IL-1 signaling modulates activation of STAT transcription ... - Nature

Articles

IL-1 signaling modulates activation of STAT transcription factors to antagonize retinoic acid signaling and control the TH17 cell–iTreg cell balance

npg

© 2015 Nature America, Inc. All rights reserved.

Rajatava Basu1,5, Sarah K Whitley1,5, Suniti Bhaumik1, Carlene L Zindl1, Trenton R Schoeb2, Etty N Benveniste3, Warren S Pear4, Robin D Hatton1 & Casey T Weaver1 Interleukin 17 (IL-17)-producing helper T cells (TH17 cells) and CD4+ inducible regulatory T cells (iTreg cells) emerge from an overlapping developmental program. In the intestines, the vitamin A metabolite retinoic acid (RA) is produced at steady state and acts as an important cofactor to induce iTreg cell development while potently inhibiting TH17 cell development. Here we found that IL-1 was needed to fully override RA-mediated expression of the transcription factor Foxp3 and induce protective TH17 cell responses. By repressing expression of the negative regulator SOCS3 dependent on the transcription factor NF-kB, IL-1 increased the amplitude and duration of phosphorylation of the transcription factor STAT3 induced by T H17-polarizing cytokines, which led to an altered balance in the binding of STAT3 and STAT5 to shared consensus sequences in developing T cells. Thus, IL-1 signaling modulated STAT activation downstream of cytokine receptors differently to control the TH17 cell–iTreg cell developmental fate. Interleukin 17 (IL-17)-producing helper T cells (TH17 cells) and CD4+ induced regulatory T cells (iTreg cells) emerge from a shared developmental axis1. While transforming growth factor-β (TGF-β) contributes to the developmental programming of both subsets, the proinflammatory cytokine IL-6 favors TH17 cell development at the expense of iTreg cell development2–6. Conversely, retinoic acid (RA), a metabolite of vitamin A produced by intestinal stromal cells and dendritic cells (DCs) that express retinaldehyde dehydrogenases7, acts in concert with TGF-β to promote expression of the transcription factor Foxp3 and Treg cell development while potently inhibiting TH17 cell development8–12. A substantial percentage of T H17 cells resident in the intestinal lamina propria have expressed Foxp3 at some point during their development, which indicates a dynamic relationship between RORγt+ TH17 cells and Foxp3 + Treg cells developing in the intestines5. Whereas IL-6 signaling induces phosphorylation of the transcription factor STAT3 that is required for expression of the transcription factor RORγt and TH17 cell development, the actions of RA are at least partially dependent on IL-2, which induces phosphorylation of STAT5 that is required for Foxp3 expression and iTreg cell development, and which suppresses TH17 cell development 9,13,14. Many DNA-binding sites targeted by STAT3 in gene loci associated with the TH17 lineage can also bind STAT5, which provides a mechanism for competitive antagonism of these trans-acting factors downstream of signaling via the IL-6 and IL-2 receptors15.

While IL-6 is important for reciprocally promoting TH17 cell development and repressing iTreg cell development, its effects in countering RA-mediated inhibition of TH17 cell development are incomplete9,14. Thus, even at saturating doses of IL-6, TH17 cell development is suppressed by RA, which suggests that additional signals might be needed to override the dominant RA signaling in the gut. IL-1 is another proinflammatory cytokine that promotes T H17 cell development while subverting Treg cell development2,16, and it has been shown to be important in the development of TH17 cells in the gut, at least at steady state17. However, other reports have suggested that signaling via IL-1β and its receptor IL-1R is dispensable in the differentiation of intestinal TH17 cells at steady state, on the basis of studies of mice deficient in the adaptor MyD88, which have impaired signaling via IL-1R and Toll-like receptors18,19. The role of IL-1 signaling during pathogen-induced intestinal inflammation has not been well studied. Therefore, both the mechanism by which IL-1 promotes intestinal TH17 cell development and the role of IL-1 in TH17 cell development during acute intestinal inflammation, particularly in disease models in which TH17 cells are protective, remain largely undefined20,21. In this study, we sought to define the role of IL-1 in regulating the developmental balance of TH17 cells and iTreg cells, with emphasis on its effects on the intestine during challenge with the enteropathogenic bacterium Citrobacter rodentium, which elicits a TH17 pathway response that is required for host protection 3,21,22. Our results indicated that IL-1 signaling was needed to enhance IL-6 signaling

1Department

of Pathology, University of Alabama at Birmingham, Birmingham, Alabama, USA. 2Department of Genetics, University of Alabama at Birmingham, Birmingham, Alabama, USA. 3Department of Cell, Developmental and Integrative Biology, University of Alabama at Birmingham, Birmingham, Alabama, USA. 4Department of Pathology and Laboratory Medicine, University of Pennsylvania, Perelman School of Medicine, Philadelphia, Pennsylvania, USA. 5These authors contributed equally to this work. Correspondence should be addressed to C.T.W. ([email protected]). Received 6 November 2014; accepted 7 January 2015; published online 2 February 2015; doi:10.1038/ni.3099

286

VOLUME 16  NUMBER 3  MARCH 2015  nature immunology

Articles the impairment of TH17 differentiation observed when DCs from mesenteric lymph nodes (MLNs) were used to activate naive CD4 + T cells (Fig. 1a,b). Moreover, IL-1β was comparable to the retinoic acid receptor (RAR) inhibitor LE450 in blocking the effects of RA (Fig. 1a,b). Accordingly, the addition of IL-1β overrode the inhibition of TH17 differentiation by RA, regardless of RA concentration (Fig. 1c,d). This result was not due to downregulation of the receptor subunits RAR or RXR, as all members of this family were either unchanged or modestly increased by IL-1 signaling, and this occurred despite partial RA-mediated down-modulation of IL-1R1, which had higher expression by developing TH17 cells than by TH0 cells (Supplementary Fig. 1). To extend the studies reported above, we investigated the effects of IL-1 on reversing the expression of Foxp3 supported by RA signaling in developing TH17 cells (Fig. 1e,f). Using TH17 cells derived from naive precursor cells from Il17fThy-1.1Foxp3GFP dualreporter mice (homozygous for sequence encoding the alloantigen Thy-1.1 knocked into the Il17f locus and for sequence encoding green fluorescent protein (GFP) knocked into the Foxp3 locus), we sorted Foxp3+IL-17F− cells and restimulated them under TH17polarizing conditions in the presence or absence of added RA. In the absence of added RA, a significant fraction of sorted Foxp3+ cells lost Foxp3 expression and expressed IL-17F. This transition was inhibited

