Human polymorphonuclear leukocytes (PMNs or neutrophils) kill invading microorganisms with reactive oxygen species (ROS) and cytotoxic granule components. PMNs from individuals with X-linked chronic granulomatous disease (XCGD) do not produce ROS, thereby rendering these individuals more susceptible to infection. In addition, XCGD patients develop tissue granulomas that obstruct vital organs, the mechanism(s) for which are unknown. To gain insight into the molecular processes that contribute to the pathophysiology of XCGD, including formation of granulomas, we compared global gene expression in PMNs from XCGD patients and healthy control individuals. Genes encoding mediators of inflammation and host defense, including CD11c, CD14, CD54, FcγR1, FcαR, CD120b, TLR5, IL-4R, CCR1, p47phox, p40phox, IL-8, CXCL1, Nramp1, and calgranulins A and B, were up-regulated constitutively in unstimulated XCGD patient PMNs. By comparing transcript levels in normal and XCGD PMNs after phagocytosis, we discovered 206 genes whose expression changed in the presence and the absence of ROS, respectively. Notably, altered Bcl2-associated X protein synthesis accompanied defective neutrophil apoptosis in XCGD patients. We hypothesize that granuloma formation in XCGD patients reflects both increased proinflammatory activity and defective PMN apoptosis, and we conclude that ROS contribute directly or indirectly to the resolution of the inflammatory response by influencing PMN gene transcription.

Human polymorphonuclear leukocytes (PMNs3 or neutrophils) are the first line of defense against invading microorganisms and ingest pathogens by a process known as phagocytosis (1). During phagocytosis, PMNs produce reactive oxygen species (ROS) to kill ingested microbes (2). Leukocytes from patients with X-linked chronic granulomatous disease (XCGD) (3) are defective in their ability to produce ROS due to abnormal or absent gp91phox, a transmembrane protein encoded by CYBB (4, 5, 6). As a result of this monogenic hereditary defect, XCGD patients acquire life-threatening bacterial and fungal infections and develop granulomas that impair the function of vital organs (7, 8, 9, 10). Although the absence of ROS logically explains the increase in the frequency and severity of pyrogenic infections in patients with XCGD, the molecular basis for formation of granulomas in these individuals is unknown.

There is significant evidence to suggest that removal of PMNs from sites of infection by apoptosis is essential for resolution of the inflammatory response (11, 12, 13, 14, 15). Importantly, recent studies have linked delayed apoptosis with accumulation of PMNs in inflammatory diseases (16). Although several mechanisms probably mediate programmed cell death in human leukocytes, ROS provide an important molecular signal that initiates or accelerates apoptosis in PMNs (17, 18). Inasmuch as inflammatory diseases can be attributed in part to delayed PMN apoptosis (16), it is possible that altered neutrophil apoptosis and the resultant unresolved inflammation contribute to formation of granulomas in XCGD patients. As an initial step toward testing this hypothesis, we studied global gene expression in PMNs from XCGD patients.

Sterile water and 0.9% sodium chloride (both Irrigation, USP) were purchased from Baxter Healthcare (Deerfield, IL). Dextran T-500 and Ficoll-Paque Plus were obtained from Amersham Pharmacia Biotech (Piscataway, NJ). Rabbit Ab specific for human serum albumin was purchased from ICN Biomedicals (Costa Mesa, CA). RPMI 1640 medium was purchased from Invitrogen (Carlsbad, CA). Latex beads (2.0 μm in diameter) were purchased from Polysciences (Warrington, PA). All reagents used for TaqMan real-time PCR were obtained from PE Applied Biosystems (Foster City, CA). Unless specified, all other reagents were obtained from Sigma-Aldrich (St. Louis, MO).

Eight patients (14–28 years of age) diagnosed clinically with XCGD participated in the microarray (six patients, XCGD1–XCGD6) or TaqMan real-time RT-PCR studies (two patients, XCGD7 and XCGD8) of PMN gene expression. In addition to clinical presentation, XCGD was confirmed by immunoblot analysis to verify the absence of gp91phox, and assays that measure superoxide-generating capacity in PMA-activated PMNs (e.g., oxidation of dihydrorhodamine 123 to rhodamine 123, reduction of ferricytochrome c or of nitro blue tetrazolium to formazan). Specific mutations for XCGD patients (where known) are as follows: XGD7, g to t mutation at the +1 splice site at beginning of intron VII, resulting in skipping of exon 7 and no protein; and XCGD8, g264a splice site mutation at the end of exon 3, resulting in no protein. None of the XCGD patients participating in this study had acute illness or infection at the time blood was drawn for the experiments.

Patients XCGD1–XCGD6 were chosen for the microarray study in part because they were not on IFN-γ therapy, and no patient had IFN-γ discontinued for the purpose of conducting this study. Patients were not receiving IFN-γ either because of a personal choice on the part of the patient or because the patient had experienced unacceptable side effects, such as intractable headaches or malaise not controlled by nonsteroidal anti-inflammatory medications. All these patients, with the exception of XCGD1, were taking 3–10 mg of prednisone once daily or every other day. Importantly, there were no notable differences in gene expression between XCGD1 (not taking prednisone) and the other five XCGD patients (taking prednisone). Additionally, patients were taking oral and/or i.v. antibacterial, antifungal, or antiviral medications at the time of the studies as follows: XCGD1, trimethoprim-sulfamethoxazole, meropenem, and posaconazole; XCGD2, vancomycin and moxifloxacin; XCGD3, azithomycin, voriconazole, levofloxin, and trimethoprim-sulfamethoxazole; XCGD4, trimethoprim-sulfamethoxazole and itraconazole; XCGD5, itraconazole, levofloxin, and azithomycin; and XCGD6, linezolid, voriconazole, acyclovir, and cefixime. Although we cannot exclude the possibility that these medications alter gene expression in neutrophils, recent reports suggest that any such effects in human cells would be limited (19, 20, 21). Furthermore, the overall expression pattern of individual genes was relatively similar among all XCGD patients (Figs. 1 and 2), a finding incompatible with the idea that changes in transcript levels are due to the diverse combinations of patient medications. For analyses of gene expression changes in PMNs after phagocytosis, it is important to note that any effects of antibacterial, antifungal, or antiviral medications would be present in both unstimulated and stimulated PMNs. Therefore, it is unlikely that there would be significant net differences in gene expression (due to prophylactic therapies) between unstimulated and stimulated PMNs.

FIGURE 1.

