant to spontaneous apoptosis. (A) Fluorescein-activated cell sorter analysis of propidium iodide-stained 24-h exudate cells ex vivo (shaded) or after 6 h of incubation at 37°C (bold line) showing high DNA fragmentation in AIRmin and BALB/c compared with AIRmax (n=3). Numbers represent the percentage of cells with fragmented DNA after 6 h. (B) Cytospin and Wright-Giemsa staining of neutrophils from 24-h exudates after 6 h of incubation at 37°C showing high numbers of apoptotic nuclei in AIRmin and BALB/c but not in AIRmax mice. Arrows indicate apoptotic nuclei. (C) Analysis of neutrophil from 24-h exudates by Annexin V/propidium iodide staining after 6 h of incubation at 37°C showing limited apoptosis in AIRmax.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Remarkably, the high inflammatory responsiveness in AIRmax seems to result from the accumulation of three convergent elements during genetic selection. They produce a higher number of neutrophils in their BM as a consequence of a higher response to granulopoietic cytokines, and they produce higher amounts of chemotactic factors in exudates to attract neutrophils. Finally, extravasated neutrophils are more resistant to spontaneous apoptosis. Together, these three elements account for the tremendous number of neutrophils at the inflammatory site observed in these mice. This phenotype has no precedent and raises the question of the link between these convergent alterations. Do they result from the effect of a predominant gene or correspond to separate allelic genes, which accumulated during selection? If one considers the first hypothesis, a link can be envisioned among the high neutrophil production in BM, the high response to IL-3, IL-5, and GM-CSF, accompanied by a strong up-regulation of CD131 on not fully mature granulocytes, the fast neutrophil maturation in response to ATRA, the high chemotactic activity in exudates, and the delayed neutrophil apoptosis. Indeed, several nuclear transcription factors are involved in granulocytic terminal differentiation including c-myb, MZF-1, PU.1, and retinoic receptors [30
]. Some of them, such as C/EBP and C/EBP, are mainly expressed in cells of the myeloid lineage [31
]. C/EBP acts temporally downstream of C/EBP in granulopoiesis but plays a critical role, as C/EBP-deficient mice, although normal at birth and fertile, fail to produce normal neutrophils and eosinophils [32
]. A number of myeloid genes contain C/EBP-binding sites in their promoter region, including those encoding for cytokine receptors such as G-CSF and GM-CSF [33
]. C/EBP-deficient mice have abnormal expression of CD62L-selectin and CD11b/CD18 integrin, two important proteins involved in neutrophil transmigration [34
] and apoptosis [12
]. C/EBP overexpression in transfected cell lines also indicates that this transcriptional activator induces the production of several chemokines [31
]. At last, the C/EBP promoter contains a retinoic acid responding element (RARE) element, suggesting that C/EBP might be one of the RAR targets [35
] providing a rational mechanism for ATRA-mediated granulopoiesis regulation [36
].
In humans, multiple C/EBP mRNA species have been detected as a consequence of two alternative promoters and alternate splicing, leading to four different proteins, which may have different transactivating activities [30
]. In mice, a single mRNA species of 1.8 kb seems detected in murine tissues and cell lines corresponding to a 34-kDa protein resembling the predominant 32-kDa form in humans [31
]. The polymorphism of murine C/EBP has not been extensively investigated, but it seems likely that mutants with functional consequences should exist as described for human C/EBP [37
]. From above, a so-far unidentified allele of the C/EBP gene, C/EBP overexpression, or expression of a form resulting from alternate splicing, from one of the other internal, translational initiation codons or from post-transcriptional modifications, might account for most of the phenotype in AIRmax mice.
This attractive hypothesis does not exclude the involvement of other separate genes. In this respect, biochemical characterization of exudate proteins might be informative. Indeed, two distinct biological activities were found. The first one allows the activation of BMC proliferation and neutrophil differentiation, alone or in combination with recombinant cytokines (Figs. 3 and 5) . The synergy observed between exudates and GM-CSF, even with increased concentrations of this cytokine (not shown), seems to exclude GM-CSF as the sole BM-stimulating factor in exudates. However, it is important that the ability to support BMC proliferation is equivalent in AIRmax and AIRmin exudates. The second activity is the induction of neutrophil chemotaxis, which is much stronger with AIRmax than with AIRmin exudates. Identification of the chemokines is a key issue under investigation. Indeed, the overall increase in neutrophil accumulation observed at the inflammatory site in AIRmax may be a result of the initial neutrophil accumulation related to a higher number of circulating leukocytes and leading to a positive feedback for further neutrophil recruitment [12 ]. In other words, the question remains as to whether the higher chemotactic activity results from a higher number of neutrophils or a higher production of chemokines by these neutrophils. This alternative is illustrated by the observation that among the chemokines accumulating in AIRmax exudates, we identified C3a and C5a. This was surprising, as polyacrylamide beads do not contain free hydroxyl or amino groups and cannot induce C activation in vitro (data not shown). However, mouse PMN can produce C3 and factor B [38 ]. Moreover, several neutrophil proteinases, including elastase and members of the cathepsin family, can generate anaphylatoxins from C3 and C5, independently of complement activation by the alternative pathway [39 40 41 42 ]. Finally, AIRmax and AIRmin seem to have similar C5 circulating levels (unpublished results). The higher number of neutrophils in AIRmax pouches might thus lead to increased proteinase release and C3 and C5 cleavage, illustrating a positive feedback regulation. Accordingly, lower amounts of high molecular weight proteins were found in AIRmax exudates compared with AIRmin (Fig. 2B and 2C) , despite the use of proteinase inhibitors during exudate analysis. High concentration of MIP-2 was also found in AIRmax exudates. It is interesting that C5a and IL-6 were shown to synergistically enhance MIP-2 production in vitro [43 ]. We found significantly higher levels of IL-6 in AIRmax exudates (data not shown). The source of IL-6 remains to be defined but represents another convergent factor leading to the AIRmax phenotype.
