Published online before print February 29, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
,
* Mycobacteria Research Laboratories, Department of Microbiology, Immunology and Pathology, Colorado State University, Fort Collins, Colorado, USA;
Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado School of Medicine, Denver, Colorado, USA;
National Jewish Medical and Research Center, Denver, Colorado, USA; and
Denver Veterans Affairs Medical Center, Denver, Colorado, USA
1Correspondence: Department of Microbiology, Immunology and Pathology, Colorado State University, Fort Collins, CO 80523-1682, USA. E-mail: D.Ordway-Rodriguez{at}colostate.edu
ABSTRACT
Chronic lung disease as a result of Mycobacterium abscessus is an emerging infection in the United States. We characterized the lung immune responses in mice and guinea pigs infected with M. abscessus. C57BL/6 and leptin-deficient ob/ob mice challenged with a low-dose aerosol (LDA) of M. abscessus did not develop an infection. However, when challenged with a high-dose aerosol (HDA), C57BL/6 and ob/ob mice developed an established infection and a pulmonary immune response consisting of an early influx of IFN-
+ CD4+ T cells; this immune response preceded the successful clearance of M. abscessus in both strains of mice, although mycobacterial elimination was delayed in the ob/ob mice. Infected guinea pigs showed an increased influx of lymphocytes into the lungs with bacterial clearance by Day 60. In contrast to the C57BL/6 and ob/ob mice and guinea pigs, IFN- knockout (GKO) mice challenged with a LDA or HDA of M. abscessus showed a progressive lung infection despite a robust influx of T cells, macrophages, and dendritic cells, culminating in extensive lung consolidation. Furthermore, with HDA challenge of the GKO mice, emergence of IL-4- and IL-10-producing CD4+ and CD8+ T cells was seen in the lungs. In conclusion, IFN- is critically important in the host defense against M. abscessus. As the number of effective drugs against M. abscessus is limited, the GKO mice provide a model for in vivo testing of novel drugs.
Key Words: T cells • dendritic cells • macrophages • cytokines
INTRODUCTION
Chronic, suppurative lung disease as a result of the rapidly growing mycobacteria, particularly Mycobacterium abscessus, is an emerging problem in the United States and many parts of the world. M. abscessus is ubiquitous in the environment, particularly prevalent in natural and municipal water sources, soil, and bioaerosols [1 , 2 ]. Chronic lung diseases as a result of M. abscessus are thus environmentally acquired infections. M. abscessus can also be acquired from contaminated medical equipment, although most of these nosocomial infections involve the skin and soft tissues [2 ].
The precise mode of transmission of environmental mycobacteria such as M. abscessus into the lungs is not known, although clinical reasoning and experimental evidence indicate that inhalation of droplet nuclei generated in bioaerosols and aspiration of colonized organisms from the upper aerodigestive tract into the lower respiratory tract occur [3
, 4
]. Two lines of experimental evidences support inhalation of droplet nuclei as a mechanism for infection into the lungs. First, it has been shown that natural bioaerosols, e.g., rivers and streams, can generate droplet sizes on the order of 10–150 µm in diameter. Upon drying, droplets that are 10–50 µm will shrink to sizes that are <5 µm, small enough to be inhaled directly into the alveoli. Man-made aerosols such as that from hot tubs, spas, and showers have very active bubble generation and thus, are likely to produce droplets of similar and even smaller sizes. Second, Falkinham and colleagues [5
] showed that environmental mycobacteria, because of their hydrophobicity, are highly concentrated in droplets, up to an order of 
5000-fold; i.e., mycobacterial concentrations in droplets generated from bioaerosols are 5000 times higher than concentrations in the original water source.
Cell-mediated immunity is central to the control of mycobacterial infections [6
7
8
9
], although most patients with chronic M. abscessus lung disease have no obvious immune defects. A central component of cell-mediated immunity is antigen presentation by infected cells, in the context of class I MHC, class II MHC, and non-MHC class molecules to CD8+, CD4+, and 
+ T cells, respectively. This collaboration activates T cells to produce IFN-, which in turn activates macrophages and dendritic cells (DC) to secrete chemokines and additional cytokines such as TNF-
[7
, 8
]. Recruitment by chemokines of additional mononuclear cells to the site of the infection leads to granuloma formation, which serves to contain mycobacterial infections and prevent their dissemination [8
]. Although an early robust response of IFN--producing Th1 cells is essential in controlling the growth of Mycobacterium tuberculosis, such inflammatory responses, if unchecked, may also contribute to the lung pathology [10
11
12
]. Hence, although immunosuppressive cytokines such as IL-4 and IL-10 are considered to increase host susceptibility to intracellular pathogens, their expression in the subacute and chronic stages of an infection may be necessary to attenuate tissue-damaging, inflammatory responses [10
, 11
].