Normal patterns of gene expression (up-regulated genes) are altered in activated PMNs from patients with XCGD. Differential gene transcription was measured with oligonucleotide microarrays after phagocytosis in human PMNs from four healthy control donors or four to six XCGD patients as indicated. The fold change for each individual is indicated by the scale bar. Time after PMN activation is indicated to the left of the boxes, which indicate the magnitude of change for each differentially expressed gene. The number of XCGD patients used at each time is as follows: PMN gene expression at 1.5 h included patients XCGD1, -2, -3, and -5; at 3 h, patients XCGD1, -2, -4, -5, and -6; and at 6 h, all XCGD patients were included. Gene expression was measured in the samples from four healthy control individuals at each time point. NC, No change.

FIGURE 1.

Normal patterns of gene expression (up-regulated genes) are altered in activated PMNs from patients with XCGD. Differential gene transcription was measured with oligonucleotide microarrays after phagocytosis in human PMNs from four healthy control donors or four to six XCGD patients as indicated. The fold change for each individual is indicated by the scale bar. Time after PMN activation is indicated to the left of the boxes, which indicate the magnitude of change for each differentially expressed gene. The number of XCGD patients used at each time is as follows: PMN gene expression at 1.5 h included patients XCGD1, -2, -3, and -5; at 3 h, patients XCGD1, -2, -4, -5, and -6; and at 6 h, all XCGD patients were included. Gene expression was measured in the samples from four healthy control individuals at each time point. NC, No change.

Close modal
FIGURE 2.

Normal patterns of gene expression (down-regulated genes) are altered in activated PMNs from patients with XCGD. Differential gene transcription was measured in PMNs from healthy control donors or XCGD patients after phagocytosis as indicated.

FIGURE 2.

Normal patterns of gene expression (down-regulated genes) are altered in activated PMNs from patients with XCGD. Differential gene transcription was measured in PMNs from healthy control donors or XCGD patients after phagocytosis as indicated.

Close modal

The XCGD patients who participated in the confirmation studies of PMN gene expression by TaqMan real-time PCR, XCGD7 and XCGD8, were undergoing IFN-γ therapy and taking trimethoprim-sulfamethoxazole at the time of the studies. In addition, patient XCGD8 was taking voriconazole. The decision to use patients undergoing IFN-γ therapy (XCGD7 and XCGD8) for the confirmation of microarray studies was entirely coincidental and based on limited patient availability. As any effects of IFN-γ would be present in both unstimulated and stimulated PMNs in patients XCGD7 and XCGD8, it is unlikely that there would be differences in gene expression (due to IFN) between unstimulated and stimulated PMNs. Notably, there was strong positive confirmation of microarray results by TaqMan real-time RT-PCR (see below), indicating that treatment of patients XCGD7 and XCGD8 with IFN-γ did not significantly alter the patterns of PMN gene expression induced by phagocytosis.

PMNs were isolated from heparinized venous blood (22) of healthy individuals or patients with XCGD in accordance with a protocol approved by the institutional review boards for human subjects at the National Institute of Allergy and Infectious Diseases and University of Iowa. All studies were conducted according to Declaration of Helsinki principles. Briefly, blood was incubated for 20 min at room temperature at a 1:1 ratio with 0.9% sodium chloride (Irrigation USP; Baxter Healthcare) containing 3.0% Dextran T-500 (Amersham Pharmacia Biotech) to sediment erythrocytes. The leukocyte-containing supernatant was centrifuged at 670 × g for 10 min and resuspended in 35 ml of 0.9% sodium chloride. The leukocyte-rich saline suspension was underlayed with 10 ml of Ficoll-Paque Plus (1.077 g/L; Amersham Pharmacia Biotech) and centrifuged at ambient temperature for 25 min at 350 × g to separate PMNs from PBMCs. PBMCs were removed by aspiration, and erythrocytes were lysed with water (Irrigation USP; Baxter Healthcare) for 15–30 s, followed by immediate mixing with 1.7% sodium chloride. Purified PMNs were centrifuged at 350 × g, resuspended in RPMI 1640 buffered with 10 mM HEPES, and enumerated by microscopy. The purity of the PMN preparations and cell viability were routinely assessed by flow cytometry (FACSCalibur; BD Biosciences, San Jose, CA). Cell preparations contained ∼99% PMNs, and all reagents used contained <25.0 pg/ml endotoxin.

Phagocytosis experiments were performed as described by Kobayashi et al. (13, 14). Briefly, PMNs (107) were combined on ice with or without IgG and C3bi-coated latex beads (8 × 107) in wells of a 12-well tissue culture plate (precoated with 20% normal human serum) and centrifuged at 350 × g for 8 min at 4°C to synchronize phagocytosis. For the purpose of these studies, activated or stimulated PMNs are defined as those that have been stimulated by phagocytosis of IgG- and C3bi-coated latex beads. After centrifugation, plates were incubated at 37°C in a CO2 incubator for up to 6 h. At the indicated times, tissue culture medium was aspirated from the plate, and PMNs were lysed directly with RLT buffer (Qiagen, Valencia, CA). Purification of PMN RNA and subsequent preparation of labeled cRNA target (12 μg) were performed as previously described (13, 14). Labeling of samples, hybridization of cRNA with Hu95Av2 oligonucleotide arrays (Affymetrix, Santa Clara, CA), and scanning were performed according to standard Affymetrix protocols (http://www.affymetrix.com/pdf/expression_manual.pdf) as described previously (13, 14, 15). Inasmuch as the control and activated cells were each placed on ice before the phagocytosis experiments, it is unlikely that there would be significant net differences in gene expression due to cold shock. Experiments were performed with blood from four healthy donors at all time points and from four to six XCGD patients at the indicated time points using a separate oligonucleotide array for each donor. Gene expression in healthy control and XCGD individuals was always compared at the same time points after phagocytosis.

Data were analyzed with Microarray Suite, version 5.0 (MAS5; Affymetrix) and GeneSpring version 5.0 (Silicon Genetics, Redwood City, CA) as described previously (13, 14, 15), but with several modifications. We judged genes to be differentially expressed in unstimulated control and XCGD cells only when 1) the gene was identified as present by MAS5 (Affymetrix) in all samples studied; 2) the difference in expression between normal and XCGD samples was ≥2-fold; and 3) the extent of difference in expression was statistically significant (p ≤ 0.05, by Student’s t test).

To compare gene expression by phagocytosing PMNs from healthy control donors and XCGD patients, we employed a two-tier procedure for screening and analysis. Two criteria were used in the first tier of evaluation. To be designated differentially expressed, a gene must 1) be identified as present by MAS5 (Affymetrix) in samples from the study group (e.g., control or XCGD donors), and 2) have an average level of expression ≥2-fold that of unstimulated PMNs from the same study group (e.g., control or X-CGD donors) at the same time point of incubation.

