Expression of IL-10 in Th memory lymphocytes is conditional on IL-12 or IL-4, unless the IL-10 gene is imprinted by GATA-3



In Th1 and Th2 memory lymphocytes, the genes for the cytokines interleukin (IL)-4 and interferon-γ (IFN-γ) are imprinted for expression upon restimulation. This cytokine memory is based on expression of the transcription factors T-bet for IFN-γ, and GATA-3 for IL-4, and epigenetic modification of the cytokine genes. In Th2 cells, expression of the cytokine IL-10 is also induced by GATA-3. Here, we show that this induction is initially not accompanied by epigenetic modification of the IL-10 gene. Only after repeated restimulation of a memory Th2 cell in the presence of IL-4, extensive histone acetylation of the IL-10 gene is detectable. This epigenetic imprinting correlates with the development of a memory for IL-10 in repeatedly restimulated Th2 cells. In Th1 cells, IL-10 expression is induced by IL-12, but the IL-10 gene lacks detectable histone acetylation. Accordingly, IL-10 expression in restimulated memory Th1 cells remains conditional on the presence of IL-12. This finding defines a potential anti-inflammatory role for IL-12 in Th1 recall responses. While in primary Th1 responses IL-12 is required to induce expression of the pro-inflammatory cytokine IFN-γ, in secondary Th1 responses IFN-γ re-expression is independent of IL-12, which still is able to induce expression of the anti-inflammatory cytokine IL-10.


chromatin immunoprecipitation assay


conserved intronic regulatory element


type 1 regulatory


The ability of T helper (Th) memory lymphocytes to maintain a memory for expression of particular cytokines (cytokine memory) is well established for the cytokines interleukin (IL)-4 and interferon-γ (IFN-γ) (reviewed in 1). Memory Th cells have the potential to re-express the memorized cytokine genes upon TCR signaling, independent of the original instructing costimulatory signal 24. Establishment of cytokine memory is accompanied by epigenetic modification of the respective gene loci, such as DNA methylation and histone acetylation 5, 6. Up-regulation of particular transcription factors, such as GATA-3 and T-bet, is critical for the establishment of a memory for expression of IL-4 and IFN-γ, respectively 7, 8. In Th2 cells, GATA-3 binds to a phylogenetically conserved sequence element (conserved intronic regulatory element, CIRE) of the first intron of the IL-4 gene, which is the site of initial Th2-specific demethylation and which is essential for memory expression of IL-4 5.

The cytokine IL-10 is expressed by many cell types. With respect to Th lymphocytes, IL-10 originally had been considered a “Th2” cytokine 9. IL-10 has been reported to support B cell differentiation into antibody-secreting cells and thus humoral immune responses 10. IL-10 limits inflammatory immune responses. IL-10-deficient mice develop spontaneous colitis due to uncontrolled immune reactions to the intestinal bacterial flora 11, and show enhanced susceptibility to parasitic and bacterial infection with increased immunopathology 12, 13. Although originally described as a Th2 cytokine, IL-10 probably is also part of the armament of CD4+CD25+ regulatory T cells 14, 15, and is the effector cytokine of adaptive type 1 regulatory (Tr1) lymphocytes 16, 17. IL-10 is expressed by human Th1 and Th2 clones 18, and murine Th1 and Th2 cells 19. Upon primary activation and differentiation into Th1 cells, individual murine Th cells express IL-2, IFN-γ and IL-10 sequentially 20.

In view of the potential of IL-10 to down-regulate antigen presentation and inhibit immune reactions, its recruitment to the functional memory of memory Th lymphocytes and immediate re-expression upon antigenic challenge would preclude efficient secondary immune responses, and thus probably should be avoided. In accordance, the establishment of a stable memory for IL-10 expression in Th2 lymphocytes requires repeated restimulation of the cells with antigen and the instructive signal IL-4, as opposed to the memory for IL-4 in Th2 cells, which is established already after primary stimulation 21. Here, we show that the retarded establishment of a memory for IL-10 in Th2 lymphocytes is accompanied by an increasing epigenetic modification of the IL-10 gene locus, spanning at least 20 kb, including the IL-10 gene itself and upstream sequences. This epigenetic modification is probably initiated by GATA-3. In Th1 cells, not expressing IL-4 or GATA-3, we show that IL-10 expression can be induced by IL-12, but could not detect any epigenetic modifications of the IL-10 gene coinciding with this induction. Re-expression of IL-10 in individual memory Th1 cells remains conditional on IL-12, avoiding preclusive expression of anti-inflammatory IL-10 in pro-inflammatory Th1 memory responses. In accordance, most IL-10-expressing, TCR-activated memory Th cells isolated ex vivo from murine spleen do not have a memory for IL-10, but re-express IL-10 conditional on costimulation with IL-4 or IL-12.