Byrd and Lyons [13
] showed that in SCID mice, intratracheal inoculation of a rough virulent strain of M. abscessus produced a persistent infection in the lungs and spleens up to a study period of 28 days. More recently, Rottman and co-workers [9
] showed that T cells, IFN-, and TNF- were important in controlling disseminated M. abscessus infection in C57BL/6 mice. However, it is important to note that the mice were infected i.v. with a large inoculum (107 organisms/mice) of M. abscessus. The route of mycobacterial infection in mice is extremely important, as with an i.v. route, the emergence of protective immunity is greatly increased, resulting in a less-pathogenic infection compared with an aerosol route [14
].
We therefore undertook this study to further characterize the host lung immune responses against M. abscessus in an aerosolized model of lung infection using relatively low concentrations of M. abscessus in three strains of mice and the guinea pig. The rationale for choosing these strains is predicated on previous studies and observations about host susceptibility to mycobacterial infections in experimental animals and in humans. C57BL/6 mice are known to be resistant to mycobacterial infections [15
] and thus, were used as resistant controls. Leptin-deficient ob/ob mice were used, as they are more susceptible to M. tuberculosis and other bacterial pathogens [16
17
18
19
20
21
]. An additional reasoning for choosing the leptin-deficient mice is based on the observation that lung diseases, due to environmental mycobacteria, including those as a result of M. abscessus, occur more frequently in patients with pre-existing, thin-body habitus [21
22
23
24
], and slender individuals, as they have smaller and fewer number of fat cells, are known to have lower levels of the adipocyte-derived hormone leptin [16
, 17
]. Although the canonical function of leptin is that of a satiety hormone, it also behaves as a cytokine with immunomodulating properties including host-defense functions against microbial pathogens. IFN- knockout (GKO) mice were also infected with M. abscessus, as these mice and those with genetic disruption for IFN-R1 have increased susceptibility to M. tuberculosis [25
]. Perhaps more germane to this study are reports showing that humans with defects in the IFN- signaling pathway are predisposed to infections caused by environmental mycobacteria [25
]. Finally, Hartley guinea pigs were infected with M. abscessus because of their increased susceptibility to M. tuberculosis and similarity of their tuberculous lesions to that seen in human cases [26
].
We found that C57BL/6 mice were resistant to a low-dose aerosol (LDA) infection of M. abscessus, and although they developed an established infection with a high-dose aerosol (HDA) challenge, these mice were able to eliminate the organisms from their lungs by Day 30 and from their spleens by Day 60. These resistant mice mounted an early lung immune response characterized by an influx of IFN--producing CD4+ and CD8+ T cells and to a lesser extent, influx of alveolar macrophages and DC. The ob/ob mice challenged with a HDA of M. abscessus cleared the organisms from their lungs by Day 60; this delayed clearance by the leptin-deficient ob/ob mice was reflected by a lower influx of IFN--producing CD4+ T cells. The infected guinea pigs developed an immune response involving an influx of macrophages and lymphocytes with bacterial clearance by Day 60. In contrast to these animals, a persistent lung infection was established in the GKO mice, despite a large influx of innate and adaptive immune cells into the lungs.
MATERIALS AND METHODS
Mice, guinea pigs, and M. abscessus
Specific, pathogen-free female C57BL/6 and GKO mice, from 6 to 8 weeks old, were purchased from the Jackson Laboratories (Bar Harbor, ME, USA). The homozygous, leptin-deficient ob/ob mice were bred from heterozygous (ob+/ob–) mice that were also purchased from the Jackson Laboratories. The ob/ob mice were identified genotypically by PCR of tail DNA or phenotypically by their obesity at 4 weeks of age. Mice were maintained in the Biosafety Level III animal laboratory at Colorado State University (Fort Collins, CO, USA) and were given sterile water, mouse chow, bedding, and enrichment for the duration of the experiments. Female outbred Hartley guinea pigs (500 g in weight) were purchased from the Charles River Laboratories (North Wilmington, MA, USA) and held under barrier conditions in a Biosafety Level III animal laboratory. The specific, pathogen-free nature of the mouse colonies was demonstrated by testing sentinel animals. The Animal Care and Usage Committee of Colorado State University approved all experimental protocols.
M. abscessus strain L948 was obtained from American Type Culture Collection (Manassas, VA, USA; Cat. #19977). It is a virulent strain, originally isolated from a patient. On solid culture medium, the colonies display a rough morphology.
Experimental infections in mice and guinea pigs
Mice were challenged with M. abscessus strain L948 using a Glas-Col (Terre Haute, IN, USA) aerosol generator, calibrated to deliver a LDA inoculum of 100 bacilli per animal or a HDA of 1000 bacilli per animal. At Days 15, 30, and 60 following infection, bacterial loads in the lungs and spleen, lung histology, and mononuclear and lymphocytic cellular expressions were determined. Guinea pigs were challenged using a Madison chamber aerosol generation device, delivering a LDA of 20 bacilli per animal or a HDA of 200 bacilli per animal. In the guinea pigs, bacterial loads were determined, and histopathology of the lungs, spleen, and lymph nodes were viewed 30 and 60 days after infection. Bacterial counts were determined by plating serial dilutions of homogenates of lungs on nutrient 7H11 agar and counting CFUs after 3 weeks incubation at 37°C. For each of the three mice strains and the guinea pigs, a total of five animals was infected for each time-point.
