Interleukin-12 primes CD4⁺ T cells for interferon-γ production and protective immunity during Mycobacterium avium infection

Abstract

Interleukin-12 (IL-12) is a crucial cytokine for the generation of a protective immune response against Mycobacterium avium infection. In contrast to infected control mice, IL-12-deficient mice were unable to control bacterial proliferation and their spleen T cells were almost unresponsive in vitro to specific antigens of M. avium. Susceptibility of mice deficient in IL-12 was similar to that of interferon-γ (IFN-γ)-deficient mice. These data indicate a crucial role of IL-12 in the development of a T-cell population able to produce IFN-γ and to mediate protection against M. avium infection. Treatment of M. avium-infected mice with IL-12 induced CD4⁺ T cells with enhanced capacity to produce IFN-γ as well as to confer increased protection against M. avium.

Introduction

Human infection caused by Mycobacterium avium is rare in immunocompetent individuals and when it occurs it is most often manifested as cervical lymphadenitis in children^1,2 or pulmonary disease in patients who usually have predisposing conditions such as chronic lung disease.³ Since the 1980s, with the increase in the number of individuals infected with human immunodeficiency virus (HIV), infections caused by M. avium have become one of the most prevalent opportunistic infections in patients with acquired immune deficiency syndrome (AIDS) in the United States,^4–7 causing considerable morbidity and mortality.^5,6,8,9 HIV-positive individuals become susceptible to M. avium infection when the number of CD4 T cells is below 100 cells per µl^4,10 and dissemination of the infection occurs very quickly, involving multiple organs.⁴ Disseminated mycobacterial infections, namely by M. avium, were also observed in individuals with point mutations in genes encoding the subunits for the receptors of interferon-γ (IFN-γ)^11–14 or interleukin-12 (IL-12),¹⁵ as well as in the IL-12 genes,¹⁶ leading to receptor deficiency or a non-functional cytokine, respectively.

IL-12 is a cytokine composed of two polypeptide chains, p40 and p35, encoded by separate genes, covalently linked to form a biologically active heterodimer, p70.^17,18 Some reports have shown that IL-12 induces proliferation of activated, but not resting, CD4 and CD8 T cells.^19–21 Moreover, IL-12 increases the production of IFN-γ by activated NK cells, CD4 and CD8 αβT cells, as well as γδT cells in response to various stimuli^{17,18,22–25} and is an essential factor for the differentiation of murine naïve T helper cells into type 1 (Th1) cells.^{17,18,25–29} In several murine models of intracellular infection, endogenous IL-12 has been shown to be crucial for the generation of a protective Th1 response during a primary infection.^30–38 Furthermore, treatment of mice with IL-12, either singly during infection^{26,32,37,39,40} or as an adjuvant in a vaccine^40–44 has been shown to protect or prevent infection by these infectious agents.

In previous studies we have demonstrated that IFN-γ is one of the molecules involved in the mechanisms of innate and acquired resistance to M. avium infection.⁴⁵ The administration of recombinant IL-12 to either immunocompetent or immunodeficient severe combined immunodeficient (SCID) or CD4-depleted mice infected with M. avium 2447 accelerated the production of IFN-γ and the protective effect of the administration of recombinant IL-12 was dependent on the production of IFN-γ.⁴⁰ In untreated mice, CD4⁺ T cells, and not CD8⁺ T cells, are the cell type involved in the mechanisms that control proliferation of M. avium.⁴⁵ Given the variety of cell types responding to IL-12, including other T-cell populations as well as natural killer (NK) cells, it is important to assess specifically the role of IL-12 in the induction of immunity mediated by CD4⁺ T cells during M. avium infections.