2.1 Imprinting of memory Th cells for IL-10 expression in vivo

CD4+ T lymphocytes from the spleen of a DO11.10 mouse were polyclonally stimulated with phorbol 12-myristate 13-acetate (PMA)/ionomycin for 3 h. In the experiment shown in Fig. 1A, 4.2% of the T cells then expressed IL-10, 79% of these together with IFN-γ. Of the 7% IFN-γ-expressing cells, 40% coexpressed IL-10. IL-10-expressing memory Th cells were isolated by magnetic cell sorting, after labeling for IL-10 secretion. The sorted IL-10+ and IL-10 cells were cultured for 6 days with irradiated antigen-presenting cells (APC) and the cognate peptide ova327–339 in the presence of blocking antibodies specific for IL-4, IL-10, IFN-γ and IL-12. The cells were then restimulated with PMA/ionomycin and analyzed for IL-10 expression after 3 h, by intracellular cytokine staining (Fig. 1A). About 16% of the originally IL-10+ cells re-expressed IL-10. Less than 1% of the originally IL-10 cells re-expressed IL-10 (∼0.4%). Of the originally IL-10+ cells, 24% expressed IFN-γ, as compared to the 75% at the time of sorting. Among the IL-10 cells, only 1.8% re-expressed IFN-γ.

Figure 1.

IL-10 memory in Th memory lymphocytes ex vivo. (A) CD4+ splenocytes from 1-year-old DO11.10 mice were restimulated with PMA and ionomycin for 3 h. The IL-10+ and IL-10 CD4+ cells were separated using the cytokine secretion assay, stained with CFSE and cultured in vitro for six further days in the presence of irradiated splenocytes derived from BALB/c mice (APC), ova327–339 peptide, rIL-2 and neutralizing antibodies to IL-4, IFN-γ, IL-12 and IL-10. For recall expression, the cells were again restimulated with PMA and ionomycin and stained intracellularly for cytokine expression or repeatedly sorted for IL-10 expression after 6-day expansion and culture under the same condition for a further 6 days. Intracellular cytokine staining was performed following 5-h restimulation with PMA and ionomycin. (B) Analysis of CFSE loss following expansion in vitro of the directly ex vivo sorted IL-10 Th cells (wk 1) shown in (A). (C) Ex vivo isolated IL-10+ and IL-10 CD62LlowCD4+ cells were isolated using the cytokine secretion assay and cultured in vitro in the presence of irradiated APC, ova327–339 peptide, rIL-2 and either under neutral (anti-IL-4, anti-IFN-γ, anti-IL-12 and anti-IL-10 antibodies), IL-4 (rIL-4, anti-IFN-γ, anti-IL-12 and anti-IL-10 antibodies) or IL-12 (rIL-12, anti-IL-4, and anti-IL-10 antibodies) conditions. Shown is the frequency of IL-10-expressing CD4+ cells following PMA and ionomycin restimulation and intracellular cytokine staining.

To determine whether the memory Th cells expressing IL-10 twice represented a subpopulation stably imprinted for IL-10 expression, these cells were again separated from the cells which had not re-expressed IL-10 (Fig. 1A). The IL-10+ and IL-10 Th cells were again expanded for another 6 days, in the presence of blocking antibodies for IL-4, IL-10, IFN-γ and IL-12. After restimulation with PMA/ionomycin, about 40% of the original and the twice IL-10-expressing cells re-expressed IL-10 again. This corresponds to 5–10% of the initially ex vivo isolated IL-10-expressing memory Th cells. Ninety percent of the IL-10-re-expressing cells and 75% of all cells express IFN-γ. Most of the cells that had not re-expressed IL-10 in the second stimulation also did not re-express it in the third stimulation (6%), demonstrating that those cells had not been stably imprinted for IL-10 expression. No differences in proliferation of cells re-expressing IL-10 or not upon restimulation were observed (Fig. 1B), excluding the alternative explanation that cells not imprinted for IL-10 re-expression had outgrown cells imprinted for IL-10 re-expression.