Histological analysis in mice and guinea pigs
The accessory lobe of the lung from each mouse and guinea pig was fixed with 10% formalin in PBS. Sections from these murine tissues were stained using H&E and acid-fast stains as previously reported [27
, 28
].
Lesion analysis in guinea pigs
In the guinea pigs, a histological rank scoring system for the lung, splenic, and lymph node lesions was applied as reported previously [28
]. This scoring system semi-quantifies the degree of inflammation in the organs measured.
Lung cell digestion in mice
Single cell suspensions were prepared as described previously [11
]. Cell suspensions from each individual mouse were incubated with mAb labeled with FITC, PE, PerCP, or allophycocyanin at 4°C for 30 min in the dark as described previously [11
]. All analyses were performed with an acquisition of at least 100,000 total events.
Intracytoplasmic cytokine staining in mice
Cells were first stained for cell surface markers as indicated above, and thereafter, the same cell suspensions were prepared for intracellular staining as described before [11
].
Statistical analysis in mice and guinea pigs
Data are presented using the mean values from five mice or guinea pigs per group and from values from replicate samples and duplicate or triplicate assays. The Student’s t-test test was used to assess statistical significance between groups of data.
RESULTS
Bacterial loads in C57BL/6, ob/ob, and GKO mice
Following LDA or HDA infection with M. abscessus of C57BL/6, ob/ob, and GKO mice, the bacterial loads in the lungs and spleens were quantified on Days 1, 15, 30, and 60 after the challenges. With a LDA challenge, no M. abscessus was recovered in the lungs or the spleen at any of the time-points in the C57BL/6 or ob/ob mice (Fig. 1A
and 1B
, 
and 
). However, with a HDA challenge, an infection was established in the C57BL/6 mice with the number of bacteria in the lungs increasing to almost 3 logs CFU by Day 15, with clearance of the organisms from the lungs by Day 30 (Fig. 1A
, •). As shown in Figure 1B
(•), C57BL/6 mice also showed a steady rise in bacterial load in the spleen, peaking at Day 30 after challenge and with rapid clearance by Day 60. With a HDA infection, the ob/ob mice demonstrated greater bacterial loads by Day 15 to nearly 4 logs CFU in the lungs and approximately 3 logs CFU in the spleen (Fig. 1A
and 1B
, 
); as shown, there was delayed clearance of bacteria in the lungs and minimal clearance in the spleen in the ob/ob mice compared with the C57BL/6 mice.
 View larger version (29K): [in a new window] |
Figure 1. Bacterial loads in C57BL/6, ob/ob, and GKO mice and guinea pigs after aerosolization with M. abscessus. C57BL/6, ob/ob, and GKO mice were challenged by LDA or HDA of M. abscessus. At Days 1, 15, 30, and 60 after challenge, bacterial loads in the lungs and spleen were quantified. (A–D) Bacterial loads in the lungs and spleens of C57BL/6 (LDA, ; HDA, •), ob/ob (LDA, ; HDA, ), and GKO mice (LDA,  ; HDA,  ). (E–G) Bacterial loads in the (E) lungs, (F) spleens, and (G) lymph nodes of Hartley guinea pigs 30 and 60 days after LDA ( ) or HDA ( ) challenges. Results are expressed as the average log10 CFU of five mice or five guinea pigs (±SE) for each time-point. Student’s t-test; *, P = 0.001, for ob/ob mice HDA compared with C57BL/6 mice HDA on Day 15 and GKO mice LDA compared with HDA on Day 60; **, P = 0.002, for guinea pigs LDA compared with HDA on Day 30.
|
). With a HDA infection, GKO mice showed nearly a 20-fold greater number of CFU in the lungs and a threefold greater number of CFU in the spleens compared with LDA infection by Day 60 (Fig. 1C
and 1D
, ).
Bacterial loads in guinea pigs
In guinea pigs exposed to a LDA or a HDA of M. abscessus, bacterial loads in the lungs, spleens, and lymph nodes were quantified at Days 30 and 60 following infection. Unlike the C57BL/6 and ob/ob mice, the guinea pigs became infected with a LDA of M. abscessus with a 3.5-log increase in bacterial growth in the lungs 30 days after challenge (Fig. 1E)
. Dissemination of M. abscessus to the spleen (Fig. 1F)
and lymph nodes (Fig. 1G)
was evident by Day 30. Guinea pigs receiving a HDA of M. abscessus showed significantly greater bacterial loads in the lungs and spleen but not in the lymph nodes compared with the LDA challenge (Fig. 1E
1F
1G)
. With LDA or HDA infection, M. abscessus organisms were eliminated from all three organs by Day 60.