In the second tier, three criteria were used to determine whether gene expression changes were different between healthy control individuals and XCGD patients after phagocytosis. These criteria were used individually or in combination. First, changes in gene expression were statistically significant in one group of donors (healthy controls or XCGD), but were not significant in the other group. Second, the difference in fold change between the groups (healthy controls or XCGD), was at least 1.0 (e.g., 2.0 vs 3.0). Third, if genes were changed by >2-fold in both the healthy control and XCGD patient groups, very limited overlap in expression changes among individuals between the two groups was allowed. In some instances, genes met criteria in the second tier at a given time point, indicating a difference in gene expression between PMNs from healthy control individuals and XCGD patients, but changes in expression were comparable at another times. Data for those genes were selectively included. In addition to the data provided in the two supplemental tables (Tables II and III, published as supplemental data on the Journal of Immunology web site), a complete set of microarray results compliant with Minimum Information About a Microarray Experiment guidelines can be found at http://www.niaid.nih.gov/dir/labs/lhbp/deleo.htm.

Phagocytosis experiments and RNA preparation for TaqMan analysis were performed using conditions identical with those used for the microarray analysis. TaqMan analysis of samples from three healthy control individuals and two patients with XCGD were performed with an ABI 7700 thermocycler (PE Applied Biosystems) as previously described (13, 14). The expression of MT1G after treatment with 100 μM ZnCl2 was measured in PMNs from three individuals at the indicated times. The correlation between TaqMan real-time RT-PCR and microarray analysis was 87.5%. That is, in 87.5% of the TaqMan real-time RT-PCR assays (21 of 24 TaqMan-microarray comparisons) in which gene expression changed at least 1.5-fold in the healthy control or patient samples, genes changed at least 1.5-fold by microarray analysis, or they were similarly unchanged (not at least 1.5-fold induced or repressed) in either the healthy control or XCGD patient samples.

PMN apoptosis after phagocytosis was measured with flow cytometry (13, 14) using a modified TUNEL assay (Apo-BRDU Apoptosis Detection Kit; BD Biosciences) as described by the manufacturer. Alternatively, cells were stained with annexin V-FITC (Annexin VFITC Apoptosis Detection Kit II; BD Biosciences) as described by the manufacturer and were fixed overnight in Cytofix/Cytoperm (BD Biosciences) before staining for BAX.

Staining for intracellular BAX (mAbs 2D2 and 6A7; Sigma-Aldrich) and metallothionein (mAb E9; DAKO, Carpinteria, CA) in PMNs from healthy control donors and patients with XCGD was performed at the indicated times after phagocytosis (BAX staining) or after treatment with 100 μM ZnCl2 (metallothionein staining). Metallothionein is a metal toxicity-responsive protein that is induced by ZnCl2 in many cell types (23). Briefly, PMNs were fixed overnight with Cytofix/Cytoperm (BD Biosciences), washed twice with Perm/Wash buffer, and stained with 20 μg/ml anti-BAX, anti-metallothionein, or IgG1 isotype control mAb (BD Biosciences) for 30 min on ice. Cells were washed twice and incubated for 30 min on ice with PE-conjugated donkey Fab′ fragments specific for mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA). After two washes, samples were analyzed using a FACSCalibur flow cytometer (BD Biosciences), and a single gate was used to eliminate debris. In some experiments cells were stained with anti-myeloperoxidase (mAb 5B8, BD Biosciences) or IgG1 isotype mAb conjugated with FITC.

Statistics were determined using Student’s t test with SigmaStat version 2.0 (SPSS, Chicago, IL) or Microsoft Excel 2002 (Microsoft, Bellevue, WA) unless indicated otherwise.

To identify constitutive differences in gene expression that underlie chronic inflammation in XCGD patients, we compared transcript levels in unstimulated PMNs from XCGD patients and healthy control individuals (Table I, and supplemental Table II on the Journal of Immunology web site, which contains the complete set of microarray data for these experiments). Notably, we found that 19 genes encoding proteins that directly mediate host defense or facilitate the inflammatory response, including IL-8, CXCL1, CD14, Toll-like receptor 5, CD11c, CD54, CD64, CD89, CCR1, calgranulins A and B, and Nramp1, were significantly up-regulated in PMNs from XCGD patients (Table I). Further, genes encoding p47phox and p40phox were up-regulated in these patients (Table I). These findings suggest that proinflammatory capacity in XCGD patients is up-regulated due to the absence of gp91phox- or ROS-dependent regulatory mechanisms and/or the long term influence of recurrent infection. Increased proinflammatory capacity in XCGD patient cells was not due to prophylaxis with an immunomodulatory agent, as none of the patients participating in the microarray studies was receiving IFN-γ therapy (see Materials and Methods). Genes encoding proteins involved in TGF-β signaling, TGF-β receptor 2, SMA- and MAD-related protein 2, and SMA- and MAD-related protein 4, were down-regulated in unstimulated PMNs from XCGD patients (Tables I and II). Inasmuch as TGF-β plays a prominent role in suppressing inflammatory responses of phagocytic cells during ingestion of apoptotic cells (24, 25), down-regulation of genes involved in TGF-β signaling in XCGD patient PMNs may contribute to pathologic sequelae of chronic granulomatous disease.

Table I.

Increased expression of proinflammatory molecules in unstimulated PMNs from XCGD patientsa

UniGeneGeneBank no.Up-Regulated Genes (vs healthy control PMNs)Fold Change
Hs.51077 Y00093 CD11c (ITGAX2.4 
Hs.75627 X06882 CD14 (CD142.5 
Hs.168383 M24283 CD54 (ICAM13.5 
Hs.169610 L05424 CD44 (CD44)b 2.4 
Hs.77424 M63835 CD64, FcγR1 (FCGR1A2.5 
Hs.193122 U43774 CD89, FcαR (FCAR2.3 
Hs.2175 M59818 CD114, GCSF receptor (CSF3R)b 2.5 
Hs.256278 AI813532 CD120b, TNF receptor, p75 (TNFRSF1B2.3 
Hs.114408 AF051151 Toll-like receptor 5 (TLR53.5 
Hs.75545 X52425 IL-4R (IL4R2.0 
Hs.158315 AF077346 IL-18R accessory protein (IL18RAP6.0 
Hs.301921 D10925 Chemokine receptor 1 (CCR12.1 
Hs.1583 M55067 p47phox (NCF12.7 
Hs.196352 AL008637 p40phox (NCF42.7 
Hs.624 M28130 IL-8 (IL83.1 
Hs.789 X54489 Chemokine ligand 1 (CXCL12.2 
Hs.182611 AI679353 Nramp1 (SLC11A1)b 5.0 
Hs.416073 AI126134 Calgranulin A (S100A82.9 
Hs.112405 W72424 Calgranulin B (S100A92.5 
Hs.48516 S82297 β2 m (β2m)c 3.2 
Hs.181244 D32129 MHC class I, A (HLA-A3.0 
Hs.277477 X58536 MHC class I, C (HLA-C4.2 
Hs.381008 X56841 MHC class I, E (HLA-E2.2 
Hs.110309 AL022723 MHC class I, F (HLA-F2.5 
Hs.418650 J04755 Ferritin, heavy polypeptide 1 (FTHP1)b 2.9 
Hs.430150 AL031670 Ferritin, light chain (FTL2.9 
    
