Lipopolysaccharide-Activated IL-10-Secreting Dendritic Cells Suppress Experimental Autoimmune Uveoretinitis by MHCII-Dependent Activation of CD62L-Expressing Regulatory T Cells

Animals

Inbred 8- to 12-wk-old C57BL/6, GFP, MHCII^−/−, B10BR, and 3A9 TCR-tg mice were provided by the Medical Research Facility at Aberdeen University Medical School. The GFP mice (on C57BL/6) have the enhanced GFP construct integrated into chromosome 14D1, and express enhanced GFP in all of their cells and tissues, as described previously (27). The MHCII^−/− (on C57BL/6) have a null insertion at the invariant chain locus that leads to defective heterodimer α/β-chain assembly and a haplotype-specific (H2^b) failure to present Ag (28, 29, 30). The 3A9 TCR-tg line (B10BR background) expresses TCR that recognizes the HEL epitope 46–61 in the context of I-A^k (31). The procedures adopted conformed to the regulations of the Animal License Act (U.K.) and to the Association for Research in Vision and Ophthalmology statement for The Use of Animals in Ophthalmic and Vision Research.

Isolation and culture of bone marrow DC

DC were prepared and cultured using a method described previously (32). In brief, the tibias and femurs were removed from mice and flushed to obtain bone marrow cells. The red cells were first lysed with ammonium chloride/Tris solution, then depleted of T cells, B cells, and MHCII⁺ cells using rat anti-mouse mAbs (CD4 (clone GK 1.5), CD8 (clone 53-6.7), CD45R, and MHCII (clone P7/7), all from BD Pharmingen, except MHCII, which was from Serotec) and Dynabeads coated with sheep anti-rat IgG (Dynal Biotech). The remaining cells were then cultured at 7.5 × 10⁵/ml (22.5 × 10⁵/well) in 12-well plates with complete RPMI 1640 supplemented with rGM-CSF (R&D Systems) (0.5 μg/100 ml medium) for 6 days, at 37°C and 5% CO₂. The culture was fed daily from day 2 by gently aspirating 75% medium and adding fresh medium described above. On day 6 of culture, the medium was withdrawn and the loosely adherent clusters were flushed, collected in PBS, and depleted of contaminating granulocytes using anti-mouse Gr1 mAb (clone RB6-8C5; BD Pharmingen) and Dynabeads. A single-cell suspension of remaining cells was then cultured immediately in 24-well plates at 10⁶ cells/well. LPS (Sigma-Aldrich; Escherichia coli) was added either immediately (IL-10-DC) or 24 h later (IL-12-DC) at 1 μg/ml. IRBP1–20 peptide/HEL was also loaded at 30 μg/ml to induce an Ag-specific response. After addition of LPS and/or peptide, the cells were cultured overnight before being harvested and washed for phenotypic analysis (flow cytometry) or adoptive transfer.

Flow cytometry

Harvested DC were surface stained for Abs against CD11c, CD11b, CD40, CD86, MHCII, CD4, CD8, TLR4-MD2, and CD14. All Abs were used as color conjugates (BD Pharmingen), and four-color FACS analysis was performed (FACSCalibur; BD Biosciences).

For the calculation of absolute Treg number and characterization of their phenotype, LN were harvested and passed through a 100-μm sieve to obtain a single-cell suspension. The cells were first blocked with serum to prevent nonspecific binding, then surface stained for Abs against CD4, CD25, CD62L, CCR7, and CD103 (BD Pharmingen, except CCR7 from eBioscience), fixed and permeabilized twice with 0.1% saponin, and stained with Ab against Foxp3 (eBioscience), and six-color FACS analysis was performed (FACS LSR; BD Biosciences). A known number of nonfluorochrome-conjugated beads (BD Pharmingen) was added to each sample for the calculation of absolute number of cells. Nonparametric analysis of the results was performed by Mann-Whitney U test (Prism). Value of p < 0.05 was considered statistically significant.

Disease induction and evaluation

EAU was induced by immunizing s.c. with 500 μg of IRBP 1–20 (GPTHLFQPSLVLDMAKVLLD; purity >95%; Sigma Genosys) emulsified in CFA supplemented with 2.5 mg/ml Mycobacterium tuberculosis strain H37Ra (Difco) at the back of hind legs. Pertussis toxin (1 μg) was also administrated i.p. at the time of IRBP immunization.

