INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Chronic granulomatous disease (CGD) is a rare genetic disorder of NADPH oxidase in which phagocytes are unable to generate reactive oxidants (12). As a result, CGD patients suffer from recurrent life-threatening bacterial and fungal infections. CGD is also characterized by abnormally exuberant tissue granuloma formation, which may manifest as inflammatory bowel disease, urinary tract obstruction, and poor wound healing or dehiscence.

The mechanisms underlying abnormal inflammatory responses in CGD are undefined. One hypothesis is that persistent granuloma formation occurs because CGD phagocytes lack the ability to effectively destroy and clear pathogens, and thus, persistent antigenic stimulation results (28). However, biopsy specimens from the exuberant granulomatous inflammatory lesions in CGD patients are usually sterile. In addition, granulomatous disease often responds to steroid therapy without the addition of antibiotics (6), suggesting that persistent infection is not the major cause of these inflammatory lesions.

In both patients and mouse models of CGD, abnormal neutrophil inflammatory responses are seen. Increased levels of neutrophil exudate were observed in X-linked CGD patients compared with healthy volunteers in a skin window model (11). Consistent with these findings, in both the p47^phox/ and X-linked mouse models of CGD, increased neutrophil peritoneal leukocytosis occurred in CGD mice versus wild-type littermates in response to intraperitoneal (i.p.) administration of the sterile irritant thioglycolate (19, 23). In vitro, reactive oxidants inactivate various proinflammatory chemotactic factors, such as leukotrienes, C5a, and N-formyl peptide (7, 15, 17). Therefore, the lack of generation of reactive oxidants by CGD phagocytes may allow for increased accumulation of chemotactic factors in tissue, leading to enhanced acute inflammatory responses.

Recently, Morgenstern et al. (22) have shown that in the X-linked mouse model of CGD, enhanced neutrophil exudation and production of proinflammatory cytokines occurred in the lungs following intratracheal inoculation with heat-killed Aspergillus fumigatus hyphae. One week postchallenge, inflammation was reduced in wild-type mice, whereas CGD mice developed pyogranulomatous lesions that persisted for at least 6 weeks. These data indicate that mechanisms other than infection are critical in driving persistent granuloma formation in CGD.

Using the p47^phox/ mouse model of CGD, we tried to address two questions: (i) is there a skewing of cytokine responses toward either the Th-1 or Th-2 pole, leading to enhanced granuloma formation in CGD mice? and (ii) are CGD mice capable of appropriately resolving tissue granulomata once they have been generated?

Mice with a targeted disruption of the p47^phox gene and wild-type littermates were challenged with either an avirulent (SmD) or a virulent (SmT) variant of Mycobacterium avium 2-151, a potent stimulus of Th-1-type responses (1, 9), or Schistosoma mansoni eggs (SME), which elicit Th-2 cytokine-mediated granulomata (5, 32-37). Tissue granuloma sizes and cytokine responses were compared.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mice. p47^phox/ mice were generated as described elsewhere (19). p47^phox/ and wild-type littermates used in the SmT and SME experiments were derived from heterozygous parents (C57/BL6 × 129 backcrossed to F₅ in a C57/BL6 lineage) at the facility of Taconic (Germantown, N.Y.). p47^phox/ and wild-type mice used in experiments with SmD were derived from homozygous parents (C57/BL6 × 129 lineage) at the National Institute of Allergy and Infectious Diseases (NIAID) animal facility (Bethesda, Md.). Mice were maintained in autoclaved cages in a specific-pathogen-free environment at the NIAID animal facility for all experiments. p47^phox/ and wild-type mice (8 to 12 weeks old) were age and sex matched for each set of experiments. All experiments were approved by the NIAID Animal Care and Use Committee.

M. avium challenge. M. avium 2-151 SmD (avirulent) and 2-151 SmT (virulent) were generously provided by Andrea Cooper, Colorado State University (Fort Collins). p47^phox/ (n = 8) and wild-type (n = 4) mice were sensitized with i.p. SmD (10⁷ CFU/mouse). Fourteen days later, mice received the same inoculum intravenously (i.v.). Mice were sacrificed by CO₂ inhalation 7 days after i.v. challenge. In experiments using the SmT strain, p47^phox/ and wild-type mice were administered a single i.v. inoculum (10⁷ CFU/mouse). Mice (n = 4 mice/group/time period) were sacrificed at 2, 5, and 8 weeks after inoculation. Mycobacterial burdens at different times after challenge were assessed by homogenizing spleens and lungs, plating serial 10-fold dilutions of homogenates on Middlebrook 7H11 agar plates, and counting bacterial colonies as previously described (9).