Lung histopathology of C57BL/6, ob/ob, and GKO mice after aerosol challenge with M. abscessus
Lung histopathology following HDA challenge was characterized in the C57BL/6, ob/ob, and GKO mice (Fig. 2
). At Day 15 after challenge, the predominant pathologic finding in all three mice strains was characterized by peribronchiolar inflammatory infiltrates (Fig. 2A
2D
and 2G
, arrows).
 View larger version (133K): [in a new window] |
Figure 2. Lung histology of C57BL/6, ob/ob, and GKO mice after aerosolization with M. abscessus. The lungs from C57BL/6, ob/ob, and GKO mice following a HDA challenge of M. abscessus were harvested on Days 15, 30, and 60, fixed, and stained with H&E. (A, D, and G) Representative lung histopathology at 15 days after M. abscessus infection in C57BL/6, ob/ob, and GKO mice, respectively. (B, E, and H) Representative lung histopathology at 30 days after M. abscessus challenge in the C57BL/6, ob/ob, and GKO mice, respectively. (C, F, and I) Representative lung histopathology at 60 days after M. abscessus challenge in the C57BL/6, ob/ob, and GKO mice, respectively. Total original magnification: A–H, 20x; I, 10x.
|
Fig. 5E
). In contrast, the lungs of the ob/ob mice, 30 days following HDA challenge, showed greater numbers of foamy cells and larger granulomatous lesions (Fig. 2E)
; by Day 60, the pulmonary lesions showed tight aggregates of lymphocytes with fewer but persistent numbers of highly vacuolated foamy cells (Fig. 2F
, and see 5F
).
 View larger version (116K): [in a new window] |
Figure 3. Lung, spleen, and lymph node histopathology and individual tissue lesion scores in Hartley guinea pigs after a LDA or HDA challenge with M. abscessus. The lungs, spleens, and lymph nodes were harvested from guinea pigs 30 and 60 days after challenge with a LDA or HDA of M. abscessus and were stained with H&E. The individual tissue lesion scores were measured and the means calculated for the lungs, spleens, and lymph nodes. (A, C, and E) Representative lung histopathology at Day 30 of the lungs, spleens, and lymph nodes, respectively, after HDA challenge. (B, D, and F) Representative lung histopathology at Day 60 of the lungs, spleens, and lymph nodes, respectively, after HDA challenge. (G–I) The lung, splenic, and lymph node histopathology mean rank score in the HDA group compared with the LDA group at 30 and 60 days after challenge. Student’s t-test; *, P = 0.001, HDA compared with LDA at Day 60. Total original magnification: A and B, 10x; C–F, 20x.
|
 View larger version (30K): [in a new window] |
Figure 4. T cell influx into the lungs of C57BL/6, ob/ob, and GKO mice after challenge with M. abscessus. Lung cells obtained from C57BL/6 and ob/ob mice challenged with a HDA of M. abscessus, and GKO mice challenged with LDA or HDA of M. abscessus were assayed by flow cytometry for percentages of CD3+CD4+, CD3+CD8+, and CD3+NK1.1 T cells on Days 15, 30, and 60 after initial infection. (A–C) The total percentage of CD3+CD4+, CD3+CD8+, and CD3+NK1.1 T cells, respectively, in the lungs of uninfected (naïve) C57BL/6 mice (lines), M. abscessus-infected C57BL/6 mice (•), and M. abscessus-infected ob/ob mice (). (D–F) The percentages of CD3+CD4+, CD3+CD8+, and CD3+NK1.1+ T cells, respectively, in the lungs of uninfected (naïve) GKO mice (lines) and GKO mice infected with a HDA () or a LDA () of M. abscessus. Results are expressed as the average percentage of five mice (±SE). Student’s t-test; *, P = 0.001, for GKO mice HDA compared with LDA on Day 30; **, P = 0.002, for C57BL/6 mice HDA compared with ob/ob mice HDA on Day 15.
|
 View larger version (52K): [in a new window] |
Figure 5. Alveolar macrophage and DC influx into the lungs of C57BL/6, ob/ob, and GKO mice after challenge with an aerosol of M. abscessus. Lung cells obtained from C57BL/6, ob/ob, and GKO mice after a challenge with M. abscessus were assayed by flow cytometry for percentages of alveolar macrophages and DC on Days 15, 30, and 60 after initial infection. (A) A representative dot-plot of lung cells obtained from a mouse primarily gated on FSCmid/high and SSCmid/high and secondarily gated according to the expression of CD11c verses CD11b markers. (B) Three representative CD11b/CD11c dot-plots of lung cells obtained 15 days after M. abscessus challenge from representative C57BL/6 (left), ob/ob (middle), and GKO (right) mice. Each dot-plot was gated into various regions including R5 as alveolar macrophages (CD11bneg/CD11chi), R6 as DC (CD11b+/CD11chi), R7 as small macrophages or monocytes CD11b+/CD11cneg (lower, middle region), and R8 as neutrophils CD11b+/hi/CD11cneg (lower, right region). (C) The mean percentages of R5 alveolar macrophages (Mo; ) and R6 DC (•) from C57BL/6 mice and mean percentages of R5 alveolar macrophages () and R6 DC () from ob/ob mice at 15, 30, and 60 days after HDA challenge with M. abscessus. (D) The mean percentages of R5 alveolar macrophages () and R6 DC () from HDA-exposed GKO and R5 alveolar macrophages () and R6 DC () from LDA-exposed GKO mice at 15, 30, and 60 days. Results are expressed as the average percentage of five mice (±SE). Student’s t-test; *, P = 0.001, for alveolar macrophage and DC. (E–G) The presence of foamy cells in the lungs of the M. abscessus-infected C57BL/6, ob/ob, and GKO mice, respectively. Total original magnification: E–G, 40x.
