CD4⁺ and CD8⁺ T Cell Interactions in IFN-γ and IL-4 Responses to Viral Infections: Requirements for IL-2

Mice

Specific pathogen-free male and female C57BL/6NTacfBR mice, and male homozygous MHC class II-deficient C57BL/6TacfBR-[KO]Aβ^b N5 or wild-type MHC class II control C57BL/6TacfBR-[KO]Aβ^b N6 mice, were purchased from Taconic, Germantown, NY. C57BL/6J-B2 m^tm1Unc and C57BL/6J-Il2^tm1Hor mice, obtained from The Jackson Laboratory (Bar Harbor, ME), were maintained at Brown University, Providence, RI, by breeding through homozygous and heterozygous matings, respectively. Mice were typed as +/+, +/−, or −/− for IL-2 gene alleles by PCR amplification. IL-2-mutated mice were used at ≤9 wk of age. Others were used at 5 to 28 wk. Mice were age, sex, and littermate matched for experiments and handled in accordance with institutional guidelines for animal care and use. Infections were initiated i.p. on day 0 with 2 × 10⁴ plaque-forming U of LCMV, Armstrong strain, clone E350. Plasma samples were collected by retroorbital harvest with heparinized Natelson blood collecting tubes and centrifugation.

Preparation of cells and conditioned media samples

Leukocytes were obtained, after teasing apart spleens, by passing through nylon mesh and lysing erythrocytes. Viable cell yields were determined by trypan blue exclusion. To enrich CD8⁺ T cells, murine CD8⁺ T cell subset column kits (R&D Systems, Minneapolis, MN) were used for negative selection per manufacturer recommendations. Recovered CD8⁺ T cells ranged from 55 to 85% purity, with <0.3% CD4⁺ T cells. Conditioned media samples were prepared by incubating 10⁷ splenic leukocytes/ml at 37°C for 24 h in media containing 10% FBS. To quantitate IL-4 or IL-2 production, rat anti-mouse IL-4 receptor (Genzyme, Cambridge, MA) or rat anti-mouse IL-2 receptor p55 (PharMingen, San Diego, CA) mAbs, respectively, were added to block factor consumption. In some cases, up to 2500 U/ml of recombinant human IL-2 (Cetus, Emeryville, CA) were added. Unless otherwise indicated, samples were prepared in the absence of additional stimuli, Ag, or mitogens. Cellfree supernatant fluids were harvested and stored at −20°C until use.

Flow cytometric analysis

Two-color flow cytometric analyses were performed as described (30, 31), using anti-CD4 R-phycoerythrin (PE)-conjugated rat mAb RM4-5, anti-CD8α FITC-conjugated rat mAb 53-6.7, anti-CD8β.2 FITC-conjugated rat mAb 53-5.8, or control Abs lacking specificities for murine determinants (all from PharMingen). Intracellular flow cytometric analyses were conducted using a published procedure (32) with modifications. Cells were resuspended at 10⁶ cells/ml in 10% FBS RPMI, and brefeldin A (Sigma, St. Louis, MO), at 1 mg/ml, was added at 0.01 ml per ml of cells for 2 h at 37°C. Where indicated, cells were stimulated with anti-CD3 for a total of 6 h at 37°C, with brefeldin A added during the last 2 h of incubation, using cluster plates previously adsorbed with purified hamster anti-mouse CD3 mAb 145-2C11 (PharMingen). Cells were collected, washed with PBS, incubated at 10⁶ cells per test for 30 min on ice with biotinylated rat anti-CD4 mAb clone RM4-5 and anti-CD8α FITC-conjugated rat mAb 53-6.7, washed with staining buffer (0.5% BSA and 0.006% NaN₃ in PBS), incubated for 30 min on ice with streptavidin-PerCP (Becton Dickinson, Mountain View, CA), washed, resuspended at 2 × 10⁶ cells/ml in PBS, and fixed with an equal volume of 4% formaldehyde in PBS. After 20 min of incubation at room temperature, cells were PBS washed, resuspended at 10⁷ cells/ml in staining buffer, stored overnight at 4°C, washed again with staining buffer, treated with permeabilization buffer (1% saponin in staining buffer; Sigma), resuspended in 25 μL of permeabilization buffer containing 300 μg/ml rat Ig (Sigma), incubated for 10 min at room temperature, and incubated an additional 20 min after addition of anti-mouse-IFN-γ R-PE-conjugated rat mAb XMG1.2 or anti-mouse-IL-4 R-PE-conjugated rat mAb 11B11 (PharMingen). To prove cytokine specificity of staining, the Abs were preincubated with either recombinant mouse IFN-γ (PharMingen) or recombinant mouse IL-4 (R & D Systems), at 0.25 μg per test, for 20 min before addition to cells. Cells were washed with permeabilization buffer followed by staining buffer, and acquired immediately at the Rhode Island Hospital of Brown University on a FACScan (Becton Dickinson, San Jose), using the LYSYS II software package, or at Brown University on a FACSCalibur, using the CELLQUEST version 1.2.2 or 3.0 software package. Argon laser output was 15 mW at 488 nm. More than 20,000 events were collected for intracellular cytokine staining analyses. Where necessary, flow cytometry standard data files were converted using CONSORT File Exchange version 1.0 and FACSConvert version 1.0 software programs, before analysis using CELLQUEST software packages.