IL-10 re-expression in Th1 cells remains dependent on IL-12

In the absence of IL-4, IL-10, IL-12 and IFN-γ, more than 90% of the IL-10+ cells isolated after restimulation ex vivo did not re-express IL-10 in later restimulations. The presence of IL-12 during in vitro expansion of cells originally having expressed IL-10, after restimulation ex vivo, drastically enhanced the frequencies of IL-10-re-expressing Th cells (Fig. 1C) from 8% in the absence of IL-12 to 18% in the presence of IL-12. IL-12 also induced IL-10 expression in IL-10 memory Th cells isolated ex vivo, which, when cultured for 6 days in the absence of IL-12 (and IL-4), showed less than 2% of IL-10-re-expressing cells upon restimulation. IL-12 enhanced this frequency to 14%.

IL-10-expressing cells isolated from cells polarized towards Th1 differentiation in vitro with IL-12, for 3 wk (Supporting Information Fig. 1), were then maintained in culture for another 6 days, in the presence of either IL-12 or IFN-γ, or both. As shown in Fig. 2A, 41 and 39% of the IL-10+ Th1 cells re-expressed IL-10 when cultured in the presence of IL-12, irrespective of the presence or absence of IFN-γ. When IL-12 was neutralized, 90% of the cells expressing IL-10 at the time of sorting failed to re-express IL-10 6 days later. The re-expression of IFN-γ did not depend on the presence of IL-12.

Figure 2.

IL-10 memory in Th1 cells is conditional. (A) Naive CD62LhighCD4+ DO11.10 cells were differentiated with rIL-12, irradiated APC and ova peptide to Th1 cells for 3 wk. The IL-10+ and IL-10 cells were separated and cultured for a further 6 days in the presence of rIL-12, rIL-12 + anti-IFN-γ antibody, or anti-IL-12 antibody. IL-10 re-expression was analyzed by 5-h stimulation with PMA and ionomycin and intracellular cytokine staining. (B) Quantification of GATA-3 expression by quantitative real-time PCR of the indicated cell population following 3-h PMA and ionomycin restimulation. (C) IL-10 expression of Th1-derived IL-10+ sorted cells following 6-day culture in neutral, IL-4, or IL-12 conditions (see Fig. 1).

IL-10 expression of Th1 cells can be induced by IL-4

The presence of IL-4 during the in vitro expansion of originally IL-10-expressing cells, isolated after ex vivo reactivation, enhanced the frequencies of cells re-expressing IL-10 upon restimulation by PMA/ionomycin 6 days later, from 8%, when cultured in the absence of IL-4 (and IL-12), to 33%. In the memory Th cells originally not expressing IL-10 after ex vivo challenge, restimulation resulted in the induction of IL-10-expressing cells from less than 2% in the absence of IL-4 and IL-12 to 20% in the presence of IL-4 (Fig. 1C). When IL-10-expressing Th1 cells, generated in vitro by stimulation with antigen and IL-12 for 6 days, were isolated and exposed to IL-4 for another 6 days, 48% of them re-expressed IL-10, as compared to 6% in the absence of both IL-4 and IL-12, and 30% in the presence of IL-12 (Fig. 2B).

The efficient induction of IL-10 expression in Th1 cells by IL-4 raises the question whether this induction is primarily dependent on STAT signaling, i.e. STAT6 of IL-4 replacing STAT4 of IL-12, or whether it is dependent on GATA-3, the Th2 transcription factor, which has been shown to be able to induce IL-10 expression in Th1 cells 7. In IL-10-expressing Th1 cells generated in and isolated from 3-wk-old polarized in vitro cultures, GATA-3 levels were not up-regulated when compared to unsorted 1- or 3-wk-old Th1 or 3-wk-old Th1-derived IL-10 cells, and 10–20 times lower than in Th cells polarized towards Th2 differentiation with IL-4 for 1 wk. The presence of IL-4 in the recultivation of 3-wk-old Th1 cells induced the expression of GATA-3 to levels comparable to Th2 cells (Fig. 2C).

The IL-10 gene shows epigenetic imprinting in Th2 but not Th1 cells

Epigenetic modification of the chromatin may be involved in the molecular establishment of a memory for re-expression of a given cytokine gene 22, 23. Methylation of cytosines or acetylation, methylation or phosphorylation of histones can be passed on from one generation to the next and contains information on the accessibility of a gene for transcription 24.

