Effects of IL-6 on the gene expression profile of CD4⁺ T cells early during activation. (A) FACS-sorted CD4⁺ T cells were activated with anti-CD3 and anti-CD28 in the absence or presence of IL-6 for 16 h. The gene expression profile from total RNA was examined by Affymetrix GeneChip analysis. In the volcano plot, open circles represent probe sets for which expression is changed more than twofold up or down after IL-6 treatment, whereas dots represent all other probe sets. The number of genes exceeding the threshold was the largest among the 10 datasets obtained by permutation of sample labels, yielding statistical significance of differential expression at P = 0.1 (the minimum value) and a false discovery rate <19% (the minimum being 10%). The blue line encloses 99% of the probe sets under the null hypothesis, obtained by permutation of sample labels. The IL-21 gene is marked with an arrow. (B) Heat maps representing the relative expression of genes among IL-6–treated and control samples. Genes were sorted within each group based on their fold change (top, highest fold induction; bottom, highest fold down-regulation). Expression statistics for each gene were centered, scaled, and mapped to a color scale. Red represents relatively low expression, whereas green represents relatively high expression. (C) Cells were purified and activated as in A, and IL-21 expression was measured by quantitative real-time RT-PCR. The means ± SEM of three experiments are shown. (D) CD4⁺ T cells were activated with anti-CD3 and anti-CD28 mAbs in the absence or presence of IL-6 for the indicated periods of time. Cell-culture supernatants were analyzed for IL-21 production by ELISA. The means ± SEM of four experiments are shown.

IL-6 is sufficient and necessary to induce IL-21 expression during the activation of naive and memory CD4⁺ T cells. (A and B) FACS-sorted CD4⁺ T cells were activated with anti-CD3 and anti-CD28 mAbs in the absence or presence of the indicated cytokines (A) or the indicated IL-6 concentrations (B). After 3 d, cell-culture supernatants were analyzed for IL-21 production by ELISA. The means ± SEM of three (A) or four (B) experiments are shown. (C) FACS-purified CD4⁺ T cells were activated with anti-CD3 and anti-CD28 mAbs in the presence of medium alone, LPS, conditioned media (CM) from LPS-stimulated mouse splenocytes, CM plus 10 μg/ml of a neutralizing anti–IL-6 (αIL-6) mAb, or CM from LPS-stimulated IL-6^−/− splenocytes. IL-21 expression was measured by quantitative real-time RT-PCR. One representative experiment out of two is shown (error bars represent SD). (D) C57BL/6 and IL-6^−/− mice were immunized s.c. with 200 μg OVA in CFA. CD4⁺ T cells prepared from spleens and lymph nodes were analyzed 6 d later for the frequency of IL-21–producing cells by ELISPOT. The means ± SD of three immunizations are shown (P < 0.05). (E) FACS-purified CD4⁺ CD25⁻ NK1.1⁻ CD44^low naive cells were activated with anti-CD3 and anti-CD28 mAbs in the absence or presence of IL-6 for the indicated periods of time. IL-21 expression was measured by quantitative real-time RT-PCR. One representative experiment out of two is shown (error bars represent SD). (F) CD4⁺ T cells were isolated from AND TCR transgenic mice, and activated with 5 μM cyt c peptide and APCs in the absence or presence of IL-6 for the indicated periods of time. IL-21 production was examined by ELISA. One representative experiment out of two is shown (error bars represent SD of a triplicate determination). (G) FACS-purified CD4⁺ CD25⁻ NK1.1⁻ CD44^high memory cells were activated with anti-CD3 and anti-CD28 mAbs in the absence or presence of IL-6 for 24 h. Cytokine expression was measured by quantitative real-time RT-PCR. Values are presented as fold induction over naive CD4⁺ T cells activated for 24 h. One representative experiment out of two is shown.

IL-6 requires IL-21 expression in CD4⁺ T cells to induce IgG1 production. (A) Total splenocytes from wild-type and IL-6^−/− mice were stimulated with anti-CD3 mAb for 7 d. IgG1 levels in the cell supernatant were measured by ELISA. Results are representative of three experiments performed in triplicate. (B) B220⁺CD19⁺ B cells and CD4⁺ T cells from the spleens of C57BL/6 mice were analyzed for IL-6Rα and IL-21R expression (bold-line histograms) by flow cytometry. Shaded histograms are control-stained cells. Representative profiles of one mouse out of five analyzed are shown. (C) Total splenocytes of IL-6^−/− mice were activated with anti-CD3 mAb and 100 ng/ml IL-6 or 100 ng/ml IL-21 where indicated. After 7 d, IgG1 levels in the supernatant were determined by ELISA. (D) CD4⁺ T cells from IL-21^−/− mice or wild-type littermates were activated with anti-CD3 mAb in the presence of B cells from IL-6^−/− mice and the indicated cytokines. After 7 d, IgG1 levels in the supernatant were determined by ELISA. (E) CD4⁺ T cells from wild-type mice were activated with anti-CD3 mAb in the presence of B cells from IL-21R^−/− mice or wild-type littermates and the indicated cytokines. After 7 d, IgG1 levels in the supernatant were determined by ELISA. The means ± SEM of a triplicate determination from one representative experiment out of three performed are shown in A and C–E.

IL-6 enhances antibody responses to inactive influenza virus immunization through IL-21. (A) Wild-type C57BL/6 mice were injected s.c. with 10⁸ egg infectious units of inactive PR8 virus alone or together with IL-6. After 30 d, serum was harvested and influenza-specific IgM and IgG1 levels were determined by ELISA. (B) IL-21^−/− mice and wild-type littermates were treated as in A, and influenza-specific IgM and IgG1 levels were determined by ELISA. Experiments were performed twice with five mice per group. Error bars indicate SD.

Role of IL-21 and IL-6 in influenza virus infection. (A) Wild-type and IL-21^−/− mice were infected i.n. with a sublethal dose (0.1 LD₅₀) of active PR8 virus, and after 2 wk, influenza-specific IgG1 and IgG2c levels in serum were determined by ELISA. (B–D) Wild-type and IL-6^−/− mice were infected i.n. with a sublethal dose (0.05 LD₅₀) of active PR8 virus. Animals were monitored daily, and the time to death was recorded as the day after infection. Kaplan-Meier survival curves were generated using Prism 4.03 software (GraphPad Software, Inc.), and the log-rank test was used to determine the significance of differences between the strains (B). Weight was also measured daily in the surviving mice. The percentage of the initial weight is shown (C). Serum levels of virus-specific IgG1 were determined in the surviving mice (six out of six for wild-type mice and two out of six for IL-6^−/− mice) 14 d after infection (D). Data represent groups of five (A) or six (B–D) mice from one experiment out of two performed. Error bars indicate SD.