IL-6 Produced by Dendritic Cells from Lupus-Prone Mice Inhibits CD4⁺CD25⁺ T Cell Regulatory Functions

Mice

C57BL/6J (B6), BALB/c, and C57BL/6-Tg (TcraTcrb)425Cbn/J (B6.OT-II) mice were obtained from The Jackson Laboratory and subsequently bred at University of Florida (Gainesville, FA). The production of the triple congenic B6.Sle1.Sle2.Sle3 (B6.TC) (15) and B6.Sle1 (3) mice has been described previously. All experiments were conducted according to protocols approved by the University of Florida Institutional Animal Care and Use Committee. Female mice were used at 6–8 or at 2–3 mo old as indicated, with mice aged matched within a same experiment.

Generation of DCs and DC phenotyping

BM single-cell suspensions were cultured in complete RPMI 1640 containing 10% FCS, 10 ng/ml GM-CSF, and 10 ng/ml IL-4 (R&D Systems). Every 2 days, 50% of the culture medium was replaced with fresh medium. After 6 days, the cultured cells were harvested and ascertained by flow cytometry, then CD11c⁺ cells were purified using magnetic microbeads according to the manufacturer’s (Miltenyi Biotec) instructions. To assess activation levels, BM-derived DCs were cultured for 24 h with medium alone, LPS (Sigma-Aldrich), CpG oligodeoxynucleotide 2336 (Coley Pharmaceutical), or TRL7 ligand Imiquimod R837 (InvivoGen), all at 1 μg/ml). For cytokine assays, CD11c⁺ DCs (10⁶ cells/ml) were stimulated with LPS for 13 h (intracellular assay) or with LPS, CpG, or TLR7L (all at 1 μg/ml) for 24 h (supernatant assay). The supernatants were harvested and stored at −80°C until assayed with commercial ELISA kits (BD Systems). To quantify apoptosis, cells were stimulated with LPS (1 μg/ml) for 24 h, stained with annexin V and 7-aminoactinomycin D (BD Biosciences Pharmingen) and examined by flow cytometry.

Flow cytometry

Single-cell suspensions were treated with FcR block (2.4G2) and then directly stained with mAbs to mouse B220 (RA3-6B2), CD3 (145-2C11), CD4 (RM4-5), CD11b (M1/70), CD11c (HL3), CD25 (7D4), CD19 (1D3), CD40 (HM40-3), CD62L (MEL14), CD80 (16-10A1), CD86 (GL1), I-A^b (AF6-120.1), NK1.1 (PK126), TER119, or isotype controls (all from BD Pharmingen). PE-conjugated anti-CD3, CD19, NK1.1, and TER119 Abs were used to gate out CD11c^low T cells, B cells, NK cells, and erythroblasts, respectively. Biotinylated Abs were revealed with streptavidin PerCP-Cy5.5. At least 30,000 cells per sample were analyzed on a BD FACSCalibur. To assay for intracellular cytokine production, GolgiPlug reagent (BD Pharmingen) was added in the last 5 h of stimulation; then cells were stained with the appropriate Abs to gate on the specific population of interest, and with anti-IL-6 Ab from an intracellular staining kit (BD Pharmingen).

T cell isolation and proliferation assays

Splenic CD4⁺ T cells were purified from 2 to 3 mo old B6, B6.TC, BALB/c or B6.OT-II mice by negative selection using the CD4⁺ T cell isolation kit (MACS; Miltenyi Biotec) and the purity was typically 90–95%. Splenic CD4⁺ CD25⁻ effector T (Teff) cells and CD4⁺ CD25⁺ Treg cells were purified from B6 using the CD4⁺CD25⁺ Regularity T cell isolation kit (MACS; Miltenyi Biotec), and the purities were 90–95% and >95%, respectively.