RESULTS IL-1b reverses RA-induced inhibition of TH17 differentiation IL-6 counteracts the effects of the RA-mediated suppression of TH17 cell development, albeit incompletely 9. In the course of investigating the role of IL-1β in promoting TH17 cell development, we found that in contrast to IL-6, IL-1β completely reversed

9.0

12.3

37.5

15.1

39.7

60

3

10

50

102 101

4

2.5

3.5

1.3 100 101 102 103 104

10

11.0

2.2

2.0

2.0

1.5

0.6

103 10

16.2

52.8

100 100 101 102 103 104

10.1

+100 nM RA

4.7

10.2

3.5

10

102

65.2

100 100 101 102 103 104 104

20

0.2

0.4

19.0 1.6

23.4

1.0

2.1

1.9

3

10

IL-17A

52.0

+ IL-1β

** IL-17F

102

Foxp3

101

80.1

0

10

70.6

68.0

100 101 102 103 104

IL-17A

IL-17F (Thy1.1) – (Foxp3) + IL-1β (Foxp3) – (IL-17A + IL-17F) + IL-1β (IL-17A + IL-17F)

d 90 80 70 60 50 40 30 20 10 0

e

– RA 55.6 6.9

3.4

19.3

5.5

70

104

** 0

1.0

* ** 10.0

* **

10

2

28.2

104

101 10

4.8

11.7

0

100 101 102 103 104

+ IL-1β

IL-17F (Thy-1.1)

100.0

RA (nm)

Foxp3 (GFP)

*

10

3

3

10

102 101

9.0

4.3 5.4

3.9

33.4

28.5

0

10

– (Foxp3) + IL-1β (Foxp3) – (IL-17F) + IL-1β (IL-17F)

f

+ RA

18.3

Foxp3 (GFP)

CD4+ cells (%)

3.8

101

0

101

+1 nM RA 2.7

3

30

10

2

Foxp3

*

4

10

– + LE540 + IL-1β

100 101 102 103 104

60 CD4+ cells (%)

100

*

40

*

c

*

Foxp3

8.9

b

+ IL-1β

+ LE540

4

10

CD4+ cells (%)

IL-17F (Thy-1.1)

a

npg

© 2015 Nature America, Inc. All rights reserved.

to fully override RA-mediated Foxp3 expression and induce protective TH17 cell responses. Through a mechanism dependent on the transcription factor NF-κB, IL-1 signaling repressed the expression of SOCS3, a feedback inhibitor of tyrosine phosphorylation of STAT3 induced by Jak kinase. Inhibition of SOCS3 expression by IL-1 signaling resulted in an increased amplitude and duration of the tyrosine-phosphorylation of STAT3 downstream of signaling by TH17-polarizing cytokines without altering the IL-2-induced tyrosine-phosphorylation of STAT5. IL-1 signaling therefore altered the balance of the binding of STAT3 versus that of STAT5 to shared consensus sequences in developing T cells, including the conserved noncoding sequence 2 (CNS2) element in the Foxp3 locus that regulates stability of Foxp3 expression, as well as target sequences in the Il17a–Il17f locus. Thus, IL-1 signaling modulated STAT activation downstream of cytokine receptors differently to control the TH17 cell–iTreg cell developmental fate.

50

*

40

*

30

*

20

**

10 0 RA

–

+

–

+

–

+

–

+

IL-17F (Thy-1.1)

Figure 1  IL-1β counteracts the RA-dependent inhibition of T H17 cell development. (a) Flow cytometry of Il17fThy-1.1 naive CD4+ T cells (CD4+CD25−CD62Lhi CD44lo) activated for 4 d with soluble anti-CD3 on IL-1β-deficient (Il1b−/−) MLN DCs under TH17-polarizing conditions alone (far left) or with the addition of the RA inhibitor LE540 (1 µM) or IL-1β (20 ng/ml), then stained for surface CD4 and Thy-1.1 (IL-17F), and for intracellular Foxp3 and IL-17A. Numbers in (or above) quadrants indicate percent cells in each throughout. (b) Frequency of cells producing only IL-17A, IL-17F (Thy-1.1) or Foxp3, among CD4 + T cells pooled from a. (c) Expression of Thy-1.1 (IL-17F) and Foxp3 by Il17fThy-1.1 naive CD4+ T cells activated for 4 d with soluble anti-CD3 and Il1b−/− DCs under TH17-polarizing conditions with or without IL-1β (left margin) in the presence or absence of increasing concentrations of RA (above plots). (d) Frequency of cells as in b among CD4+ T cells pooled from c. (e) Flow cytometry of cells obtained by activation of Il17fThy-1.1Foxp3GFP naive CD4+ T cells for 4 d with plate-bound anti-CD3 and soluble anti-CD28 under T H17-polarizing conditions (primary culture), followed by sorting of the Thy-1.1 −GFP+ fraction of cells by flow cytometry (far left) for further culture for 4 d under TH17-polarizing conditions with or without IL-1β (20 ng/ml) in the presence or absence of RA (1 nM) (secondary culture), then by surface staining of Thy-1.1 (IL-17F). (f) Frequency of Foxp3+ (GFP) and IL-17F+ (Thy-1.1) cells among CD4+ T cells pooled from secondary cultures in e. Each symbol represents an individual culture; small horizontal lines indicate the mean. *P < 0.05 and **P < 0.01 (two-tailed unpaired t-test). Data are from one experiment representative of three independent experiments with similar results (a,c) or two independent experiments (e) or are pooled from three independent experiments with nine samples (n = 9) per group (b) or twelve samples (n = 12) per group (d) (mean and s.e.m. in b,d) or from two independent experiments with six samples (n = 6) per group (f).

nature immunology  VOLUME 16  NUMBER 3  MARCH 2015

287

Articles

5

9

13

b

Il7fThy-1.1

2.0

Thy-1.1

Counts/s (×105)

Il7f + α-Thy-1.1 Il1r1–/– Il1r1–/– Il7fThy-1.1 + α-Thy-1.1

Luminescence (×103) 1 2 3 4 5

c

Il17fThy-1.1 Il17fThy-1.1 + α-Thy-1.1 Il1r1–/– Il1r1–/– Il17fThy-1.1 + α-Thy-1.1

600

** ** **

** ** **

1.5

** ** **

1.0

** **

0.5 0

5

7

9

11

**

6

400 300 200

**

100

**

** UI

1

3

5

7

10

* 13

**

4 3 2

**

1 0

UI

Time (d)