Mice were sacrificed 28 days post-IRBP immunization by asphyxiation in CO₂, and eyes harvested were immediately fixed in 2.5% phosphate-buffered glutaraldehyde and then embedded in glycol methacrylate resin for sectioning. Sections (3 μm) of each globe were taken at seven different levels. The sections were subsequently stained with H&E. Severity of disease was scored on a scale of 0 (no disease) to 4 (maximum disease) in half-point increments, according to a semiquantitative scoring system described previously (33). Nonparametric analysis of the EAU grades was performed by Mann-Whitney U test (Prism). Value of p < 0.05 was considered statistically significant.

Adoptive transfer experiments

A total of 10⁶ DC was injected s.c. into the nape of the neck of each mouse. For studying the effects of DC on EAU, mice were immunized with IRBP 24 h after DC transfer. For the study of DC-induced Tregs, the LN (submandibular, superficial cervical, deep cervical, and inguinal) were harvested 3 days post-DC transfer.

Cytokine secretion and intracellular signaling analyses

Cytokine profiles in DC culture supernatants were evaluated using the cytometric bead multiplex array system (BD Biosciences) for IL-12 and IL-10, and cytokine ELISAs (R&D Systems) for IL-2 and TGF-β. Phosphorylation of intracellular ERK1 was measured using ERK1 ELISA (R&D Systems).

In vitro proliferation and suppression assays

LN were harvested from 3A9 TCR-tg mice, and single-cell suspensions were made by passing tissues through 70/100-μm-pore cell strainers. CD4⁺CD25⁺/CD25⁻ cells were isolated using the MACS isolation kit (Miltenyi Biotec). IL-10-DC were prepared from non-tg B10BR mice, and then pulsed or not pulsed with 30 μg/ml HEL overnight before use. To test the ability of DC to expand Tregs/Teffs, 10⁴ isolated CD4⁺CD25⁺/CD25⁻ T cells were plated out per well on a 96-well flat-bottom plate. IL-10- DC^+/−HEL were added at various quantities, as appropriate. A total of 100 U/ml IL-2 (PeproTech) and/or 30 μg/ml HEL was added per well for stimulation. [³H]Thymidine was added after 72-h culture and measured 12–16 h later. To test the suppressive ability of in vitro DC-expanded Tregs, 3A9 TCR-tg Tregs were expanded as above at 1:1 DC:Treg ratio for 5 days before harvest. Tregs were then isolated by depletion of the CD11c⁺ cells using magnetic microbeads (Miltenyi Biotec). A total of 10⁴ freshly isolated 3A9 TCR-tg CD4⁺CD25⁻ cells was plated per well on 96-well flat-bottom plate with appropriate number of DC-expanded Tregs, 5 × 10⁴ CD4⁺/CD8⁺ depleted and mitomycin-treated splenocytes, and 30 μg/ml HEL. [³H]Thymidine was added at 72 h, and uptake by proliferating lymphocytes was measured 12–16 h later.

In vivo proliferation of CFSE-labeled lymphocytes

LN were collected from 3A9 TCR-tg mice, and single-cell suspensions were made by passing tissues through 70/100-μm cell strainers, as above. The cells were then labeled with 5 μM CFSE (Molecular Probes) and injected i.v. at 10⁷ cells per animal (day 0). Animals were then transferred with 10⁶ IL-10-DC^+HEL the following day (day 1) and immunized with 500 μg of HEL in CFA supplemented with 2.5 mg/ml mycobacterium s.c. in the back of hind legs. In addition, 1 μg of pertussis toxin was also administered i.p. on day 2. Tissues were harvested on day 5 for FACS analyses of CFSE dilution. To assess the activation of CFSE-labeled Teffs, the harvested tissues were stained with various fluorochrome-conjugated Abs, and the Foxp3⁻ CD4⁺ cells were evaluated for their CD25 expression as a marker of T cell activation.