All DCs used in the proliferation and suppression assays were obtained from a 6-day BM culture from 2- to 3-mo-old mice. For allogeneic T cell proliferation assays, BALB/c CD4⁺ T cells (5 × 10⁵) were labeled with CFSE (Invitrogen Life Technologies) and incubated with B6 or TC BM-derived DCs (5 × 10⁴) in a 37°C incubator with 5% CO₂ in 0.5 ml of complete RPMI 1640 (10% FCS, 2 mM l-glutamine, 10 mM HEPES, 100 U/ml penicillin, 100 μg/ml streptomycin, 50 μM 2-ME). For OVA-specific proliferation assays, CFSE-labeled B6.OT-II CD4⁺ T cells (5 × 10⁵) were incubated with B6 or TC BM-derived DCs (5 × 10⁴) prepulsed with 1 μg/ml OVA_323–339 peptide (ISQAVHAAHAEINEAGR, synthesized at the University of Florida Protein Core) for 2 h in complete RPMI 1640. After 72 h of culture, the degree of CD4⁺ T cell division was gauged from the serial dilution of CFSE.

For syngeneic assays, irradiated (30 Gy) B6 or TC BM DCs (1 × 10⁴ or 1 × 10³) were cocultured with CD4⁺ or CFSE-stained CD4⁺CD25⁻ Teff cells (1 × 10⁵), with or without CD4⁺CD25⁺ Treg cells (5 × 10⁴) at a Treg:Teff ratio ranging from 0:1 to 1:8 in wells precoated with anti-CD3 (1 μg/ml; BD Biosciences) for 72 h in complete RPMI 1640. T cell proliferation was measured by adding 1 μCi of [³H]thymidine to each well for the last 16 h of culture. In some assays, anti-IL-6 (MP5-20F3; BD Pharmingen) was added at 5 μg/ml to neutralize IL-6. An irrelevant matched Ab (murine IgG1) was used at the same concentration as control. In other assays, IL-6 (R&D Systems) was added in graded amounts as indicated in the text. IL-2 and IL-6 levels were measured in supernatants after 48 h (IL-2) or 72 h of culture by sandwich ELISA (BD Biosciences).

Statistics

Statistical significance was always evaluated between the values obtained for the lupus-prone mice and B6 controls. After verification of normal distribution with D’Agostino and Pearson omnibus normality tests, we used either one-tailed t tests when two groups were compared (B6 vs (B6 vs B6.TC and B6.Sle1)). Statistical significance was obtained when p was ≤0.05.

DCs are more abundant and less mature in B6.TC mice

The percentage and absolute number of CD11c⁺ cells were significantly higher in all the lymphoid organs of B6.TC mice (Fig. 1⇓, A and B, and Table I⇓). On average, the percentage of CD11c⁺ cells was 2.67 and 3.60 times higher in B6.TC spleen and cervical lymph nodes than in B6, respectively. The difference was even more striking with absolute numbers (3.68- and 4.17-fold higher for the spleen and the lymph node, respectively), due to the lymphoid hyperplasia that is characteristic of lupus-prone mice, including B6.TC (15). Interestingly, the percentage of DCs was significantly decreased in the peritoneal cavity, mostly likely as a consequence of the expansion of B-1a cells in this compartment (18). Both numbers (Fig. 1⇓, C and D, and Table I⇓) and percentages (data not shown) of plasmacytoid (CD11c^lowB220⁺; pDCs) and myeloid (CD11c^highCD11b⁺ B220⁻; MDCs) DCs were 3–4 times higher in the spleen and lymph nodes of B6.TC mice. In the BM, the number of MDCs was also significantly higher in B6.TC, but the number of pDCs was ∼1.5 times lower than in B6. This latter finding suggests an enhanced recruitment of pDCs to the secondary lymphoid organs, which parallels the relative paucity of circulating pDCs and recruitment to inflammatory sites in SLE patients (39). DC accumulation in lymphoid organs was age-dependent, as the differences were not significant at 2–3 mo of age, although individual B6.TC mice already showed higher values (data not shown). A higher percentage of CD11c⁺ cells were also generated from B6.TC BM cultures, and this phenotype was observed in both old (Fig. 1⇓E and Table I⇓) and young mice (data not shown). Moreover, these BM-derived DCs showed a significantly decreased apoptotic death in response to LPS stimulation (Fig. 1⇓F).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Higher numbers and different distribution of DC subsets in TC mice. Percentage (A) and numbers (B) of CD11c⁺LIN⁻ cells in BM, spleen (Sp), cervical lymph node (LN), peritoneal cavity (PC), and thymus (Th). Numbers of CD11c⁺ B220⁺LIN⁻ pDCs (C) and CD11c⁺B220⁻LIN⁻ MDCs (D) in the BM, spleen, and lymph node. E, Percentage of CD11c⁺ cells in BM cells cultured with GM-CSF and IL-4 for 6 days. F, Percentage of annexin V⁺ apoptotic BM culture-derived CD11c⁺ cells after 24 h of stimulation with LPS. Individual B6 or B6.TC 8- to 10-mo-old mice are represented by filled and open symbols, respectively (n = 8–11). Significance between B6 and TC for each culture condition was evaluated with t tests. ∗, p < 0.05; ∗∗, p < 0.001; ∗∗∗, p < 0.001.