13

**

5

1

3

5

7

10

13

Time (d)

+

–

Il1r1+/+ (CD45.1 ) Il1r1–/– (CD45.1 )

f

+/+

considerably by RA, which supported the retention of Foxp3 expression and suppressed IL-17 expression. Similar to its effects on developing TH17 cells, IL-1 signaling overrode the effects of RA to support Foxp3 expression and repress IL-17 expression. Thus, signaling via the IL-1 receptor overrode RA-mediated induction of iTreg cell programming to promote and stabilize TH17 cell programming and thereby shift the TH17 cell–iTreg cell balance in favor of TH17 cell development. IL-1R1 deficiency impairs the development of TH17 cells To explore the role of IL-1 signaling in modulating the T H17 cell–iTreg cell balance in vivo, we assessed infection by the intestinal pathogen C. rodentium, protection against which is mediated by the TH17 pathway3,21,22. It has been suggested on the basis of studies of steady-state intestinal T H17 cell development that conflicting reports on the role of IL-1 have arisen due to use of the stimulation of recovered T cells ex vivo with the phorbol ester PMA and ionomycin17,23. We therefore used IL-1R1-sufficient and IL-1R1-deficient Il17fThy-1.1Foxp3GFP mice to directly examine gene expression ex vivo without requirement for an in vitro recall response induced by stimulation with PMA plus ionomycin or with antibody to the invariant signaling protein CD3 (anti-CD3)24. Because Il17f expression occurs early in TH17 cell development, at which time it is dominant over Il17a expression24, the Il17fThy-1.1 reporter model provides a sensitive ‘readout’ of early gene expression in TH17 cells. Moreover, the Il17fThy-1.1 reporter model enables specific depletion 288

IL-17F+ cells (%)

IL-17F (Thy-1.1)

Il1r1 Figure 2  IL-1R signaling is required NS 0.1 0 0.1 0 Il1r1–/– 40 for the host-protective TH17 and ** 4 10 TH17 cell–iTreg cell balance in vivo. SPL 30 103 (a) Serial whole-body imaging 16.7 17.5 2 95.5 10 94.1 of Il17fThy-1.1, Il1r1−/− and Il1r1−/− 20 4 10 Il17fThy-1.1 mice inoculated with a 101 32.4 0.3 2.5 0.1 luminescent strain of C. rodentium 103 100 10 100 101 102 103 104 ** ** NS LPL and left untreated or treated with an * 102 CD45.1 + IL-17F cell–depleting mAb to Thy-1.1 0 101 UI Inf UI Inf UI Inf (α-Thy-1.1), imaged 5, 9 or 13 d after infection (above images). 19.0 78.1 100 SPL MLN LP 100 101 102 103 104 (b) Kinetics of the colonization of mice as in a at 5–13 d after Foxp3 infection, presented as counts per second. (c,d) Quantitative enzyme-linked immunosorbent assay (ELISA) of IL-17A (c) and IFN-γ (d) in supernatants of cultured homogenates of colonic tissue collected from Il17fThy-1.1 and Il1r1−/− mice left uninfected (UI) or 1–13 d after inoculation with C. rodentium. (e) Flow cytometry (right) of Il1r1+/+ (CD45.1+) and Il1r1−/− (CD45.1−) splenic lymphocytes (SPL) and colonic lamina propria lymphocytes (LPL) from recipient Tcrb−/− mice reconstituted with a mixture of 3 × 106 Il1r1+/+Il17fThy-1.1 (CD45.1+CD45.2+) and Il1r1−/−Il17fThy-1.1 (CD45.2+) donor CD4+ T cells, followed by infection of recipients with C. rodentium 2 weeks after reconstitution (protocol, Supplementary Fig. 2d) and analysis, 7 d later, of the expression of Thy-1.1 (IL-17F) and intracellular Foxp3 (with gating on activated CD4+ T cells). Far left, numbers above outlined areas indicate percent CD45.1+ (Il1r1+/+) (right) or CD45.1− (Il1r1−/−) (left) TCRβ+ donor cells in the recipient mice. (f) Frequency of IL-17F+ (Thy-1.1+) cells among Il1r1+/+ and Il1r1−/− populations from the spleen, MLNs and lamina propria of the large intestine (LP) of recipient Tcrb−/− mice reconstituted as in e and left uninfected (UI) or infected (Inf) with C. rodentium 2 weeks after reconstitution. NS, not significant; *P < 0.05 and **P < 0.01 (two-tailed unpaired t-test). Data are from one experiment representative of two (anti-Thy-1.1 treated Il1r1−/− mice) or three (untreated Il17fThy-1.1 or Il1r1−/− mice or anti-Thy-1.1 treated Il17fThy-1.1 mice) independent experiments with similar results (a), are pooled from two or three independent experiments with nine to eleven mice per group (b; mean and s.e.m.) or two independent experiments with six mice per group (f; mean and s.e.m.), are from two independent experiments with six mice per group (c,d; mean and s.e.m.) or are representative of one of two similar independent experiments (e). TCRβ

© 2015 Nature America, Inc. All rights reserved.

7

500

0

Il17fThy-1.1 Il1r1–/–

Time after infection (d)

e

npg

d Il17fThy-1.1 Il1r1–/–

IFN-γ (ng/ml)

Time (d):

