Interleukins 27 and 6 induce STAT3-mediated T cellproduction of interleukin 10

Jason S Stumhofer1, Jonathan S Silver1, Arian Laurence2, Paige M Porrett3, Tajie H Harris1,Laurence A Turka3, Matthias Ernst4, Christiaan J M Saris5, John J O’Shea2 & Christopher A Hunter1

Interleukin 10 (IL-10) has a prominent function in regulating the balance between protective and pathological T cell responses.

Consistent with that activity, many sources of this cytokine are found in vivo, including from myeloid cells and a variety of T cell

subsets. However, although there are many pathways that regulate innate production of IL-10, the factors that govern its

synthesis by the adaptive response are poorly understood. Here we report that IL-27 and IL-6 induced T helper type 1 and

type 2 cells, as well as T helper cells that produce IL-17, to secrete IL-10. This effect was dependent on the transcription

factors STAT1 and STAT3 for IL-27 and on STAT3 for IL-6. Our studies identify a previously unknown pathway that allows the

immune system to temper inflammatory responses.

Interleukin 10 (IL-10) was initially described as a cytokine associatedwith T helper type 2 (TH2) cells that inhibited the production ofinterferon-g (IFN-g) by TH1 cells1,2. It was subsequently recognizedthat this was an indirect effect and that its ability to temper TH1 cellfunction was due to its capacity to antagonize accessory cell activity.Thus, IL-10 inhibits macrophage production of proinflammatorycytokines such as IL-1, tumor necrosis factor and IL-12 and decreasesthe expression of costimulatory and major histocompatibility complexmolecules required for optimal T cell activity3. Although IL-10 has avariety of biological properties, one of its main functions in vivo is tolimit inflammatory responses, consistent with its inhibitory effects onantigen-presenting cells, a function first noted in initial reportsdemonstrating that IL-10-deficient mice (Il10–/– mice) spontaneouslydevelop inflammatory bowel disease4. Subsequent studies of mousemodels of sepsis, infectious disease and autoimmunity have extendedunderstanding of the function of IL-10 in the regulation of innate andadaptive responses associated with the activities of TH1 and TH2 cellsas well as T helper cells that produce IL-17 (TH-17 cells)3,5,6. In thecontext of infectious disease, there are several examples of how theabsence of IL-10 leads to enhanced resistance to pathogens but alsoresults in the development of an aberrant immune response and severeimmune-mediated disease7,8. Such studies have shown the centralinvolvement of IL-10 in maintaining a balance between protectiveimmunity and the development of pathology.

Given the importance of IL-10 in limiting inflammation, it isperhaps not unexpected that there are many cellular sources of thisimmune modulator; these include macrophages and dendritic cellsstimulated with microbial products. In addition, although IL-10 was

initially characterized as a ‘TH2 cytokine’, it is now recognized that ‘Tr1cells’ (induced regulatory helper T cells that rely on IL-10 fordifferentiation and function)9, CD25+ regulatory T cells (Treg

cells)10,11 and CD25– Treg cells12,13, as well as TH1 cells14,15, also secreteIL-10 in certain conditions. The relative importance of these differentsubsets as sources of IL-10 has been a longstanding issue16,17, but thelink between Treg cells and IL-10 has dominated this area of research.Although there is an established literature on the presence ofIFN-g+IL-10+ helper T cells in a variety of disease settings18,19, severalstudies (for example, infection with Toxoplasma gondii, as wellas a nonhealing model of Leishmania major20,21) have emphasizedthe contribution of IL-10-dependent immune suppression byCD4+CD25–Foxp3– T cells that also produce IFN-g. Yet despiteextensive evidence supporting the importance of T cell–derivedIL-10 in limiting inflammation, the events that induce the productionof IL-10 by T cells have remained unclear16,17.

IL-10 is not the sole mediator used by the immune system to limitinflammation; and the list of additional molecules that independentlylimit inflammation (such as CTLA-4, BTLA and PD-1) continues togrow. IL-27, a heterodimeric cytokine composed of the subunits EBI3and p28 (ref. 22), has been described as an antagonist of several T cellfunctions. Although initially identified as a factor that promotes thedevelopment of TH1 cells23, IL-27 has been subsequently reported tolimit TH1, TH2 and TH-17 responses in various models of infectionand autoimmunity24–26. Indeed, studies have shown that, like Il10–/–

mice, Il27ra–/– mice infected with T. gondii develop a lethal CD4+

T cell–mediated response characterized by excessive production ofproinflammatory cytokines, large areas of necrosis in the liver and

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

Received 7 June; accepted 15 October; published online 11 November 2007; corrected online 16 November 2007 (details online); doi:10.1038/ni1537

1Department of Pathobiology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA. 2Molecular Immunology andInflammation Branch, National Institute of Arthritis and Muskoskeletal and Skin Diseases, National Institutes of Health, Bethesda, Maryland 20892. 3Department ofMedicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA. 4Ludwig Institute for Cancer Research, Parkville, Victoria 3050, Australia. 5Department ofInflammation Research, Amgen, Thousand Oaks, California 91320, USA. Correspondence should be addressed to C.A.H. (chunter@vet.upenn.edu).

NATURE IMMUNOLOGY ADVANCE ONLINE PUBLICATION 1

A R T I C L E S

the presence of severe lymphocytic infiltrates in multiple organs7,27.The similarities of these phenotypes not only emphasize the impor-tance of these anti-inflammatory cytokines in regulating an ongoingimmune response but also suggest a potential link between these twoimmune modulators.

Here we describe experiments we undertook to better understandthe effect of IL-27 on T cells. Results obtained assaying 67 solubleimmune mediators in the presence or absence of IL-27 showed thatwhereas IL-27 inhibited many cytokines associated with TH1, TH2 andTH-17 cells, it unexpectedly also promoted the production of IL-10.This observation was mirrored in vivo. We found that T cells fromIL27ra–/– mice chronically infected with T. gondii had less capacity tomake IL-10. In vitro studies showed that IL-27 enhanced the produc-tion of IL-10 by CD4+ T cells in TH1 and TH2 but not TH-17conditions. Moreover, IL-27 promoted the secretion of IL-10 byinducible regulatory T cells. However, most IL-10+ T cells inducedby IL-27 did not express the transcription factor Foxp3, indicating thatIL-27 stimulated IL-10 production by many T cell populations.Furthermore, stimulation of T cells with IL-27 plus transforminggrowth factor-b (TGF-b) resulted in an additive effect on theproduction of IL-10, and that IL-6 (which, like IL-27, signals throughglycoprotein 130), when combined with TGF-b, was also a potentinducer of IL-10 in CD4+ T cells. We also found that the ability ofIL-27 to stimulate IL-10 production was independent of the intra-cellular signaling molecule STAT4 and the transcription factor T-betbut was dependent on activation of transcription factors STAT1 andSTAT3, whereas IL-6 production required only STAT3. Our datacollectively provide new insight into the cytokine environment thatpromotes T cell production of IL-10 and the molecular events thatunderpin this regulatory pathway.

RESULTS

IL-27 induces T cell production of IL-10

Although studies have emphasized the ability of IL-27 to inhibit theproduction of many proinflammatory cytokines by T cells, we used ascreen in an attempt to identify additional targets for IL-27. Weactivated naive CD4+ T cells from C57BL/6 mice with antibody to theT cell receptor (antibody to CD3 (anti-CD3)) and anti-CD28 in thepresence of accessory cells in nonpolarizing conditions (anti-IFN-gplus anti-IL-4) with or without IL-27. After 3 d of cell culture, thesupernatants were assayed with a rodent multi-analyte profile for apanel of 67 secreted immune products. Consistent with publishedreports, the addition of IL-27 to these nonpolarized cultures led todecreases in many cytokines associated with TH1 responses (IFN-g),TH2 responses (IL-5) and TH-17 responses (IL-17), as well asgranulocyte-macrophage colony-stimulating factor, IL-1b, IL-3, the

chemokines CCL3 (MIP-1a) and CCL4 (MIP-1b) and lymphotactin(Fig. 1a). In addition, several other cytokines, including IL-18,IL-6 and IL-7, and chemokines, including, CCL2 (MCP-1), CCL7(MCP-3), macrophage colony-stimulating factor and matrix metallo-proteinase 9, were unaltered by this treatment (SupplementaryTable 1 online). We also noted an increase of 1,000-fold in theproduction of IL-10 in these IL-27-treated cultures (Fig. 1a andSupplementary Table 1).