The expression of MHC class II (I-A) and costimulatory molecules CD80 and CD86 was significantly lower in immature BM-derived DCs in B6.TC mice (Fig. 2⇓). CD40 expression also tended to be lower, but without reaching significant difference. DCs from both strains responded vigorously to LPS, but the expression levels of these four activation markers remained significantly lower in TC than in B6 DCs. Stimulation through TLR9 maintained a significant difference between B6 and TC DCs for the percentage of CD80⁺ DCs (Fig. 2⇓A) and I-A level of expression (data not shown). Stimulation through TLR7L abrogated the difference for all four activation markers. Interestingly, DCs tended to respond more to TLR stimulation than B6, reaching statistical significance for CD80, CD86, and I-A after TLR7 stimulation and for CD86 after stimulation through all three TLR. The efficacy of CpG and TLR7L at the dose used in this experiment was confirmed by their significant induction of IL-6 production (Fig. 5⇓B). The results obtained with BM-derived DCs were paralleled ex vivo, with CD80 expression significantly decreased on pDCs and MDCs in the spleen and lymph node of B6.TC mice as compared with B6 (p < 0.01, Fig. 2⇓F). Bone marrow pDCs and MDCs showed, however, a significant increase expression of these activation markers as compared with B6 (p < 0.05). Interestingly, CD80 expression in B6 significantly increased as pDCs and MDCs exit the BM toward the spleen (p < 0.05 for both pDCs and MDCs) and the lymph nodes (p < 0.001 for pDCs and p < 0.05 for MDCs). In contrast, the B6.TC expression level of this activation marker either did not change (pDCs) or decreased (MDCs for the lymph nodes, p < 0.001) when the DCs exited the BM to the secondary lymphoid tissues (Fig. 2⇓D). Similar results were obtained for I-A^b and CD86 (data not shown). As for DC numbers, activation levels were similar between young and old mice for BM-derived DCs in B6.TC mice, but not in ex vivo lymphoid organs, in which no difference was found between young B6 and B6.TC mice (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

DCs are more immature in B6.TC mice. Expression of activation markers in BM-derived TC (○) and B6 (•) cultured for 24 h in medium alone (med), LPS, CpG, or TLR7L of I-A^b. A, CD80; B, CD86; C, CD40; D, I-A. Gates used in A–D are shown on the respective histograms in E. Significance between B6 and TC for each culture condition was evaluated with t tests. ∗, p < 0.05; ∗∗, p < 0.001; ∗∗∗, p < 0.001. E, Representative FACS histograms of I-A^b, CD80, CD86, and CD40 in immature BM-derived CD11c⁺ cells. Black and gray lines represent TC and B6 DCs, respectively. Gray-filled histograms show the isotype controls. F, Means and SEs of CD80 mean fluorescence intensity (mfi) on pDCs (rectangles) and MDCs (circles) in the BM, spleen (Spl), and lymph node (LN). Mice 8–10 mo old, filled symbols for B6 and open symbols for B6.TC, n = 8–11. For each tissue and each DC subset, the mean value was significantly different between B6 and B6.TC at least at p < 0.05 as determined with t tests.