IL-17A (pg/ml)

a

of IL-17F-producing cells in vivo by administration of monoclonal antibody (mAb) to Thy-1.1 (ref. 25). Depletion of IL-17F-producing cells from C. rodentium–inoculated Il17fThy-1.1 single- reporter mice through the use of mAb to Thy-1.1 during the peak of infection (3–7 d after infection21; data not shown) resulted in impaired clearance of bacteria and heightened injury of the intestinal mucosa (Fig. 2a,b and Supplementary Fig. 2a,b). Infection of mice deficient in IL-1R1 (Il1r1−/− mice) showed similarly impaired host protection with a concomitant decrease in colonic production of IL-17A but not of interferon-γ (IFN-γ) (Fig. 2c,d). Those results were similar to published findings obtained with mice deficient in the IL-17 receptor component IL-17RA and were consistent with the possibility of a role for IL-1 in promoting TH17 pathway–dependent host protection21. Depletion of IL-17-producing cells from Il1r1−/−Il17fThy-1.1 mice did not significantly alter this result (Fig. 2a,b), which suggested that a major effect of IL-1R1 deficiency during this period of the infection was mediated by its actions on these cells. Production of IL-22 in the infected colon was reduced in IL-1R1-deficient mice, but only at the peak time point (day 7 after infection; Supplementary Fig. 2c). We observed a transient increase in Foxp3+CD4+ T cells in the large intestine and draining lymph nodes of IL-1R1-sufficient mice during infection (days 3–5; Supplementary Fig. 3a,b), as well as a shift in the balance of IL-17+ cells versus Foxp3+ cells recovered from the large intestine of infected IL-1R1-deficient mice (Supplementary Fig. 3c,d). Therefore, we posited that infection-induced IL-1 signaling in CD4+ T cells might be needed to overcome the Treg cell VOLUME 16  NUMBER 3  MARCH 2015  nature immunology

Articles

1.7

Inf TCRβ-deficient

IL-1R1-deficient 0.1

0.4

0.0

WT 0.7

IL-1R1-deficient 0.0

0.4

0.0

SPL 27.5 1.3

0.1

32.6 1.0

0.7

6.7 3.3

0.0

8.9 0.6

b

0.1

c

WT IL-1R1-deficient

IL-17F (Thy-1.1)

43.4

1.4

20.5 12.3

0.5

3.2 95.5

0.8

14.2 47.6

1.4

103 10

2

10

1

10

0

0.5

29

0.1

Foxp3+ cells (%)

10.1 104

15.1

100 101 102 103 104

Foxp3 (GFP)

Veh 0.4

**

*

0.4

0.6

0.3

68.7 4.1

7.9

MLN

30 20

68.3

**

** **

104

**

10

LP

0 UI

Inf

UI Inf MLN

SPL

UI

Inf

IL-17F (Foxp3 ) cells (%)

8.7

3.2

67.5 55

10.3

103 10

2

101 100

LPL

72.6

9.5

100 101 102 103 104

Foxp3 (GFP)

e 70

Veh LE135

**

+

+

Figure 3  In the absence of IL-1β, in vivo blockade of RA facilitates the 60 + Veh iTreg cell–to–TH17 cell conversion during enteropathogenic bacterial 50 infection. (a) Expression of Thy-1.1 (IL-17F) and Foxp3-GFP on splenic 40 and MLN cells and LPLs isolated from Tcrb−/− recipients left uninfected 30 6 + or infected with C. rodentium 2 d before transfer of 2.5 × 10 Foxp3 20 ** IL-17F− donor CD4+ T cells (collected from C. rodentium–infected Il1r1+/+ 10 + LE135 −/− Thy-1.1 GFP Foxp3 mice (IL1R1-deficient) and sorted mice (WT) or Il1r1 Il17f 0 on day 2 after infection; protocol, Supplementary Fig. 3e), analyzed by flow cytometry 8 d after infection of recipients, without restimulation ex vivo, 70 ** Veh LE135 with gating on TCRβ+ cells. (b) Frequency of IL-17F+ (Thy-1.1+) cells and 60 Veh 50 LE135 50 Foxp3-GFP+ cells as prepared in a. (c) Expression of Thy-1.1 (IL-17F) and 40 40 Foxp3 (GFP) on CD4+CD25+ T cells from the spleen, MLN and lamina 30 30 propria of Il1r1−/−Il17fThy-1.1Foxp3GFP mice inoculated with 2 × 109 ** 20 ** 20 colony-forming units of C. rodentium and given vehicle alone (Veh) or ** 10 10 ** the RA inhibitor LE135 by gavage on days 3–7 after infection, analyzed 0 0 by flow cytometry on day 8 after infection, without restimulation ex vivo. SPL MLN LPL Epi Infl Total + + + (d) Frequency of IL-17F (Thy-1.1 ) cells and Foxp3-GFP cells among the + + + + −/− Thy-1.1 GFP Foxp3 mice Foxp3 (GFP ) or activated (CD4 CD25 ) fraction of splenic and MLN cells and LPLs of C. rodentium–infected Il1r1 Il17f treated with vehicle or LE135. Each symbol represents an individual mouse; small horizontal lines indicate the mean. (e) Histopathology of hematoxyolin- and eosin-stained colonic tissue sections derived from vehicle- or LE135-treated Il1r1−/−Il17fThy-1.1Foxp3GFP mice at day 10 after infection with C. rodentium. (f) Histopathology scores of colons from mice as in e, assessing the epithelium (Epi) and inflammation (Infl), as well as the aggregate histopathology score (Total). *P < 0.05 and **P < 0.01 (two-tailed unpaired t-test). Data are from one experiment representative of three (a) or two (c,e) independent experiments with similar results (a,c,e) or are pooled from three (b) or two (d,f) independent experiments with ten (b), six (d) or four (f) mice per group (b,d,f; mean and s.e.m.).

programming that is favored in intestinal tissues by the production of RA, as was observed ex vivo. To test this hypothesis, we obtained congenically marked CD4+ T cells from IL-1R1-sufficient (CD45.1+) Il17fThy-1.1 mice or IL-1R1-deficient (CD45.1−) Il17fThy-1.1 mice and transferred those cells into recipient mice deficient in the T cell antigen receptor (TCR) β-chain (Tcrb−/− mice). We either left recipient mice untreated (uninfected) or inoculated them with C. rodentium (infected), then assessed the frequency of Foxp3+ cells and IL-17F+ cells. Although a large majority of the transferred T cells were unreactive to C. rodentium antigens, assessment of the frequency of Foxp3+ or IL-17F+ CD4+ T cells among the pool of recently activated cells in the lamina propria of the large intestine showed a marked shift toward higher IL-17F expression by wild-type T cells than by IL1R1-deficient T cells (>6-fold), with a reciprocal decrease in the frequency of Foxp3+ T cells (>2.5-fold) (Fig. 2e,f and Supplementary Fig. 2d). In contrast, there were no significant differences in frequency of Foxp3+ or IL-17F+ cells recovered from the spleens of recipient mice. These findings were consistent with a defect in the transition of activated T cells from iTreg cells to TH17 cells in the absence of IL-1 signaling. nature immunology  VOLUME 16  NUMBER 3  MARCH 2015

f

Histopathologic score

+

+

IL-17F (CD25 ) cells (%)

© 2015 Nature America, Inc. All rights reserved.