Here, we analyzed the epigenetic state of the IL-10 gene by chromatin immunoprecipitation, with an antibody recognizing acetylated residues of the N-terminal tail of core histone H4. We focused on regions of the gene with high phylogenetic conservation (Fig. 3A). Th1 cells polarized for 3 wk with IL-12 and restimulated with PMA/ionomycin display little acetylation of histone H4 of regions located 9.9 and 0.36 kb upstream and 1.8 and 6.1 kb downstream of the transcriptional start, as shown in Fig. 3B. This acetylation is comparable to the H4 acetylation of the CIRE 5 of the IL-4 gene in the same Th1 cells. One-week-old Th2 cells, when restimulated with PMA/ionomycin, showed a twofold increase of histone H4 acetylation at the regions –9.9, –0.36 and +1.8 kb, but not at +6.1 kb of the IL-10 gene. This acetylation of H4 at the IL-10 locus is maintained in unstimulated, 3-wk-old Th2 cells. After restimulation, these 3-wk-old Th2 cells increase H4 acetylation another twofold. Analysis of histone H3 lysine 4 trimethylation of the proximal promoter (–0.36kb) showed that in 3-wk-old Th2 cells also this modification could be detected following restimulation with PMA and ionomycin (Supporting Information Fig. 2).

Figure 3.

Histone H4 acetylation of the IL-10 gene in Th2 and Th1 cells. (A) Comparison of the genomic sequences of the murine and the human IL-10 locus. The height of the peaks indicates the degree of sequence conservation. Non-coding sequences with a conservation degree of 80% or more are shaded in grey, coding sequences in black. Non-filled areas indicate regions with conservation above 50%. The regions analyzed for histone H4 acetylation are as indicated. (B, C) Naive CD62LlowCD4+ DO11.10 cells were differentiated in the presence of IL-4 for 1 or 3 wk to Th2 cells or to Th1 cells for 3 wk with rIL-12. Either in the resting state or following restimulation with PMA and ionomycin for 3 h (B), or following separation of 1- or 3-wk-old Th2 cells into IL-10+ and IL-10 cells (C), the cells were used for chromatin immunoprecipitation with anti-acetylated histone H4 antibody. The immunoprecipitated DNA was quantified by real-time PCR using primers specific for the indicated region. The precipitated DNA was normalized to the amount of input DNA. By chromatin immunoprecipitation, the acetylation state of histone H4 at the indicated region was analyzed and quantified by real-time PCR.

Acetylation of H4 of the IL-10 gene locus extended as far as 20 kb upstream of the transcriptional start. There, the degree of histone acetylation was 3–5-fold higher in 3-wk-old Th2 cells as compared to 1-wk-old Th2 cells (Fig. 3C). Between IL-10-expressing and -non-expressing cells there was no significant and consistent difference in the degree of acetylation, at both time points. Again, the region of +6.1 kb showed the lowest degree of histone acetylation. In the 1- or 3-wk-old Th2 cells, the IFN-γ promoter showed no histone acetylation.

The IL-10 gene contains at least five potential GATA-3 binding sites in regions of phylogenetic conservation (Fig. 4A). Of these sites, only site 2 binds GATA-3 efficiently, as shown by chromatin immunoprecipitation of 1-wk-old Th2 cells expressing IL-10 (Fig. 4B–D). In 1-wk-old Th1 cells, GATA-3 binding could not be detected at that site. Separation of the Th2 cells into IL-10-secreting and -non-secreting cells led to the enrichment of GATA-3-bound DNA of site 2 with the IL-10-expressing cells (Fig. 4D). IL-10-expressing cells also contained about twofold more GATA-3 per cell when compared to the non-expressing cells (Fig. 5B), and both fractions contained significantly higher levels of GATA-3 than Th1 cells (compare to Fig. 2C). The isolated IL-10-expressing cells were cultured for another 6 days in the presence of anti-IL-4 antibody and the IL-10-expressing cells remaining were again separated from the cells which had not re-expressed IL-10 (Fig. 5A). These cells were then again analyzed for expression of GATA-3 by quantitative real-time PCR. Fig. 5B shows that cells which had maintained their IL-10 expression despite the presence of anti-IL-4 antibody had up-regulated GATA-3 expression compared to the first week (threefold). The 2-wk IL-10-non-expressing cells that had been sorted out as IL-10+ cells after the first week did not show up-regulation to GATA-3. Thus, an increase in expression of GATA-3 and an increased binding of GATA-3 to the promoter of IL-10 is a hallmark of IL-10-expressing Th2 cells, and distinguishes them from IL-10-expressing Th1 cells.

Figure 4.

GATA-3 binds to the IL-10 promoter of IL-10-expressing Th2 cells. (A) The conserved putative GATA-3 binding sites in the IL-10 gene. (B) One-week-old Th2 cells were restimulated with PMA and ionomycin and subjected to chromatin immunoprecipitation with anti-GATA-3 antibody. The regions analyzed correspond to the regions indicated in (A). (C) The putative GATA-3 binding sites in the region corresponding to region 2. (D) One-week-old Th1 or Th2 cells restimulated for 1 wk with PMA and ionomycin. The Th2 cells were further segregated into IL-10+ and IL-10 cells. A chromatin immunoprecipitation with anti-GATA-3 antibody was performed and the DNA comprising the IL-10 promoter (–360) was quantified. n.d. = not detectable.