|
Histopathology of the lungs, spleens, and lymph nodes of guinea pigs after LDA or HDA challenges with M. abscessus
Following exposure of guinea pigs to LDA or HDA of M. abscessus, the lungs, spleens, and lymph nodes were harvested at Days 30 and 60 following infection. The severity of pulmonary inflammation was quantified using a lesion-scoring system developed in the guinea pig that evaluates extent of disease [28
]. Shown in Figure 3A
3B
3C
3D
3E
3F
, are representative histopathological sections for the lungs, spleens, and pulmonary lymph nodes 30 and 60 days after HDA challenge with M. abscessus. Histological sections for LDA infections are not shown but were quantified (Fig. 3G
3H
3I
, open bars). Figure 3A
shows well-formed, granulomatous lung lesions at Day 30 following a HDA infection; as shown in Figure 3G
, no histopathological differences were found in the rank scoring of lung tissues between LDA and HDA at 30 days. However, at 60 days, guinea pigs infected with a HDA challenge (Fig. 3B)
had more severe granulomatous inflammation in the lungs than those infected by LDA (Fig. 3G)
. The inflammatory response was characterized by sheets of epithelioid macrophages and organized aggregates of lymphocytes that infiltrated septal walls and filled alveolar spaces.
Guinea pigs exposed to a HDA of M. abscessus demonstrated splenic (Fig. 3C) and lymph node (Fig. 3E) lesions at 30 days, which were comprised of paracortical or subcapsular accumulations of epithelioid macrophages and hyperplasia of lymphoid follicles, progressing to more severe granulomatous inflammation by Day 60 (Fig. 3D and 3F) . As shown in Figure 3H and 3I , by Day 60, the splenic and lymph node mean rank scores in animals exposed to a HDA infection were significantly higher than those exposed to a LDA challenge.
Characterization of T cell subset expression during M. abscessus infection
A comparative flow cytometric analysis of T cell populations from the lungs of the infected mice was conducted as described previously [11
, 27
, 29
]. The T cells were primarily gated on viable forward-scatter (FSC)low versus side-scatter (SSC)low lymphocytes and secondarily gated on CD3+ T cells. Changes in the total percentage of CD3+CD4+, CD3+CD8+, and CD3+NK1.1+ T cells were monitored over the course of infection. As shown in Figure 4A
4B
4C
(•), the total percentage of CD3+CD4+, CD3+CD8+, and CD3+NK1.1+ T cells in the lungs of C57BL/6 mice after a HDA exposure of M. abscessus infection peaked at Day 15 and then diminished by Day 30 compared with the uninfected ("naïve") mice. In contrast, the emergence of CD3+CD4+ T cells in the lungs was delayed in the ob/ob mice, peaking by Day 30 and then subsiding by Day 60 (Fig. 4A
, ). There was little or no influx of CD3+CD8+ and CD3+NK1.1+ T cells in the ob/ob mice (Fig. 4B
and 4C
, ). T cell expression of uninfected C57BL/6 mice (naïve) is shown as line graphs in Figure 4A
4B
4C
.
In contrast to the C57BL/6 and ob/ob mice, infected GKO mice displayed a more robust and sustained T cell response (Fig. 4D
4E
4F
, , ). GKO mice exposed to a HDA of M. abscessus showed an even greater total percentage of CD3+CD4+ T cells and CD3+CD8+ T cells at Days 30 and 60 in the lungs compared with the LDA-challenged mice, with minimal decline even at 60 days post-challenge (Fig. 4D
and 4E)
. However, the percentages of CD3+NK1.1+ T cells in the lungs of LDA and HDA groups of GKO-infected mice significantly increased between Days 15 and 30 and waned to basal, uninfected levels by Day 60 (Fig. 4F)
. T cell expression of uninfected GKO mice (naïve) is shown as line graphs in Figure 4D
4E
4F
.