ELISA

For quantitation of IL-2 and IL-4 proteins, sandwich ELISAs were performed using protocols from PharMingen. Immulon 4 96-well ELISA plates (Dynatech, Chantilly, VA) were coated with rat anti-mouse IL-2 mAb clone JES6-1A12 or rat anti-mouse IL-4 mAb clone BVD4-1D11 (PharMingen) in coating buffer (0.1 M NaHCO₃, pH 8.2), PBST washed, incubated with 5% FBS-PBS, washed, and then samples, recombinant mouse IL-2 (PharMingen) or recombinant mouse IL-4 (Life Technologies, Gaithersburg, MD) standards were added in duplicate. After incubation overnight at 4°C, plates were washed, incubated with biotinylated rat anti-mouse IL-2 mAb clone JES6-5H4 or biotinylated rat anti-mouse IL-4 mAb clone BVD6-24G2 (PharMingen) for 45 min at room temperature, washed, incubated for 30 min with avidin-peroxidase, washed, and ABTS ((2,2′-azino-di-[3-ethylbenzthiazoline sulfonate(6)]); Kirkegaard & Perry Laboratories, Gaithersburg, MD) substrate added. Quantitation of IFN-γ by ELISA was performed as described (33) using rat anti-mouse IFN-γ mAb clone XMG1.2 for capture, polyclonal rabbit anti-mouse IFN-γ antisera (gift from Dr. Philip Scott, University of Pennsylvania, Philadelphia, PA) for detection, peroxidase-conjugated donkey anti-rabbit Ig (Jackson ImmunoResearch, West Grove, PA), and recombinant mouse IFN-γ (PharMingen) standards. At multiple intervals following ABTS addition, absorbances were detected at 410 nm using a Dynatech MR-4000 plate reader, or at 405 nm using a SpectraMax 250 (Molecular Devices, Sunnyvale, CA) with the SOFTmax Pro software for data analysis. Limits of detection for IL-2, IL-4, and IFN-γ ELISAs were 5.2 pg/ml (1 U of IL-2 equivalent to 26.5 pg of IL-2), 4.9 to 9.8 pg/ml, and 9.8 to 19.7 pg/ml, respectively.

Virus-specific Ab titers

LCMV-specific plasma Ab titers were quantitated as described (34, 35). Briefly, viral Ag was used to coat plates and bound LCMV-specific Abs were detected by peroxidase-conjugated Abs to mouse Ig. LCMV-specific Ab titer was defined as reciprocal of the dilution of plasma resulting in an absorbance at least twice the SD as samples from uninfected mice.