Further analysis with intracellular staining and flow cytometryshowed that when stimulated in similar conditions followed byrestimulation with phorbol 12-myristate 13-acetate (PMA) and iono-mycin, the number of IL-10+ CD4+ T cells and IL-10+ CD8+ T cellsincreased in a dose-dependent way (Fig. 1b,c and SupplementaryFig. 1 online). Although these results indicated that a similar numberof CD4+ and CD8+ T cells made IL-10 in response to IL-27, theamount of IL-10 in the supernatants of cultures enriched for CD4+

T cells was higher than that in cultures containing CD8+ T cells,consistent with differences in the mean fluorescence intensity (MFI) ofthe cells by flow cytometry (Fig. 1b,c). Further examination of naiveCD4+ T cells (CD4+CD25–CD44–CD62L+) sorted by flow cytometryindicated that these cells could be stimulated with IL-27 to secreteIL-10 (Supplementary Fig. 1). Consequently, we used CD4+ T cells inmost of our subsequent studies to examine the factors that regulatethe production of IL-10.

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

IL-1

0

CD4

CD4

0

1

2

3

4.0 27

62 135

Cha

nge

(%)

1,500

1,000

500100

0

–100

GM-C

SFIF

N-γ

MIP

-1α

MIP

-1β

TNFIL

-6IL

-17

Lym

phot

actin

IL-1

βIL

-10

IL-2 IL

-3

IL-27

IL-27

IL-27

1.0 24

135 91

CD8

IL-1

0 (n

g/m

l)IL

-10

(ng/

ml)

1.0

0.75

0.5

0.25

0

CD8

a

b

c

IL-1

0

α-IFN-γ + α-IL-4

α-IFN-γ + α-IL-4

α-IFN-γ +α-IL-4

α-IFN-γ +α-IL-4

IL-27

Figure 1 IL-27 promotes the production of IL-10 by CD4+ and CD8+ T cells.

(a) Rodent multi-analyte profile bioassay, presented as the percent change

between cells cultured in nonpolarizing conditions (a-IFN-g plus a-IL-4) and

those stimulated with IL-27. GM-CSF, granulocyte-macrophage colony-

stimulating factor; TNF, tumor necrosis factor. (b,c) Flow cytometry of

intracellular IL-10 (left) and ELISA of IL-10 production (right) in CD4+

T cells (b) and CD8+ T cells (c) isolated from the spleens and lymph nodes

of C57BL/6 mice, activated with anti-CD3 and anti-CD28 in nonpolarizing

conditions in the presence or absence of IL-27 for 4 d in the case of the

CD4+ T cells or for 3 d in the case of the CD8+ T cells. Then the cells werestimulated for 4 h with PMA and ionomycin in the presence of brefeldin A;

ELISAs were done after 72 h of stimulation (error bars, s.d.). Numbers in

boxes (left) indicate percent IL-10+ cells; numbers adjacent to boxes

indicate MFI. Data are representative of three independent experiments with

similar results.

2 ADVANCE ONLINE PUBLICATION NATURE IMMUNOLOGY

A R T I C L E S

T cell production of IL-10 requires the IL-27 receptor

The studies described above indicated that IL-27 could enhance T cellproduction of IL-10. To assess the function of the IL-27 receptor(IL-27R) in these events, we activated splenocytes from wild-typeor Il27ra–/– mice in nonpolarizing conditions in the presence orabsence of IL-27 and assayed their production of IL-10. Whereascells from wild-type mice had enhanced secretion of IL-10 in responseto IL-27, cells from Il27ra–/– mice did not (Fig. 2a). Indeed, even basalIL-10 in these supernatants was lower than that of wild-type controls.Additionally, restimulation of wild-type brain mononuclear cells(BMNCs) with soluble toxoplasma antigen in the presence of IL-27resulted in a significant augmentation of IL-10 (Fig. 2b), indicting thatthe capacity of parasite-specific effector cells to make IL-10 wasenhanced by IL-27. To determine whether this interaction wasinvolved in the regulation of inflammatory responses in vivo, we

used an experimental system in which we chronically infected wild-type and Il27ra–/– mice with T. gondii28. In these studies, restimulationof preparations of BMNCs and splenocytes from chronically infectedwild-type mice immediately after isolation showed the presence ofCD4+ T cells that produced IL-10. In contrast, there was a notabledefect in IL-10 for cells from the brains and spleens of chronicallyinfected Il27ra–/– mice (Fig. 2c). Further examination of the pheno-type of CD4+ T cells isolated from the brain showed that whereas therewas no difference in the percentage of cells making IFN-g, there werefewer cells producing both IFN-g and IL-10 in the Il27ra–/– mice thanin wild-type mice (Fig. 2d). Conversely, in separate experiments, therewas a higher percentage of IL-17-producing CD4+ T cells in the brainsof Il27ra–/– mice than in the brains of wild-type mice, and neitherstrain of mouse had a distinct population of cells making both IL-17and IL-10 (Fig. 2d). These results collectively suggest a prominent

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

α-IFN-γ + α-IL-4 IL-270

1

2

3

4

5

6WT

WT

Il27ra–/– Il27ra–/–

Il27ra–/–

WT

IL-1

0 (n

g/m

l)

IL-1

0 (n

g/m

l)

IL-1

0

1.5 *

1.0

0.5

0STAg STAg + IL-27

CD4

BMNC Splenocytes

3.4 9

1.3 0.6

c d

IFN

-�

IL-1

7

15 1.1

0.5

16 <0.1

0.3

0.1 <0.1

3

1.1 <0.1

1

ba

BMNC

IL-10 IL-10

Figure 2 Cells produce less IL-10 in the absence

of IL-27R signaling. (a) ELISA of IL-10 in

supernatants of wild-type (WT) C57BL/6 or

Il27ra–/– CD4+ T cells grown in nonpolarizing

conditions in the presence or absence of IL-27.

Data are representative of three independent

experiments with similar results (error bars, s.d.).

(b) ELISA of IL-10 in supernatants of wild-typeBMNCs (n ¼ 9 mice) restimulated for 48 h

in vitro with soluble toxoplasma antigen (STAg)

in the presence or absence of IL-27. *,

P ¼ 0.0012. Data represent three pooled

experiments. (c) Flow cytometry of intracellular

IL-10 in BMNCs and splenocytes from wild-type

and Il27ra–/– mice chronically infected with

T. gondii; cells were stimulated 5 h ex vivo with

PMA and ionomycin in the presence of brefeldin

A. Numbers in boxes indicate percent IL-10+

cells. Data are representative of three

independent experiments with similar results.

(d) Flow cytometry of the production of IFN-gand IL-10 (left) or IL-17 and IL-10 (right) by

CD4+ T cells isolated from the brains of wild-type

and Il27ra–/– mice. Numbers in quadrants

indicate percent cells in each. Data are

representative of four experiments.

CFSE

TH1 TH2

TH1 + IL-27 TH2 + IL-27 TH-17 + IL-27

TH-17

IL-1

0

24 h 48 h 72 h 96 h

0.5 0.5 3.0 4.5

2224100.8

19 19 31

35 34 32

CD4

IL-1

0

60 58 57

84 119 87

a b

+ IL-27

α-IFN-γ + α-IL-4

Figure 3 CD4+ T cells make IL-10 in response to IL-27 in TH1 and TH2 but not TH-17 conditions. (a) CFSE dilution analysis of CD4+ T cells isolated from

C57BL/6 mice and activated with anti-CD3 and anti-CD28 in nonpolarizing conditions in the presence or absence of IL-27 (time, above plots) before being

stained for intracellular IL-10. Plots are gated on CD4+ T cells. (b) Flow cytometry of CD4+ T cells isolated from C57BL/6 mice and activated with anti-CD3

and anti-CD28 in TH1-, TH2- or TH-17-polarizing conditions in the presence or absence of IL-27 before being stained for intracellular IL-10. Numbers in

boxes indicate percent IL-10+ cells (a,b); numbers adjacent to boxes indicate MFI (b). Data are representative of three independent experiments with

similar results.