SME challenge. p47^phox/ and wild-type mice were sensitized by i.p. administration of 5,000 SME followed by i.v. challenge with 5,000 eggs 10 days later as previously described (33). In this model, SME are trapped within the lung vasculature and generate perivascular pulmonary granulomata. Generation of granulomata is dependent on Th-2-type cytokine production, and granuloma size is markedly reduced by pretreatment with interleukin-12 (IL-12), a potent inducer of Th-1-type cytokine responses (33-35, 37). On days 8, 20, and 32 following i.v. challenge, mice (n = 4 to 5 mice/group/time period) were sacrificed by CO₂ inhalation.

Histopathology. In M. avium experiments, spleens and lungs were harvested and weighed, and a segment of tissue was fixed in 10% buffered formalin solution for histopathology. Paraffin-embedded sections were Fite stained to detect acid-fast bacteria and visualized with hematoxylin and eosin (H&E) staining. In SME experiments, lungs were removed and fixed in Bouin's solution. Tissues were embedded in paraffin, and sections were stained with H&E. Mean granuloma volumes were evaluated as previously described (36). All histopathology slides were read in a blinded fashion.

Cytokine studies. In M. avium SmD experiments, spleens were collected in phosphate-buffered saline (PBS; Biofluids, Rockville, Md.) and pooled by genotype, and single-cell suspensions were prepared. Following hypotonic lysis of contaminating erythrocytes in ACK lysing buffer (BioWhittaker, Walkersville, Md.), cells were washed twice with PBS and resuspended at 10⁶ cells/ml in RPMI-1640 supplemented with 10% fetal bovine serum (HyClone, Logan, Utah), 20 mM HEPES (BioWhittaker), 2 mM sodium pyruvate (Biofluids), nonessential amino acids (1% by volume) (Biofluids), 2 mM glutamine (Biofluids), and penicillin (100 U/ml)-streptomycin (100 µg/ml) solution (Sigma, St. Louis, Mo.). One-milliliter aliquots were transferred to 24-well plates (Costar, Cambridge, Mass.) and incubated with medium alone, freeze-thawed M. avium antigen (Mag; 0.3 µg/ml) prepared as described (9), or concanavalin A (ConA) (5 µg/ml) (Sigma) for 3 days at 37°C under a 5% CO₂ atmosphere, after which supernatants were harvested and frozen at 20°C. Concentrations of IL-5, IL-10, and gamma interferon were determined by enzyme-linked immunosorbent assay (ELISA) as previously described (36).

phox/

phox/

Statistics. Student's t test (StatWorks, Malvern, Pa.) was used to compare continuous variables. A P value of <0.05 was considered significant.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

M. avium challenge. Spleens from p47^phox/ and wild-type mice challenged with the avirulent M. avium SmD were moderately enlarged (176 ± 81 versus 217 ± 97 mg, respectively; the spleen weight in unstimulated mice was approximately 100 mg). Histology showed normal spleen architecture without bacilli on Fite staining (data not shown).

phox/

View larger version (14K): [in this window] [in a new window]   FIG. 1.   Cytokine responses in mice challenged with the avirulent SmD strain of M. avium. Splenocytes from wild-type (n = 4) and p47phox/ (n = 8) mice challenged with M. avium were restimulated in vitro with medium (open bar), ConA (solid bar), or Mag (hatched bar). p47phox/ and wild-type mice generated similar polar Th-1-type cytokine responses (high levels of gamma interferon (IFN-) and basal levels of IL-5 and IL-10). Data are means ± standard deviations (SD).

phox/

phox/

phox/

View larger version (10K): [in this window] [in a new window]   FIG. 2.   Spleen and lung weights in p47phox/ and wild-type mice challenged with the virulent M. avium SmT. Marked splenomegaly developed in p47phox/ (open triangles) and wild-type (open circles) mice at 5 and 8 weeks after challenge. Lung weight was significantly greater in p47phox/ mice (solid triangles) than in wild-type mice (solid circles) only at 8 weeks (P < 0.05). Four mice per group per time period were used. Data are means ± SD.

phox/

phox/

phox/

phox/

View larger version (104K): [in this window] [in a new window]   FIG. 3.   Spleen and lung histopathology in p47phox/ mice at 8 weeks after challenge with the virulent SmT strain of M. avium. (A and B) Representative spleen section stained with H&E at low-power (×50) (A) and high-power (×200) (B) magnifications, showing disruption of the normal spleen architecture caused by massive histiocytic infiltration. (C) Lung section showing well-demarcated granulomata composed of macrophages, lymphocytes, and occasional giant cells (H&E stained; magnification, ×200). Histopathology was similar for p47phox/ and wild-type mice.