0.5

87.9

d

npg

LE135 0.5

SPL

**

40

MLN

LPL

100 90 80 70 60 50 40 30 20 10 8 4 0

IL-17F (Thy-1.1)

UI TCRβ-deficient WT

IL-17F+ cells (%)

a

In vivo blockade of RA compensates for IL-1 signaling deficiency To further investigate the possible transition of Foxp3+ precursor cells into IL-17-producing effector cells, we tracked the fate of Foxp3+IL17F− CD4+ T cells isolated from infected IL-1R1-sufficient and IL-1R1-deficient Il17fThy-1.1Foxp3GFP mice after their transfer into infected TCRβ-deficient recipient mice. In accord with the possibility of a role for IL-1 signaling in promoting the silencing of Foxp3 and induction of IL-17, we recovered a significantly greater frequency of Foxp3+ T cells from the recipients of IL-1R1-deficient Foxp3 + T cells than from the recipients of IL-1R1-sufficient Foxp3+ T cells, with a deficit in IL-17+ T cells (Fig. 3a,b and Supplementary Fig. 3e). We then determined whether the impaired TH17 cell development in the IL-1R1-deficient mice was due to a failure to overcome RA-mediated repression. We treated C. rodentium–infected IL-1R1-deficient Il17fThy-1.1Foxp3GFP mice with an antagonist of RAR or vehicle following inoculation with C. rodentium (Fig. 3c,d). Blockade of RAR signaling in infected IL-1R1-deficient mice significantly reversed the deficit in IL-17-producing T cells in both the lamina propria of the large intestine and MLNs but not in the spleen; this blockade in the IL-1R1-deficient mice resulted in decreased bacterial loads associated 289

Articles 9.4

2.8

16.5

2.3

25.6

2.2

b

– + IL-1β

+ IL-2 + IL-1β + IL-2

c

30 1.6

0.6

17.7 1.5

0.7

13.8 2.6

1.1

8.1 2.2

0.6

Foxp3

103

+ IL-1β

102 101

29.2

27.6

100 100 101 102 103 104

28.8

30.6

IL-17F (Thy-1.1)

Cells (%)

20.8 104

*

25 20

*

15

* *

10 5 0

1

10

d

2 1

*

5.9

43.6

1.3

3.4

1.0

+ RA + IL-1β 9.5

5.4

WT 51.4 104

1.5

1.4

6.5 7.6

1.7

5.1

– + + +

– + – –

– + + +

Stat5fl/fl Cd4-Cre

f

0.7

6.2

1.7

1

102 101

62.6

100 100 101 102 103 104

IL-17A

50.3

*

10 RA (nM)

100

41.2

69.8

WT 70 60 50 40 30 20 10 0 50

+

– + – –

*

*

0

100

60.9

44.1

103

Foxp3

0 RA IL-1β

*

1 10 RA (nM)

+

* Cd122 (fold)

Cd25 (fold)

© 2015 Nature America, Inc. All rights reserved.

npg

*

*

+

+ RA + α-IL-2

+ RA 9.2

3

IL-17F

*

*

7 6 5 4 3 2 1 0

100

e 4

7 6 5 4 3 2 1 0

0 1 10 100

+

0

+ IL-1β

0

IL-2: 0 1 10100 Foxp3

rhIL-2 (U/ml):

IL-2 (ng/ml)

35

IL-2 (ng/ml)

2.8

IL-17 cells (%)

8.4

Foxp3 cells (%)

a

fl/fl

Stat5

Cd4-Cre

*

* * *

40 30 20 10 0

+ RA

+ RA + RA + α-IL-2 + IL-1β

Figure 4  IL-1β counteracts RA-driven, IL-2- and STAT5-dependent repression of T H17 cell development. (a) Expression of IL-17F (Thy-1.1) and Foxp3 by naive CD4+ T cells (CD4+CD25−CD62LhiCD44lo) obtained from OT-II Il17fThy-1.1 mice (which have transgenic expression of an ovalbumin-specific TCR) and activated for 4 d with ovalbumin peptide and IL-1β-deficient DCs under TH17-polarizing conditions with or without IL-1β, and with or without (0) the addition of increasing doses (wedge) of recombinant human IL-2 (rhIL-2) at day 4. (b) Frequency of cells producing only Foxp3 or IL-17F, among OT-II CD4+ T cells polarized as in a. (c) Quantitative ELISA of IL-2 in supernatants of Il17fThy-1.1 naive CD4+ T cells activated with plate-bound anti-CD3 and soluble anti-CD28 under TH17-polarizing conditions in the absence (left) or presence (right) of IL-1β, with or without the addition of increasing concentrations (1–100 nM) of RA. Each symbol represents an individual culture; small horizontal lines indicate the mean. (d) Quantitative RT-PCR analysis of Cd25 and Cd122 mRNA in Il17fThy-1.1 naive CD4+ T cells activated for 60 h with plate-bound anti-CD3 and soluble anti-CD28 under TH17-polarizing conditions in the presence (+) or absence (−) of IL-1β, with (+) or without (−) the addition of RA (1 nM); results are presented relative to those of cells cultured under TH0 conditions (no cytokines). (e) Expression of IL-17A and Foxp3 by CD8−CD4+ thymocytes isolated from wild-type C57BL/6 mice (WT) or mice with T cell–specific deletion of STAT5 (Stat5fl/flCd4-Cre), then purified by magnetic-activated cell sorting and activated for 4 d with soluble anti-CD3 and Il1b−/− splenic DCs under TH17-polarizing conditions with or without the addition of various combinations (above plots) of RA (1 nM), anti-IL-2 (10 µg/ml) and/or IL-1β (20 ng/ml). (f) Frequency of IL-17A+ or Foxp3+ CD4+ T cells from mice as in e. *P < 0.01 (two-tailed unpaired t-test). Data are from one experiment representative of three independent experiments with similar results (a,e), are pooled from three (b,f) or two (c) independent experiments with nine samples per group (n = 9) (b), six samples (c) or ten samples per group (n = 10) (f) (mean and s.e.m. in b,c,f), or are representative of two independent experiments with six samples per group (n = 6) (d; mean and s.e.m.).