NATURE IMMUNOLOGY ADVANCE ONLINE PUBLICATION 3

A R T I C L E S

function for IL-27 in promoting T cell production of IL-10 inthe setting of chronic infection–induced inflammation associatedwith T. gondii.

IL-27 induces IL-10 in TH1 and TH2 but not TH-17 conditions

Although we used neutral conditions for the in vitro studiesdescribed above, the data from mice infected with T. gondii indicatedinvolvement of IL-27 in the development of IL-10-producing cellsduring a TH1-dominated response. Therefore, we did additionalin vitro studies to determine at what point after T cell activationIL-10 was produced and to assess the ability of IL-27 to promoteIL-10 in conditions that favored the development of TH1, TH2 orTH-17 cells. Analysis of IL-10 production by CD4+ T cells in responseto IL-27 over a 4-day period showed that the cells began makingIL-10 at 48 h after activation and that the numbers of IL-10+ cellspeaked at 72 h and were maintained over 96 h (Fig. 3a). That resultcorrelated with the expression pattern of IL-10 mRNA after stimula-tion with IL-27 (Supplementary Fig. 2 online). In addition, todetermine if the IL-10-producing T cells were actively proliferating,we labeled the cells with the cytosolic dye CFSE (carboxyfluoresceindiacetate succinimidyl diester) and then stimulated them withIL-27; CFSE dilution of most CD4+ T cells producing IL-10 indicatedthat the cells were actively dividing. Additionally, in the presenceof IL-27, CD4+ T cells showed decreased proliferation in nonpolari-zing conditions after 72 h of activation. Moreover, the pattern of IL-10production and rate of T cell proliferation in these experimentscorrelated with the expression profile of the IL-27R on recentlyactivated T cells29.

Consistent with published reports15, we found that in TH1 condi-tions (IL-12 plus anti-IL-4), CD4+ T cells that made IL-10 weregenerated. Furthermore, we also noted that the addition of IL-27increased the number of IL-10+ cells (Fig. 3b). Likewise, in TH2conditions (IL-4 plus anti-IFN-g), many IL-10+ CD4+ T cells werepresent, similar to published reports, and the addition of IL-27resulted in many more IL-10+ cells as well as a higher MFI forIL-10, as determined by flow cytometry. Unexpectedly, polarizationof CD4+ T cells in TH-17 conditions (TGF-b plus IL-6) resulted inthe presence of the largest population of T cells that producedIL-10 relative to all other conditions. However, when IL-27 was

added, there was no further increase in thepercentage of IL-10+ cells (Fig. 3b). The lackof an IL-27-dependent increase in the num-ber of IL-10+ cells was not the result of largeamounts of endogenous IL-27 in the cultures,because CD4+ T cells from wild-type andIl27ra–/– mice produced similar numbers ofIL-10+ cells in TH-17 conditions (data notshown). These data collectively indicate thatIL-27 promotes IL-10 production most pro-minently in TH1 and TH2 conditions but notin TH-17 conditions.

IL-27 affects cells producing ‘dual

cytokines’

Although IL-27 promoted IL-10 productionin TH1 and TH2 conditions and many IL-10+

CD4+ T cells were produced after TH-17polarization, it was unclear whether theIL-10+ cells also produced other cytokinesassociated with the three helper T cell subsets.To evaluate this, we stimulated CD4+ T cells

in TH1, TH2 or TH-17 conditions and then stained the cells for intra-cellular IL-10 plus IFN-g, IL-13 or IL-17, respectively. When stimu-lated in TH1 conditions, most IL-10-producing T cells also stainedpositively for IFN-g, but this population of ‘double producers’ was stilla minority relative to cells producing only IFN-g (Fig. 4a). Theaddition of IL-27 did not reduce the percentage of IFN-g+ cells;instead, it resulted in an increase in the percentage of IFN-g+IL-10+

CD4+ T cells. In TH2 conditions, approximately 50% of the IL-10+

cells also made IL-13 (Fig. 4a). The addition of IL-27 increased thepercentage of IL-10+ cells and caused a concurrent reduction in thepercentage of IL-13+IL-10+ cells and IL-13 ‘single producers’.

Unexpectedly, analysis of the production of IL-17 and IL-10 byT cells cultured with IL-6 plus TGF-b (TH-17 conditions) showed thepresence of three distinct populations of T cells: ‘single producers’ ofIL-17 or IL-10 and a population of IL-17+IL-10+ cells (Fig. 4a).Similar to published reports28, the addition of IL-27 inhibited theexpression of IL-17 but did not increase the percentage of cellsexpressing IL-10. Given the presence of accessory cells in thesecultures, it was possible that the inhibition of the production ofIL-17 by IL-27 was the result of its ability to induce IL-10 secretion.However, when we used CD4+ T cells from Il10–/– mice, IL-27 stillinhibited IL-17 production (Fig. 4b).

TGF-b enhances the generation of IL-10+ CD4+ T cells by IL-27

The findings that TH-17 cells produced substantial IL-10 and thatIL-27 was unable to enhance IL-10 production in these conditionssuggested that TGF-b or IL-6 could be involved in the regulation ofthese events. We thus examined the effects of TGF-b on CD4+ T cellsand found that unlike IL-27, TGF-b alone resulted in a modestincrease in IL-10 expression, but when combined with IL-27, TGF-bhad an additive effect, leading to an increase in the percentage ofIL-10+ cells as well as an increase in the MFI, as measured byflow cytometry (Fig. 5a,b). In addition, whereas exogenous TGF-bdid increase the percentage of IL-10+ cells, neutralization of endogen-ous TGF-b did not eliminate the ability of IL-27 to promote IL-10but led to a modest reduction in the percentage of IL-10+ cells (datanot shown).

As TGF-b can convert CD4+CD25– T cells into CD4+CD25+

Treg cells that express Foxp3 (refs. 30,31), it was possible that the

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

IL-10

IL-1

7

IFN

-γ

IL-1

3

IL-1

7

35 13

2

16 8

7

15 15

16

38 <0.1

<0.1

5 <0.1

0.2

0.6 1.5

33

4 4

24

35

6

22

a b

IL-10 IL-10 IL-10

+ IL-27

No IL-27

TH1 TH2 TH-17

TH-17

TH-17 + IL-27

Il10–/–

Figure 4 IL-27 induces the generation of IFN-g+IL-10+ CD4+ T cells in TH1 conditions. (a) Flow

cytometry of CD4+ T cells isolated from C57BL/6 mice and activated with anti-CD3 and anti-CD28 in

TH1-, TH2- or TH-17-polarizing conditions in the presence or absence of IL-27 before being stained for

intracellular IL-10 and for the ‘signature’ helper T cell–associated cytokines IFN-g, IL-13 and IL-17.

(b) Flow cytometry of CD4+ T cells isolated from Il10–/– mice and activated with anti-CD3 and anti-

CD28 in TH-17-inducing conditions in the presence or absence of IL-27 before being stained for

intracellular IL-10 and IL-17. Plots are gated on CD4+ T cells; numbers in quadrants indicate percent

cells in each. Data are representative of three (a) or two (b) independent experiments.