Differential cytokine secretion by bone marrow-derived DC (BMDC) in response to LPS correlates with TLR4/CD14 expression levels and ERK1 signaling

We have previously demonstrated that LPS-activated BMDC secrete a different set of cytokines by varying the timing of LPS activation of freshly purified immature DC from B10RIII (H2^k) mice (17, 23). In this study, we confirm this result in C57BL/6 (H2^b) mice: freshly purified BMDC are CD11C^+lowCD11b⁺CD4⁻CD8a⁻ expressing low levels of CD40, MHCII, and CD86 (Fig. 1⇓, A and B, and data not shown). Freshly purified DC cultured immediately with LPS for 24 h (IL-10-DC) or cultured for 48 h with LPS added after a delay of 24 h (IL-12-DC) (see legend to Fig. 1⇓ for definition of IL-10 and IL-12 DC) led to equal levels of up-regulation of CD40, MHCII, and CD86, but to differential secretion of IL-10 (immediate addition of LPS) and IL-12 (delayed addition of LPS), as shown previously (Fig. 1⇓C). Interestingly, IL-12-DC secreted higher levels of IL-2 than IL-10-DC or immature DC, and also produced significant levels of TGF-β, unlike IL-10-DC or immature DC, both of which failed to secrete TGF-β (Fig. 1⇓C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Differentially LPS-activated DC are phenotypically similar, but functionally different. A, FACS analyses showing the purity of DC (90% CD11c⁺) and all DC coexpress CD11b, indicating a myeloid cell type. B, FACS histograms showing the expression of cell surface markers on differentially activated DC. Non-LPS-stimulated immature DC after 24-h (dotted green line) or 48-h incubation (dotted red line) both express low levels of CD40, MHCII, CD83, and CD86, whereas IL-10 (solid green line) and IL-12 (solid red line) LPS-activated DC express equally up-regulated levels of all markers. C, Although immature DC failed to secrete any of the cytokine examined, DC activated with LPS immediately after preparation secrete high IL-10 and low IL-12 (IL-10DC), whereas DC activated 24 h after initial purification produce low IL-10 and high IL-12 (IL-12DC). IL-10 and IL-12 DC both secrete IL2, whereas only IL-12 DC secrete TGF-β.

Because LPS activates innate immune cells via TLR4 and CD14, we next investigated the effects on TLR4 and CD14 expression as well as the subsequent intracellular signaling events that are involved in the LPS-induced activation of IL-10-/IL-12-DC, and whether these correlate with their differential cytokine production. Contrary to previous reports in which LPS induces the up-regulation of TLR4 on DC, leading to their maturation and activation (34), our flow cytometric data showed that the expression of TLR4-MD2 complex was significantly down-regulated on both IL-10-DC and IL-12-DC upon LPS stimulation (Fig. 2⇓A). Interestingly, although both IL-10- and IL-12-DC showed an increase in their surface expression of CD14 upon LPS stimulation, the increase was significantly greater in the IL-12-DC compared with the IL-10-DC. In addition, the level of phosphorylated ERK1 (MAPK downstream of TLR4 signaling (35)) was greatly enhanced in IL-12-DC, but not in IL-10-DC. It is possible, therefore, for the full activation of DC and IL-12 production, both the expression of surface CD14 and the up-regulation of ERK1 phosphorylation are required.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

IL-10/IL-12 secretion by differentially LPS-activated DC correlates with the expression of TLR4 and CD14 and their intracellular level of phosphorylated ERK1. A, The level of TLR4 and CD14 expression on different types of DC. The expression of TLR4 is greatly reduced after LPS activation (II (IL-10-DC) (solid green line) and IV (IL-12-DC) (solid red line) compared with their immature counterparts (I (broken green line) and III (broken red line)), whereas the expression of CD14 is greatly increased on IL-12-DC (IV) compared with IL-10-DC (II). B, Only LPS-activated IL-12-DC produce a significant level of phosphorylated ERK1.

The s.c. transfer of LPS-activated IL-10-secreting DC inhibits the induction of EAU in a MHCII-dependent manner

To investigate the role of Ag presentation by IL-10-DC, we prepared BMDC from MHCII^−/− mice, in which a null mutation has been inserted at the invariant chain locus, resulting in defective heterodimer α/β-chain assembly and a haplotype-specific (H2^b) failure to present Ag (28, 29, 30).