Another set of independent genes might be that controlling neutrophil apoptosis. The mechanisms involved in this spontaneous process remain incompletely understood [29 ]. Several cytokines and chemotactic agents such as GM-CSF, G-CSF, IL-1ß, or chemokines involving mitogen-activating protein kinases, such as extracellular-regulated kinase or phosphatidylinositol-3K, have been shown to delay neutrophil apoptosis [44 45 46 ]. Yet, the regulation of apoptosis in extravasated neutrophils is poorly known and seems different from that in circulating ones [47 ]. Cytokines and chemokines present in AIRmax exudates might be responsible for the delayed apoptosis observed in extravasated neutrophils from this line. CD11b/CD18 ß2 integrin was shown involved in neutrophil apoptosis after transmigration [12 ]. As seen in Figure 1C , CD11b level on AIRmax, AIRmin, and BALB/c PMN was similar, suggesting that Mac-1-mediated death signal may not be responsible for the interline difference. This, however, does not exclude the differential intervention of a downstream effector in this apoptotic pathway.
In several respects, neutrophil homeostasis in AIRmin seems rather similar to that seen in BALB/c mice but different from AIRmax. The two mouse lines were submitted to opposite selective pressures. It is thus possible that genes selected to achieve low inflammatory response are different from those leading to the AIRmax phenotype. Genetics studies indicated that the AIRmax/AIRmin interline difference would involve seven to 11 distinct loci [15 ]. If one considers BALB/c as a reference strain, the number of loci separating AIRmax and AIRmin from BALB/c might be different. Our present results indicate that several gene combinations can deeply affect neutrophil homeostasis. Given the high resistance of AIRmax animals in several models of infection by intracellular pathogens [17 ] and experimental carcinogenesis [48 , 49 ], these mice represent a unique model for the study of neutrophil homeostasis and its consequence on the susceptibility to major diseases. Such incidental gene combinations might exist in the human population with major consequences on health. The rare HIV-infected patients with very low CD4 counts but resistant to opportunist infections [50 51 52 ] might represent human AIRmax equivalents.
Received January 23, 2003; revised June 2, 2003; accepted June 3, 2003.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Adriouch, S. Hubert, S. Pechberty, F. Koch-Nolte, F. Haag, and M. Seman NAD+ Released during Inflammation Participates in T Cell Homeostasis by Inducing ART2-Mediated Death of Naive T Cells In Vivo J. Immunol., July 1, 2007; 179(1): 186 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Laragione, N. C. Yarlett, M. Brenner, A. Mello, B. Sherry, E. J. Miller, C. N. Metz, and P. S. Gulko The Arthritis Severity Quantitative Trait Loci Cia4 and Cia6 Regulate Neutrophil Migration into Inflammatory Sites and Levels of TNF-{alpha} and Nitric Oxide J. Immunol., February 15, 2007; 178(4): 2344 - 2351. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Ha, F. Hua, D. Grant, Y. Xia, J. Ma, X. Gao, J. Kelley, D. L. Williams, J. Kalbfleisch, I. W. Browder, et al. Glucan phosphate attenuates cardiac dysfunction and inhibits cardiac MIF expression and apoptosis in septic mice Am J Physiol Heart Circ Physiol, October 1, 2006; 291(4): H1910 - H1918. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. F. Di Pace, S. Massa, O. G. Ribeiro, W. H. K. Cabrera, M. De Franco, N. Starobinas, M. Seman, and O. C. M. Ibanez Inverse genetic predisposition to colon versus lung carcinogenesis in mouse lines selected based on acute inflammatory responsiveness Carcinogenesis, August 1, 2006; 27(8): 1517 - 1525. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Villamor and W. W. Fawzi Effects of Vitamin A Supplementation on Immune Responses and Correlation with Clinical Outcomes Clin. Microbiol. Rev., July 1, 2005; 18(3): 446 - 464. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