4 ADVANCE ONLINE PUBLICATION NATURE IMMUNOLOGY

A R T I C L E S

inclusion of TGF-b in the culture conditions would favor the expan-sion of Treg cell populations and that IL-27 would promote thesecretion of IL-10 by Treg cells. To test that hypothesis, we obtainedCD4+ T cells from chimeric mice with a reporter gene encodinggreen fluorescent protein (GFP) ‘knocked in’ at the translationstart site of Foxp3 (Foxp3GFP mice)32 and activated the cells withanti-CD3 and anti-CD28 in nonpolarizing conditions in the presenceof TGF-b, IL-27 or a combination of bothcytokines. After 72 h, few Foxp3GFP+ cellswere present in cultures without TGF-b;however, the addition of TGF-b resulted inthe generation of a large number ofFoxp3GFP+ CD4+ T cells, with less than 10%making IL-10 (Fig. 5c). When we culturedCD4+ T cells in the presence of IL-27, therewas no expansion of the number ofFoxp3GFP+ cells, but 50% of the Foxp3GFP+

cells made IL-10. However, most of the IL-10-producing T cellsgenerated in response to IL-27 were Foxp3GFP– (20% versus 1.4%).Finally, TGF-b combined with IL-27 resulted in a decrease of almost70% in the number of Foxp3GFP+ cells relative to that of culturescontaining TGF-b alone. This observation was similar to publisheddata showing that IL-27 inhibits TGF-b-driven induction of Foxp3(refs. 33,34). Nevertheless, as seen with IL-27 alone, nearly 50% of theFoxp3GFP+ cells made IL-10, but most of the IL-10-producing CD4+

T cells remained Foxp3GFP–, indicating that the effects of IL-27 onIL-10 production were not specific to Foxp3+ Treg cells. These datacollectively establish that TGF-b has a synergistic effect with IL-27 inpromoting increased numbers of IL-10+ CD4+ T cells, an effect that isnot due to the development of increased numbers of Foxp3+ Treg cellsin the cultures.

Although TGF-b could have enhanced the capacity of IL-27 tostimulate IL-10, TGF-b alone could not have accounted for the large

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

IL-1

0

IL-27 TGF-β TGF-β + IL-27

CD4

2 15 5 28

70415244

5

4

3

2

1

0

TGF-βIL

-27

TGF-β +

IL-2

7

IL-1

0 (n

g/m

l)

IL-1

0

TGF-β

IL-27 TGF-β + IL-27

6 0.3 5 3

1.4 30

20 1.4

1.4

30 5

6

a

b c

Foxp3GFP

P < 0.0001

P < 0.01

α-IFN-γ + α-IL-4

α-IFN-γ + α-IL-4

α-IFN-γ

+ α-IL

-4

Figure 5 TGF-b augments the IL-27-driven production of IL-10 by CD4+

T cells. (a,b) Flow cytometry (a) and ELISA (b) of intracellular IL-10 in

CD4+ T cells isolated from C7BL/6 mice (n ¼ 10) and activated for 4 d

with anti-CD3 and anti-CD28 in nonpolarizing conditions in the presence or

absence of IL-27, TGF-b or a combination of both, then stimulated for 4 h

with PMA and ionomycin in the presence of brefeldin A. (a) Numbers in

boxes indicate percent IL-10+ cells; numbers adjacent to boxes indicate

MFI. (b) Analysis of culture supernatants at 72 h after stimulation. (c) Flowcytometry of intracellular IL-10 and GFP in CD4+ T cells isolated from

Foxp3GFP mice and activated for 3 d with anti-CD3 and anti-CD28 in

nonpolarizing conditions in the presence or absence of IL-27, TGF-b or a

combination of both, then stimulated for 4 h with PMA and ionomycin in

the presence of brefeldin A. Plots are gated on CD4+ T cells; numbers in

quadrants indicate percent cells in each. Data are representative of three

(a,b) or two (c) independent experiments (mean + s.d., b).

CD4

IL-1

0

TGF-βIL-6 TGF-β + IL-6

3 5 6 33

2P < 0.01

1

0

IL-1

0 (n

g/m

l)

p-S

TAT

1p-

STA

T3

CD4

CD4

1.4 88 86 69 37

17 51 40 1.4

IL-6

IL-27

IL-6

IL-27

1 72 65 25 7

24 68 58 6

Unstimulated

Unstimulated

5 min 30 min 60 min

5 min 30 min 60 min 3 h

3 h

a

c

b

d

49 51 36 62

α-IFN-γ + α-IL-4

TGF-βIL

-6

TGF-β +

IL-6

α-IFN-γ

+

α-IL-4

Figure 6 IL-6 acts in synergy with TGF-b to

promote IL-10 production. (a,b) Flow cytometry

(a) and ELISA (b) of intracellular IL-10 in CD4+

T cells isolated from C57BL/6 mice (n ¼ 3) and

activated for 4 d with anti-CD3 and anti-CD28

in nonpolarizing conditions in the presence or

absence of IL-6, TGF-b or a combination of both,

then stimulated for 4 h with PMA and ionomycin

in the presence of brefeldin A. (a) Numbers in

boxes indicate percent IL-10+ cells; numbers

adjacent to boxes indicate MFI. (b) Analysis of

culture supernatants at 72 h after stimulation.(c,d) Flow cytometry of intracellular

phosphorylated STAT1 (p-STAT1; c) or STAT3

(p-STAT3; d) in CD4+ T cells purified from

C57BL/6 mice and left unstimulated (far left

plots) or stimulated with IL-6 or IL-27 (time,

above plots). Numbers in boxes indicate percent

CD4+ T cells positive for phosphorylated STAT1

(c) or phosphorylated STAT3 (d). Data are

representative of three independent experiments

(mean + s.d., b).

NATURE IMMUNOLOGY ADVANCE ONLINE PUBLICATION 5

A R T I C L E S

number of IL-10+ T cells present in TH-17 conditions. Therefore, todetermine if IL-6, a type I cytokine that shares structural homologyand a receptor subunit with IL-27, could also promote IL-10 produc-tion, we incubated CD4+ T cells with IL-6 in nonpolarizing condi-tions. In these experiments, as seen with TGF-b, the addition ofIL-6 resulted in only a modest increase in IL-10 (Fig. 6a,b). Yet whencombined, TGF-b and IL-6 acted in synergy to promote the develop-ment of many IL-10+ CD4+ T cells. As IL-27-dependent upregulationof IL-10 was most apparent in TH1-polarizing conditions, we exam-ined the effect of IL-6 on IL-10 in TH1 and TH2 conditions. UnlikeIL-27, however, IL-6 did not increase IL-10 production after TH1polarization (Supplementary Fig. 3 online). In contrast, in TH2differentiation conditions, IL-6 had an additive effect on the numberof IL-10+ cells produced (Supplementary Fig. 3).

The generation of IL-10+ CD4+ T cells requires STAT1 and STAT3

The activation of specific STAT proteins in CD4+ T cells is associatedwith the differentiation of T cells into distinct helper T cell lineages.Like other factors, IL-27 has been shown to activate manySTAT proteins, including STAT1 (and as a consequence T-bet),STAT3 and, to a lesser extent, STAT4, whereas IL-6 activates mainlySTAT3 and, to a lesser extent, STAT1. To determine the kinetics withwhich IL-27 and IL-6 activate STAT1 and STAT3, we stimulatedpurified CD4+ T cells with each cytokine for 3 h and then monitoredthe phosphorylation of STAT1 and STAT3 by intracellular antibodystaining and flow cytometry. The addition of either cytokine causedphosphorylation of STAT1 and STAT3; however, IL-6 led to faster and

longer phosphorylation than did IL-27 (Fig. 6c,d). Although theproduction of phosphorylated STAT1 after IL-27 stimulation laggedbehind that of IL-6, the percentage of cells positive for STAT1phosphorylation and the duration of STAT1 phosphorylation weresimilar in response to the two cytokines. IL-6 had an even greatereffect on STAT3, with rapid induction of STAT3 phosphorylation suchthat at 5 min of stimulation, approximately 90% of the T cells werepositive for STAT3 phosphorylation and a high percentage of thesecells was maintained over a 3-hour time period. In contrast, a muchsmaller percentage of CD4+ T cells were positive for STAT3 phos-phorylation in response to IL-27 and this population of cells was notmaintained at 3 h after stimulation.