We found that LPS-early activated DC from MHCII^−/− mice (MHCII^−/− IL-10-DC) also secreted a large amount of IL-10 and minimal IL-12 (Fig. 3⇓A). However, s.c. transfer of MHCII^−/− IL-10-DC failed to suppress EAU, whereas IRBP peptide-pulsed wild-type (WT) IL-10-DC suppress the disease significantly (Fig. 3⇓, B–D). This suggests that IL-10 secretion by DC alone is not sufficient for the induction of immune tolerance. More importantly, the data show that Ag-specific engagement of the TCR is required to inhibit disease.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

WT IL-10-DC suppress the induction of EAU in a MHCII-dependent manner. Although both early LPS-activated WT and MHCII^−/− DC (WT and MHCII^−/− IL-10-DC) secrete high IL-10, MHCII^−/− IL-10-DC fail to suppress EAU, whereas WT IL-10-DC can significantly suppress the disease. A, MHCII^−/− IL-10-DC produce similar levels of IL-10 and IL-12 compared with WT IL-10-DC. B, 10⁶ WT or MHCII^−/− IL-10-DC were s.c. inoculated in the nape before IRBP immunization in the hindlegs to induce EAU. Each full circle represents one mouse. Data shown are representative of two different experiments showing similar results. C and D, Histopathology of harvested eyes. C, MHCII^−/− IL-10-DC, extensive damage of the entire retina, including photoreceptor destruction, vasculitis, vitritis, and granuloma formation. D, WT IL-10-DC, most of the retina is intact, with minor signs of inflammation such as retinal folds.

Both WT and MHCII^−/− IL-10-DC induce the expansion of CD4⁺CD25⁺Foxp3⁺ Tregs at the cervical dLN

We have previously suggested that the suppressive function of IL-10-producing DC may be due to their ability to expand Tregs (23), because the s.c. injections of IRBP peptide-pulsed IL-10-producing DC into the neck led to an increase in the percentage of CD4⁺CD25⁺GITR⁺ cells in the cervical dLN compared with PBS control, an effect that appeared to peak 48–72 h after DC transfer. In this study, we have investigated this further by assessing the absolute number of CD4⁺CD25⁺Foxp3⁺ Tregs at the dLN after s.c. inoculation of DC.

We have found that inoculation of WT IRBP peptide-pulsed IL-10 DC correlated with an increase in the absolute number of CD4⁺CD25⁺Foxp3⁺ Tregs selectively at the site-specific skin-draining cervical LN (Fig. 4⇓, A and B), and not at other sites, e.g., at the juxtaposed submandibular eye dLN (Fig. 4⇓C) (36). This supports the notion that the suppression of EAU induced by IL-10-DC is associated with expansion of Tregs in the dLN. However, MHCII^−/− IL-10-DC also correlated with an increase in the population of CD4⁺CD25⁺Foxp3⁺ Tregs at the dLN (Fig. 4⇓, A and B), despite the fact that they failed to suppress EAU (Fig. 3⇑). This indicates that Treg expansion alone at this site was insufficient to inhibit EAU, and suggested that there may be a defect in the trafficking of Tregs in MHCII^−/− IL-10-DC-treated mice to the distal site of Teff priming.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

WT or MHCII^−/− IL-10-DC are both associated with an increase in the absolute numbers of CD4⁺CD25⁺Foxp3⁺ Tregs at the DC-dLN. The absolute numbers of CD4⁺CD25⁺Foxp3⁺ Tregs are increased by both WT or MHCII^−/− IL-10-DC at the DC-draining superficial cervical (A) (∗∗, p = 0.0043, 0.0076) and deep cervical (B) (∗∗, p = 0.0043, 0.0011) LN, but not at the non-dLN, i.e., the submandibular (C) (p = 0.409, 0.3496) (see Fig. 5⇓). WT or MHCII^−/− IL-10-DC were adoptively transferred s.c. in the neck 3 days before the harvest of tissues. The absolute numbers of CD4⁺CD25⁺Foxp3⁺ Tregs induced by WT IL-10-DC, MHCII^−/− IL-10-DC, and PBS control were detected using flow cytometry. Data shown are representative of two different experiments showing similar results.