Enzyme-linked immunospot (ELISPOT) assay

ELISPOT assays to detect IL-4-producing cells were performed essentially as previously described for detection of IFN-γ-producing cells (29), with modifications. Immulon 4 96-well ELISA plates (Dynatech) were coated with purified rat anti-mouse IL-4 mAb clone BVD4-1D11 (PharMingen), with or without 30 μL of 10 μg/ml purified hamster anti-mouse CD3 mAb (PharMingen), in coating buffer (0.1 M Na₂CO₃, 0.1 M NaHCO₃, pH 8.5). Plates were washed with PBST and PBS, incubated with 10% FBS-RPMI for 1 h at room temperature, washed, and cell dilutions in 100-μL vol of 10% FBS-RPMI were added in duplicate wells. After incubation at 37°C for 24 to 30 h, plates were washed, incubated with biotinylated rat anti-mouse IL-4 mAb clone BVD6-24G2 (PharMingen) at 37°C for 1 h, washed again, and streptavidin-alkaline phosphatase (Zymed, S. San Francisco, CA) diluted 1000-fold in 5% FBS-PBST was added at 100 μL per well and incubated at 37°C for 30 min. Plates were washed, substrate solution containing 1 mg/ml 5-bromo-4-chloro-3-indolyl-phosphate disodium salt (Sigma), 0.6% (w/v) SeaPlaque agarose (FMC BioProducts, Rockland, ME) in 0.1 M 2-amino-2-methyl-1-propanol alkaline buffer (Sigma) added, incubated at room temperature overnight, and blue precipitates enumerated under low power magnification (×10).

Statistical analysis

The two-tailed Mann-Whitney U test was performed where indicated using the statistical software package DATA DESK 5.0 (Ithaca, NY). Unless otherwise indicated, mean ± SE are shown. Results identified as having less than (<) means had sample values below the level of detection for the assay. These were calculated as averages of readings at the level of detection for those below detection with those above detection. The results are reported as less than mean averages.

Ex vivo IL-2, IFN-γ, and IL-4 production in normal mice

The kinetics of T cell cytokine production elicited during acute LCMV infections of normal C57BL/6 mice were characterized. To evaluate responses to stimuli received in vivo, splenic leukocytes isolated at various times after infection were cultured in media without additional stimulation for only 24 h. The ex vivo-released IL-2, IFN-γ, and IL-4 proteins were quantitated by ELISA. Addition of Abs to IL-2 or IL-4 receptors to block factor consumption was necessary for optimal detection of IL-2 and IL-4. Consistent with previous work from this laboratory (20, 28), IL-2 induction was evident after infections (Fig. 1⇓A). Cells (10⁷) from day 7-infected mice were induced to produce 47 ± 5 pg of IL-2. Production, however, approached maximal levels of 900 pg with cells from day 14-infected mice and continued to be elevated for 21 days after infection (M. Ruzek and C. Biron, unpublished observations). IFN-γ and IL-4 levels also were induced, with peak levels produced by cells from day 11 or 14 infected mice (Fig. 1⇓, B and C). Cytokine production on a per spleen basis showed similar trends except that, because of maximal cell yields on day 11 after infection, maximal levels of both IFN-γ and IL-4 production occurred at this time. Thus, LCMV infections induce production of IL-2, IFN-γ, and IL-4.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Virus-induced cytokine production in the presence or absence of CD8⁺ T cells. Ex vivo splenic leukocytes from uninfected or LCMV-infected immunocompetent C57BL/6 (A–C) or C57BL/6 β₂m^−/− (D–F) mice were used to condition media for 24 h in the absence of additional stimuli and in the presence (filled squares) or absence (open squares) of anti-cytokine receptor Abs, as described in Materials and Methods. Day 0 samples were prepared from uninfected mice. Other samples were prepared from mice infected on day 0 through days 3, 5, 7, 9, 11, 14, and 21 after infection as indicated. IL-2 (A, D), IFN-γ (B, E), and IL-4 (C, F) production were quantitated by ELISA. Results from multiple experiments were combined. Numbers of mice used for each of the experimental points shown were as follows: A, 2 to 12; B, 3 to 30; C, 3 to 26; D, 3 to 6; E, 1 to 11; and F, 4 to 9. Limits of detection are indicated by dotted lines.