The fact that BM-derived TC DC phenotypes were identical between 2-, 3-, and 6- to 8-mo-old mice, the young age group being anti-dsDNA negative and the old one corresponding to an age at which the great majority of B6.TC mice produce anti-dsDNA Abs (15), indicates that these phenotypes are not a consequence of the disease but precede the overt manifestations of humoral autoimmunity. The manifestation of these phenotypes in vivo is amplified with age, as most of the autoimmune phenotypes described in murine models. Overall, our results showed that B6.TC mice produce more DCs, which are found in greater abundance in lymphoid tissues, and that these DCs have a defect in maturation as compared with B6 DCs.

TC DCs induce a greater proliferative response in CD4⁺ T cells

We compared the effect of TC and B6 DCs on CD4⁺ T cells in three different experimental conditions. In all assays, T cells and DCs were obtained from young 2- to 3-mo-old mice. In an allogeneic mixed lymphocyte reaction, BALB/c CD4⁺ T cells responded more vigorously to TC than B6 DCs (Fig. 3⇓A). In a syngeneic mixed lymphocyte reaction, B6 CD4⁺ T cells proliferated significantly more to anti-CD3 stimulation in the presence of TC DCs than B6 DCs (Fig. 3⇓B). The difference was not due to the genetic disparity between the B6 and B6.TC strains that could induce alloreactivity between TC DCs and B6 T cells, because the same results were obtained using either B6 or TC CD4⁺ T cells (Fig. 3⇓C). Therefore, we used the more abundant B6 mice as the source of T cells for the rest of the experiments. Finally, we tested the stimulatory effects of the TC DCs in an Ag-specific assay. CD4⁺ T cells purified from B6.OT-II mice, which carry a transgenic TCR specific for the OVA_323–339 peptide (40), proliferated significantly more to TC than to B6 DCs pulsed with the OVA peptide (Fig. 3⇓, D and E). These results indicated that TC DCs induce a greater proliferative response in CD4⁺ T cells regardless of the nature of the stimulation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

TC DCs induce a higher proliferation in CD4⁺ T cells. A, CD4⁺-gated CFSE dilution in BALB/c CD4⁺ splenocytes stimulated with TC (right) and B6 (left) DCs for 72 h. Representative plots from four replicates per strain. B, Proliferation of CD4⁺ T cells stimulated with two ratios of B6 or TC DCs to B6 CD4⁺ T cells (1:10 or 1:100). C, Comparison of B6 (left) and TC (right) CD4⁺ T cell proliferation stimulated with B6 or TC DCs. For B and C, irradiated B6 or TC BM-derived DCs were cocultured with splenic CD4⁺ T cells and anti-CD3 for 72 h and CD4⁺ T cell proliferation was measured by [³H]thymidine incorporation in the last 18 h of culture. Means and SE of five mice per group. D, Representative histogram showing CFSE profiles of B6.OT2 CD4⁺ splenocytes after 72 h of stimulation with B6 (gray line) or TC (black line) DCs pulsed with OVA peptide. E, Comparison of the M1 and M2 + M3 CFSE peaks, as defined in D. For B, C, and E, filled bars represent B6, and open bars represent B6.TC. Means and SE of four mice per group. t tests: ∗∗, p < 0.01; ∗∗∗, p < 0.001).