Down-Regulated Genes (vs healthy control PMNs)    
Hs.1244 M38690 CD9 −4.3 
Hs.40034 X16983 CD49D (ITGA4−2.8 
Hs.1722 M28983 IL-1α (IL1α) −4.7 
Hs.82028 D50683 TGF-β receptor 2 (TGFβR2−2.3 
Hs.158324 U28694 Chemokine (C-C motif) receptor 3 (CCR3−2.3 
Hs.285115 Y10659 IL-13Rα1 (IL13RαI−2.7 
Hs.68876 M75914 IL-5Rα (IL5Rα) −2.0 
Hs.166156 AF025529 Leukocyte Ig-like receptor (LILRA1−2.3 
Hs.23262 AI142565 Ribonuclease K6 (RNASE6−2.5 
Hs.88974 X04011 gp91phox (CYBB−3.1 
Hs.865 M22995 Rap1a (RAP1A−3.2 
Hs.75772 X03348 Glucocorticoid receptor β −2.1 
UniGeneGeneBank no.Up-Regulated Genes (vs healthy control PMNs)Fold Change
Hs.51077 Y00093 CD11c (ITGAX2.4 
Hs.75627 X06882 CD14 (CD142.5 
Hs.168383 M24283 CD54 (ICAM13.5 
Hs.169610 L05424 CD44 (CD44)b 2.4 
Hs.77424 M63835 CD64, FcγR1 (FCGR1A2.5 
Hs.193122 U43774 CD89, FcαR (FCAR2.3 
Hs.2175 M59818 CD114, GCSF receptor (CSF3R)b 2.5 
Hs.256278 AI813532 CD120b, TNF receptor, p75 (TNFRSF1B2.3 
Hs.114408 AF051151 Toll-like receptor 5 (TLR53.5 
Hs.75545 X52425 IL-4R (IL4R2.0 
Hs.158315 AF077346 IL-18R accessory protein (IL18RAP6.0 
Hs.301921 D10925 Chemokine receptor 1 (CCR12.1 
Hs.1583 M55067 p47phox (NCF12.7 
Hs.196352 AL008637 p40phox (NCF42.7 
Hs.624 M28130 IL-8 (IL83.1 
Hs.789 X54489 Chemokine ligand 1 (CXCL12.2 
Hs.182611 AI679353 Nramp1 (SLC11A1)b 5.0 
Hs.416073 AI126134 Calgranulin A (S100A82.9 
Hs.112405 W72424 Calgranulin B (S100A92.5 
Hs.48516 S82297 β2 m (β2m)c 3.2 
Hs.181244 D32129 MHC class I, A (HLA-A3.0 
Hs.277477 X58536 MHC class I, C (HLA-C4.2 
Hs.381008 X56841 MHC class I, E (HLA-E2.2 
Hs.110309 AL022723 MHC class I, F (HLA-F2.5 
Hs.418650 J04755 Ferritin, heavy polypeptide 1 (FTHP1)b 2.9 
Hs.430150 AL031670 Ferritin, light chain (FTL2.9 
    
Down-Regulated Genes (vs healthy control PMNs)    
Hs.1244 M38690 CD9 −4.3 
Hs.40034 X16983 CD49D (ITGA4−2.8 
Hs.1722 M28983 IL-1α (IL1α) −4.7 
Hs.82028 D50683 TGF-β receptor 2 (TGFβR2−2.3 
Hs.158324 U28694 Chemokine (C-C motif) receptor 3 (CCR3−2.3 
Hs.285115 Y10659 IL-13Rα1 (IL13RαI−2.7 
Hs.68876 M75914 IL-5Rα (IL5Rα) −2.0 
Hs.166156 AF025529 Leukocyte Ig-like receptor (LILRA1−2.3 
Hs.23262 AI142565 Ribonuclease K6 (RNASE6−2.5 
Hs.88974 X04011 gp91phox (CYBB−3.1 
Hs.865 M22995 Rap1a (RAP1A−3.2 
Hs.75772 X03348 Glucocorticoid receptor β −2.1 
a

Gene expression in human PMNs from five XCGD patients was compared with expression in cells from four healthy control individuals. For all genes except CYBB and RAP1A, p ≤ 0.04 for changes in PMN gene expression from XCGD patients vs healthy control cells (by Student’s t test). The common name of the protein encoded by the gene precedes the gene abbreviation (in parentheses).

b

Data represented by two separate probe sets on the array.

c

Data represented by three separate probe sets on the array.

To test the hypothesis that ROS influence (directly or indirectly) patterns of gene transcription in human neutrophils, we next compared changes in gene expression in activated PMNs from healthy control individuals and XCGD patients (Figs. 1 and 2, and supplemental Table III on the Journal of Immunology web site). We chose receptor-mediated phagocytosis as a model system for PMN activation based on the relevance of this process in patients with recurrent infection, i.e., in XCGD. Moreover, previous studies indicate that phagocytosis induces global changes in PMN gene expression (13, 14, 15). Fifty-two genes that were differentially transcribed in activated PMNs from XCGD patients remained unchanged in neutrophils from control individuals (Figs. 1 and 2, and Table III). By comparison, 154 genes that were up- or down-regulated in activated PMNs from healthy control individuals were not changed significantly in cells from XCGD patients or were changed to a lesser extent (Figs. 1 and 2, and Table III). Inasmuch as the signature biochemical defect in cells from patients with XCGD is inability to produce superoxide, changes in neutrophil gene expression between healthy control individuals and XCGD patients after phagocytosis probably reflected the presence and the absence of ROS, respectively. Of note, genes encoding proteins that moderate responses to oxidants were up-regulated only in activated PMNs from healthy control individuals (Fig. 1). Consistent with the data from unstimulated XCGD patient cells (Tables I and II), genes encoding several molecules that mediate host defense or the inflammatory response, such as CCL3, CCL4, and IL-1R, were up-regulated significantly in activated neutrophils from XCGD patients (Fig. 1). Previous studies have demonstrated that the expression of chemokines and receptors for inflammatory molecules are regulated by cell redox status (26, 27). Importantly, up-regulation of these genes is consistent with the idea that they potentiate chronic inflammation in individuals with XCGD.