Materials and methods

Reagents, cytokines and antibodies

Mycobacterial growth media were purchased from Difco (Sparks, MD). Dulbecco's modified Eagle medium (DMEM), RPMI-1640 medium, HEPES buffer, fetal calf serum (FCS), sodium pyruvate, penicillin/streptomycin, l-glutamine and protein G columns were purchased from Gibco Life Technologies (Paisley, UK). Tween-80, oleic acid, 2-mercaptoethanol (2-ME), potassium bicarbonate, bovine serum albumin (BSA), concanavalin A (ConA) and incomplete Freund's adjuvant (IFA) were purchased from Sigma (St. Louis, MO). The IFN-γ-specific immunoglobulin G1 (IgG1) monoclonal antibodies (mAb) were obtained from the hybridomas AN18 (DNAX, CA) and R4-6A2 (American Type Culture Collection; ATCC, Manassas, VA). Hybridomas were grown in ascites in HSD nude mice primed with IFA. Antibodies were purified by affinity chromatography using a protein G–agarose column followed by dialysis against phosphate-buffered saline (PBS). Fluorescein isothiocyanate (FITC)-conjugated rat anti-mouse CD4, phycoerythrin (PE)-conjugated rat anti-mouse CD8 and PE-conjugated rat anti-mouse CD3 were purchased from Pharmingen (San Diego, CA). Recombinant mouse IL-12 (rmIL-12) was kindly supplied by Dr S. Wolf from Genetics Institute (Cambridge, MA).

Mice

Specific pathogen-free BALB/c female mice were purchased from the Gulbenkian Institute for Science (Oeiras, Portugal). IL-12 p40 gene-disrupted (IL-12 p40^–/–) and IFN-γ gene-disrupted (IFN-γ^–/–) mice were purchased from Jackson Laboratories (Bar Harbor, ME) in a C57BL/6 background. As control, heterozygous mice (IL-12 p40^+/–) with a C57BL/6 background or C57BL/6J mice were used. Outbred HSD nude mice were purchased from the Gulbenkian Institute for Science and used to raise ascites from hybridomas. These mice were kept in sterile housing conditions in cages provided with high efficiency particulate air (HEPA) filter-bearing caps. All mice were used at 6–8 weeks of age.

Bacterial infections

Mycobacterium avium strain 2447 (an AIDS isolate obtained from Dr F. Portaels, Institute of Tropical Medicine, Antwerp, Belgium) was grown in Middlebrook 7H9 medium containing 0·04% Tween-80 at 37° until the mid-log phase of growth. Bacteria were harvested by centrifugation and resuspended in a small volume of saline containing 0·04% Tween-80. The suspension was briefly sonicated with a Branson sonifier (15 seconds at 50 W) to disrupt bacterial clumps, diluted and frozen in aliquots at −70° until use. Mice were infected intravenously (i.v.) with 10⁶ colony-forming units (CFU) of M. avium 2447 through a lateral tail vein. At different time-points, mice were killed by cervical dislocation and the organs were collected, homogenized, serially diluted in a 0·04% Tween-80 solution in distilled water and plated onto Middlebrook 7H10 agar medium. The plates were incubated for 1–2 weeks at 37° and the number of CFU was counted. The data are expressed as the mean of the log₁₀ number of CFU recovered per organ ± standard deviation of the mean (for four or five animals).

Treatment of mice with rmIL-12

Treatment with rmIL-12 was performed by administration of 0·4 µg of rmIL-12 per animal intraperitoneally (i.p.) every other day from the beginning of infection and during the whole period of experimental infection. Control mice were injected i.p. with PBS.

Antigen preparations

Antigen preparations were obtained as described elsewhere.⁴⁶ Briefly, M. avium 2447 was grown for 2 weeks in Sauton medium supplemented with 0·5% pyruvate and 0·5% glucose, and the culture supernatants were obtained by centrifugation of the bacteria and filtration through 0·45 µm filters. Such culture supernatants were concentrated 100-fold through ultrafiltration using filters with a molecular mass cut-off of 3000. The resulting preparation was then subjected to 80% ammonium sulphate precipitation followed by extensive dialysis against PBS. This reagent was therefore named culture filtrate protein from M. avium (CFP). Mycobacterium avium envelope proteins (MEP) were obtained from bacteria after sonicating the pellet in PBS containing 0·1% Tween-80, 1 mm MgCl₂ (Merck, Darmstadt, Germany) and 1 mm benzamidine (Sigma) with pulses of 1 min at maximum power, keeping the sample in ice during the procedure. The sonicate was centrifuged to discard intact mycobacteria (30 min at 2700 g) and the supernatant was dialysed against PBS. The suspension was then ultracentrifuged for 2 hr at 150 000 g and the pellet, containing the envelope proteins, was resuspended in PBS.