MHCII Ag presentation by DC is required to significantly increase the trafficking of Tregs to the distal inguinal LN

As indicated above, because Treg expansion appeared unaffected by the lack of functional MHCII on the inducing DC, yet Ag presentation through MHCII was necessary for EAU suppression, we speculated that specific Ag may be required to induce homing signals in the expanded Treg population. Our experimental design of IL-10-DC s.c. inoculated at one site (cervical dLN) and Ag immunization at another site (inguinal dLN) allowed us to explore this hypothesis.

We first wished to verify that the s.c. inoculation of IL-10-DC in the nape of the neck was associated with selective trafficking of the inoculated DC to the site-specific (superficial cervical, deep cervical) dLN. IL-10-DC were cultured from congenic GFP tg mice (see Materials and Methods) and inoculated s.c. in the nape, and various lymphoid tissues were harvested on days 1, 3, and 6 after adoptive transfer. We found that there was a progressive accumulation of IL-10-DC in the draining superficial and deep cervical LN over 6 days (Fig. 5⇓, A and B), but no migration of DC to other secondary lymphoid tissues such as the eye dLN (submandibular, Fig. 5⇓C) or the distal inguinal LN (Fig. 5⇓D).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

IL-10 DC selectively track to their dLN. Subcutaneously administered GFP IL-10-DC (10⁶ cells at the nape) selectively traffic to the superficial cervical (A) and deep cervical (B) LN. No GFP DC could be detected at the eye dLN (submandibular (C)) or the distal inguinal LN (D). Tissues were harvested at days 1, 3, and 6 after DC transfer, and GFP cells were detected using flow cytometry.

We next examined whether s.c. inoculation of IL-10-DC expands CD4⁺CD25⁺Foxp3⁺ Tregs at distal non-DC dLN as well as the DC-draining cervical LN (Fig. 5⇑). If this were so, it would indicate that the IL-10-DC-induced population of Tregs at the site-specific dLN had the capacity to traffic to distal LN. Our data showed that s.c. inoculation of WT IL-10-DC is associated with a significant increase in the population of CD4⁺CD25⁺Foxp3⁺ Tregs at distal sites, whereas MHCII^−/− IL-10-DC failed to do so (Fig. 6⇓A). This suggests that presentation of specific Ag is required to permit trafficking of Tregs to distant sites and particularly to the site of T cell activation (the inguinal dLN of the immunization site).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

WT IL-10-DC, but not MHCII^−/− IL-10-DC, correlates with an increase in the population of Tregs at the distal inguinal LN and the expression of CD62L on Tregs at the DC-dLN. A, 10⁶ WT or MHCII^−/− IL-10-DC were inoculated s.c. in the neck 3 days before the harvest of cells from the distal inguinal LN. The absolute numbers of CD4⁺CD25⁺Foxp3⁺ Tregs after s.c. transfer of WT IL-10-DC, MHCII^−/− IL-10-DC, and PBS control were detected using flow cytometry. B, Only WT IL-10-DC is associated with a significant increase in the expression of CD62L on CD4⁺CD25⁺Foxp3⁺ Tregs. The cervical dLN were harvested 3 days after the s.c. injection of 10⁶ WT or MHCII^−/− IL-10-DC in the neck.

This was further supported by data showing that the percentage of CD4⁺CD25⁺Foxp3⁺ Tregs expressing CD62L was significantly increased when IRBP peptide-loaded WT IL-10-DC were s.c. inoculated, compared with PBS controls (p < 0.05), whereas MHCII^−/− IL-10-DC failed to associate with a significant increase in CD62L expression on Tregs (Fig. 6⇑B). Because CD62L is a lymphoid-homing molecule previously shown to be critical for the in vivo suppressive function of Tregs (25, 37, 38, 39, 40, 41), our data would imply that, in addition to the increase in the population of Ag-specific Tregs, MHCII Ag presentation by tolerogenic IL-10-DC appears to mediate the appropriate trafficking of Tregs to distal non-DC dLN via the induction of their CD62L expression.