View this table: [in this window] [in a new window]   TABLE 1.   Mycobacterial burdens in spleens and lungs of wild-type and p47phox/ mice following challenge with M. avium SmT

phox/

SME challenge. Following in vitro stimulation with ConA or SEA, draining pulmonary lymph node cells from SME-challenged p47^phox/ and wild-type mice generated mixed Th-1- and Th-2-type cytokine responses (Fig. 4). By day 32, a reduction in Th-2-type cytokine responses occurred in p47^phox/ and wild-type lymph node cells following in vitro stimulation with SEA but not with ConA; this decay in Th-2-type cytokine generation paralleled the decay in granuloma size (Fig. 5). Mean granuloma volumes were similar in p47^phox/ and wild-type mice at each of three different times following SME challenge (Fig. 5). In both groups, granulomata receded as a function of the time from SME challenge. By day 32 postchallenge, the sizes of granulomata in both p47^phox/ and wild-type mice were markedly reduced compared with sizes at days 8 and 20 (P < 0.05), and granulomata demonstrated reduced cellularity and increased fibrosis compared with those at earlier times (Fig. 6).

View larger version (22K): [in this window] [in a new window]   FIG. 4.   Cytokine production in mice challenged with SME. p47phox/ and wild-type mice (n = 4 to 5 mice/group/time period) were sacrificed 8, 20, or 32 days post-SME challenge. Draining pulmonary lymph nodes were restimulated in vitro with either SEA or ConA. Mononuclear cells from wild-type (solid bars) and p47phox/ (hatched bars) mice generated similar mixed Th-1- and Th-2-type cytokine responses. Reductions in the generation of Th-2-type cytokines (IL-4 and IL-10) in response to SEA were observed on day 32 in both p47phox/ and wild-type mice corresponding to the regression of granulomata observed histologically. IFN-, gamma interferon; NA, not available.

View larger version (11K): [in this window] [in a new window]   FIG. 5.   Pulmonary-granuloma sizes (means ± standard errors of the means) following SME challenge. p47phox/ and wild-type mice were sacrificed at 8, 20, and 32 days after SME challenge (n = 4 to 5 mice/group/time period), and mean pulmonary-granuloma volumes were determined. The mean volumes of granulomata in p47phox/ (circles) and wild-type (squares) mice were similar at each time point. In both p47phox/ and wild-type mice, the mean granuloma size decayed as a function of time following SME challenge and was significantly reduced at day 32 compared with days 8 and 20 (P < 0.05).

View larger version (64K): [in this window] [in a new window]   FIG. 6.   Histopathology (with H&E staining) demonstrating regression of pulmonary granulomata in p47phox/ mice after SME challenge. (A) Representative lung section at day 8 after SME challenge, showing highly cellular granulomata composed of lymphocytes, macrophages, and giant cells. (B) Day 32 post-SME challenge. Pulmonary granulomata are smaller than on day 8 and show reduced cellularity and increased fibrosis. Original magnification, ×200.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

This is the first study that specifically evaluates T-cell-mediated cytokine responses in experimental models of granuloma formation in CGD. We found that p47^phox/ mice generated appropriate Th-1- and Th-2-type cytokine responses to challenge with M. avium and SME, respectively. We initially hypothesized that p47^phox/ mice would be less able than wild-type mice to resolve granulomatous inflammation once it had been established. However, SME-elicited pulmonary granulomata in p47^phox/ mice regressed normally at days 20 and 32 after i.v. challenge, and normal healing of granulomata, characterized by fibrosis and loss of inflammatory cells, occurred. These data suggest that T-cell-mediated cytokine responses and granuloma resolution are normal in p47^phox/ mice. Therefore, abnormally exuberant granulomatous inflammation in CGD likely does not arise from primary abnormalities in T-cell phenotypes.

Phagocyte-derived reactive oxidants do not appear to influence granuloma formation in M. avium infection, but other phagocyte products have been shown to affect both host defense and pathologic responses in this model. In vitro studies have shown that the low-virulence M. avium strain 1983 stimulated high levels of superoxide and tumor necrosis factor alpha (TNF-) production in mouse peritoneal macrophages, whereas the highly virulent 25291 strain induced low levels of these products (27). In vivo, granuloma formation and production of gamma interferon and TNF- were impaired in M. avium-infected SCID mice and in mice depleted of CD4⁺ cells (16). Knockout mice lacking the TNF- receptors or inducible nitric oxide synthase were able to control M. avium infection as well as wild-type mice but showed significantly less splenomegaly and reduced splenic mononuclear cell infiltrate compared to wild-type mice (9). In contrast, mice with a disruption of the gamma interferon gene had far greater M. avium burdens than wild-type mice but had reduced splenomegaly and less-pronounced pathology (9). Gamma interferon augments the production of TNF-, nitric oxide, and reactive oxidants by macrophages (reviewed in reference 13). These data indicate that gamma interferon-mediated TNF- and nitric oxide generation plays an important role in the formation of splenic granulomata in response to experimental M. avium infection. In contrast, our data show that NADPH oxidase-derived oxidants do not significantly alter the host response to M. avium. The mean lung weight was slightly greater in p47^phox/ mice than in wild-type mice at week 8 after challenge (Fig. 2). However, histopathology and cytokine expression data were similar, indicating that NADPH-derived reactive oxidants likely do not modulate the pulmonary inflammatory response to M. avium. There was no significant difference in spleen and lung M. avium burden between p47^phox/ and wild-type mice (Table 1).