Alveolar macrophage and DC expression during M. abscessus infection
Flow cytometric analysis was performed to quantify the relative percentages of alveolar macrophages and DC in the lungs of C57BL/6, ob/ob, and GKO mice infected with M. abscessus using a strategy reported previously [30
]. Shown in Figure 5A
is a typical dot-plot of lung cells obtained from a representative mouse, primarily gated on FSCmid/high and SSCmid/high and secondarily gated according to the expression of CD11c versus CD11b markers. Furthermore, each dot-plot was gated into various regions as reported previously [29
, 30
]. Shown in Figure 5B
are three representative CD11b/CD11c dot-plots of lung cells from C57BL/6, ob/ob, and GKO mice 15 days post-M. abscessus challenge. As can be seen, in comparison with the control C57BL/6 mice, ob/ob mice showed increased percentages of region 8 (R8) neutrophils, and GKO mice had higher percentages of R5 alveolar macrophages and R6 DC.
The absolute increase in the influx of alveolar macrophages and DC in C57BL/6 and ob/ob mice was small—on the order of 
3%—in response to LDA (data not shown) or HDA (Fig. 5C)
M. abscessus infection. In the C57BL/6 mice, the influx of alveolar macrophages increased by Day 15 and declined minimally by Days 30 and 60 (Fig. 5C
, ). The influx of DC was even less than the alveolar macrophages (Fig. 5C
, •). In contrast, the ob/ob mice showed a delayed influx of alveolar macrophages and DC (Fig. 5C
, and , respectively). By Day 60, the percentage of DC in the ob/ob mice has declined to near basal levels, and the macrophages remained elevated.
In contrast to C57BL/6 and ob/ob mice, infected GKO mice had a significantly greater influx of macrophages and DC into the lungs (Fig. 5D)
. Compared with mice that received the LDA challenge (Fig. 5D
, and ), GKO mice exposed to a HDA (Fig. 5D
, • and ) had an even greater influx of alveolar macrophages and DC from Days 15 to 60. Thus, during chronic infection, GKO mice infected with M. abscessus showed increased bacterial loads and a waning T cell population in the affected organs but greater influx of macrophages and DC, resulting in greater lung consolidation (Fig. 2)
.
When sections from all the experimental groups of infected mice were compared at 60 days after challenge, foamy cells were a prominent finding in the lung granulomas. The foamy cells present in the C57BL/6 mice (Fig. 5E) had smaller cytoplasms and were fewer in number than those present in the ob/ob mice (Fig. 5F) . The GKO mice showed the largest number of aggregates of highly vacuolated foamy cells (Fig. 5G) .
Expression of cytokines in CD4+ and CD8+ T cells during M. abscessus infection
We next determined the expression of cytokine-producing T cells in the course of M. abscessus infection in the mice strains, as the relative expression of Th1 or Th2 cytokines may impact their ability to control the infection. T cells were primarily gated on viable FSClow versus SSClow lymphocytes and secondarily gated on SSClow versus CD4+ T cells or CD8+ T cells. Figure 6A
is an example of a dot-plot for CD4+ T cells. Subsequently, changes in the total percentage of IFN-, IL-4, IL-10, TNF-, and/or CD25-positive CD4+ and CD8+ T cells were monitored over the course of infection in C57BL/6 and ob/ob mice with HDA of M. abscessus and in the GKO mice with LDA and HDA infection.
 View larger version (26K): [in a new window] |
Figure 6. Expression of intracellular IFN-, IL-4, IL-10, and TNF- in CD4+ and CD8+ T cells in the lungs of C57BL/6, ob/ob, and GKO mice infected with M. abscessus. Lung cells obtained from C57BL/6, ob/ob, and GKO mice after HDA challenge with M. abscessus were assayed by flow cytometry for percentages of IFN-, IL-4, IL-10, TNF-, and CD25-positive CD4+ and CD8+ T cells on Days 15, 30, and 60 after initial infection. (A) A representative histogram of T cells, which were primarily gated on viable FSClow versus SSClow lymphocytes and then on CD4+. (B) The total percentage of CD25+IFN- (IFN-y+)-expressing CD4+ T cells (•) and CD8+ T cells () in the lungs of the C57BL/6 mice during the course of HDA M. abscessus infection. Also, the total percentages of CD25+IFN-+-expressing CD4+ T cells () and CD8+ T cells () in the lungs of the ob/ob mice challenged with HDA infection. (C) The influx of CD4+ T cells expressing IL-4 (), IL-10 (), TNF- (), or CD25 (•) in the GKO mice challenged with a HDA of M. abscessus. (D) The influx of CD8+ T cells expressing IL-4 (), IL-10 (), TNF- (), or CD25 (•) in the GKO mice challenged with a HDA of M. abscessus.
|
-expressing CD4+ T cells (•) and CD8+ T cells () peaked by Day 15 and declined by Day 30 in the lungs following M. abscessus infection, coinciding with bacterial clearance. The ob/ob mice had significantly less influx of CD4+IFN-+ T cells in the lungs at Day 15 (Fig. 6B
, ) and showed a 50% reduction in CD4+CD25+IFN-+ expression by Day 30 and thereafter, dropped further as bacterial clearance was achieved. The modest influx of CD8+CD25+IFN-+ T cells in the ob/ob mice was similar to the C57BL/6 mice, peaking on Day 15 and returning back to basal levels by Day 30 (Fig. 6B
, ). This pattern suggests that the ultimate control of bacterial growth, albeit delayed, in the ob/ob mice was predominantly achieved by CD4+ T cells. As would be expected for the GKO mice, there was absence of CD4+CD25+IFN-+ or CD8+CD25+IFN-+ T cells (data not shown).