Recent studies by Cramer et al. (28) underscore the importance of transcription factors in phagocyte-mediated inflammation. However, relatively little is known about how transcription factors contribute to chronic inflammation. After phagocytosis, 32 genes encoding transcription factors or proteins that bind RNA or DNA were differentially regulated in normal PMNs and not in the cells from XCGD patients (Figs. 1 and 2, and Table III). For example, genes encoding SFRS protein kinase 1 (SRPK1), lipoma HMGIC fusion partner-like 2 (LHFPL2), CBF1-interacting corepressor (CIR), oligodendrocyte lineage transcription factor 2 (RACK17), zinc finger protein 147, and zinc finger protein 254 were up-regulated only in normal cells (Fig. 1). Although specific functions have not been ascribed to some of these genes (e.g., LHFPL2 and ZNF254), it is known that SRPK1 is associated with U1-snRNP in apoptotic cells and is activated early during apoptosis to phosphorylate SR proteins (29), which regulate the splicing of select mRNAs. CSDA (cold shock domain protein 2), BCAS2 (breast carcinoma amplified sequence 2), CFDP1, WTAP, and EGFL5 were differentially expressed in cells from patients with XCGD, but not in normal neutrophils (Figs. 1 and 2, and Table III). There is no reported function for these genes in human PMNs, although cold shock domain protein 2 may repress the GM-CSF promoter (30), and breast carcinoma amplified sequence 2 is probably associated with the spliceosome complex (31). Further studies are necessary to determine whether these genes contribute, directly or indirectly, to the pathophysiology of XCGD or chronic inflammation in the patients.

There were significant changes in the expression of 11 genes involved in apoptosis in normal PMNs not observed in cells from XCGD patients (Figs. 1 and 2, and Table III). Notably, genes encoding modulator of apoptosis-1 (MOAP1) and BAX were significantly up-regulated in normal PMNs, but not in cells from patients with XCGD (Fig. 1, green arrow). The finding that BAX was up-regulated only in normal human PMNs after phagocytosis suggests that the protein is regulated at least indirectly by ROS. As it is possible that the induction of BAX contributes to a difference in PMN apoptosis between healthy control individuals and XCGD patients, we tested the hypothesis that up-regulation of BAX accompanies PMN apoptosis (below).

We identified 13 apoptosis and cell fate-related genes that were differentially expressed in PMNs from XCGD patients, but were not significantly changed in normal cells (Figs. 1 and 2). Genes encoding four distinct isoforms of metallothionein (MT1X, MT1G, MT1F, and MT3) were significantly up-regulated in PMNs from patients with XCGD and not in cells from healthy control individuals (Fig. 1). We observed that MT1X, MT1G, and MT1F were not expressed in resting human PMNs (data not shown). However, metallothioneins are induced by many factors, including hypoxia, in other cell types (32, 33), and notably, metallothioneins protect against apoptosis (34, 35). Although we obtained insufficient numbers of cells to test whether metallothionein protein was up-regulated in neutrophils from patients with XCGD, we found that up-regulation of metallothionein correlated closely with induction of MT1G in normal human PMNs (Fig. 3).

FIGURE 3.

Increased expression of metallothionein 1G (MT1G) correlates with up-regulation of metallothionein protein. A, Induction of MT1G in human PMNs. Expression of the metallothionein 1G (MT1G) gene was determined by TaqMan real-time PCR at 0, 4.5, 9, and 20 h after treatment with 100 μM zinc chloride. Results shown are representative of data from three separate individuals, and each TaqMan experiment was assayed in triplicate. B, Up-regulation of metallothionein protein in human PMNs. Intracellular expression of metallothionein was determined by flow cytometry after treatment with 100 μM zinc chloride. Metallothionein was detected with an mAb that recognizes metallothionein isoforms I and II (αMT). Isotype, IgG1 isotype control Ab. Results are representative of two or three experiments.

FIGURE 3.

Increased expression of metallothionein 1G (MT1G) correlates with up-regulation of metallothionein protein. A, Induction of MT1G in human PMNs. Expression of the metallothionein 1G (MT1G) gene was determined by TaqMan real-time PCR at 0, 4.5, 9, and 20 h after treatment with 100 μM zinc chloride. Results shown are representative of data from three separate individuals, and each TaqMan experiment was assayed in triplicate. B, Up-regulation of metallothionein protein in human PMNs. Intracellular expression of metallothionein was determined by flow cytometry after treatment with 100 μM zinc chloride. Metallothionein was detected with an mAb that recognizes metallothionein isoforms I and II (αMT). Isotype, IgG1 isotype control Ab. Results are representative of two or three experiments.

Close modal

We next used TaqMan real-time RT-PCR to verify changes in expression detected by microarray analysis (Fig. 4). We selected 12 genes representative of the entire microarray dataset for confirmation (Fig. 4). These genes were assigned to five different functional categories and included several key genes encoding proteins that participate in apoptosis and/or survival. There was strong correlation (87.5%) between the TaqMan and microarray gene expression data (Fig. 4), consistent with previous studies (13, 14, 15).

FIGURE 4.

Confirmation of microarray data by TaqMan real-time RT-PCR. TaqMan confirmation of microarray results. Genes (n = 12) identified as differentially expressed by oligonucleotide microarrays were selected for confirmation by TaqMan real-time PCR after phagocytosis. TaqMan data represent the mean fold change in gene expression from three healthy control donors (red) and two XCGD patients (green). Microarray data are the mean fold change in four healthy control donors (black) and six XCGD patients (blue). TaqMan data for GPNMB was scaled down by a factor of 3 to fit on the graph.

FIGURE 4.

Confirmation of microarray data by TaqMan real-time RT-PCR. TaqMan confirmation of microarray results. Genes (n = 12) identified as differentially expressed by oligonucleotide microarrays were selected for confirmation by TaqMan real-time PCR after phagocytosis. TaqMan data represent the mean fold change in gene expression from three healthy control donors (red) and two XCGD patients (green). Microarray data are the mean fold change in four healthy control donors (black) and six XCGD patients (blue). TaqMan data for GPNMB was scaled down by a factor of 3 to fit on the graph.