To further investigate the function of the Janus kinase–STATsignaling pathway in the induction of IL-10 by IL-27, we used micedeficient in individual STAT proteins. The ability of IL-27 to inhibitIL-17 has been attributed before to its ability to activate STAT1,whereas the function of IL-27R signaling in the promotion of TH1differentiation has been attributed mainly to activation of T-betthrough STAT1-dependent as well as STAT1-independent mechanisms.Therefore, to determine if IL-27-mediated production of IL-10+ cellsinvolved these proteins, we obtained CD4+ T cells from Stat1–/– andTbx21–/– (T-bet-deficient) mice and stimulated the cells in nonpolar-izing conditions in the presence or absence of IL-27. Few CD4+ T cellsfrom Stat1–/– mice were able to produce IL-10 in response to IL-27(Fig. 7a), whereas the absence of T-bet did not affect the ability ofIL-27 to promote IL-10+ T cells (Fig. 7b). To assess the function ofSTAT3, we obtained STAT3-deficient CD4+ T cells from transgenic

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

CD4

IL-1

0

IL-1

0

IL-1

0

CD4

IL-1

0

IL-1

0

IL-1

0

34 31 31 3

10 4 6 2

31 36

523

4 1

5 3

32 2 20 26

9

WT WTStat1–/–Stat3+/fl Stat3–/–Tbx21–/–

WT WTStat4–/– Stat3+/flStat1–/– Stat3–/–

0.7 7 9

+ IL-27 + IL-27 + IL-27

+ IL-27

IL-6

TGF-β +IL-6

TGF-β +IL-6

IL-6

a c

e

b

d f

CD4 CD4

CD4 CD4

α-IFN-γ +α-IL-4

α-IFN-γ +α-IL-4

α-IFN-γ +α-IL-4

α-IFN-γ +α-IL-4

Figure 7 STAT-dependent induction of IL-10. (a–d) Flow cytometry of intracellular IL-10 in CD4+ T cells isolated from Stat1–/– mice (a), Tbx21–/– mice (b),

Stat3+/fl and Stat3–/– mice (c), or (d) Stat4–/– mice (and wild-type C57BL/6 mice, a,b,d), then activated with anti-CD3 and anti-CD28 in nonpolarizing

conditions in the presence or absence of IL-2. (e,f) Flow cytometry of intracellular IL-10 in CD4+ T cells isolated from wild-type C57BL/6 and Stat1–/– mice

(e) or Stat3+/fl and Stat3–/– mice (f), then activated with anti-CD3 and anti-CD28 in nonpolarizing conditions with IL-6 in the presence or absence of TGF-b.

Numbers in boxes indicate percent IL-10+ cells. Data are representative of three independent experiments.

6 ADVANCE ONLINE PUBLICATION NATURE IMMUNOLOGY

A R T I C L E S

mice expressing Cre driven by the Cd4 promoter (CD4-Cre mice) anda loxP-flanked (floxed) Stat3 allele (Stat3–/–)35 and stimulated the cellsin the presence or absence of IL-27. We found that IL-27 stimulationof Stat3–/– CD4+ T cells produced fewer IL-10+ cells than did IL-27stimulation of CD4+ T cells from Stat3+/fl littermate control mice(Fig. 7c). The data obtained with STAT1-deficient and STAT3-deficient mice collectively indicate that both transcription factors arerequired for IL-27-induced production of IL-10+ CD4+ T cells.

IL-27 function has also been linked to the activity of STAT4; thus,we next evaluated the response of Stat4–/– CD4+ T cells to IL-27stimulation. In nonpolarizing conditions in the presence of IL-27, theabsence of STAT4 did not block the IL-27-induced production ofIL-10+ CD4+ T cells (Fig. 7d), indicating a STAT4-independentmechanism. However, notably, when we cultured Stat4–/– CD4+

T cells in TH1 conditions, fewer IL-10+ T cells were produced inresponse to IL-12 even when IL-27 was added, relative to resultsobtained with wild-type cells (Supplementary Fig. 4 online). Alsonotably, in the absence of STAT1, STAT3 or STAT4, basal IL-10was lower than that of wild-type cells cultured in nonpolarizingconditions (Fig. 7a,c,d). Thus, CD4+ T cells from any of theSTAT-deficient mice had less capacity to respond to endogenouslyproduced cytokines (IL-27, IL-6 or IL-12) in these assays, whichfurther emphasized the importance of these cytokines in directingT cell production of IL-10. Finally, we assessed the responses of CD4+

T cells from Stat1–/– or CD4-Cre–Stat3–/– mice to either IL-6 alone orIL-6 in combination with TGF-b. Although stimulation of eitherStat1–/– or Stat3–/– CD4+ T cells with IL-6 resulted in few IL-10+

cells (Fig. 7e,f), stimulation of Stat1–/– CD4+ T cells with IL-6 plusTGF-b resulted in the production of IL-10+ T cells similar to that ofStat1+/+ CD4+ T cells (Fig. 7e). In contrast, Stat3–/– CD4+ T cells wereunresponsive to IL-6 plus TGF-b, whereas Stat3+/+ CD4+ T cellsproduced IL-10+ cells (Fig. 7f). These findings indicate that STAT3but not STAT1 signaling is required by IL-6 for the initiation of IL-10production in TH-17 conditions.

DISCUSSION

After its original description as a cytokine associated with TH2 cells,IL-10 is now recognized to have many innate and adaptive sources andto function as a global inhibitor of many types of immune responses.Nevertheless, despite the early appreciation that T cells are a chiefsource of IL-10, many questions have remained about the factors thatregulate IL-10 expression in these lymphocytes. Initial work in thisarea linked IL-12 to the development of IFN-g+IL-10+ cells15,18; ourfindings here showing that in TH1 conditions, STAT4 was involved inthose events support that work. Other studies have shown that chronicstimulation of human and mouse T cells in the presence of IL-10results in the generation of a population of regulatory helper T cells(Tr1 cells) that secrete large amounts of IL-10 and can amelioratecolitis9. Still other work has shown that repeated stimulation of naivehuman and mouse CD4+ T cells in vitro with dexamethasone plusvitamin D3 promotes the production of IL-10+ Treg cells36. Alterna-tively, the studies we have presented here have shown that even aftershort-term stimulation, IL-27 and IL-6 independently induced theproduction of IL-10 by T cells in a variety of polarizing conditions,although IL-6 plus TH1-polarizing conditions did not. Our resultsidentify a previously unknown pathway that promotes the productionof IL-10 and reinforce the idea of the complex relationship betweenthe pro- and anti-inflammatory properties of the IL-6 and IL-12family members.

Even though IL-27 was first described on the basis of its ability topromote TH1 responses, it is now recognized that this type I cytokine

is a negative regulator of the intensity and duration of T cellresponses27,37. The broad anti-inflammatory effects of IL-27 havebeen attributed to its ability to antagonize helper T cell functionsthrough inhibition of the production of IFN-g, IL-2, IL-4 and IL-17(ref. 26). However, in many experimental settings, the phenotype ofIl27ra–/– mice is very similar to that of Il10–/– mice27,28,38. For example,when infected with T. gondii, both Il10–/– mice and Il27ra–/– micedevelop a lethal CD4+ T cell–mediated inflammation acutely asso-ciated with dysregulated TH1 responses7,27 but develop altered TH-17responses in chronic disease28,39. Related to those reports, it has beenestablished that CD4+CD25–Foxp3–IL-10+ T cells are required for theprevention of toxoplasma-induced pathology21. Along with the datawe have presented here, such findings suggest a model in which one ofthe functions of IL-27 is to promote T cell production of IL-10 thathelps to limit T cell–mediated pathology during infection. Presumablythis regulatory pathway would not be restricted to toxoplasmosis, butthe enhanced inflammation noted in Il27ra–/– mice in a variety ofinfectious and inflammatory settings26,40 may be attributed at least inpart to defective IL-10 responses.

Although there are many cellular sources of IL-10, there is a limitednumber of studies that have defined the lineage-specific requirementsfor Il10 transcription. In macrophages, microbial products andimmune complexes can induce IL-10, and mitogen-activated proteinkinase and the transcription factors NF-kB and Sp1 have been linked tothe transcriptional regulation of Il10 (refs. 41–43). For T cells, muchless is know about the molecular events that control IL-10 synthesis,although in TH2 cells, the Jun transcription factors have been linkedto these events, and the transcription factor GATA-3 is associatedwith the remodeling and stability of the Il10 locus required for Il10transcription44. As IL-27 ‘antagonizes’ GATA-3 expression45, it seemsunlikely that GATA-3 accounts for the ability of IL-27 and IL-6 topromote Il10 transcription in TH1- and TH17-polarizing conditions.Instead, our data presented here linked mainly STAT3, but alsoSTAT1 and STAT4, to the cytokine-mediated induction of IL-10 inCD4+ T cells. This observation is consistent with the presence ofSTAT-binding sites in the Il10 promoter and a published report thatIFN-a can induce the recruitment of STAT1 and STAT3 to transactivatean Il10 reporter46.