Figure 5.

GATA-3 expression in IL-10-producing Th2 cells. (A) One-week-old Th2 cells were segregated into IL-10-expressing and -non-expressing cells. The IL-10-expressing cells were cultured for 6 days in the presence of APC, ova327–339 peptide, rIL-2 and neutralizing antibodies to IL-4, IFN-γ, IL-12 and IL-10. After restimulation with PMA and ionomycin, the IL-10-expressing cells were again separated from the IL-10-non-expressing cells. (B) Quantification of GATA-3 expression levels by quantitative real-time PCR of the IL-10-expressing and -non-expressing Th cells sorted from the 1-wk-old Th2 culture and of IL-10-expressing and -non-expressing Th cells sorted after 1 wk of culture of IL-10+ Th cells as indicated and shown in (A).


IL-10 is a potent regulator of immune reactivity, mostly by virtue of its ability to inhibit antigen presentation (reviewed in 25). Although initially described to be part of the cytokine repertoire of Th2 cells suppressing Th1 responses 26, expression by Th1 cells and reciprocal suppression of Th2 responses has been demonstrated 18. We show here that, when generated in vitro, Th2 cells establish a stable IL-10 memory for re-expression of IL-10 only when restimulated repeatedly in the presence of IL-4. Establishment of a stable IL-10 memory in Th2 cells is accompanied by elevated expression levels of the transcription factor GATA-3, binding of GATA-3 to the IL-10 promoter and increased acetylation of the core histone H4 of the IL-10 locus. IL-10 expression in Th1 cells is induced by IL-12. However, IL-10 re-expression remains dependent on the continued presence of IL-12, and no histone acetylation of the IL-10 locus is observed.

Th lymphocytes expressing IL-10 upon restimulation ex vivo were readily detectable among splenic Th cells of aged mice. In these cells, IL-10 expression was closely linked to the expression of IFN-γ, resembling the cytokine expression pattern of Tr1 cells 27. Of these cells, 70–80% failed to maintain IL-10 expression in vitro, unless they were cultured in the presence of IL-12 or IL-4. Our data demonstrate that unlike the memory for the effector cytokines IL-4 and IFN-γ, the re-expression of IL-10 remains conditional on costimulation by IL-4 or IL-12. In Th2 cells the IL-10 memory is eventually stabilized by GATA-3 and probably also by epigenetic modification of the IL-10 locus. The vast majority of IL-10-expressing cells ex vivo does not have a stable memory for IL-10 re-expression. This may reflect the requirement to exclude the regulatory cytokine IL-10 from the cytokine memory of Th cells, in order not to preclude secondary immune responses. It also points to an anti-inflammatory role of IL-12 in secondary immune responses (Fig. 6). In view of the rapid and IL-12-independent re-expression of IFN-γ by memory Th1 cells, IL-12 is then mainly required to induce IL-10 expression of Th1 cells, to shut off antigen presentation and end the immune reaction.

Figure 6.

The pro- and anti-inflammatory roles of IL-12.

For human T lymphocytes, IL-12 has been implicated in the regulation of IL-10 expression in vitro2831. Other induction signals of IL-10 expression in vitro include repeated stimulation of naive CD4+ Th lymphocytes in the presence of IL-10 32, costimulation through CD2 33 or CD46 34, 35, stimulation with immature DC 36, 37, costimulation with type 1 interferons 3840 or sequential stimulation with IL-12 and IL-4 17. Which of the signals operate in vivo remains unclear. Apart from this, it is not clear whether or not IL-10 is part of the functional memory of Th1, Tr1 or Th2 memory cells. This question is of considerable practical interest since it impacts on the immunotherapeutic use of IL-10-expressing cells.