Presentation of specific Ag is required for the optimal expansion of Ag-specific Tregs, as well as for their Ag-specific suppressive function and CD62L expression

To demonstrate Ag specificity, we turned to the HEL tg mouse model because there is no equivalent TCR-tg model for IRBP peptide. Using the CD4⁺ 3A9 TCR-tg mouse (in which the majority of CD4⁺ T cells express the 3A9 TCR that specifically recognize the HEL epitope 46–61 in the context of I-A^k (31)), we show that the proliferation and activation of Ag-specific Teffs can be regulated by LPS-activated IL-10-DC.

First, we investigated the ability of IL-10-DC to expand Ag-specific T cells in vitro. To study the role of specific Ag, we have used HEL-pulsed IL-10-DC, in the presence or absence of the cognate Ag HEL and/or IL-2. We have isolated the separate populations of Tregs/Teffs from the LN of TCR-tg mice according to their CD25 expression using MACS magnetic columns, and confirmed by flow cytometry that this method gives a good purity of Tregs (∼70% Foxp3⁺) and Teffs (>90% Foxp3⁻) (Fig. 7⇓A). By culturing these separate populations of Tregs/Teffs with different quantities of IL-10-DC (pulsed or not pulsed with HEL), we have found that HEL-pulsed IL-10-DC can induce the proliferation of both Tregs and Teffs (Fig. 7⇓, B and C). Furthermore, the level of proliferation directly correlates with the numbers of DC in culture, and both CD4⁺CD25⁺ (Treg) and CD4⁺CD25⁻ (Teff) populations require the presence of HEL for maximum proliferation. Of particular interest is the strong proliferation of Tregs induced by IL-10-DC and HEL (Fig. 7⇓B) because these cells have previously been shown to be anergic and failed to proliferate unless exogenous IL-2 is added in vitro. However, we have shown above that IL-10-DC secrete significant amounts of IL-2 (Fig. 1⇑C), which probably provides the stimulus for overcoming Treg anergy, allowing them to proliferate in response to specific Ag, i.e., HEL. Interestingly, the combination of both HEL and IL-2 appears to have an additive effect on the amount of Treg proliferation. In contrast, non-HEL-pulsed IL-10-DC incubated with IL-2 failed to induce any significant proliferation of Teffs, although HEL-pulsed IL-10-DC with IL-2 together seems to induce proliferation that is higher than that induced by HEL-pulsed IL-10-DC alone. Note that there appears to be a decline in proliferation at 1:1 DC:Treg ratio compared with 1:2 ratio, which may be explained by excessive TCR stimulation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

IL-10-DC require cognate Ag for the optimal expansion of 3A9 T cells. CD4⁺CD25^high Tregs and CD4⁺CD25⁻ populations were isolated from 3A9 lymphocytes using MACS isolation. A, The expression of Foxp3 in the two cell populations confirm that CD25^high Tregs are mostly Tregs, whereas most CD25⁻ cells are Foxp3⁻. B and C, When cultured with HEL-pulsed or non-Hel-pulsed LPS-activated IL-10-DC at various DC:T cell ratios, the data show that although IL-2 can induce some proliferation of Tregs, as previously documented (B), the presentation of HEL Ag greatly improves the optimal expansion of both populations of Tregs (B) and Teffs (C). Data shown are representation of three experiments showing similar results.

In addition to testing the ability of IL-10-DC to induce the proliferation of Tregs described above, we have also harvested the cocultured Tregs and used flow cytometry to phenotype these cells. Although we found all sets of Tregs maintain similar level of Foxp3 after culture irrespective of the treatment (data not shown), interestingly, only those Tregs that have been expanded in the presence of HEL retained their expression of CD62L, compared with freshly isolated TCR-tg Tregs, or DC or DC plus IL-2 cocultured Tregs (Fig. 8⇓A). This is in good agreement with the data above, which suggested that the presentation of specific Ag via MHCII by DC is required to up-regulate the expression of CD62L (Fig. 6⇑). We propose this to be a key function of tolerizing DC, because the up-regulation of CD62L expression on the expanded Treg population allows them entry to other lymphoid tissues, where they can exert their suppressive function.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Specific Ag (HEL) is required for in vitro expansion of CD4⁺CD25⁺ Tregs that retain their expression of CD62L and their ability to mediate Ag-specific suppression. A, Flow cytometry analyses showing the requirement of specific Ag HEL to retain the expression of CD62L on DC-expanded 3A9 Tregs. B, Suppression assays demonstrating that DC-induced Tregs must be expanded in the presence of HEL to retain its potent Ag-specific suppressive function.