Stimulation of spleen cells in vitro

Mice were killed on day 15 or day 60 of infection and spleens were collected. Half of the organ was used to prepare single cell suspensions. Spleen cells were depleted of red blood cells by incubation in haemolytic buffer (155 mm NH₄Cl, 10 mm KHCO₃, pH 7·2) for 5–10 min at room temperature and thoroughly washed. Spleen cells were resuspended in RPMI-1640 containing 10% heat-inactivated FCS, HEPES (10 mm), penicillin (100 U/ml), streptomycin (100 µg/ml), l-glutamine (1%), and 2-ME (0·05 mm) (complete RPMI) and cultured in triplicate at a density of 2 × 10⁵ cells per well in duplicate 96-well plates either with no further stimulus or in the presence of ConA (5 µg/ml), CFP (4 µg/ml), or MEP (4 µg/ml) at 37° in a 7% CO₂ atmosphere. Supernatants were harvested from triplicate cultures after 72 hr and stored at −20° until used for measurement of IFN-γ by enzyme-linked immunosorbent assay (ELISA).

Purification and culture of CD4⁺ T cells

Pools of nucleated spleen cells (5 × 10⁶ cells/ml) were incubated in tissue culture Petri dishes at 37° in a 7% CO₂ atmosphere for 3 hr to remove adherent cells. Non-adherent cells were centrifuged, resuspended in PBS containing 0·1% glucose, 1% BSA and 5 mm ethylenediaminetetraacetic acid (EDTA) pH 7·2 at a concentration of 1 × 10⁸ cells per ml, and incubated with microbeads coated with anti-CD4 mAb (clone L3T4; Miltenyi Biotec, Germany) at 4° for 20 min. Cells were washed, resuspended in PBS/glucose/BSA/EDTA buffer (10⁸ cells/500 µl) and CD4 T spleen cells were positively selected by using MiniMACS separation columns (Miltenyi) as described in the instructions from the manufacturer. CD4⁺ T cells [85–94% pure on fluorescence activated cell sorter (FACS) analysis] were washed, counted, and resuspended in complete RPMI at 2 × 10⁶/ml. CD4⁺ T cells were stimulated in 96-well plates with irradiated (3000 rads) nucleated spleen cells as APC (1 × 10⁷ cells/ml) plus ConA (5 µg/ml), CFP (4 µg/ml), or MEP (4 µg/ml) at 37° in 7% CO₂ atmosphere. Supernatants were collected from triplicate cultures after 72 hr and frozen at −20° until the measurement of IFN-γ by ELISA.

Adoptive transfer of T cells

Spleens of non-infected mice or mice that had been infected for 3 weeks with M. avium strain 2447 and treated or not with rmIL-12 were aseptically collected and gently teased in RPMI-1640 medium containing 2% FCS (AT medium). Spleen cells were pooled (from five animals), washed and red blood cells were lysed as described above. Purified CD4⁺ T spleen cells were obtained as described above. Alternatively, T-cell-enriched populations were prepared as described previously.⁴⁰ Cells were washed twice in AT medium, counted and injected i.v. in 0·5 ml of the same medium into recipient mice (four animals per group) that had been irradiated with a Cs source (500 rads/mouse) 24 hr earlier. After 2 hr, recipient mice were challenged with 10⁶ CFU of the same M. avium strain. Mice were killed 30 days later, and viable counts were performed on the spleens and livers as described above. The protection achieved was calculated by subtracting the geometric mean of CFU in the organs of mice receiving immune CD4⁺ or enriched T cells from that in mice receiving non-immune CD4⁺ or enriched T cells. The percentage of CD4⁺ T spleen cells in the final cell suspension of selected CD4⁺ cells varied from 80 to 83% and the number of CD4⁺ T cells infused into recipient animals was approximately 7 × 10⁶ CD4⁺ T cells per mouse.

Flow cytometry

To analyse the CD4⁺ T-cell-enriched spleen cells from uninfected and infected mice used in culture experiments and in adoptive transfer, 10⁶ cells from the CD4⁺ T-cell pools were incubated for 30 min at 4° in a microtitre plate with PE-conjugated rat anti-mouse CD8 or PE-conjugated rat anti-mouse CD3 and FITC-conjugated rat anti-mouse CD4 mAbs. Cells were washed twice with staining medium (PBS containing 0·1% sodium azide and 3% FCS) and resuspended in staining medium containing propidium iodide (PI) to allow exclusion of dead cells. Flow cytometric analysis was performed with a FACScan apparatus (Becton Dickinson, Mountain View, CA) equipped with PCLysisII software. For each sample, 10 000 events were analysed.