with protection from mucosal injury and less weight loss, but such blockade in their IL-1R1-sufficient counterparts did not (Fig. 3e,f and Supplementary Fig. 4). Collectively, these findings identified a role for pathogen-induced IL-1 signaling in overriding the iTreg cell programming favored by RA-producing intestinal DCs to induce protective TH17 cell responses to microbial antigens. IL-1b counters RA-driven repression of TH17 cells RA suppresses TH17 cell development in favor of iTreg cell development at least in part via an IL-2-dependent mechanism9,14. To determine whether IL-1 signaling might also modulate IL-2-dependent suppression of TH17 differentiation, we assessed the effect of adding increasing concentrations of IL-2 to cultures of naive CD4+ T cells activated by antigen under TH17-polarizing conditions, with or without the addition of IL-1β. In the absence of added IL-1β, increasing concentrations of IL-2 suppressed the frequency of IL-17+ cells, with a reciprocal increase in the frequency of Foxp3+ cells (Fig. 4a,b), as reported14. IL-1β abrogated the suppressive effects of IL-2, regardless of the concentration of IL-2 (Fig. 4a,b). We obtained comparable effects of IL-1β on IL-17 expression and Foxp3 repression under DC-free conditions of TH17 polarization, wherein the induction 290

of higher concentrations of endogenously produced IL-2 favored less-robust TH17 cell development and higher expression of Foxp3 (Supplementary Fig. 5 and data not shown). Notably, IL-23 signaling had no comparable effect on the TH17 cell–iTreg cell balance during the early stages of differentiation (3–4 d), when IL-1 signaling induced an enhanced TH17 response, and the addition of both IL-23 and IL-1β did not augment the effect of IL-1β alone (Supplementary Fig. 5b,c and data not shown). So, in line with its effects on overriding the RA-mediated repression of TH17 differentiation, IL-1β antagonized the effects of IL-2, and this effect was independent of a requirement for co-signaling via IL-23. The similar effects of IL-1 signaling on the reversal of RA- and IL-2mediated inhibition of TH17 differentiation were consistent with our finding that the production of IL-2 by T cells was enhanced about threefold by the addition of RA, at all concentrations tested (Fig. 4c). This result indicated that the RA-mediated repression of TH17 differentiation was due at least in part to enhanced IL-2 production. Consistent with the actions of IL-2 in inducing higher expression of the inducible components of IL-2R, the addition of RA also upregulated the expression of IL-2Rα (CD25) and IL-2Rβ (CD122) (Fig. 4d). Notably, neither the enhanced IL-2 production nor the upregulation VOLUME 16  NUMBER 3  MARCH 2015  nature immunology

Articles

RA + IL-1β

IL-1β

IL-6

Time (h):

c

IL-6 + IL-1β

Time (h):

63.3

STAT3

10 30 60

10 30 60

IL-23

IL-1β

IL-23 + IL-1β

5

ND

1.0

2.3

k2

i k2

Ja +

IL-23 + IL-1β

pY-STAT3 2.5

0 Time (min):

8.0

2.5

7.9

12.5

2.2

10.0 7.5 5.0 2.5

STAT3 0

0

30 60120 30 60120 30 60120 IL-23

IL-1β

*

i

i κB IL-23

*

– IL-23

IL-23 + IL-1β

IL-23 +IL-1β

N – F+ κBi Jn + k2i Ja k2 i

10 30 60

–

*

0

g

5.0

**

*

10

f

*

7.5

**

15

Socs3 mRNA (fold)

1

0

20

N F-

Socs3 mRNA (fold)

2

5

of the expression of components of the IL-2 receptor by RA was significantly affected by IL-1β (Fig. 4c,d), which suggested that the principal effect of IL-1β in blocking the effects of RA was probably due to interference with IL-2 signaling downstream of receptor binding of IL-2. Accordingly, blockade of IL-2 signaling reversed the reciprocal effects of RA on the TH17 cell–iTreg cell developmental balance, similar to the effects induced by IL-1β (Fig. 4e,f). Furthermore, in CD4+ T cells deficient in STAT5 expression (and thus IL-2 signaling), the effects of RA in repressing IL-17 expression in favor of Foxp3 expression were similarly overridden and were not further augmented by IL-1β signaling (Fig. 4e,f). Thus, nature immunology  VOLUME 16  NUMBER 3  MARCH 2015

0

0

3.

1.

5

0.

0

3.

5

1.

0.

0

0

3.

5

1.

0.

0

0.

0

0

3.

5

1.

0

0.

0

3.

5

1.

0

0.

0

3.

5

1.

0.