Here, we show that in Th2 but not in Th1 memory cells, the IL-10 gene locus is epigenetically modified. Acetylation of histone H4 stretches over more than 20 kb and includes a functional GATA-3 binding site in the promoter of IL-10. GATA-3 is up-regulated in cells with a memory for IL-10. Histone modifications have been linked to memory for the cytokines IL-4 and IFN-γ (5, 41, 42 and reviewed in 43). Histone acetylation has not only been associated with transcriptional activation 44 but is also regarded as an epigenetic determinant of long-term transcriptional competence 45, 46. For the IL-4 gene, epigenetic modifications and GATA-3 binding to a critical regulatory element have been shown to be essential for IL-4 memory 5, and GATA-3 is sufficient to induce this chromatin remodeling 1, 7. Conditional deletion of GATA-3 abrogates IL-10 induction and expression in Th2 cells 47, 48. Shoemaker et al. have shown that GATA-3 is involved in the remodeling of the IL-10 locus (49 and Supporting Information Fig. 3). We show here that this remodeling of the IL-10 locus correlates with the IL-10 memory of Th2 cells in that it is much more pronounced in 3-wk-old than in 1-wk-old Th2 cells. Interestingly, Th2 cells expressing and not expressing IL-10 showed roughly equal histone acetylation of the IL-10 locus, but IL-10-expressing cells had about twice as much GATA-3. Whether or not this had an impact on the expression of IL-10 in the restimulation remains unclear. The binary decision of an individual Th cell with a memory for IL-10 to re-express IL-10, likewise for IL-4 50, could be a stochastic process regulated by TCR-dependent recruitment of GATA-3 to the IL-10 promoter, in close interaction with NFAT 51, 52 or AP-1 53, 54. The molecular basis of the differential induction of a memory for re-expression of IL-4 versus IL-10 in Th2 cells remains elusive. Is it the threshold of GATA-3 concentration, the requirement of an additional transcription factor indirectly up-regulated by GATA-3, for establishment of the IL-10 memory, or the epigenetic disclosure of an as yet unknown regulatory element of the IL-10 gene?

In Th1 memory cells, expression of IL-10 remains dependent on IL-12 or IL-4. The reactivity towards IL-4 ceases upon repeated restimulation 17. Th1 cells restimulated up to three times in the presence of IL-12 will no longer react to IL-4 by expression of IL-10. IL-10 expression of Th1 cells as induced by IL-4 reflects a reprogramming of the cell and includes epigenetic modification of the IL-10 locus 49. In the absence of IL-4, Th1 cells require IL-12 to re-express IL-10, as we show here. IL-12 does not induce detectable epigenetic modification of the IL-10 locus. GATA-3 is not up-regulated in IL-10-expressing Th1 cells. The exclusion of IL-10 from the functional memory of Th1 cells probably is essential in order not to block secondary Th1 responses a priori. A fast memory expression of IL-10 should prevent an effective immune response, as has been shown for mice expressing IL-10 under control of the IL-2 promotor 55. Assenmacher et al. have shown that during the primary response of a Th1 lymphocyte, IL-10 is expressed late following the expression of IL-2 and IFN-γ 20, ensuring a proper time window for the effector phase, before down-regulation of the immune response.

IL-12 seems to play an ambivalent role in the control of inflammatory immune responses in that its action is pro-inflammatory in the primary response while it is anti-inflammatory in secondary responses. During primary antigenic challenge, pathogen-induced IL-12 of APC is essential to induce the sequential expression of IFN-γ and IL-10. In secondary Th1 responses, IFN-γ re-expression of Th1 memory cells is rapid and independent of IL-12. IL-12 expression by APC, induced by this IFN-γ, and signals from the pathogen now could induce the re-expression of IL-10 by the Th1 memory cell, and thus the down-regulation of the secondary response (Fig. 6). It will be interesting to learn whether this hypothesis is confirmed by analysis of in vivo recall immune responses.

Our findings have important consequences for the use of ex vivo isolated and/or expanded IL-10-expressing cells as regulatory cells in autologous or allogeneic cellular immunotherapy of autoimmunity, chronic inflammation, allergy and graft tolerance. IL-10 expression at the site of action is vital for the success of such a therapeutic approach. The present results imply that cells with a stable memory for expression of IL-10 upon restimulation with antigen require imprinting of the IL-10 locus and up-regulation of GATA-3 47, 48. Alternatively, IL-10-inducing conditions have to be provided at the site of requested action, such as the presence of IL-12 or IL-4. Such cells may be preferred tools for targeted immunotherapy since they restrict immunosuppression locally and in time.

Materials and methods


BALB/c and OVA-TCRtg/tg DO11.10 mice (kind gift of Dennis Y. Loh and Kenneth Murphy, Washington University School of Medicine, St. Louis, MO) were bred under specific pathogen-free conditions in our animal facility. The mice were sacrificed by cervical dislocation. All animal experiments were performed in accordance with institutional, state, and federal guidelines.