Close modal

BAX plays a key role in the execution of apoptosis in eukaryotic cells (36) and probably contributes to neutrophil-mediated inflammation (16). We measured BAX in human PMNs to determine whether changes in the gene encoding BAX were reflected directly by increased protein (Fig. 5). PMNs constitutively expressed BAX (Fig. 5,A, arrowheads). However, there was an increase in BAX (inducible BAX (iBAX)) protein over time that was accelerated significantly after phagocytosis (iBAX was 20.1 ± 7.3% in unstimulated PMNs vs 43.3 ± 5.0% in activated cells at 9 h, p < 0.0002; Fig. 5,B). Importantly, cells expressing iBAX had a concomitant increase in surface expression of Annexin-V (Fig. 5,C, arrows), a marker for programmed cell death (Fig. 5,C). The observation that there was heterogeneous induction of BAX after phagocytosis may be a reflection of neutrophil populations that differ slightly in age, as emigration of neutrophils from the bone marrow is a continuous and thus asynchronous process. Differences in BAX expression and subsequent apoptosis due to cell maturity could play an important role in modulating host defense in conditions such as sepsis, where there is an efflux of immature neutrophils from bone marrow (37, 38, 39). Although PMNs from patients with XCGD constitutively expressed BAX, induction of BAX in these patients was not accelerated significantly by phagocytosis (6.8 ± 6.4% induced BAX in CGD PMNs (p = 0.3 vs unstimulated cells) compared with 21.7 ± 8.0% induced BAX in PMNs from normal individuals (p = 0.006 vs unstimulated cells); Fig. 5 D). Taken together, these findings suggest that ROS play a role in regulating BAX synthesis in human PMNs and are consistent with studies demonstrating that BAX expression is increased during neutrophil apoptosis induced by phagocytosis of Mycobacterium tuberculosis (40).

FIGURE 5.

BAX is induced during PMN apoptosis. BAX expression is induced by ROS and accompanies apoptosis in activated PMNs. A, Intracellular BAX was measured in unstimulated and stimulated (Phagocytosis) PMNs from healthy individuals as indicated. Green histograms represent cells stained with Ab specific for myeloperoxidase in the same experiment. B, Quantitation of intracellular BAX in unstimulated and stimulated (Phagocytosis) PMNs. Results are from five or six separate experiments. ∗, p < 0.001 vs unstimulated PMNs (by Student’s t test). C, Cells stained with Annexin VFITC and anti-BAX Ab. Note the strong positive correlation between percentages of cells staining with annexin V and BAX. D, BAX is not induced by PMN phagocytosis in patients with CGD. The percentage of BAX induced by phagocytosis was quantitated in PMNs from five healthy individuals or in two XCGD patients (red circles) and one patient with autosomal deficiency of p22phox (blue circle). ∗, p = 0.35 vs CGD patients. cBAX, constitutively expressed BAX.

FIGURE 5.

BAX is induced during PMN apoptosis. BAX expression is induced by ROS and accompanies apoptosis in activated PMNs. A, Intracellular BAX was measured in unstimulated and stimulated (Phagocytosis) PMNs from healthy individuals as indicated. Green histograms represent cells stained with Ab specific for myeloperoxidase in the same experiment. B, Quantitation of intracellular BAX in unstimulated and stimulated (Phagocytosis) PMNs. Results are from five or six separate experiments. ∗, p < 0.001 vs unstimulated PMNs (by Student’s t test). C, Cells stained with Annexin VFITC and anti-BAX Ab. Note the strong positive correlation between percentages of cells staining with annexin V and BAX. D, BAX is not induced by PMN phagocytosis in patients with CGD. The percentage of BAX induced by phagocytosis was quantitated in PMNs from five healthy individuals or in two XCGD patients (red circles) and one patient with autosomal deficiency of p22phox (blue circle). ∗, p = 0.35 vs CGD patients. cBAX, constitutively expressed BAX.

Close modal

We next investigated apoptosis in PMNs from the six XCGD patients used in the microarray studies to determine whether global changes in patient gene expression, including diminished transcription of BAX, reflected altered cell fate. Compared with cells from healthy control donors, apoptosis in PMNs from XCGD patients was reduced significantly after phagocytosis (e.g., 36.5 ± 9.4 vs 12.2 ± 5.2% at 6 h for neutrophils from control and XCGD donors, respectively; Fig. 6). Furthermore, there was no significant difference in apoptosis between unstimulated and stimulated PMNs from XCGD patients (e.g., 11.4 ± 3.1 vs 12.2 ± 5.2% at 6 h for unstimulated and stimulated cells, respectively; Fig. 6). These findings suggest that ROS produced during phagocytosis by normal PMNs accelerate apoptosis and are consistent with studies by Coxon et al. (41). Taken together, the data provide strong evidence that altered gene expression in PMNs from XCGD patients underlies defective apoptosis.

FIGURE 6.

PMN apoptosis is defective in XCGD patients. Apoptosis was measured in unstimulated human PMNs (lower panels) or after phagocytosis (upper panels) in cells from healthy control donors (CTL) or XCGD patients as indicated. Results are from ≥10 healthy control donors (n = 10 at 3 and 9 h, n = 14 at 6 h) and five or six XCGD patients (n = 6 at 3 and 6 h, n = 5 at 9 h) in stimulated cells (Phagocytosis). ∗, p < 0.001 vs unstimulated cells; ∗∗, p = 0.04 vs CTL at 3 h or p < 0.001 vs CTL at 6 and 9 h (by Student’s t test).

FIGURE 6.

PMN apoptosis is defective in XCGD patients. Apoptosis was measured in unstimulated human PMNs (lower panels) or after phagocytosis (upper panels) in cells from healthy control donors (CTL) or XCGD patients as indicated. Results are from ≥10 healthy control donors (n = 10 at 3 and 9 h, n = 14 at 6 h) and five or six XCGD patients (n = 6 at 3 and 6 h, n = 5 at 9 h) in stimulated cells (Phagocytosis). ∗, p < 0.001 vs unstimulated cells; ∗∗, p = 0.04 vs CTL at 3 h or p < 0.001 vs CTL at 6 and 9 h (by Student’s t test).