0

h

0

+

Figure 5  NF-κB-dependent repression of SOCS3 by IL-1β enhances the amplitude IL-1β IL-23 + IL-1β IL-1β IL-23 + IL-1β IL-23 IL-23 and duration of STAT3 phosphorylation. Time (h): (a) Intracellular tyrosine-phosphorylated STAT5 and STAT3 in wild-type (C57BL/6) pY-STAT3 naive CD4+ T cells left unstimulated (US) 1.0 7.1 5.3 4.5 0.1 0.5 0.4 11.2 11.9 10.8 1.0 7.8 5.5 2.9 1.4 1.6 1.7 6.3 5.5 3.0 or activated (Stim) for 4 d with plate-bound anti-CD3 and soluble anti-CD28 under STAT3 TH17-polarizing conditions with or without the WT addition of RA alone or RA plus IL-1β (above plots), Relafl/fl Cd4-Cre followed by the addition of IL-6 and IL-2 for 20 min and analysis by flow cytometry. Numbers above bracketed lines indicate percent cells with tyrosine-phosphorylated (pY-) STAT5 (top row) or STAT3 (bottom row). (b) Immunoblot analysis (top) of tyrosine-phosphorylated STAT3 and total STAT3 in lysates of naive CD4 + T cells cultured for 4–5 d under TH17-polarizing conditions, then restimulated for various times (above lanes) with IL-6 or IL-1β alone or both IL-6 and IL-1β. Below, quantification of tyrosine-phosphorylated STAT3 in stimulated cells relative to that in unstimulated cells (stim/US). (c) Immunoblot analysis of STAT3 phosphorylated at Tyr705 in naive CD4+ T cells cultured for 4 d under TH17-polarizing conditions, then restimulated for various times (above lanes) with IL-23 or IL-1β alone or both IL-23 and IL-1β, assessed as in b. (d) ELISA of active Jak2 (phosphorylated at Tyr1007 and Tyr1008) and total Jak2 in lysates of naive CD4+ T cells cultured under TH17-polarizing conditions and activated as in c, results for tyrosine-phosphorylated Jak2 were normalized to total Jak2 and results are presented relative to those obtained for unstimulated T H17 cells, set as 1. (e) Quantitative RT-PCR analysis of Socs3 transcripts in naive CD4+ T cells cultured under TH17 conditions as in c and restimulated for various times (horizontal axis) with IL-23 and/or IL-1β; results for Socs3 were normalized to those of the gene encoding β2-microglobulin and are presented relative to those of unstimulated cells (calculated by the change-in-threshold (∆∆Ct) method). (f) Immunoblot analysis (as in c) of STAT3 phosphorylated at Tyr705 and total STAT3 in lysates of naive CD4+ T cells polarized under TH17 conditions, without inhibitory pretreatment (−) or pretreated for 1 h with various signaling inhibitors (‘i’), then left unstimulated (−) or treated with IL-23 and/or IL-1β. Numbers below lanes indicate integrated density value of tyrosinephosphorylated STAT3 relative to that of total STAT3 (average values of pooled data). (g) Quantitative RT-PCR analysis (as in e) of Socs3 transcripts in naive CD4+ T cells polarized under TH17 conditions, pretreated with inhibitors as in f, then restimulated for 45 min with IL-23 and/or IL-1β. (h) Immunoblot analysis (as in c) of STAT3 phosphorylated at Tyr705 and total STAT3 in lysates of wild-type and Relafl/flCd4-Cre naive CD4+ T cells polarized as in c (numbers between lanes as in f; red (IL-23 + IL-1β) indicates values significantly different (P < 0.01) from the corresponding values for cells stimulated with IL-23 alone). ND, not detectable; *P < 0.05 and **P < 0.01 (two-tailed unpaired t-test). Data are from one experiment representative of two (a,f,h) or three (b,c) independent experiments with similar results (a–c,f,h; mean and s.e.m. in b,c) or are pooled from two (d,g) or three (e) independent experiments with six samples per group (n = 6) (d) or nine samples (n = 9) per group (e,g) (mean and s.e.m. in d,e,g). 0.

© 2015 Nature America, Inc. All rights reserved.

90.7

e

*

3

0 Time (min): 0

npg

49.1

pY-STAT3

pY-Jak2/Jak2

d

47.4

10

–

100 80 60 40 0.2 20 0 1 0 2 3 4 10 10 10 10 10

*

15

pY-STAT3 (stim/US)

pY-STAT5

STAT3

**

20

Jn

72.1

pY-STAT3 (stim/US)

46.4

IL-23 + IL-1β

pY-STAT3

pY-STAT3 10.6

IL-1β

IL-23

+

RA

+

–

0. 0 0. 2 0. 5 1. 0 2. 0 4. 0 0. 2 0. 5 1. 0 2. 0 4. 0 0. 2 0. 5 1. 0 2. 0 4. 0

b

Stim US

0. 0 0. 5 1. 0 3. 0 0. 5 1. 0 3. 0 0. 5 1. 0 3. 0

a

although RA acted to enhance IL-2 production and expression of the inducible components of the IL-2 receptor by T cells, the reversal of RA-mediated effects by IL-1β signaling appeared to be mediated mainly through interference of IL-2 signaling and not through altered expression of IL-2 or its receptor. IL-1b enhances phosphorylated STAT3 via repression of SOCS3 To explore mechanisms by which IL-1 might interfere with the RA-mediated modulation of IL-2 signaling, we assessed the tyrosine-phosphorylation of STAT5 following stimulation of T cells under TH17-polarizing conditions with or without the addition 291

Articles

– 20.7 Rosa26TdTomato fl/+ Control

npg

© 2015 Nature America, Inc. All rights reserved.

TdTomato fl/+

+TAT-Cre

0.7

42.7

0.5

35.2

b

RA + IL-1β

RA + IL-23 0.2

13.7

100

12.3

2.5

2.1

0.8

15.2

40

78

0

100 101 102 103 104 TdTomato

15.6

0.6

12.0

0.3

11.4

2

10

1

10

19.2

100 0 1 2 3 4 10 10 10 10 10 IL-17A

13.3

14.7

17.1

15

*

*

*

*

10 5 50

0.5

103

fl/fl

Socs3 +Tat-Cre

20

1.4

Foxp3+ cells (%)

104 14.5

NS

20

0

80 60

Socs3fl/fl +Control Socs3fl/fl +Tat-Cre

25

Socs3fl/fl +Control

Foxp3

Events (% of max)

Rosa26

RA

IL-17+ cells (%)

a

40 30 20

NS

10 0

–

RA

RA RA + IL-23 + IL-1β

Figure 6  Deletion of SOCS3 abrogates the IL-1-dependent reversal of RA’s repression of T H17 cell development. (a) Expression of Foxp3 and IL-17A in Socs3fl/fl naive CD4+ T cells cultured, with (+Tat-Cre) or without (+Control) Tat-Cre peptide, under T H17-polarizing conditions only (−) or with the addition various combinations (above plots) of RA, IL-1β and IL-23 (right), and expression of the red fluorescent protein TdTomato from Rosa26TdTomato fl/+ reporter mice (with a loxP-flanked ‘stop’ cassette preceding sequence encoding TdTomato knocked into one of the the ubiquitously expressed Rosa26 alleles) given no Tat-Cre peptide (Control) or treated with Tat-Cre peptide (+Tat-Cre) and cultured similarly under TH17 conditions (far left). Number above bracketed line (far left) indicates percent TdTomato + cells (an indicator of Cre-mediated deletion efficiency). (b) Frequency of IL-17A+ or Foxp3+ cells among CD4+ T cells treated as in a. *P < 0.01 (two-tailed unpaired t-test). Data are from one experiment representative of four (a, far left) or two (a, right) independent experiments (a) or are pooled from two independent experiments with 12 samples (n = 12) per group (b; mean and s.e.m.).