Cytokine assays

IFN-γ levels in culture supernatants were quantified by a two-site sandwich ELISA using IFN-γ-specific affinity-purified mAbs (R4-6A2 as capture and biotinylated AN-18 as detecting antibodies) and a standard curve was generated with known amounts of recombinant mouse IFN-γ (Genzyme, CA). The sensitivity of the assay was 80 pg/ml.

Proliferation assay

Cell proliferation was measured by [³H]TdR incorporation. Briefly, proliferative responses were assessed after 48 hr of culture in a humidified atmosphere of 7% CO₂ in air. Cultures were pulsed for 18–20 hr before harvesting with 0·5 µCi [³H]TdR (Amersham, UK) per well, and incorporation of [³H]TdR was measured by liquid scintillation. Results are expressed as mean counts per minute (c.p.m.) of triplicate cultures ± 1 standard deviation.

Statistical analysis

Statistical analysis was performed using the Student's t-test.

Results

Mice deficient in the p40 subunit of IL-12 (IL-12 p40^–/–) and their littermate controls (IL-12 p40^+/–) were infected i.v. with 10⁶ CFU of M. avium 2447 and the bacterial load was quantified at different time-points (Fig. 1a). As shown, IL-12 p40^–/– mice had no ability to control bacterial proliferation whereas IL-12 p40^+/– mice started to restrict mycobacterial growth between 2 and 4 weeks after infection. The susceptibility of the IL-12 p40^–/– mice was also compared to that of IFN-γ^–/– animals. As shown in Fig. 1(b), mycobacterial loads at 90 days of infection were higher in these two mutant strains than in wild-type C57BL/6 mice (P < 0·01 for both organs and for both strains). There was little difference between bacterial counts in the two mutated strains. At 60 days of infection, spleen cells from IL-12 p40^–/– and IL-12 p40^+/– mice infected with M. avium as well as non-infected IL-12 p40^+/– mice were collected and stimulated in vitro with conA or with specific antigens from M. avium. As shown in Fig. 1(c), ConA-induced secretion of IFN-γ by spleen cells from infected IL-12 p40^–/– mice was dramatically reduced when compared to those from heterozygous control mice. Splenocytes from infected IL-12 p40^–/– mice were unresponsive in vitro to specific antigens of M. avium (MEP and CFP) whereas cells from infected controls responded with significant production of IFN-γ. The proliferative response to M. avium-specific antigens or ConA was identical for spleen cells from either infected IL-12 p40^–/– or IL-12 p40^+/– mice (not shown). At 60 days of infection there were no significant differences in spleen cell number between IL-12 p40^–/– mice (1·1 × 10⁸ ± 3·8 × 10⁷ cells) and IL-12 p40^+/– mice (1·3 × 10⁸ ± 1·9 × 10⁷ cells) or in the levels of IFN-γ found in spleen homogenates of those animals (IL-12 p40^–/–, 423·8 ± 84·8 pg/ml; IL-12 p40^+/–, 456·2 ± 45·2 pg/ml).

Figure 1.

Failure of IL-12 p40^–/– mice to control M. avium infection and of their T cells to respond to specific antigens. IL-12 p40^–/– mice and their IL-12 p40^+/– littermates (a) and (c) or wild-type C57BL/6, IL-12 p40^–/– and IFN-γ^–/– mice (b) were infected i.v. with 10⁶ CFU of M. avium 2447. (a) Number of viable bacteria present in the organs of IL-12 p40^–/– (filled symbols) and IL-12 p40^+/– (open symbols) mice. Each time-point represents the geometric mean of the CFU from four or five animals. (b) Number of viable bacteria present in the organs of C57BL/6 (open bars), IL-12 p40^–/– (solid bars), and IFN-γ^–/– (hatched bars) mice at 90 days of infection. Each value represents the geometric mean of the CFU from four or five animals. (c) In vitro production of IFN-γ by spleen cells from uninfected (stippled bars) or infected mice at 60 days of infection (IL12 p40^+/–, open bars; IL12 p40^–/–, solid bars) after stimulation with ConA or specific antigens of M. avium (MEP and CFP) for 72 hr. Values from non-stimulated spleen cells were below the detection limit of the assay and were not represented. Data are presented as the means ± SD of triplicates from individual mice for each group (four mice per group). Statistically significant differences between the two groups of infected mice are labelled as **P < 0·01.

