Generalized Immunological Decline during Long-Term Experimental Infection with Mycobacterium avium

Mycobacterial infections are known to impair the immunological function of the host. Development of tuberculosis is often associated with depression of cellular immunity, as shown by a loss of the tuberculin skin test reaction (7), a decrease in interleukin-2 (IL-2) secretion and in IL-2 receptor expression (16), and a reduction in lymphocyte proliferation (10). In pulmonary tuberculosis, the predominant form of the disease, 17 to 25% of newly diagnosed patients are unresponsive to purified protein derivative (PPD) skin testing (4), and 40 to 60% of patients have low blastogenic responses to PPD (9). Failure by peripheral blood mononuclear cells (PBMCs) from human tuberculosis patients to produce gamma interferon (IFN-γ) has been attributed by some to the sequestration of Mycobacterium tuberculosis-reactive T cells at the site of infection (2, 5).

We have developed an experimental system that may reflect some of the events during chronic mycobacterial infection. Thus, intranasal infection of C57BL/10 mice with a virulent strain of the facultative intracellular pathogen Mycobacterium avium causes a slowly progressive disease culminating in the death of the animals about 40 weeks postinfection (6). A marked decline in the number of both CD4⁺ and CD8⁺ T cells in the spleen by 40 weeks postinfection was accompanied by an increase in apoptosis among CD4⁺ and CD8⁺ T cells (6). Furthermore, this process was associated with a terminal decline in the percentage of CD4⁺ and CD8⁺ T cells in the spleen producing the key protective cytokine IFN-γ measured by intracellular cytokine staining (6).

Because infection spanned 10 months in the 2-year life span of a mouse, it was important to determine the impact, if any, of aging on the decline in immunological function. It is widely believed that immune responsiveness gradually decays with age, rendering the individual more susceptible to infectious diseases (15). These diseases include tuberculosis, which remains relatively common in elderly humans (11). Furthermore, the question of whether deficient IFN-γ production by spleen cells during M. avium infection stemmed from the sequestration of antigen-responsive T cells to the lung, as suggested in human infection, is relevant (2, 5). Here, we investigate the production of IFN-γ by old and young mice at different stages of infection and at another anatomical site, namely the lungs, where bacterial numbers exceed 10⁹, compared with 10⁷ in spleen (6).

Initially to confirm that splenic T cells taken late in infection had an intrinsic inability to recall IFN-γ production in vitro, T cells taken from mice at different stages of infection were purified by passage through nylon wool (8). T cells (2 × 10⁶) were cultured with 2 × 10⁶ normal irradiated splenocytes as antigen-presenting cells (APCs) and with 5 × 10⁶ live M. avium CFU. After 72 h, T cells from 10-week-infected mice produced almost twice as much IFN-γ as the same number of T cells from 40-week-infected mice, as measured by bioassay of the supernatant (12) (Table 1) (P < 0.01). Interestingly, the cells from 40-week-infected mice also showed a fourfold reduction in IFN-γ production in response to concanavalin A (ConA), suggesting the deficiency was nonspecific.

The question of the specificity of the declining response was further addressed by using the response to another intracellular pathogen. Mice were infected intravenously with 5 × 10³Listeria CFU and allowed 3 weeks to completely resolve infection. They were then intranasally infected with 10⁵M. avium cells for a further 38 weeks. Mice were sacrificed, and splenocytes were cultured for 72 h with either live M. avium CFU or heat-killed Listeria (HKL) CFU, as shown in Table 2. In response to live M. avium CFU, spleen cells from 38-week M. avium-infected mice showed a reduced capacity to produce IFN-γ, whether or not they were previously infected with Listeria (Table 2). This reduction in IFN-γ production in long-term M. avium infection was nonspecific, since mice infected successively with Listeria and M. avium produced only 39 IU/ml in response to HKL, compared with 209 IU of IFN-γ per ml from mice immunized with Listeria 41 weeks before.

It was conceivable that the decline in IFN-γ production observed 35 to 40 weeks after infection was a consequence of aging and not infection per se. To rule this out, mice of different ages were infected intranasally with M. avium, and the levels of production of IFN-γ were compared. As shown in Table 3, there were no significant differences in the M. avium-specific recall of IFN-γ from mice infected 6 weeks previously at different ages. Importantly, mice infected for 38 weeks that were the same age as their 6-week-infected counterparts demonstrated a significantly reduced capacity to produce IFN-γ. Thus, the decline in IFN-γ production was related to infection and was not simply a product of aging.