Ex vivo IL-2, IFN-γ, and IL-4 production in the absence of CD8⁺ or CD4⁺ T cells

Mice genetically lacking functional histocompatibility molecules were used to evaluate consequences of CD8⁺ and CD4⁺ T cell deficiencies for cytokine responses. Effects of CD8⁺ T cells were examined in C57BL/6-β₂-microglobulin (β₂m)^−/− mice rendered CD8⁺ T cell deficient as a result of MHC class I expression deficiency. At day 7 after LCMV infection, splenic leukocyte populations from β₂m^−/− mice were induced to produce IL-2 levels comparable to those of cells from immunocompetent C57BL/6 mice, i.e., 80 ± 18 pg per 10⁷ cells (Fig. 1⇑D). However, IL-2 production by cells from β₂m^−/− mice fell to baseline values on days 11 or 14 after infection. Likewise, IFN-γ production by cells from LCMV-infected β₂m^−/− mice was normal on days 5 through 9, but dramatically reduced by day 11 (Fig. 1⇑E). IL-4 production remained undetectable using cells isolated from β₂m^−/− mice at any of the infection times examined (Fig. 1⇑E). Despite decreases in cell yields following infection in β₂m^−/− mice, cytokine production on a per spleen basis showed similar kinetic trends. To examine effects of CD4⁺ T cells, responses were evaluated in Aβ^b−/− mice rendered CD4⁺ T cell deficient as a result of MHC class II expression deficiency. Splenic leukocytes from Aβ^b−/− mice produced no detectable IL-2 and IFN-γ, and reduced levels of IL-4 (Table I⇓). Splenic cell yield increases were induced during infections of Aβ^b−/− mice, and cytokine production on a per spleen basis showed similar kinetic trends. These results indicate that although detectable IL-4 can be induced without CD4⁺ T cells and detectable IL-2 and IFN-γ can be induced without CD8⁺ T cells, optimal induction of all three cytokine responses to viral infections, as measured by ex vivo production, requires in vivo presence of both responses.

Production of virus-specific Ab titers in the absence of CD8⁺ or CD4⁺ T cells

As IL-2, IFN-γ, and IL-4 contribute to induction and modulation of B cell and Ab responses, virus-specific Ab responses were evaluated. Plasma anti-LCMV Ab titers were induced in normal C57BL/6 mice by day 7, continued to increase on day 11, and remained elevated on day 21 after infection (Fig. 2⇓A). In contrast, anti-LCMV Ab titers in CD8⁺ T cell-deficient β₂m^−/− mice were not apparent on day 7, and were induced to approximately 2-log lower levels on days 11 and 21. CD4⁺ T cell-deficient Aβ^b−/− mice also demonstrated impaired induction of anti-LCMV Ab titers on both days 7 and 11 (Fig. 2⇓B). These results show the expected CD4⁺ T cell requirement for optimal T-dependent Ab production. In addition, they indicate that during infections, the in vivo presence of CD8⁺ T cells associated with optimal induction of cytokines promotes conditions leading to high titer detectable Ab production.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Production of virus-specific Ab titers in the absence of CD8⁺ or CD4⁺ T cells. Plasma from uninfected or LCMV-infected C57BL/6-β₂m^−/− (A) or C57BL/6-Aβ^b−/− (B), and appropriate C57BL/6 immunocompetent mice were used to measure LCMV-specific Ab titers by ELISA as described in Materials and Methods. Numbers of mice per group for results shown were 3 to 9 in A and 3 to 16 in B. Limits of detection are indicated by dotted lines; §, mean below limit of detection. Differences between results were significant with β₂m^−/− or Aβ^b−/− ≠ C57BL/6 controls; * p < 0.05, ** p < 0.005.