It is notable that whereas IL-6 and IL-27 signal through glycoprotein130, activate STAT1 and STAT3 and can promote the production ofIL-10+ CD4+ T cells, only IL-27 can downregulate production ofIL-2 and IL-17, whereas IL-6 promotes the development and activityof TH-17 cells. These observations are consistent with a growing bodyof literature emphasizing seemingly contradictory effects of STAT mole-cules on the differentiation and function of helper T cells. Since themid-1990s, STAT4 and STAT6 have been recognized as key transcrip-tion factors that promote TH1 and TH2 development47,48, and sub-sequent studies have linked STAT3 to TH-17 polarization49,50. It is nowbecoming apparent which STAT proteins mediate the effects of IL-27in T cells. Thus, the ability of IL-27 to induce STAT1 can ‘antagonize’TH-17 development, whereas STAT1 and STAT3 are required for theinduction of IL-10 by IL-27. In contrast, the ability of IL-6 to promoteTH-17 activity and IL-10 is STAT3 dependent. A likely explanation forthese distinct effects is that although the receptors for IL-6 and IL-27both contain glycoprotein 130, there are unique IL-6Ra and IL-27Rachains. It remains to be formally tested whether these differences arisefrom the arrangement of STAT1-STAT3 heterodimers in response toIL-27 rather than the formation of mainly STAT3 homodimers inresponse to IL-6 when it acts in conjunction with TGF-b. Alternatively,it is possible that the stronger, more persistent pattern of STAT3phosphorylation in response to IL-6 may be sufficient to promote

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

NATURE IMMUNOLOGY ADVANCE ONLINE PUBLICATION 7

A R T I C L E S

IL-10 expression, whereas IL-27 requires the combination of STAT1and STAT3; this idea requires further exploration.

Although the focus of our work here was on the ability of IL-6 andIL-27 to promote IL-10, perhaps as important is the observation thatTGF-b also influenced this pathway. On the basis of (in part) thepresence of T cell–mediated inflammation in the TGF-b-deficientmice51 and the ability of TGF-b to directly inhibit innate and adaptiveproduction of IFN-g1,52–54, it has been assumed that TGF-b is an anti-inflammatory cytokine. With the realization that TGF-b has a pro-minent function in the development of Treg and TH-17 cells and, asshown here, the production of IL-10 by non–Treg cells, it remainsunclear whether it directs T cell differentiation or is a shared centralregulator of T cell activity that is modulated by cytokines (IL-12, IL-6and IL-27) present in the environment that determine cell fate.Nevertheless, our studies have emphasized another facet of thecomplex biology of TGF-b and its involvement in suppressingcell-mediated immunity.

Although our studies here have identified IL-27 and IL-6 as factorsthat promote T cell production of IL-10, one of the larger issues relatesto whether this observation indicates the development of distinctT cell subsets. When TH1 and TH2 cells were first described, the totaldiversity of T cell phenotypes and whether other T cell types existin vivo were questioned55. One possible interpretation of our data hereis that helper T cell subsets can be defined by their ability to produceIFN-g, IL-4 and IL-17 alone or in combination with IL-10. So far, ourattempts to generate stable populations of IL-10-producing T cellsin vitro with IL-27 have proven unsuccessful, although there isevidence that stable TH1 clones derived from humans are able toproduce large amounts of IL-10 and IFN-g14,15,18,19,56. Although manyinterpretations of these preliminary data are possible, one is that theability to secrete IL-10 is not a hallmark of distinct T cell subsets butinstead that cytokines such as IL-27 and IL-6 represent ‘modifiers’ ofthe main helper T cell subsets that enable them to make IL-10 in thecontext of chronic inflammation. This may be one mechanism thatallows the establishment of appropriate helper T subsets required todeal with different classes of pathogens but provides each of thesedistinct effector cells with a mechanism to regulate their own inflam-matory activities.

METHODSMice and parasites. C57BL/6, BALB/c, Stat4–/– and Tbx21–/– mice were from

Jackson laboratories. Foxp3GFP mice have been described32, as have CD4-

Cre-Stat3–/– mice57. Mice were housed and bred in specific pathogen–free

facilities in the Department of Pathobiology at the University of Pennsylvania

in accordance with institutional guidelines.

The ME49 strain of T. gondii was prepared from chronically infected CBA/ca

mice, and mice used in experiments were infected intraperitoneally with

20 cysts. Il27ra–/– and C57BL/6 wild-type control mice were treated with sulfa-

diazine (200 mg/l; Sigma) in their drinking water for 2 weeks beginning on day

5 after infection to allow the Il27ra–/– mice to progress to a chronic stage of

infection. Soluble toxoplasma antigen was prepared from tachyzoites of the RH

strain as described58. BMNCs from chronically infected mice were isolated in

accordance with a published protocol39.

The generation of IL-10-producing T cells. CD4+ and CD8+ T cells were

isolated from splenocyte samples and lymph nodes that were depleted of CD8+

and NK1.1+ cells to enrich for CD4+ T cells, or were depleted of CD4+ and

NK1.1+ cells to enrich for CD8+ T cells, by magnetic bead separation

(Polysciences). Cells were plated in 96-well round-bottomed plates (Costar)

at a density of 5 � 106 cells per ml. Cells were stimulated with anti-CD3 (1 mg/

ml; clone 145-2C11; eBioscience) and anti-CD28 (1 mg/ml; clone 37.51;

eBioscience). For the production of IL-10+ T cells, cultures were supplemented

with recombinant mouse IL-27 (50 ng/ml; Amgen) or human TGF-b (1 ng/ml;

R&D Systems) alone or in combination with IL-27. Additionally, IFN-g and

IL-4 were neutralized in the nonpolarized cultures with anti-IFN-g (10 mg/ml;

XMG1.2) and anti–IL-4 (10 mg/ml; 11B11; NCI Preclinical Repository). In

some cases, T cells were cultured in TH1 conditions (recombinant IL-12 (10 ng/

ml; eBioscience) plus anti-IL-4 (10 mg/ml)), TH2 conditions (recombinant IL-4

(8 ng/ml; eBioscience) plus anti-IFN-g (10 mg/ml)) or TH-17 conditions

(TGF-b (1 ng/ml; R&D Systems) and IL-6 (10 ng/ml; eBioscience) plus anti-

IFN-g (10 mg/ml) and anti-IL-4). CD8+ T cells were collected on day 3; CD4+

T cells were supplemented with fresh medium and reagents on day 3 and were

collected on day 4. T cells were then restimulated with PMA and ionomycin

plus brefeldin A (Sigma). A FACSCalibur (BD Biosciences) or BDFACS CantoII

(BD Biosciences) was used for flow cytometry, and data were analyzed with

FlowJo software (Tree Star). For intracellular staining of GFP, cells were stained

with polyclonal anti-GFP (14-6774-81; eBioscience), followed by a second

stain with fluorescein isothiocyanate–conjugated rat anti-rabbit (111-096-144;

Jackson Immunoresearch).

Rodent multi-analyte profile. Supernatants from CD4+ T cells cultured in

nonpolarizing conditions in the presence or absence of IL-27 were analyzed by

Rules-Based Medicine with a rodent multi-analyte profile. With automated

pipetting, an aliquot of each sample was introduced into one of the capture

microsphere ‘multiplexes’ (multiplex panels) of the profile. These mixtures of

sample and capture microspheres were mixed thoroughly and were incubated

at for 1 h at 25 1C. ‘Multiplexed cocktails’ of biotinylated reporter antibodies for

each ‘multiplex’ were then added by a robot and, after being thoroughly mixed,

were incubated for an additional hour at 25 1C. ‘Multiplexes’ were developed

with an excess of streptavidin-phycoerythrin solution, which was thoroughly

mixed into each ‘multiplex’ and incubated for 1 h at 25 1C. The volume of each

‘multiplexed’ reaction was reduced by vacuum filtration and the volume was

increased by dilution into matrix buffer for analysis. A Luminex 100 was used

for analysis, and the resulting data stream was interpreted with proprietary data

analysis software developed at Rules-Based Medicine and licensed to Qiagen

Instruments and Upstate Biotechnology. For each ‘multiplex’, both calibrators

and controls were included on each microtiter plate. Eight-point calibrators

were assayed in the first and last column of each plate and controls were

included in duplicate. Testing results were determined first for the high,

medium and low controls for each ‘multiplex’ to ensure proper assay perfor-

mance. The unknown values for each analyte located in a specific ‘multiplex’

were determined with four and five parameters and with weighted and

nonweighted curve-fitting algorithms included in the data analysis package.