The following antibodies were either conjugated in-house or purchased as indicated: FITC-conjugated or biotinylated anti-CD4 (GK1.5), biotinylated anti-DO11.10 OVA-TCR (KJ1.26), PE-conjugated anti-CD62L (MEL14; Caltag, Hamburg, Germany), AlexaFluor488-conjugated anti-IL-4 (11B11; Caltag), PE- and allophycocyanin-conjugated anti-IL-10 (JES5–16E3; BD PharMingen, San Diego, CA), Cy5-conjugated anti-IFN-γ (AN18.17.24). For neutralization of IL-4, IFN-γ, IL-12 and IL-10, anti-IL-4 (11B11), anti-IFN-γ (AN18.17.24), anti-IL-12 (C17.8) and anti-IL-10 (JES5–2A5) antibodies were used at 10 µg/mL.

Isolation of IL-10-expressing memory cells ex vivo

Splenic CD4 cells from 1-year-old ex-breeder DO11.10 mice were isolated with the mouse CD4 T cell isolation kit (Miltenyi Biotec, Bergisch-Gladbach, Germany). Naive CD62Lhi cells were depleted using mouse anti-CD62L microbeads (Miltenyi Biotec). The CD4+CD62Llo memory cells were restimulated with 10 ng/mL PMA (Sigma Chemicals, St. Louis, MO) and 1 µg/mL ionomycin (Sigma Chemicals). Viable IL-10-expressing cells were isolated as described 21. Specificity of the IL-10 secretion assay was confirmed by intracellular staining and quantitative PCR (Supporting Information Fig. 4).

Cell culture

Naive CD4+CD62L+ lymphocytes from 6–8-wk-old DO11.10 mice were isolated as described 23. Cell cultures were set up with 3 × 106 cells/mL in complete RPMI 1640 (Invitrogen, Karlsruhe, Germany) supplemented with 10% fetal calf serum (Sigma), 100 U/mL penicillin, 0.1 mg/mL streptomycin, 0.3 mg/mL glutamine (Invitrogen) and 10 µM β-mercaptoethanol. The cognate peptide ova323–339 was added at 0.5 µM. Irradiated (30 Gy) BALB/c splenocytes were used as APC at a ratio of 5 : 1. For Th1 differentiation, cells were stimulated in the presence of recombinant (r) IL-12 (5 ng/mL; R&D Systems, Minneapolis, MN) and anti-IL-4 antibody for 6 days. For Th2 differentiation, cells were stimulated in the presence of IL-4 (100 ng/mL, culture supernatant of HEK293T cells transfected with murine IL-4 cDNA), anti-IL-12 and anti-IFN-γ antibodies.

Sorted IL-10-expressing cells were stained with carboxyfluorescein succinimidyl ester (CFSE) as described in 23 and cultured for 6 days with ova323–339 peptide, fresh APC and rIL-2 (10 ng/mL, R&D Systems) in the presence of IL-4 (100 ng/mL), anti-IL-12, anti-IFN-γ and anti-IL-10 antibodies, or in the presence of rIL-12 (5 ng/mL), anti-IL-4 and anti-IL-10 antibodies, or in the presence of anti-IL-4, anti-IL-12, anti-IFN-γ and anti-IL-10 antibodies.

FACS analysis of intracellular cytokines

T cells were restimulated with PMA/ionomycin, and brefeldin A (5 µg/mL; Sigma) was added after 2 h. After 4 h of restimulation, cells were washed, fixed with 2% formaldehyde in PBS for 15 min at room temperature, and stained for CD4 or OVA-TCR and for intracellular cytokines as described 56 (Supporting Information Fig. 1, 5). Cytometric analysis was performed with a FACSCalibur using CellQuest (BD Bioscience, Heidelberg, Germany) and FlowJo (TreeStar, Ashland, OR) software.

RNA quantification

Total RNA was prepared using NucleoSpin RNA II (Macherey-Nagel, Düren, Germany) and quantified with the NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, USA). For cDNA preparation, the TaqMan reverse transcription kit (Applied Biosystem, Darmstadt, Germany) was used. Total RNA (400 ng), 2 µL 10× reverse transcription buffer, 4.4 µL 25 mM MgCl2, 4 µL 10 mM dNTPs, 0.5 µL random hexamer, 0.5 µL oligo-dT primer, 0.4 µL RNase inhibitor, 0.5 µL reverse transcriptase, and RNase-free H2O to a final volume of 20 µL were incubated at 25°C for 10 min, at 48°C for 40 min and at 95°C for 5 min. The quality of the cDNA was controlled by the ability to amplify the housekeeping gene β-actin. The PCR was performed with the primers β-act up (5′-gTg ggg CgC CCC Agg CAC CA-3′) and β-act do (5′-CTC CTT AAT gTC ACg CAC gAT TTC-3′) at an annealing temperature of 62°C. For the quantification of GATA-3 by quantitative real-time PCR, the following primers were used: GATA-3 S, 5′-CCT ACC ggg TTC ggA TgT AAg T-3′ and GATA-3 R, 5′-AgT TCg CgC Agg ATg TCC-3′. For the normalization of the cDNA, the transcripts for the housekeeping gene ubiquitin ligase UBCH5 was quantified: UBCH5F1, 5′-TCT TgA CAA TTC ATT TCC CAA CAg-3′ and UBCH5BB2, 5′-TCA ggC ACT AAA ggA TCA TCT gg-3′. The cDNA quantification was repeated in three independent experiments, always showing similar results.