We next assessed the role of IL-12 in the differentiation of CD4⁺ T cells during M. avium infection. BALB/c mice were inoculated intravenously with 10⁶ CFU of M. avium 2447 and were treated every other day with either rmIL-12 (0·4 µg/animal) or PBS i.p. At 15 days of infection total spleen cells from non-infected mice as well as infected mice, treated or not with rmIL-12, were stimulated in vitro with ConA or specific antigens of M. avium 2447 for 72 hr. Spleen cells from infected mice treated with IL-12 produced 3·9 times more IFN-γ when stimulated in vitro with ConA than those from infected control mice (Fig. 2a). However, the production of IFN-γ in response to specific mycobacterial antigens by spleen cells from the former mice was very low when compared with that in response to ConA. Additionally, the production of IFN-γ induced by CFP was very similar for spleen cells from infected mice regardless of the IL-12 treatment. However, flow cytometric analysis showed that the fraction of CD4⁺ (13·5%) and CD8⁺ (6·2%) cells in spleens of infected mice treated with rmIL-12 was lower when compared to those found in spleens of infected control mice (17·8% CD4⁺ cells and 10·7% CD8⁺ cells). Thus, the higher percentage of inflammatory cells, namely macrophages, induced by treatment with rmIL-12⁴⁰ could be interfering negatively with the production of IFN-γ by T cells stimulated with specific mycobacterial antigens. Additionally, the absolute number of responding cells was also lower in the treated group. CD4⁺ cells were purified from the pools of spleen cells from the previous experiment and cultured with irradiated spleen cells plus ConA or CFP for 72 hr and the levels of IFN-γ in the supernatants were measured. As shown in Fig. 2(b), CD4⁺ T cells from spleens of IL-12-treated infected mice produced large amounts of IFN-γ in response to both ConA and CFP. In contrast, CD4⁺ T cells from spleens of infected control mice secreted 4·9 times less IFN-γ in response to ConA and 19 times less in response to CFP. FACS analysis showed that the number of CD4⁺ T cells in culture was similar for all groups of mice (9·1 × 10⁴ cells for non-infected mice; 8·5 × 10⁴ cells for IL-12-treated infected mice; 9·4 × 10⁴ cells for infected control mice).

Figure 2.

Effect of rmIL-12 administration on the ability of spleen cells or purified CD4⁺ T cells from spleens of M. avium infected mice to produce IFN-γ after stimulation in vitro with M. avium-specific antigens. BALB/c mice were infected i.v. with M. avium and treated i.p. with rmIL-12 (hatched bars) or PBS (open bars) every other day during the 15 days of the experimental infection. At 15 days of infection, total spleen cells (a) or purified CD4⁺ T spleen cells (b) obtained from uninfected mice or infected mice were cultured in the presence of ConA (5 µg/ml), CFP (4 µg/ml), or medium alone. Supernatants from stimulated and unstimulated cell cultures were harvested at 72 hr and the amount of IFN-γ was measured by ELISA. IFN-γ levels in culture supernatants of unstimulated cells were below the detection limit of the assay and were not represented. Values are reported as means ± SD of triplicates from four individual mice for each group or pool of CD4⁺ T spleen cells from mice for each group (n = 4). The statistical significance of the differences between rmIL-12 versus PBS treatments is shown as **P < 0·01.