Intracellular flow cytometric analyses for cytokine expression in T cell subsets

Earlier in situ hybridization and ex vivo production studies from this laboratory have shown that although IL-2 mRNA transcripts can be detected in both CD4⁺ and CD8⁺ T cell subsets from normal LCMV-infected mice, CD4⁺ T cells account for virtually all detected IL-2 released in culture (20, 28). Intracellular flow cytometric analyses for cytokine expression were conducted to ascertain whether CD8⁺ and CD4⁺ T cells contributed directly or indirectly to optimal cytokine production, and to extend characterization of cell subset cytokine expression to IFN-γ and IL-4. Preliminary experiments using published techniques with brefeldin A (32), to allow accumulation of cytokine within synthesizing cells, demonstrated marginally detectable IFN-γ, but not IL-2 or IL-4, proteins in cells from infected but not uninfected mice. To enhance analyses and quantitate numbers of cells in vivo primed to express cytokine, expression was facilitated by short-term cell incubation with immobilized anti-CD3. This step enhanced detection of IFN-γ and allowed modest detection of IL-4, but not IL-2, protein. Thus, studies of intracellular cytokine expression were limited to IFN-γ and IL-4. Specificity of staining was demonstrated by cold competition with soluble unlabeled cytokine.

High levels of cytoplasmic IFN-γ protein were detected on day 7 after LCMV infection (Fig. 3⇓A). Both single positive CD8⁺ and CD4⁺ T cells were shown to express IFN-γ; averages from multiple samples demonstrated that these populations accounted for 80 to 90% of cells expressing the factor (Fig. 3⇓B). Proportionally more CD8⁺ T cells synthesized IFN-γ; in the experiment shown with n = 3, CD8⁺ T cells (Fig. 3⇓C) averaged 54% (±6) IFN-γ positive (Fig. 3⇓D), whereas the CD4⁺ T cells (Fig. 3⇓C) averaged only 27% (±5) positive (Fig. 3⇓E). In multiple experiments after LCMV infection, the percentages of CD8⁺ T cells that were IFN-γ positive ranged from 45 to 84% on day 7, and 39 to 78% on day 11 after infection, whereas those for CD4⁺ T cells ranged from 16 to 52% for day 7, and 12 to 56% for day 11. Because of preferential subset expansion, CD8⁺ T cells accounted for approximately sevenfold more of the IFN-γ-expressing cells in the spleen than did CD4⁺ T cells (Fig. 4⇓, A and B). Although cytoplasmic IL-4 protein expression was detectable under these conditions (Fig. 3⇓, F and G), both proportions of expressing cells and intensity of staining remained low. Thus, conditions and reagents available for use were not sensitive enough for reliable quantitation of IL-4-expressing cells. This was consistent with detection of IFN-γ but not IL-4 production in media without blocking of cytokine consumption using Abs to the receptor, and greater than fourfold excess peak production of IFN-γ compared with IL-4 (Fig. 1⇑C). Analyses with cells isolated on day 11 after infection yielded similar results (Fig. 4⇓, A and B, and data for IL-4 not shown). Thus, although these experiments provided limited information about cellular sources of IL-4, they conclusively demonstrate that high proportions and numbers of CD8⁺ T cells are activated and are the predominant cell type induced to directly synthesize IFN-γ during the infection.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Intracellular flow cytometric analyses for virus-induced cytokine expression in T cell subsets. Splenic leukocytes from day 0 uninfected or day 7 LCMV-infected C57BL/6 mice were treated for 6 h with anti-CD3 Abs and for 2 h with brefeldin A as described in Materials and Methods. Flow cytometric analyses are shown for the following: intracellular IFN-γ of total viable populations from day 0 and day 7 LCMV-infected mice (A), CD4⁺ and CD8⁺ T cells in IFN-γ-positive populations on day 7 (B), total CD4⁺ and CD8⁺ T cells on day 7 (C), and IFN-γ (D, E) or IL-4 (F, G) expression by day 7 CD8⁺CD4⁻ (D, F) or CD4⁺CD8⁻ (E, G) T cells. Control intracellular staining shown in D through G are results with cells from uninfected day 0 mice and with cells from day 7-infected mice in the presence of unlabeled cytokine block. Percentages given are for an individual experiment shown with the background subtracted.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Quantitation of effects of virus-induced IFN-γ expression in T cell subsets. Flow cytometric analyses, as described in Figure 3⇑, were performed on cells from day 0 (uninfected), or day 7 and 11 LCMV-infected C57BL/6 mice. CD8⁺CD4⁻ (filled bars) or CD4⁺CD8⁻ (open bars) cells positive for IFN-γ expression were quantitated, with backgrounds in samples blocked by unlabeled IFN-γ subtracted, as either relative proportions of total splenic leukocytes (A) or absolute numbers recovered per spleen (B). Results are means of two day 0, and seven each of days 7 and 11, mice per group.