Intracellular staining for phosphorylated STAT1 and STAT3. CD4+ T cells

were purified from C57BL/6 mice with a CD4+ isolation kit (Milltenyi).

Purified CD4+ T cells (1 � 106) were incubated with IL-6 or IL-27 for 5, 30,

60 or 180 min. Cells were then fixed for 10 min at 37 1C with 2% (wt/vol)

paraformaldehyde. After being fixed, cells were made permeable for 30 min on

ice with 90% (vol/vol) methanol, then were stained for CD4 and phospho-

rylated STAT1 and STAT3. Antibodies to phosphorylated tyrosine residues of

STAT1 (clone 4a) and STAT3 (clone 4/P-STAT3) were from BD Pharmingen.

Statistics. A paired Student’s t-test was used to determine statistical signifi-

cance, and P values of less than 0.05 were considered significant.

Note: Supplementary information is available on the Nature Immunology website.

ACKNOWLEDGMENTSWSX-1-deficient (Il27ra–/–) mice were provided by C. Saris (Amgen);BALB/c Il10–/– mice were originally obtained from R. Coffman (DNAX);Stat1–/– mice were provided by P. Scott (University of Pennsylvania);Foxp3GFP mice were provided by L. Turka (University of Pennsylvania); andCD4-Cre–Stat3–/– mice were provided by J.J. O¢Shea (National Institutes ofHealth). Supported by the National Institutes of Health (AI42334 and AI41158to C.H.; AI 43620 to L.T.; 1-T32-AI-055428 to J.S.S.; and 1-T32-AI-07532 toJ.S.), the Scholler Foundation, the State of Pennsylvania, and the NationalHealth and Medical Research Council of Australia.

AUTHOR CONTRIBUTIONSJ.S.S. and C.A.H. contributed to all studies; J.S. and M.E. were involved in theanalysis and interpretation of the p-STAT1 and p-STAT3 signaling experiments;

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

8 ADVANCE ONLINE PUBLICATION NATURE IMMUNOLOGY

A R T I C L E S

T.H.H. was involved in the analysis of IL-10 expression by real-time quantitativePCR; A.L. and J.J.O. contributed to the studies of Stat3–/– CD4+ T cells; andP.M.P. and L.A.T. contributed to the studies of the Foxp3GFP mice.

COMPETING INTERESTS STATEMENTThe authors declare competing financial interests: details accompany the full-textHTML version of the paper at http://www.nature.com/natureimmunology/.

Published online at http://www.nature.com/natureimmunology

Reprints and permissions information is available online at http://npg.nature.com/

reprintsandpermissions

1. Fiorentino, D.F., Bond, M.W. & Mosmann, T.R. Two types of mouse T helper cell. IV. Th2clones secrete a factor that inhibits cytokine production by Th1 clones. J. Exp. Med.170, 2081–2095 (1989).

2. Moore, K.W. et al. Homology of Cytokine synthesis inhibitory factor (IL-10) to theEpstein-Barr virus gene BCRFI. Science 248, 1230–1234 (1990).

3. Moore, K.W., de Waal Malefyt, R., Coffman, R.L. & O’Garra, A. Interleukin-10 and theinterleukin-10 receptor. Annu. Rev. Immunol. 19, 683–765 (2001).

4. Kuhn, R., Lohler, J., Rennick, D., Rajewsky, K. & Muller, W. Interleukin-10-deficientmice develop chronic enterocolitis. Cell 75, 263–274 (1993).

5. Yen, D. et al. IL-23 is essential for T cell-mediated colitis and promotes inflammationvia IL-17 and IL-6. J. Clin. Invest. 116, 1310–1316 (2006).

6. Grunig, G. et al. Interleukin-10 is a natural suppressor of cytokine production andinflammation in a murine model of allergic bronchopulmonary aspergillosis. J. Exp.Med. 185, 1089–1099 (1997).

7. Gazzinelli, R.T. et al. In the absence of endogenous IL-10, mice acutely infected withToxoplasma gondii succumb to a lethal immune response dependent on CD4+ T cellsand accompanied by overproduction of IL-12, IFN-g and TNF-a. J. Immunol. 157,798–805 (1996).

8. Hunter, C.A. et al. IL-10 is required to prevent immune hyperactivity during infectionwith Trypanosoma cruzi. J. Immunol. 158, 3311–3316 (1997).

9. Groux, H. et al. A CD4+ T-cell subset inhibits antigen-specific T-cell responses andprevents colitis. Nature 389, 737–742 (1997).

10. Suffia, I.J., Reckling, S.K., Piccirillo, C.A., Goldszmid, R.S. & Belkaid, Y. Infected site-restricted Foxp3+ natural regulatory T cells are specific for microbial antigens. J. Exp.Med. 203, 777–788 (2006).

11. Zhang, X. et al. IL-10 is involved in the suppression of experimental autoimmune en-cephalomyelitis by CD25+CD4+ regulatory T cells. Int. Immunol. 16, 249–256 (2004).

12. Jonuleit, H., Schmitt, E., Schuler, G., Knop, J. & Enk, A.H. Induction of interleukin 10-producing, nonproliferating CD4+ T cells with regulatory properties by repetitivestimulation with allogeneic immature human dendritic cells. J. Exp. Med. 192,1213–1222 (2000).

13. Vieira, P.L. et al. IL-10-secreting regulatory T cells do not express Foxp3 but havecomparable regulatory function to naturally occurring CD4+CD25+ regulatory T cells.J. Immunol. 172, 5986–5993 (2004).

14. Del Prete, G. et al. Human IL-10 is produced by both type 1 helper (Th1) and type 2helper (Th2) T cell clones and inhibits their antigen-specific proliferation and cytokineproduction. J. Immunol. 150, 353–360 (1993).

15. Gerosa, F. et al. Interleukin-12 primes human CD4 and CD8 T cell clones for high pro-duction of both interferon-g and interleukin-10. J. Exp. Med. 183, 2559–2569 (1996).

16. O’Garra, A. & Vieira, P. TH1 cells control themselves by producing interleukin-10. Nat.Rev. Immunol. 7, 425–428 (2007).

17. Trinchieri, G. Interleukin-10 production by effector T cells: Th1 cells show self control.J. Exp. Med. 204, 239–243 (2007).

18. Gerosa, F. et al. CD4+ T cell clones producing both interferon-g and interleukin-10predominate in bronchoalveolar lavages of active pulmonary tuberculosis patients. Clin.Immunol. 92, 224–234 (1999).

19. Pohl-Koppe, A., Balashov, K.E., Steere, A.C., Logigian, E.L. & Hafler, D.A. Identifica-tion of a T cell subset capable of both IFN-g and IL-10 secretion in patients withchronic Borrelia burgdorferi infection. J. Immunol. 160, 1804–1810 (1998).

20. Anderson, C.F., Oukka, M., Kuchroo, V.J. & Sacks, D. CD4+CD25–Foxp3– Th1 cells arethe source of IL-10-mediated immune suppression in chronic cutaneous leishmaniasis.J. Exp. Med. 204, 285–297 (2007).

21. Jankovic, D. et al. Conventional T-bet+Foxp3– Th1 cells are the major source of host-protective regulatory IL-10 during intracellular protozoan infection. J. Exp. Med. 204,273–283 (2007).

22. Pflanz, S. et al. IL-27, a heterodimeric cytokine composed of EBI3 and p28 protein,induces proliferation of naive CD4+ T cells. Immunity 16, 779–790 (2002).

23. Chen, Q. et al. Development of Th1-type immune responses requires the type I cytokinereceptor TCCR. Nature 407, 916–920 (2000).

24. Batten, M. et al. Interleukin 27 limits autoimmune encephalomyelitis by suppressing thedevelopment of interleukin 17-producing T cells. Nat. Immunol. 7, 929–936 (2006).