TC BM-derived DCs inhibit CD4⁺ CD25⁺ Treg-mediated suppression

We have previously shown that the Sle1 locus was associated with a reduced number of CD4⁺ CD25⁺ CD62L⁺ Foxp3⁺ Treg (8). The percentage of CD4⁺ T cells that expressed both CD25 and CD62L was significantly decreased in B6.TC spleens as compared with B6 (2.06 ± 0.27 vs 5.27 ± 0.31%, respectively; p < 0.0001), and this reduction was comparable with that we have described in B6.Sle1 mice (8), suggesting that Sle2 and Sle3 do not contribute to this phenotype. To assess whether TC DCs affect Treg functions without the potential confounding factor of B6 and TC Treg disparity, we used the same B6-derived CD4⁺ CD25⁻ (Teff) and Treg in all assays_. As expected, B6 BM-derived DCs supported Treg-mediated suppression of Teff proliferation and IL-2 production (Fig. 4⇓). Confirming the results presented earlier with total CD4⁺ T cells, TC DCs were associated with a greater Teff survival as measured by the percentage of CFSE-positive CD4⁺ cells, and a greater IL-2 production (Fig. 4⇓, B and C, left bars). Teff proliferation did not differ, however, in the presence of B6 or TC Dcs (data not shown). In the presence of Tregs, TC DCs resulted in a decreased inhibitory capability, especially in the lower Treg:Teff ratios (Fig. 4⇓A) and a still significantly higher percentage of live CD4⁺ T cells (Fig. 4⇓B). Finally, the production of IL-2 was reduced to 38.28% in the presence of TC DCs as compared with 21.72% in the presence of B6 DCs (p < 0.03) (Fig. 4⇓C). Overall, these results demonstrate that TC DCs impede the regulatory function of Treg.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

CD4⁺ CD25⁺ T cell-mediated suppression of CD4⁺CD25⁻ T cells is inhibited by TC BM-derived DCs. B6 and TC mice BM-derived DCs were used as APCs in standard suppression assays with the indicated CD4⁺CD25⁺: CD4⁺CD25⁻ B6 T cell ratios. A, Inhibition of CD4⁺CD25⁻ T cell proliferation measured by [³H]thymidine incorporation relative to a 0:1 Treg:Teff ratio (i.e., in the absence of Treg). B, Percentage of CFSE-positive CD4⁺CD25⁻, as a measure of cell survival; C, IL-2 production measured in the culture supernatant. ▪, B6; □, B6.TC DCs. Means and SEs of four mice per group. t tests: ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001). Data representative of three independent experiments performed on 2- to 3-mo-old mice.

IL-6 is a critical factor for TC DC inhibition of Treg-mediated suppression

It has been shown that DC-produced IL-6 inhibits Treg-mediated suppression in a TLR4-dependent manner (38). We assessed whether IL-6 was also responsible for the TC DC effect on Treg cells. Purified BM-marrow derived CD11c⁺ DCs produced significantly more IL-6 in B6.TC than in B6 mice (an average of 2-fold more when exposed to LPS (1.75-fold increase), CpG (1.80-fold increase) and TLR7L (1.60-fold increase) (Fig. 5⇓, A and B). When cocultured with Teff cells, TC DCs produced also significantly more IL-6 than B6 DCs (Fig. 5⇓E, left bars). The addition of Treg cells decreased the amount of IL-6 produced by both TC and B6 DCs, which corresponds to the recently described modulation of immature DC function by Treg (41), but IL-6 levels were still significantly higher for TC DCs (Fig. 5⇓E, right bars). The addition of a neutralizing anti-IL-6 Ab to the suppression assays restored Treg inhibitory effect on T cell proliferation in the presence of TC DCs, but it had no effect in the presence of B6 DCs (Fig. 5⇓C). The effect of anti-IL-6 was dose dependent with a maximum effect reached starting at 2.5 μg/ml (Fig. 5⇓F). Interestingly, blocking IL-6 had no effect on the proliferation of Teff cells alone, in the presence of either B6 or TC DCs (data not shown), although large amounts of IL-6 were produced in these cultures (Fig. 5⇓E). We confirmed that the addition of exogenous IL-6 (0.1 μg/ml) abrogated Treg suppression with B6 DCs as previously reported (38), and this effect was similar with TC DCs (Fig. 5⇓D). Interestingly, higher concentrations of IL-6 promoted a significantly higher proliferation in the presence of TC DCs than B6. The spontaneous production of IL-6 by TC DCs in the presence of T cells (Fig. 5⇓E) may be sufficient to not only inhibit Treg suppressive ability but also induce substantial proliferation. Overall, these results suggest that the high level of IL-6 produced by TC DCs exerts a specific function in overcoming Treg inhibitory functions.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