IFN-γ production by separated CD8⁺ T cells

The aforementioned experiments, examining cytoplasmic expression with cells from immunocompetent mice and ex vivo cytokine production without additional stimulation with cells from both immunocompetent mice and T cell subset-deficient mice, did not distinguish between ex vivo roles in facilitating reciprocal subset responses and in vivo roles for promoting and/or expanding cytokine-expressing T cells. To evaluate the contribution of other cell populations in facilitating ex vivo cytokine production, highly purified cells activated in an immunocompetent context were prepared and examined for ex vivo cytokine production in the absence of additional stimuli. Under these conditions, CD8⁺ T cell production of IFN-γ was low. In one representative experiment, total cell populations from day 11 LCMV-infected C57BL/6 mice produced 1952 pg, but purified day 11 CD8⁺ T cell populations produced only 138 pg of IFN-γ per 10⁷ cells. Thus, populations of cells with only 20 to 30% CD8⁺ T cells made >10-fold higher levels of IFN-γ than did populations of cells that were ≥90% CD8⁺ T cells. These results mirror the reduced IFN-γ production by cells from CD4⁺ T cell-deficient mice (Table I⇑) and indicate that, although CD8⁺ T cells are the predominant IFN-γ source, they require additional signals for sustained IFN-γ production.

To examine cellular roles for promoting and/or expanding cytokine-expressing T cells in vivo, cytoplasmic IFN-γ expression by populations activated in the absence of either CD8⁺ or CD4⁺ T cell responses was examined. Averages of multiple experiments demonstrated that infection induced changes in proportions and numbers of CD4⁺ T cells in the CD8⁺ T cell-deficient β₂m^−/− mice and of CD8⁺ T cells in the CD4⁺ T cell-deficient Aβ^b−/− mice similar to those changes during infections of immunocompetent mice (Table II⇓). On day 7 after infection, CD4⁺ T cells in immunocompetent and β₂m^−/− mice were, respectively, 9 and 12% with similar total cell numbers per spleen, and CD8⁺ T cells in immunocompetent and Aβ^b−/− mice were significantly increased to, respectively, 28 and 38%, with about fourfold rises in total cell numbers. Cytoplasmic IFN-γ staining (Fig. 5⇓) demonstrated that priming and expansion for cytokine expression in the present T cell subsets was within the range of observed responses of both subsets in immunocompetent mice (Fig. 5⇓). Low proportions of IFN-γ-expressing cells were induced in β₂m^−/− mice (Fig. 5⇓A), and these were predominantly CD4⁺ T cells (Fig. 5⇓, B and C), whereas high proportions of IFN-γ-expressing cells were induced in Aβ^b−/− mice (Fig. 5⇓D), and these were primarily CD8⁺ T cells (Fig. 5⇓, E-G). Remarkably, averages of IFN-γ-positive CD4⁺ T cells were 12% (±0.2) in β₂m^−/− (Fig. 5⇓C) and 9% (±1) in immunocompetent mice, and those of IFN-γ-positive CD8⁺ T cells mice were 18% (±3) (Fig. 5⇓E) in Aβ^b−/− compared with 27% (±2) in immunocompetent mice, used in the same experiments. Moreover, total numbers of virus-induced IFN-γ-expressing CD4⁺ T cells in the CD8 cell-deficient β₂m^−/− mice and IFN-γ-expressing CD8⁺ T cells in the CD4 cell-deficient Aβ^b−/− mice were similar to those in immunocompetent mice (Fig. 6⇓). Thus, inductions and expansions of T cell subsets primed to make IFN-γ were independent of the presence or absence of the reciprocal subset in vivo. Taken together with the ex vivo production studies, these results demonstrate that CD4⁺ T cells are not required for in vivo responses, but do promote continued IFN-γ production by CD8⁺ T cells in culture.