SME-elicited granulomata are composed of macrophages, lymphocytes, eosinophils, and rare neutrophils. Granuloma formation in this model is dependent on the balance between Th-1- and Th-2 type cytokines: IL-4 drives granuloma formation, whereas IL-12 reduces it (5, 32-35, 37). The chemokine receptor CCR1, which is expressed in neutrophils, macrophages, lymphocytes, and eosinophils, and which binds the chemoattractant MIP-1 as well as other related CC chemokines, is also involved in granuloma formation (14). Mice with a targeted disruption of the CCR1 gene generated smaller granulomata than wild-type mice following SME challenge. This was associated with a relative increase in gamma interferon production and decrease in IL-4 production from draining pulmonary lymph nodes stimulated in vitro with SEA. Thus, granulomata generated in the SME model are regulated by a complex interplay of cytokine and chemokine signaling. The ability of p47^phox/ mice to elaborate and resolve established tissue granulomata after SME challenge indicates that NADPH oxidase function is not required for either the generation or normal downregulation of granulomatous inflammation in this model.

Reactive oxidants are known to induce apoptosis in several cellular systems (e.g., neutrophils, lymphocytes, and fibroblasts) either directly or through generation of other toxic intermediates, such as peroxynitrite anion (3, 8, 10, 18, 26, 30, 31). Coxon et al. (8) showed that phagocytosis-induced apoptosis in human neutrophils is blocked by the flavocytochrome inhibitor diphenylene iodonium (an NADPH oxidase inhibitor) and is impaired in CGD neutrophils. In another study, the hydrogen peroxide scavenger catalase was shown to inhibit the ability of activated mouse peritoneal macrophages to suppress ConA-mediated lymphocyte proliferation (20), suggesting that macrophage-derived reactive oxidants may have an antiproliferative effect on lymphocytes. Classic granulomata are composed primarily of macrophages and lymphocytes and are dependent on generation of lymphocyte-derived proinflammatory cytokines. Thus, in CGD, the absence of phagocyte-derived reactive oxidants could be hypothesized to cause persistent lymphocyte viability within tissue granulomata and consequent abnormal downregulation of the inflammatory response in some settings. In the SME model, normal resolution of granulomata occurred in p47^phox/ mice, indicating that loss of cellularity, whether through apoptosis or other mechanisms, was independent of phagocyte-derived reactive oxidants.

CGD neutrophils are defective in catabolizing the chemotactic factors leukotriene B₄, C5a, and N-formyl peptide in vitro due to the inability to generate reactive oxidants (7, 15, 17). Leukotriene B₄ can augment cytokine expression in T cells committed to either the Th-1 or Th-2 phenotype (2, 21) as well as monocyte-derived proinflammatory cytokines (24, 25). The relevance of leukotrienes to granulomatous disease was suggested by Sharon and Stenson (29), who showed that colonic mucosa from patients with inflammatory bowel disease contained ~50-fold-higher levels of leukotriene B₄ than mucosa from patients without inflammatory bowel disease. Since granulomatous colitis commonly occurs in CGD, it is possible that the lack of reactive oxidants may lead to the accumulation in tissue of various chemotactic factors leading to persistent granulomata. Persistent granulomata in CGD may also relate to impaired digestion of phagocytosed products, as suggested by Segal et al. (28). Other reactive oxidant-generating systems, such as xanthine oxidase, may be able to compensate for the lack of NADPH oxidase activity in certain types of tissue granuloma formation but not in others.

Pyogranulomata, such as those elicited in the X-linked mouse model of CGD following intratracheal challenge with heat-killed A. fumigatus hyphae (22), are composed of neutrophils, macrophages, and lymphocytes. We have observed similar pulmonary pyogranulomata in p47^phox/ mice, but not in wild-type littermates, challenged with Aspergillus nidulans (4). The granulomata elicited by M. avium and SME are composed mainly of macrophages and lymphocytes, with few or no neutrophils.

The abnormally exuberant granulomata in CGD are not induced by using the above models. Therefore, inflammatory dysregulation in CGD is probably stimulus dependent and not an invariant feature of the disease. Granulomatous inflammation in CGD does not appear to result from a primary abnormality in lymphocyte activation and cytokine production but instead may be linked to other cell types (e.g., neutrophils) that either are absent from, or are minor components in, M. avium and SME granulomata.