Close modal

Several lines of evidence suggest that increased proinflammatory capacity and decreased PMN turnover in patients with XCGD contribute to the pathophysiology of XCGD, including granuloma formation. First, PMNs from XCGD patients have increased expression of proinflammatory molecules and decreased expression of genes encoding anti-inflammation mediators compared with PMNs from healthy controls (Tables I and II, and Fig. 1). Increased proinflammatory capacity could contribute to the recruitment and activation of immune cells and thereby prolong inflammation in CGD patients. Second, a comparison of activated PMNs from XCGD patients and normal individuals identified genes encoding key regulators of apoptosis, transcription, inflammation, host defense, and other processes accompanying apoptosis that were differentially expressed in the absence or the presence of ROS, respectively. Among these regulators was the proapoptotic protein BAX, whose expression was not induced significantly in activated neutrophils from XCGD patients. Third, XCGD PMNs have defective apoptosis after phagocytosis that could delay resolution of inflammation during infection. Therefore, the lack of induction or repression of key cell fate-related genes in XCGD patients (e.g., BAX) and the associated defective apoptosis of XCGD PMNs has important implications for our understanding of the pathogenesis of granulomas in CGD. Notably, these granulomas form at sites of infection to which neutrophils have been recruited, and these cells predominate in granulomatous tissue and abscesses resected from patients with CGD (not shown).

We used gene expression profiling to generate a global view of how PMN-derived oxidants probably contribute to the resolution of inflammation in humans. Taken together, our findings suggest that altered PMN apoptosis after phagocytosis results in delayed resolution of inflammation, which facilitates the formation of granulomas in XCGD patients. Insight into defects in resolution of neutrophil-mediated inflammation revealed by our genomic studies of XCGD PMNs is probably applicable to other types of chronic inflammatory diseases. Importantly, these studies identified potential gene targets for prophylaxis and treatment of chronic inflammatory processes facilitated by PMNs.

We thank P. Woltz and S. Foster for assistance with collecting venous blood from XCGD patients, K. Leidel, and M. Goedken for technical assistance, and J. M. Musser for critical review of the manuscript.

3

Abbreviations used in this paper: PMN, polymorphonuclear leukocyte; BAX, Bcl2-associated X protein; CGD, chronic granulomatous disease; iBAX, inducible BAX. ROS, reactive oxygen species; XCGD, X-linked chronic granulomatous disease.