Little (<1%) or no IL-4- or IL-10-producing CD4+ and CD8+ T cells were found in the lungs of C57BL/6 or ob/ob mice infected by HDA challenge or of GKO mice infected by LDA challenge (data not shown). In contrast, GKO mice exposed to a HDA of M. abscessus demonstrated increased influx of CD4+ T cells that expressed IL-4, IL-10, TNF-, and CD25 by 30 days and further increased by 60 days (Fig. 6C)
. There was also increased influx of CD8+ T cells that produced IL-4, IL-10, TNF-, and CD25 by 30 days following infection (Fig. 6D)
; although the CD8+TNF-+ T cell number declined at Day 60, CD8+ T cells expressing IL-4, IL-10, and CD25 increased further at this later time-point.
DISCUSSION
This study shows that the GKO mice can be productively infected with M. abscessus by an aerosol exposure. Despite an early, increased influx of CD4+, CD8+, and NK1.1+ T cells in the lungs of the GKO mice infected with M. abscessus, the lack of IFN- makes these T cells functionally ineffective. These results support a primary role for IFN--expressing CD4+ and CD8+ T cells in controlling and clearing M. abscessus infection. The increased susceptibility of the GKO mice supports published studies, which show that humans with defects in the IFN-–IL-12 axis, including those with defects in IFN- expression, IFN- R1 or IFN- R2 function, STAT1 expression, or IL-12p40 expression, have an increased susceptibility to mycobacterial infections [25
]. Recently, Rottman et al. [9
] showed that T cell immunity, IFN-, and TNF- were essential for optimal control of M. abscessus infections in C57BL/6 mice. However, there are important differences between their study and ours, which raise new, interesting questions. First, they i.v.-infected their mice with a large inoculum of M. abscessus (1x107 organisms per mice). Despite this large inoculant, they were unable to detect M. abscessus in the lungs or any obvious lung pathology, even in the susceptible IFN-R1 knockout animals. Thus, although their murine model of i.v. infection revealed important, immunological findings, it does not reflect what is most concerning in human infections; i.e., M. abscessus causes primarily a lung infection. Second, in their IFN-R1-deficient mice, there was a progressive decrease in the number of M. abscessus recovered in the liver and spleen during the follow-up period of 90 days after inoculation. In contrast, we saw an increasing number of organisms in the liver and spleen in the GKO mice over a follow-up period of 60 days after aerosolization. These differences between their study and ours is likely a result of differences in the routes of infection, as it has been shown that the i.v. model of mycobacterial infection is less pathogenic than the aerosol model of infection, owing to a more rapid induction of systemic immunity in the former [14
]. Thus, our combined experimental results indicate that i.v. infection of M. abscessus in mice is less pathogenic than the aerogenic route of infection.
Another significant finding in our study was that the larger inoculum (HDA) of M. abscessus in the GKO mice elicited a significant Th2 response, whereas the smaller inoculum (LDA) did not. As M. abscessus is ubiquitous in the environment, perhaps this finding may in part explain why some individuals develop lung disease, whereas others do not. Indeed, it is not inconceivable and actually quite plausible that the size of the infectious dose is an important determinant of whether an infection becomes established or not. Although the inoculum size of M. abscessus for the individual patient who acquires such lung disease will likely never be known, our experimental studies were conducted with reasonable infectious doses of M. abscessus (100–1000 organisms per mouse and 20–200 organisms per guinea pig) and by a realistic route of infection.
An additional difference between the GKO-infected mice and the other mice tested was the increased inflammatory response in the GKO mice, as measured by an increased influx of T cells, alveolar macrophages, and DC. As we showed previously, increased numbers of DC, characterized by their foamy appearance, may lead to mycobacterial persistence by providing a safe, intracellular niche for the micoorganisms [27
, 29
, 30
]. Moreover, macrophages can produce a variety of immunoregulatory cytokines, such as TNF-, IL-1β, IL-6, TGF-β1, and GM-CSF, which may influence mycobacterial killing but may also produce disordered inflammation within the lungs [31
, 32
].
Epidemiological studies involving a large number of subjects have shown that in healthy individuals, the body mass index inversely correlated with the risk of reactivation tuberculosis [33 , 34 ]; i.e., thin subjects are more likely to develop active tuberculosis. This finding was corroborated by a recent study from Hong Kong, which showed that obesity was associated with a lower risk of active pulmonary tuberculosis [35 ]. This increased susceptibility to reactivation tuberculosis in thin individuals, who have decreased fat stores, may be a result of a relative deficiency of leptin. Thus, most human subjects who are relatively leptin-deficient are thin because of diminished fat stores; in contrast, leptin-deficient mice have a genetic defect for their leptin deficiency and somewhat pardoxically develop obesity when food is freely available, as they have lost the satiety function of leptin.