Chromatin immunoprecipitation assay

Th1 and Th2 cells were fixed with 1% formaldehyde for 10 min at room temperature. The chromatin was sheared to 200–1000 bp of length by sonication with five pulses of 10 s at 30% power (Bandelin, Berlin, Germany). In the case of the GATA-3 chromatin immunoprecipitation assay (ChIP), the chromatin was precleared for 2 h with protein G microbeads (Miltenyi Biotec) and then incubated with anti-GATA-3 antibody (5 µg/mL, HG3–31; Santa Cruz Biotechnology, Santa Cruz, CA) overnight followed by incubation with protein G microbeads for 1 h. For the acetyl-H4 ChIP, the chromatin was precleared with protein A microbeads (Miltenyi Biotec) and incubated with anti-acetyl-H4 antibody (Upstate Biotechnology, Charlottesville, VA) overnight, followed by incubation with protein A microbeads for 1 h. Washing steps were performed on µ columns (Miltenyi Biotec). Washing was performed sequentially with high-salt, low-salt, LiCl, and TE buffer. The chromatin precipitate was eluted with 1%SDS, 0.1 M NaHCO3. Cross-links were reversed by incubation at 65°C for 4 h in the presence of 0.2 M NaCl, and the DNA was purified with NucleoSpin Extract II (Macherey-Nagel). The amount of immunoprecipitated DNA was determined by real-time PCR with a LightCycler (Roche Applied Science, Mannheim, Germany) using FASTStart SYBR Green Master (Roche Applied Science).The relative amount of DNA was calculated with E(crossing point input – crossing point sample), where E represents the reaction efficiency, determined by serial dilution of DNA. The regions analyzed were selected based on their high degree of phylogenetic conservation and partly based in addition on previous description in the context of regulation of IL-10 in Th cells 51, 54, 57.

The following primers were used: mIL10 –20000 up, 5′-ACA AAA CgC Tag gCT TgA TAC Ag-3′ and mIL10 –20000 do, 5′-gTg AgC AgC Tag gTT CTT gAg AC-3′; mIL10 –9900 up, 5′-ATg TCC TgC CAC TCA ATg AAg-3′ and mIL10 –9900 do, 5′-CTC Tgg AAg TgC CAT TCT gTA Ag-3′, mIL10 –360 up, 5′-gTT CTg gAA Tag CCC ATT TAT CC-3′ and mIL10 –360 do, 5′-CTA AAg AAC Tgg TCg gAA TgA AC-3′; mIL10 +1800 up, 5′-gCA Tag CCT TCC TgT TAT TTg Tg-3′ and mIL10 +1800 do, 5′-AAg AgA TgA ATg TTT TCC TgT gC-3′; mIL10 +6100 up, 5′-ggA TAA ggg gAA ATA ATg AgC Tg-3′ and mIL10 +6100 do, 5′-CCT TCC gAT ggT TAC TTT AAC Tg-3′; IFNγ promoter up, 5′-ATg gTT CAA gTC TgC ACC CAT AgC-3′ and IFNγ promoter do, 5′-CTC ATA CCC ACA TgT ggC TAA ggC-3′. The chromatin immunoprecipitation of the various cell populations was repeated in three independent experiments, always showing the same result.

In silico genomic DNA analysis

The genomic sequences for the IL-10 locus of Mus musculus and Homo sapiens were obtained from UCSC Genome Bioinformatics ( The exact location of the sequences used is chr1:133852257–133909380 for the murine IL-10 locus and chr1:204013647–204069425 for the human locus. The sequences were screened for evolutionarily conserved regions by comparison of the genomic sequences using the Vista tool at 58. To identify putative binding sites for DNA binding factors, the conserved DNA sequences were submitted to Matinspector at 59.


This work was supported by the Deutsche Forschungsgemeinschaft (SFB 421).


  1. 1


  2. 2


  3. 3


  4. 4


  5. 5


  6. 6