IL-6 produced by TC DCs contributes to their inhibition of Treg functions. Representative intracellular staining (A) and supernatant measurement (B) of IL-6 production by CD11c⁺ magnetic bead-isolated DCs from B6, or TC BM cultures stimulated with 1 μg/ml LPS, CpG, or TLR7L for 24 h. C, Anti (α)-IL-6 blocking Ab reverses the inhibition of Treg function mediated by TC DCs. Inhibition assays were conducted with 1:2 CD4⁺CD25⁺:CD4⁺CD25⁻ B6 T cell ratio, and results are expressed as percentage of inhibition relative to a Treg:Teff ratio of 0:1. Anti-IL-6 or isotype control (IC, 5 μg/ml each) was added as indicated. D, Proliferation in 1:2 Treg:Teff cultures in the presence of graded amount of IL-6. The results are expressed as percentage proliferation of the 0:1 Treg:Teff ratio. E, IL-6 production by B6 or TC DCs in the inhibition assays performed as described above. The gray bars show the level of detected IL-6 when anti-IL-6 has been added to inhibition assays shown in C. Means and SEs of threemice per group. F, Dose-dependent restoration of Treg inhibitory function by anti-IL-6 with 1:2 (black) or 0:1 (gray) CD4⁺CD25⁺:CD4⁺CD25⁻ B6 T cell ratio in the presence of TC DCs. t tests: ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001). Means and SEs of three mice per group.

Sle1 contributes to TC DC inhibition of Treg-mediated suppression

Sle3-expressing DCs induce hyperactivation and higher proliferation in CD4⁺ T cells (13). Although Sle3-expressing DCs produce a high level of proinflammatory cytokines, their production of IL-6 is not different from that of B6. Because we have shown that IL-6 is a key factor in TC DC inhibition of Treg-mediated suppression, we deduced that Sle3 expression was not sufficient to mediate this phenotype. IL-6 is produced at high levels by Sle1-expressing B cells and macrophages, and is linked to STAT3 and ERK activation (42). We therefore postulated that Sle1 contributed to the TC DCs phenotype. Similarly to TC DCs, a greater percentage (Table I⇑) and number (data not shown) of DCs were found in the lymphoid organs of B6.Sle1 mice, and Sle1 BM-cultures produced significantly more DCs than B6. All parameters measured were, however, intermediate between the B6 and B6.TC values, suggesting that Sle1 is not the only contributor to the increased TC DC number and percentage in lymphoid organs. Contrary to TC DCs, however, Sle1-BM-derived immature DCs expressed a normal to elevated level of activation markers as shown for CD86 (Fig. 6⇓A). The same results were obtained for CD80 and I-A, with and without LPS stimulation (data not shown). Similarly again to TC DCs, Sle1 DCs showed a significantly decreased apoptotic response (Fig. 6⇓B) and an increased production of IL-6 (Fig. 6⇓C) when exposed to LPS. This increased IL-6 production was associated with an inhibition of Treg functions, because B6 Treg cell ability to suppress B6 Teff cell proliferation were similarly inefficient in the presence of either TC or Sle1 DCs (Fig. 6⇓D). Furthermore, neutralizing of IL-6 significantly increased Treg suppressive effect in the presence of Sle1 DCs (Fig. 6⇓D). Overall, although both Sle3 and Sle1 are expressed in TC DCs, it is Sle1 that plays a major role in TC DC inhibition of Treg-mediated suppression through IL-6 production.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

TC DC immunophenotypes partially map to the Sle1 locus. A, Expression of CD86 in immature CD11c⁺ cells derived from 6-day BM cultures. B, Percentage of annexin V⁺ apoptotic BM culture-derived CD11c⁺ cells after 24 h of stimulation with LPS. C, IL-6 production by CD11c⁺ DCs from BM cultures stimulated with 1 μg/ml LPS for 24 h. D, Proliferation in syngeneic cultures containing 1:2 CD4⁺CD25⁺:CD4⁺CD25⁻ B6 T cell ratio with DCs from the strains indicated on the graph in the presence or absence of anti-IL-6. Results are expressed as percentage of inhibition relative to a Treg:Teff ratio of 0:1. Dunnett’s multiple comparison tests with B6 values as controls. ∗, p < 0.05; ∗∗, p < 0.01. All data presented in this figure were obtained with 2- to 3-mo-old mice.