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Intracellular flow cytometric analyses for virus-induced IFN-γ expression in CD8⁺ and CD4⁺ T cell-deficient mice. Splenic leukocytes from day 0 uninfected or day 7 LCMV-infected C57BL/6 β₂m^−/− and C57BL/6 Aβ^b−/− mice were treated for 6 h with anti-CD3 Abs and for 2 h with brefeldin A as described in Materials and Methods. The following flow cytometric analyses with cells from β₂m^−/− mice are shown: intracellular IFN-γ of total viable populations from day 0 and 7 LCMV-infected (A), total CD4⁺ and CD8⁺ T cells on day 7 (B), and IFN-γ in CD4⁺CD8⁻ T cells on day 7 (C). The following flow cytometric analyses with cells from Aβ^b−/− mice are shown: intracellular IFN-γ of total viable populations from days 0 and 7 LCMV-infected (D), CD4⁺ and CD8⁺ T cells in IFN-γ-positive populations on day 7 (E), total CD4⁺ and CD8⁺ T cells on day 7 (F), and IFN-γ in CD8⁺CD4⁻ T cells on day 7 (G). Percentages given are results for individual experiments.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Quantitation of in vivo CD8⁺ or CD4⁺ T cell absence effects on virus-induced IFN-γ expression in reciprocal T cell subsets. Flow cytometric analyses were performed on cells handled as described in Figure 5⇑. CD8⁺CD4⁻ (filled bars) or CD4⁺CD8⁻ (open bars) cells positive for IFN-γ expression were quantitated as absolute numbers recovered per spleen. C57BL/6 control (A), C57BL/6 β₂m^−/− (B), and C57BL/6 Aβ^b−/− (C) mice were all analyzed in a single experiment. Results shown are means of the following numbers of mice per group: A, 3 each on days 0, 7, and 11; B, 2 each on days 0, 7, and 11; and C, 3 each on days 0 and 7, and 4 on day 11.

Induction of IFN-γ and IL-4 production in the absence of IL-2

CD4⁺ T cells are major, but do not appear to be the only, producers of IL-2 during LCMV infection (20, 28, 29). IL-2 may contribute to optimal IFN-γ production by supporting the in vivo expansion and priming phases of producing cells and/or by facilitating the maintenance of cytokine production from activated cells. To determine whether IL-2 was required for IFN-γ and IL-4 production during viral infection, mice rendered IL-2 deficient by genetic mutation were LCMV infected, and cytokine production was measured in ex vivo cell-conditioned media. In contrast to cells from control IL-2^+/+ or IL-2^+/− mice, those from IL-2^−/− mice produced little to no detectable IFN-γ (Fig. 7⇓A) or IL-4 (Fig. 7⇓B). IFN-γ and IL-4 production on a per spleen basis showed the same trends. The concurrent deficiencies in cytokine production were accompanied by impaired virus-specific Ab induction, with >2-log less plasma anti-LCMV Ab titers in IL-2^−/− mice by day 9 after infection (Fig. 7⇓C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Virus-induced cytokine production and virus-specific Ab titers in the absence of IL-2. Splenic leukocytes from day 0 uninfected, or LCMV-infected, for 7, 9, and 11 days, IL-2^+/+ (filled bars), IL-2^+/− (striped bars), or IL-2^−/− (open bars) mice were used to condition media for 24 h in the absence of additional stimuli. Shown are ELISA quantitations of IFN-γ (A) in samples prepared in the absence of added Abs, or IL-4 (B) in samples prepared in the presence of anti-IL-4 receptor Abs. The numbers of mice per group for results shown in A and B were as follows: day 0: 6, +/+; 8, +/−; 5, −/−. Day 7: 5, +/+; 6, +/−; 2, −/−. Day 9: 3, +/+; 5, +/−; 3, −/−. Day 11: 7, +/+; 5, +/−; and 4, −/−. Plasma LCMV-specific Ab titers (C) were also quantitated by ELISA. The numbers of mice per group for results shown in C were as follows: day 0: 2, +/+; 3, +/−; 3, −/−. Day 7: 5, +/+; 6, +/−; 2, −/−. Day 9: 3, +/+; 5, +/−; 3, −/−. Limits of detection are indicated by dotted lines; §, mean below limit of detection. Differences between results were significant with IL-2^−/− ≠ IL-2^+/+; * 0.05 <p < 0.10; ** p < 0.05.