1
Jones, S. L., F. P. Lindberg, E. J. Brown.
1999
. Phagocytosis. W. F. Paul, ed.
Fundamental Immunology
4th Ed.
997
. Lippincott-Raven, New York.
2
Nauseef, W. M., R. A. Clark.
2000
. Granulocytic Phagocytes. G. L. Mandel, and J. E. Bennett, and R. Dolin, eds. In
Basic Principles in the Diagnosis and Management of Infectious Diseases
Vol. 1
:
89
. Churchill Livingstone, New York.
3
Berendes, H., R. A. Bridges, R. A. Good.
1957
. A fatal granulomatosus of childhood.
Minn. Med.
40
:
309
.
4
Segal, A. W., O. T. Jones.
1978
. Novel cytochrome b system in phagocytic vacuoles of human granulocytes.
Nature
276
:
515
.
5
Royer-Pokora, B., L. M. Kunkel, A. P. Monaco, S. C. Goff, P. E. Newburger, R. L. Baehner, F. S. Cole, J. T. Curnutte, S. H. Orkin.
1986
. Cloning the gene for an inherited human disorder–chronic granulomatous disease: on the basis of its chromosomal location.
Nature
322
:
32
.
6
Parkos, C. A., M. C. Dinauer, A. J. Jesaitis, S. H. Orkin, J. T. Curnutte.
1989
. Absence of both the 91kD and 22kD subunits of human neutrophil cytochrome b in two genetic forms of chronic granulomatous disease.
Blood
73
:
1416
.
7
Malech, H. L., W. M. Nauseef.
1997
. Primary inherited defects in neutrophil function: etiology and treatment.
Semin. Hematol.
34
:
279
.
8
Lekstrom-Himes, J. A., J. I. Gallin.
2000
. Immunodeficiency diseases caused by defects in phagocytes.
N. Engl. J. Med.
343
:
1703
.
9
Halamish, A., A. Klar, D. Shoseyov, G. Blinder, H. Hurvitz.
2001
. Corticosteroid therapy reversed progressive chronic granulomatous lung disease following deterioration on interferon-γ treatment.
Pediatr. Pulmonol.
32
:
257
.
10
Chin, T. W., E. R. Stiehm, J. Falloon, J. I. Gallin.
1987
. Corticosteroids in treatment of obstructive lesions of chronic granulomatous disease.
J. Pediatr.
111
:
349
.
11
Savill, J..
1997
. Apoptosis in resolution of inflammation.
J. Leukocyte Biol.
61
:
380
.
12
Savill, J. S., A. H. Wyllie, J. E. Henson, M. J. Walport, P. M. Henson, C. Haslett.
1989
. Macrophage phagocytosis of aging neutrophils in inflammation: programmed cell death in the neutrophil leads to its recognition by macrophages.
J. Clin. Invest.
83
:
865
.
13
Kobayashi, S. D., J. M. Voyich, C. L. Buhl, R. M. Stahl, F. R. DeLeo.
2002
. Global changes in gene expression by human polymorphonuclear leukocytes during receptor-mediated phagocytosis: cell fate is regulated at the level of gene expression.
Proc. Natl. Acad. Sci. USA
99
:
6901
.
14
Kobayashi, S. D., J. M. Voyich, K. Braughton, F. R. DeLeo.
2003
. Down-regulation of proinflammatory capacity during apoptosis in human polymorphonuclear leukocytes.
J. Immunol.
170
:
3357
.
15
Kobayashi, S. D., K. R. Braughton, A. R. Whitney, J. M. Voyich, T. G. Schwan, J. M. Musser, F. R. DeLeo.
2003
. Bacterial pathogens modulate an apoptosis differentiation program in human neutrophils.
Proc. Natl. Acad. Sci. USA
100
:
10948
.
16
Dibbert, B., M. Weber, W. H. Nikolaizik, P. Vogt, M. H. Schoni, K. Blaser, H. U. Simon.
1999
. Cytokine-mediated Bax deficiency and consequent delayed neutrophil apoptosis: a general mechanism to accumulate effector cells in inflammation.
Proc. Natl. Acad. Sci. USA
96
:
13330
.
17
Fadeel, B., A. Ahlin, J. I. Henter, S. Orrenius, M. B. Hampton.
1998
. Involvement of caspases in neutrophil apoptosis: regulation by reactive oxygen species.
Blood
92
:
4808
.
18
Gardai, S., B. B. Whitlock, C. Helgason, D. Ambruso, V. Fadok, D. Bratton, P. M. Henson.
2002
. Activation of SHIP by NADPH oxidase-stimulated Lyn leads to enhanced apoptosis in neutrophils.
J. Biol. Chem.
277
:
5236
.
19
Friccius, H., H. Pohla, M. Adibzadeh, P. Siegels-Hubenthal, A. Schenk, G. Pawelec.
1992
. The effects of the antifungal azoles itraconazole, fluconazole, ketoconazole and miconazole on cytokine gene expression in human lymphoid cells.
Int. J. Immunopharmacol.
14
:
791
.
20
Tian, B., Y. Zhang, B. A. Luxon, R. P. Garofalo, A. Casola, M. Sinha, A. R. Brasier.
2002
. Identification of NFκB-dependent gene networks in respiratory syncytial virus-infected cells.
J. Virol.
76
:
6800
.
21
Rae, J. M., M. D. Johnson, M. E. Lippman, D. A. Flockhart.
2001
. Rifampin is a selective, pleiotropic inducer of drug metabolism genes in human hepatocytes: studies with cDNA and oligonucleotide expression arrays.
J. Pharmacol. Exp. Ther.
299
:
849
.
22
Boyum, A..
1968
. Isolation of mononuclear cells and granulocytes from human blood. Isolation of mononuclear cells by one centrifugation, and of granulocytes by combining centrifugation and sedimentation at 1 g.
Scand. J. Clin. Lab. Invest.
97
:(Suppl.):
77
.
23
Butcher, H., W. Kennette, O. Collins, J. Demoor, J. Koropatnick.
2003
. A sensitive time-resolved fluorescent immunoassay for metallothionein protein.
J. Immunol. Methods
272
:
247
.
24
Fadok, V. A., D. L. Bratton, A. Konowal, P. W. Freed, J. Y. Westcott, P. M. Henson.
1998
. Macrophages that have ingested apoptotic cells in vitro inhibit proinflammatory cytokine production though autocrine/paracrine mechanisms involving TGF-β, PGE2, and PAF.
J. Clin. Invest.
101
:
890
.
25
McDonald, P. P., V. A. Fadok, D. Bratton, P. M. Henson.
1999
. Transcriptional and translational regulation of inflammatory mediator production by endogenous TGF-β in macrophages that have ingested apoptotic cells.
J. Immunol.
163
:
6164
.
26
Saccani, A., S. Saccani, S. Orlando, M. Sironi, S. Bernasconi, P. Ghezzi, A. Mantovani, A. Sica.
2000
. Redox regulation of chemokine receptor expression.
Proc. Natl. Acad. Sci. USA
97
:
2761
.
27
Michalec, L., B. K. Choudhury, E. Postlethwait, J. S. Wild, R. Alam, M. Lett-Brown, S. Sur.
2002
. CCL7 and CXCL10 orchestrate oxidative stress-induced neutrophilic lung inflammation.
J. Immunol.
168
:
846
.
28
Cramer, T., Y. Yamanishi, B. E. Clausen, I. Forster, R. Pawlinski, N. Mackman, V. H. Haase, R. Jaenisch, M. Corr, V. Nizet, et al
2003
. HIF-1α is essential for myeloid cell-mediated inflammation.
Cell
112
:
645
.
29
Kamachi, M., T. M. Le, S. J. Kim, M. E. Geiger, P. Anderson, P. J. Utz.
2002
. Human autoimmune sera as molecular probes for the identification of an autoantigen kinase signaling pathway.
J. Exp. Med.
196
:
1213
.
30
Coles, L. S., P. Diamond, F. Occhiodoro, M. A. Vadas, M. F. Shannon.
1996
. Cold shock domain proteins repress transcription from the GM-CSF promoter.
Nucleic Acids Res.
24
:
2311
.
31
Neubauer, G., A. King, J. Rappsilber, C. Calvio, M. Watson, P. Ajuh, J. Sleeman, A. Lamond, M. Mann.
1998
. Mass spectrometry and EST-database searching allows characterization of the multi-protein spliceosome complex.
Nat. Genet.
20
:
46
.
32
Murphy, B. J., K. R. Laderoute, R. J. Chin, R. M. Sutherland.
1994
. Metallothionein IIA is up-regulated by hypoxia in human A431 squamous carcinoma cells.
Cancer Res.
54
:
5808
.
33
Murphy, B. J., G. K. Andrews, D. Bittel, D. J. Discher, J. McCue, C. J. Green, M. Yanovsky, A. Giaccia, R. M. Sutherland, K. R. Laderoute, et al
1999
. Activation of metallothionein gene expression by hypoxia involves metal response elements and metal transcription factor-1.
Cancer Res.
59
:
1315
.
34
Kondo, Y., J. M. Rusnak, D. G. Hoyt, C. E. Settineri, B. R. Pitt, J. S. Lazo.
1997
. Enhanced apoptosis in metallothionein null cells.
Mol. Pharmacol.
52
:
195
.
35
Kang, Y. J., Z. X. Zhou, G. W. Wang, A. Buridi, J. B. Klein.
2000
. Suppression by metallothionein of doxorubicin-induced cardiomyocyte apoptosis though inhibition of p38 mitogen-activated protein kinases.
J. Biol. Chem.
275
:
13690
.
36
Oltvai, Z. N., C. L. Milliman, S. J. Korsmeyer.
1993
. Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death.
Cell
74
:
609
.
37
Ardron, M. J., J. C. Westengard, T. F. Dutcher.
1994
. Band neutrophil counts are unnecessary for the diagnosis of infection in patients with normal total leukocyte counts.
Am. J. Clin. Pathol.
102
:
646
.
38
Seebach, J. D., R. Morant, R. Ruegg, B. Seifert, J. Fehr.
1997
. The diagnostic value of the neutrophil left shift in predicting inflammatory and infectious disease.
Am. J. Clin. Pathol.
107
:
582
.
39
van Eeden, S. F., Y. Kitagawa, M. E. Klut, E. Lawrence, J. C. Hogg.
1997
. Polymorphonuclear leukocytes released from the bone marrow preferentially sequester in lung microvessels.
Microcirculation
4
:
369
.
40
Perskvist, N., M. Long, O. Stendahl, L. Zheng.
2002
. Mycobacterium tuberculosis promotes apoptosis in human neutrophils by activating caspase-3 and altering expression of Bax/Bcl-xL via an oxygen-dependent pathway.
J. Immunol.
168
:
6358
.
41
Coxon, A., P. Rieu, F. J. Barkalow, S. Askari, A. H. Sharpe, U. H. von Andrian, M. A. Arnaout, T. N. Mayadas.
1996
. A novel role for the β2 integrin CD11b/CD18 in neutrophil apoptosis: a homeostatic mechanism in inflammation.
Immunity
5
:
653
.