It is increasingly appreciated that fat tissues and particularly leptin, produced by adipocytes, have immunomodulatory functions [35
36
37
]. A key immune function of leptin is that it differentiates naive (undifferentiated) T0 cells toward the IFN--producing Th1 phenotype [38
]. In this regard, leptin has been shown to have host defense function against bacterial pathogens including Klebsiella pneumoniae and Streptococcus pneumoniae [19
20
21
]. Furthermore, Weiland et al. [18
] showed in an intranasal model of M. tuberculosis infection, that leptin-deficient mice had reduced IFN- expression and greater susceptibility to M. tuberculosis than C57BL/6 mice. In humans, low leptin levels seen in patients with active tuberculosis may further impair host immunity, creating a vicious cycle of cachexia and cellular immunodeficiency [39
40
41
42
]. We found that the leptin-deficient ob/ob mice had an intermediate susceptibility to M. abscessus; i.e., the clearance of M. abscessus in the lungs and spleens was delayed and intermediate to that of the C57BL/6 and GKO mice. This increased susceptibility of the ob/ob mice was associated with a delayed and decreased influx of IFN--producing CD4+ T cells into the lungs (Figs. 4A
and 6B)
. Perhaps a greater susceptibility to M. abscessus was not seen in the leptin-deficient mice because of the presence of other immunomodulatory adipokines in fat tissues [36
]. Nevertheless, our findings support the hypothesis that relative leptin deficiency in thin individuals may increase susceptibility to M. abscessus and other mycobacterial infections.
The finding of peribronchiolar infiltrates in the infected mice correlates well with the inflammatory bronchiolitis seen in nearly all patients with M. abscessus lung disease [1 ]. The lung granulomas of GKO mice have a larger lymphocytic core, greater number of highly vacuolated foamy cells, increased thickening of the parenchymal walls, and more rapid progression of pathology than C57BL/6 or ob/ob mice. Similar pathology associated with M. abscessus infection in humans and SCID mice has been reported [13 ]. Increased granuloma size correlates with poorer disease outcome in animal models of tuberculosis [8 , 27 , 29 ] and human tuberculosis [43 ]. As reported previously for M. tuberculosis, the greater numbers of highly vacuolated foamy cells in the granulomatous lesions of the GKO mice and to a lesser extent, in the ob/ob mice may allow for intracellular bacterial persistence [40 , 44 ].
Despite the lack of productive infection in the C57BL/6 and ob/ob mice and in the guinea pig, considerable lung inflammation was seen in all three. However, the greatest inflammatory response was seen in the susceptible GKO mice, suggesting that this may be an element of M. abscessus pathogenesis. In addition to the increased influx of alveolar macrophages and DC, the increased inflammatory response in the GKO mice may be a result of the appearance of immunosuppressive cellular phenotypes, as corroborated by the increased influx of CD4+CD25+ and CD8+CD25+ T cell expressing IL-4 and IL-10 with HDA infection. Immunosuppressive T cells have also been shown to contribute to the acute lung damage caused by mycobacterial infection [11 ].
In conclusion, C57BL/6 and ob/ob mice were effective in clearing M. abscessus aerosol infection, although the clearance was less efficient in the leptin-deficient ob/ob mice. Guinea pigs were slightly more susceptible but were able to eventually clear the infection. In contrast, GKO mice were unable to eliminate the mycobacteria at the end of the follow-up period of 60 days following aerosol challenge. With a higher inoculum of M. abscessus in the GKO mice, there was an emergence of a Th2-immunosuppressive response. Persistent M. abscessus infection in the GKO mouse, which is already used for screening potential drugs for tuberculosis [45 ], provides a suitable model for testing new agents against this highly drug-resistant organism.
ACKNOWLEDGEMENTS
This work was supported by National Institutes of Health grants AI-040091 (I. M. O.), AI-44072 (I. M. O.), HL-66112 (E. D. C.) and the Potts Memorial Foundation. We thank Veronica Gruppo (Colorado State University) for culturing the M. abscessus and Dr. Joseph O. Falkinham III (Virginia Polytechnic Institute and State University, Blacksburg, VA, USA) for helpful discussions.
Received October 17, 2007; revised January 16, 2008; accepted February 10, 2008.
REFERENCES
interferon, and tumor necrosis factor in immune control of the rapid grower Mycobacterium abscessus in C57BL/6 mice Infect. Immun. 75,5898-5907 axis in protective immunity to mycobacteria Kotb, M. Calandra, T. eds. Cytokines and Chemokines in Infectious Diseases Handbook ,151-161 Humana Totowa, NJ, USA. ], IL-1[β] and IL-6 expression in ovine paratuberculosis Vet. Immunol. Immunopathol. 49,331-345[CrossRef][Medline] interferon gene-disrupted mice Antimicrob. Agents Chemother. 47,783-785
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