Previous work from this laboratory has established that the decreased IFN-γ production in IL-2^−/− mice is due in part, but not entirely, to decreased frequencies of IFN-γ-expressing cells as quantitated by ELISPOT (29). The decreased IL-4 production seen in IL-2^−/− mice at late times after infection could also be attributed to lower frequencies of IL-4-expressing cells. Numbers of IL-4-expressing cells by day 11 after infection reached 19,575 ± 2,987 (n = 7) in IL-2^+/+ mice and 20,783 ± 1,052 (n = 5) in IL-2^+/− mice but were significantly less at 7,049 ± 1,300 (n = 4) in IL-2^−/− mice (p < 0.05). These results indicate that IL-2 functioned in vivo during viral infections to promote expansion of not only IFN-γ-expressing cells, but also IL-4-expressing cells.

Characterization of the IL-2 requirement for CD8⁺ T cell IFN-γ production

Earlier studies from this group have demonstrated the importance of IL-2 for in vivo CD8⁺ T cell expansion (29). To determine whether IL-2 also acts on primed CD8⁺ T cells to promote continued production of IFN-γ, the effects of IL-2 addition were examined during the cytokine production phase in culture. The addition of 2500 U/ml of IL-2 did induce IFN-γ produced into conditioned media by CD8⁺ T cells prepared from uninfected immunocompetent mice (Fig. 8⇓A). The value, however, was lower than that produced by total day 11 populations. In contrast, although highly purified CD8⁺ T cells activated in the immunocompetent mice failed to produce detectable IFN-γ in culture, the addition of 2500 U/ml of IL-2 to these cells restored their ability to produce IFN-γ and resulted in enhanced production to approximately 1300 pg/10⁷ cells (Fig. 8⇓A). Interestingly, although total cells and purified CD8⁺ T cells from day 11 CD4⁺ T cell-deficient Aβ^b−/− mice did not produce IFN-γ on their own in culture, the addition of 2500 U/ml of IL-2 to the purified CD8⁺ T cells resulted in greatly enhanced IFN-γ production to 6000 pg/10⁷ cells (Fig. 8⇓B). IL-2 did not appear to act by improving cell viabilities, as recoveries remained approximately 50% after a 24-h incubation in the presence or absence of added factor. Taken together with the earlier studies, these results indicate that IL-2 acts both to promote the in vivo expansion and priming of IFN-γ-expressing CD8⁺ T cells and to facilitate continued CD8⁺ T cell IFN-γ production.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Characterization of IL-2 requirement for CD8⁺ T cell IFN-γ production. CD8⁺ T cells enriched from uninfected or day 11 LCMV-infected C57BL/6 (A) or Aβ^b−/− (B) mice were used, in the presence or absence of 2500 U/ml of IL-2, to condition media as described in Materials and Methods. IFN-γ production was quantitated by ELISA. Limits of detection are indicated by dotted lines. CD8⁺ T cell preparations ranged from 78 to 86% purity with no detectable CD4⁺ T cell contamination. ND = not done.