To test the ability of the CD4⁺ T-cell population developed in spleens of IL-12-treated and infected mice to induce protective immunity, a passive transfer of CD4⁺ T-cell-enriched spleen cells from infected animals treated or not treated with rmIL-12 was performed and compared to the effects of the transfer of non-selected T-cell-enriched spleen cells. Donor mice were infected i.v. with 10⁶ CFU of M. avium 2447 for 3 weeks. During the period of infection, half of the animals were given rmIL-12 (0·4 µg/animal) and the other half was treated with PBS. At 21 days post-infection, spleen cells from these immune animals as well as spleen cells from non-immune controls were collected, enriched for CD4⁺ T cells or for whole T cells and infused into irradiated (500 rads) recipient animals. The latter were then challenged with 10⁶ CFU of the same M. avium strain and mycobacterial growth was evaluated 30 days later. As shown in Fig. 3, the CD4⁺ T-cell population from spleens of immune mice treated with rmIL-12 conferred to recipient animals a significant and substantial protection against M. avium infection, in both liver and in spleen (equal to 1 log) contrasting with the poor protection afforded by the CD4⁺ T cells from infected mice that were not treated with IL-12. Protection afforded by the former cells was significantly higher than that conferred by the latter cells (P < 0·001 for both organs). The ability of the CD4⁺ T cells to protect was similar to that of the non-selected whole T-cell populations, suggesting that CD4⁺ T cells are the most important cell population, if not the only one, to protect against M. avium infection upon adoptive transfer. In independent experiments, depletion of CD8⁺ T cells did not affect the ability of immune T cells to protect, whereas protection was completely ablated by CD4⁺ T-cell depletion (data not shown).

Figure 3.

The administration of rmIL-12 enhances the generation of a protective CD4⁺ T-cell population. The log₁₀ protection afforded by M. avium-immune whole T cells (open bars) or purified CD4⁺ T cells (stippled bars) from rmIL-12-treated and PBS-treated mice is shown. Five mice were used per group of donors and four were used per group of recipient animals. The total numbers of CD4⁺ T cells adoptively transferred to the irradiated recipient mice were 6·9 × 10⁶ from uninfected, 6·5 × 10⁶ from PBS-treated infected and 6·9 × 10⁶ from IL-12-treated and infected group. Statistically significant protection (*P < 0·05 and **P < 0·01) was found with all immune cells. No statistically significant differences were found when comparing protection afforded by purified CD4⁺ T cells versus whole T cells.

Discussion

We used IL-12 gene-disrupted mice to show that IL-12 is a major player in protective immunity to M. avium. Similar data were previously obtained with the use of neutralizing mAbs^30,36 and gene-disrupted animals.⁴⁷ Although the absence of IL-12 and IFN-γ led to similar increases in susceptibility at 90 days of infection, we cannot exclude the possibility that IL-12 may induce additional effector molecules, leading to protective immunity. Indeed, we have already suggested that mechanisms of immunity to M. avium exist that involve signals other than IFN-γ.⁴⁸ In any case, the protective activity of any additional IL-12-induced molecule or mechanism depends on the concomitant activity of IFN-γ as administration of recombinant IL-12 to IFN-γ^–/– mice did not protect or was even slightly deleterious.⁴⁰ Also, a role for an inducer cytokine in addition to IL-12 may still be envisaged. We believe that such a putative inducer plays a minor role and therefore has no major redundancy with regards to IL-12.

Splenic T cells from mice deficient in IL-12 had an impaired Th1 recall immune response to M. avium-specific antigens in vitro. According to the well known role of IL-12 on T-cell populations,^{17,19–21,26–29} these results suggest that a lack of endogenous IL-12 in this system leads to an impaired development of M. avium-specific IFN-γ-producing T cells. Given the strict requirement for IFN-γ in the development of protective immunity to M. avium, this defective priming of immune IFN-γ-producing T cells could, on its own, explain the increased susceptibility to infection. However, we cannot exclude that other protective effector functions of the same or distinct T cells may as well be induced by IL-12.

We specifically assessed that IL-12 acts on the CD4⁺ T-cell subset. This cell population had a higher capacity to secrete IFN-γ and conferred higher levels of protection upon adoptive transfer when exogenous recombinant IL-12 was given in vivo during its induction by infection. Thus, IL-12 can specifically promote specific immunity dependent on CD4⁺ T cells in addition to other possible effects on the innate immune system, namely NK cells,⁴⁹ or on bystander responses.⁵⁰ These data are particularly pertinent with regards to the possible use of this cytokine in vaccination protocols.^40,51

In summary, our results highlight the potential use of IL-12 in modulating protective immunity to mycobacteria by acting on the major cell population responsible for mediating immunity against this opportunistic pathogen, the CD4⁺ T lymphocyte.