25. Amadi-Obi, A. et al. TH17 cells contribute to uveitis and scleritis and are expanded byIL-2 and inhibited by IL-27/STAT1. Nat. Med. 13, 711–718 (2007).

26. Kastelein, R.A., Hunter, C.A. & Cua, D.J. Discovery and biology of IL-23 and IL-27:related but functionally distinct regulators of inflammation. Annu. Rev. Immunol. 25,221–242 (2007).

27. Villarino, A. et al. The IL-27R (WSX-1) is required to suppress T cell hyperactivityduring infection. Immunity 19, 645–655 (2003).

28. Stumhofer, J.S. et al. Interleukin 27 negatively regulates the development of inter-leukin 17-producing T helper cells during chronic inflammation of the central nervoussystem. Nat. Immunol. 7, 937–945 (2006).

29. Villarino, A.V. et al. Positive and negative regulation of the IL-27 receptor duringlymphoid cell activation. J. Immunol. 174, 7684–7691 (2005).

30. Chen, W. et al. Conversion of peripheral CD4+CD25– naive T cells to CD4+CD25+

regulatory T cells by TGF-b induction of transcription factor Foxp3. J. Exp. Med. 198,1875–1886 (2003).

31. Rich, S., Seelig, M., Lee, H.M. & Lin, J. Transforming growth factor b1 costimulatedgrowth and regulatory function of staphylococcal enterotoxin B-responsive CD8+ Tcells.J. Immunol. 155, 609–618 (1995).

32. Bettelli, E. et al. Reciprocal developmental pathways for the generation of pathogeniceffector TH17 and regulatory T cells. Nature 441, 235–238 (2006).

33. Neufert, C. et al. IL-27 controls the development of inducible regulatory T cells andTh17 cells via differential effects on STAT1. Eur. J. Immunol. 37, 1809–1816(2007).

34. Korn, T. et al. IL-21 initiates an alternative pathway to induce proinflammatory TH17cells. Nature 448, 484–487 (2007).

35. Chiarle, R. et al. Stat3 is required for ALK-mediated lymphomagenesis and provides apossible therapeutic target. Nat. Med. 11, 623–629 (2005).

36. Barrat, F.J. et al. In vitro generation of interleukin 10-producing regulatory CD4+ T cellsis induced by immunosuppressive drugs and inhibited by T helper type 1 (Th1)- andTh2-inducing cytokines. J. Exp. Med. 195, 603–616 (2002).

37. Hamano, S. et al. WSX-1 is required for resistance to Trypanosoma cruzi infection byregulation of proinflammatory cytokine production. Immunity 19, 657–667 (2003).

38. Rosas, L.E. et al. Interleukin-27R (WSX-1/T-cell cytokine receptor) gene-deficient micedisplay enhanced resistance to Leishmania donovani infection but develop severe liverimmunopathology. Am. J. Pathol. 168, 158–169 (2006).

39. Wilson, E.H., Wille-Reece, U., Dzierszinski, F. & Hunter, C.A. A critical role for IL-10 inlimiting inflammation during toxoplasmic encephalitis. J. Neuroimmunol. 165, 63–74(2005).

40. Hunter, C.A. New IL-12-family members: IL-23 and IL-27, cytokines with divergentfunctions. Nat. Rev. Immunol. 5, 521–531 (2005).

41. Brightbill, H.D., Plevy, S.E., Modlin, R.L. & Smale, S.T. A prominent role for Sp1 duringlipopolysaccharide-mediated induction of the IL-10 promoter in macrophages.J. Immunol. 164, 1940–1951 (2000).

42. Liu, Y.W., Chen, C.C., Tseng, H.P. & Chang, W.C. Lipopolysaccharide-induced trans-criptional activation of interleukin-10 is mediated by MAPK- and NF-kB-inducedCCAAT/enhancer-binding protein d in mouse macrophages. Cell. Signal. 18,1492–1500 (2006).

43. Lucas, M., Zhang, X., Prasanna, V. & Mosser, D.M. ERK activation following macro-phage FcgR ligation leads to chromatin modifications at the IL-10 locus. J. Immunol.175, 469–477 (2005).

44. Shoemaker, J., Saraiva, M. & O’Garra, A. GATA-3 directly remodels the IL-10 locusindependently of IL-4 in CD4+ T cells. J. Immunol. 176, 3470–3479 (2006).

45. Lucas, S., Ghilardi, N., Li, J. & de Sauvage, F.J. IL-27 regulates IL-12 responsivenessof naive CD4+ T cells through Stat1-dependent and -independent mechanisms. Proc.Natl. Acad. Sci. USA 100, 15047–15052 (2003).

46. Ziegler-Heitbrock, L. et al. IFN-a induces the human IL-10 gene by recruiting both IFNregulatory factor 1 and Stat3. J. Immunol. 171, 285–290 (2003).

47. Hou, J. et al. An interleukin-4-induced transcription factor: IL-4. Stat. Sci. 265,1701–1706 (1994).

48. Jacobson, N.G. et al. Interleukin 12 signaling in T helper type 1 (Th1) cells involvestyrosine phosphorylation of signal transducer and activator of transcription (Stat)3 andStat4. J. Exp. Med. 181, 1755–1762 (1995).

49. Chen, Z. et al. Selective regulatory function of Socs3 in the formation of IL-17-secreting T cells. Proc. Natl. Acad. Sci. USA 103, 8137–8142 (2006).

50. Mathur, A.N. et al. Stat3 and Stat4 direct development of IL-17-secreting Th cells.J. Immunol. 178, 4901–4907 (2007).

51. Shull, M.M. et al. Targeted disruption of the mouse transforming growth factor-b1 generesults in multifocal inflammatory disease. Nature 359, 693–699 (1992).

52. Espevik, T. et al. Inhibition of cytokine production by cyclosporin A and transforminggrowth factor b. J. Exp. Med. 166, 571–576 (1987).

53. Silva, J.S., Twardzik, D.R. & Reed, S.G. Regulation of Trypanosoma cruzi infections invitro and in vivo by transforming growth factor b (TGF-b). J. Exp. Med. 174, 539–545(1991).

54. Hunter, C.A., Bermudez, L., Beernink, H., Waegell, W. & Remington, J.S. Transforminggrowth factor-b inhibits interleukin-12-induced production of interferon-g by naturalkiller cells: a role for transforming growth factor-b in the regulation of T cell-independent resistance to Toxoplasma gondii. Eur. J. Immunol. 25, 994–1000(1995).

55. Mosmann, T.R. & Coffman, R.L. TH1 and TH2 cells: different patterns of lymphokinesecretion lead to different functional properties. Annu. Rev. Immunol. 7, 145–173(1989).

56. Cohen, S.B. et al. High level of interleukin-10 production by the activated T cellpopulation within the rheumatoid synovial membrane. Arthritis Rheum. 38, 946–952(1995).

57. Lee, C.K. et al. STAT3 is a negative regulator of granulopoiesis but is not required forG-CSF-dependent differentiation. Immunity 17, 63–72 (2002).

58. Sharma, S.D., Mullenax, J., Araujo, F.G., Erlich, H.A. & Remington, J.S. Western blotanalysis of the antigens of Toxoplasma gondii recognized by human IgM and IgGantibodies. J. Immunol. 131, 977–983 (1983).

©20

07 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reim

mu

no

log

y

NATURE IMMUNOLOGY ADVANCE ONLINE PUBLICATION 9

A R T I C L E S

©20

07 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

urei

mm

unol

ogy

Corrigendum: Interleukins 27 and 6 induce STAT3-mediated T cell production of interleukin 10Jason S Stumhofer, Jonathan S Silver, Arian Laurence, Paige M Porrett, Tajie H Harris, Laurence A Turka, Matthias Ernst, Christiaan J M Saris, John J O’Shea & Christopher A HunterNat. Immunol. doi:10.1038/ni1537; corrected 16 November 2007

In the version of this article initially published online, reference 32 is cited incorrectly. The correct reference is “Bettelli, E. et al. Reciprocal devel-opmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature 441, 235–238 (2006)” (originally reference 57). The error has been corrected for all versions of the article.

CORR IGENDA