In addition to the ability of DC-expanded Tregs to maintain their expression of Foxp3 and CD62L, it is also important that they retain their ability for Ag-specific suppression. We have therefore isolated the in vitro DC-expanded Tregs by depleting CD11C⁺ cells, then using these Tregs to suppress the proliferation of HEL-induced 3A9 CD25⁻ Teffs. By adding different quantities of Tregs to Teffs, we investigated the suppressive function of HEL-specific Tregs. Our data show that although the presence of DC plus IL-2 alone can expand Tregs (Fig. 7⇑B), this set of expanded Tregs only showed minimal suppressive ability, i.e., slightly more than that of freshly isolated unstimulated Tregs (Fig. 8⇑B). In contrast, both DC plus HEL- and DC plus HEL plus IL-2-expanded Tregs are equally potent at inhibiting the proliferation of 3A9 Teffs upon presentation of the cognate Ag; even at a ratio of 1:64 Treg:Teff, both these expanded Tregs suppress proliferation by 50% (Fig. 8⇑B). These data demonstrate that, although IL-2 plus HEL appears to have an additive effect on the level of Treg proliferation (Fig. 7⇑B), it is the activation by specific Ag, i.e., HEL, that is essential to determine the ability of the expanded Tregs to retain their Ag-specific suppressive function.

IL-10-DC^+HEL administered to one LN drainage site (the cervical nodes) inhibit priming of HEL-specific Teffs induced at a distal LN (the inguinal node)

Finally, we wished to demonstrate this suppressive activity of IL-10-DC in vivo in an Ag-specific model. To achieve this, we first transferred CFSE-labeled 3A9 lymphocytes to non-tg B10BR mice before the administration of DC at the one site of drainage (cervical nodes; see Fig. 5⇑ above), followed by the activation of 3A9 Teffs by HEL/CFA immunization at the draining site of the inguinal LN.

Our data clearly show that the s.c. administration of IL-10-DC at the neck is associated with a strong reduction in both the number and activation of proliferating 3A9 Teffs in vivo (Fig. 9⇓). It is evident that although this effect can be observed in all the LN studied, the effect is far more pronounced in the inguinal LN, i.e., the LN draining the site of HEL/CFA immunization and therefore the site of T cell priming. We believe the data provide compelling evidence that the administration of IL-10-DC^+Ag at one site, i.e., the neck, which we have demonstrated traffic only to the site-specific dLN (cervical) (Fig. 5⇑), can potently reduce the activation of Teffs at a separate site, i.e., Ag-draining inguinal LN. These data also correlate with the up-regulation of CD62L expression on Tregs, and we suggest that CD62L expression, which has previously been demonstrated as a requirement for Treg trafficking (25, 37, 38, 39, 40, 41), is induced in an Ag-specific manner by tolerizing IL-10-DC; thus, these two events are probably causally linked.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9.

The s.c. transfer of IL-10-DC^+HEL at the neck is associated with a strong reduction in both the number and activation of proliferating 3A9 Teffs induced at the distal inguinal LN. CFSE-labeled 3A9-TCR lymphocytes were transferred to non-tg B10BR mice, followed by the s.c. transfer of HEL-pulsed IL-10-DC at the neck and the active immunization of HEL/CFA at the inguinal. Lymphoid tissues were harvested, and CFSE cells were detected by flow cytometry 3 days later; data shown were gated on CD4⁺ T cells. Our data show that CD4⁺ T cell proliferation (CFSE dilution (A)) and activation (CD25 expression (B)) are most pronounced, as expected, at the Ag-draining inguinal LN, and the s.c. transfer of HEL-pulsed IL-10-DC can greatly reduce this effect. C, Mean values of i, the total number of CFSE cells and ii, the percentage of CFSE-proliferating cells expressing CD25 (upper left quadrants of B) at various LN studied from the HEL/CFA-immunized or IL-10-DC^+HEL-treated and HEL/CFA-immunized animals. Data shown are representative of two separate experiments with two to four animals per group.