Since the deficient immune response of PBMCs from tuberculosis patients has been attributed to sequestration of IFN-γ-producing cells to the site of infection, intracellular cytokine staining was used to distinguish IFN-γ-producing CD4⁺ and CD8⁺ T cells at another anatomical site in the mice, namely the lungs, the major site of infection. Measurement of the production of IFN-γ from in vitro-cultured lung cells as described above for spleens was ruled out, because they produced inordinately low titers of IFN-γ (data not shown). Because cells recovered from the lung comprised over 70% macrophages and other large cells, it is likely that this was due to either a relative shortage of T cells in cultures or suppression by products of activated macrophages (1, 13). Lungs were removed from groups of three mice at different stages of infection and subjected to enzymatic digestion to generate single-cell suspensions (14). Cells were stimulated with immobilized anti-CD3 monoclonal antibody (MAb) (2.5 μg of protein G-purified antibody 145-2C11 per ml) in the presence of 2 μM monensin (Pharmingen) for 6 h, and then stained with either phycoerythrin (PE)-conjugated anti-CD4 (H129.19) or anti-CD8 (53-6.7) MAb, fixed and permeabilized, and stained with fluorescein isothiocyanate (FITC)-conjugated anti-IFN-γ MAb (XMG1.2). Because the majority of lung cells were Fc receptor-bearing macrophages, high levels of nonspecific staining made it difficult to analyze whole populations of lung cells by flow cytometry. Therefore, an analysis gate was set around the lymphocyte population based on forward and side light-scattering properties. In all experiments, cells stained with the appropriate isotype control or unstained cells and cells stained separately with each fluorochrome were included to optimize compensation settings and set gates (6).

The total number of cells in the lung more than doubled during the first 6 weeks of M. avium infection (Fig. 1A). During this time, the percentage of CD4⁺ T cells within the lymphocyte gate rose from 16.5% ± 0.5% to 35.7% ± 0.1% (Fig. 1B). The absolute number of these cells expanded from (41.2 ± 1.3) × 10⁴ to (207.1 ± 1.0) × 10⁴ (Fig. 1C). In contrast, there was only a modest increase in the proportion of CD8⁺ T cells in the lung during the first 6 weeks of infection, although the absolute number of these cells also rose significantly. From that time on, there was a steady decline in the number of CD4⁺ and CD8⁺ T cells, which preceded by many weeks the appearance of substantial tissue damage. Eventually, the lungs of 40-week-infected mice contained comparable T-cell numbers to uninfected controls. By this stage of infection, the lungs contained fewer cells in total than those of healthy animals.

Figure 2 depicts the data from groups of three mice stained for intracellular IFN-γ at different times postinfection. In the lungs of uninfected animals, <1% of CD4⁺ T cells and 3% of CD8⁺ T cells stained positive for IFN-γ (Fig. 2B). The response in the lung peaked around 8 weeks of infection, when 9.0% ± 0.5% of gated cells produced IFN-γ (Fig. 2A), representing approximately 50 × 10⁴ IFN-γ-producing cells (Fig. 2C). Of these, over 90% were CD4⁺ T cells, with 22% of this subset producing IFN-γ (Fig. 2B). This compared to just 7% of CD8⁺ T cells producing IFN-γ at the same stage, this proportion remaining relatively constant throughout infection. After the peak, the number of CD4⁺ producers declined progressively with infection, falling to 5 × 10⁴ by 40 weeks (Fig. 2D). This was partly due to the decline in T-cell numbers; however, there was also a clear twofold reduction in the percentage of CD4⁺ T cells that produced IFN-γ at this stage of infection (Fig. 2B). Statistical comparisons were not possible for this observation, because the low numbers of cells recovered made it necessary to pool the lungs of individual mice. However, similar reductions in the percentage of IFN-γ-producing CD4⁺ T cells were observed in all five experiments undertaken, confirming that there was a genuine decline in IFN-γ production even at the central site of infection.

These changes in IFN-γ production in the lungs were generally similar to those previously reported in the spleen (6). This indicates that the final decline in IFN-γ production, previously reported among spleen cells, was not due to sequestration of specifically reactive T cells to the lung, an explanation often put forward to explain the deficient IFN-γ production by PBMCs from tuberculosis patients (2, 5). One marked difference in IFN-γ production between lung and spleen was the role of CD8⁺ T cells. Although in the spleen, CD8⁺ cells accounted for 23% of IFN-γ-producing cells, in the lung, they only amounted to 10%. This may account for the lack of an effect of CD8 depletion on the course of infection (3, 14). Indeed, we have preliminary evidence that, although CD4⁺ IFN-γ-producing cells are antigen specific, the CD8⁺ cells may be bystanders.

In conclusion, the previously described loss of IFN-γ-producing T cells from the spleen 10 months after M. avium infection was nonspecific, but was not due to the relatively advanced age of the mice. Importantly, this loss was not due to sequestration of reactive T cells to the primary site of infection in the lung.

• Open in new tab
• Download powerpoint

FIG. 1.

Changes in the percentage and absolute number of T cells in the lung during infection. Lung cells were pooled from groups of three mice killed at different stages of infection and stained for expression of CD4 and CD8 cell surface markers. An analysis gate was set around the lymphocyte population based on forward and side scatter. The total number of lung cells (A) and the percentage (B) or absolute number (C) of cells expressing CD4 or CD8 are shown. The data are representative of five experiments performed.

• Open in new tab
• Download powerpoint

FIG. 2.

Changes in the percentage and absolute number of IFN-γ-producing cells in the lung during infection. Lung cells pooled from groups of three mice sacrificed at different stages of infection were stained for expression of CD4 and CD8 cell surface markers and intracellular IFN-γ. An analysis gate was set around the lymphocyte population based on forward and side scatter. The results represent either the percentage of IFN-γ-producing cells within the lymphocyte population (A) or among CD4⁺ and CD8⁺ T cells (B) or the absolute number of IFN-γ-producing cells within the same populations (C and D). The data are representative of five experiments performed.

• Copyright © 2002 American Society for Microbiology