of either RA or RA plus IL-1β. Consistent with its effects on enhancing the expression of IL-2 and its receptor, RA enhanced the tyrosine-phosphorylation of STAT5 in TH17 cells, a result that was not altered substantially by concurrent IL-1 signaling (Fig. 5a). In contrast, RA had no effect on IL-6-induced tyrosine-phosphorylation of STAT3, whereas the addition of IL-1 resulted in a substantial increase in tyrosine-phosphorylated STAT3 (Fig. 5a). The effect of IL-1 signaling on differentiating TH17 cells resulted in both increased and more-sustained tyrosine-phosphorylation of STAT3 (Fig. 5b) and was not limited to IL-6-induced tyrosine-phosphorylation of STAT3, as we obtained similar results when STAT3 was activated downstream of signaling by IL-21 (Supplementary Fig. 6) or IL-23 (Fig. 5c). Therefore, while IL-1 signaling failed to directly repress the RA-mediated enhancement of IL-2-induced tyrosine-phosphorylation of STAT5, it did alter the ratio of tyrosine-phosphorylated STAT5 to tyrosine-phosphorylated STAT3 through its enhancement of IL-6-, IL-21- and IL-23-induced tyrosine-phosphorylation of STAT3, which established an indirect mechanism by which IL-1 might act in concert with each of the STAT3-activating TH17-pathway cytokines to override RA-mediated, IL-2-dependent repression of TH17 cell development. Because of its failure to affect RA-mediated effects on the TH17 cell–iTreg cell balance during early differentiation (Supplementary Fig. 5), we chose IL-23 to further investigate mechanisms by which IL-1 signaling might alter the abundance of phosphorylated STAT3. We initially assessed the effects of IL-1 upstream of the tyrosinephosphorylation of STAT3. The addition of IL-1 significantly increased and sustained the rapid tyrosine-phosphorylation of Jak2 induced by activation of the IL-23 receptor (Fig. 5d). The autophosphorylation of Jak2 that is induced by heterodimerization of the IL-23 receptor activates the tyrosine kinase activity of Jak2, which in turn acts to tyrosine-phosphorylate STAT3 that is recruited to the receptor complex. The tyrosine-phosphorylation of Jak2 and STAT3 is competitively inhibited by SOCS3, which is rapidly induced by STAT3 activation in a negative feedback loop26. We reasoned, therefore, that IL-1 might modulate the phosphorylation of Jak2 and/or STAT3 through effects on SOCS3 expression. Indeed, while IL-1 signaling 292

alone had no effect on the expression of Socs3 mRNA, it significantly inhibited the expression of Socs3 mRNA induced by IL-23 (Fig. 5e), which indicated that IL-1-dependent inhibition of Socs3 expression contributed to the enhanced tyrosine-phosphorylation of STAT3. NF-κB and mitogen-activated protein kinase signaling cascades are activated downstream of ligation of the IL-1 receptor. We therefore assessed the effects of specific inhibitors of NF-κB and the kinases Jnk2, Erk and p38 on IL-1-induced modulation of the tyrosinephosphorylation of STAT3 and modulation of Socs3 mRNA; we included an inhibitor of Jak2 as a control. Notably, of the inhibitors of IL-1 signaling examined, only the inhibitor of NF-κB blocked the enhancement of tyrosine-phosphorylation of STAT3 induced by IL-23, and this was associated with reversal of the suppression of Socs3 expression (Fig. 5f,g and Supplementary Fig. 7). This result suggested that IL-1-induced activation of NF-κB might repress SOCS3 expression as a mechanism for enhancing the abundance of tyrosine-phosphorylated STAT3. Accordingly, TH17 cells derived from mice with specific deficiency in the NF-κB component RelA in CD4+ T cells (Relafl/flCd4-Cre mice) showed no enhancement in the induction of tyrosine-phosphorylation of STAT3 by co-signaling via IL-1 and IL-23 (Fig. 5h). That finding was in agreement with a published study of a hepatocyte cell line, in which after a transient, modest early enhancement in the induction of SOCS3 expression, IL-1 repressed IL-6-induced expression of SOCS3 via NF-κB27. In addition to its effects on the tyrosine-phosphorylation of STAT3, we found that IL-1 signaling increased the phosphorylation of Ser727 of STAT3 via the p38 kinase pathway (Supplementary Fig. 7a,b), an effect that has been shown to potentiate the actions of the STAT3 homodimer28. However, this effect was more transient than the IL-1-induced tyrosine phosphorylation of STAT3 and was independent of a requirement for co-signaling via IL-23 (or IL-6) (Supplementary Fig. 7a,b). This suggested that any effects on the serine-phosphorylation of STAT3 would probably be secondary to the principal effects on tyrosine-phosphorylation modulated by the NF-κB–SOCS3 axis. Finally, we directly established a role for IL-1-mediated repression of SOCS3 in overriding the RA-mediated inhibition of TH17 differentiation, as the inhibitory effect of RA was largely mitigated by specific VOLUME 16  NUMBER 3  MARCH 2015  nature immunology

Articles b

WT TH0 US WT TH17

WT TH0 US WT TH17

WT TH17 + RA

**

4

**

2

0

© 2015 Nature America, Inc. All rights reserved.

**

**

**

npg

**

II17a–5

** II17a P II17a+10

**

**

II17a+28 II17f–5

STAT5-deficient TH17 + RA

** **

8 6 4

** **

**

**

** **

**

**

2

** II17f P

0

II17a–5

II17a P II17a+10

II17a+28 II17f–5

II17f P

2.5 STAT3 binding (% of input)

* 6

WT TH17 + RA + IL-1β

10

STAT5-deficient TH17 + RA

STAT5 binding (% of input)

STAT3 binding (% of input)

8

WT TH17 + RA

WT TH17 + RA + IL-1β

**

2.0 1.5 1.0

** 0.5 0

6 STAT5 binding (% of input)

a

**

5 4 3 2 1 0

Foxp3 CNS2

Figure 7  IL-1β reverses the RA-mediated binding of STAT5 to the Il17a-Il17f and Foxp3 loci. Chromatin-immunoprecipitation of wild-type or Stat5fl/fl Cd4-Cre (STAT5-deficient) purified CD4+CD8− thymocytes that were activated for 4 d with plate-bound anti-CD3 and soluble anti-CD28 and cultured under TH0 conditions (no cytokines) only or under TH17 conditions (IL-6 plus TGF-β) alone or with the addition of various combinations of RA and IL-1β (key), then left unstimulated (TH0) or stimulated with IL-6 (STAT3 analysis) or IL-6 plus IL-2 (STAT5 analysis) (T H17), followed by PCR analysis of the binding of STAT3 and STAT5 to a region 5 kb upstream of the Il17a transcription start site (TSS) (Il17a–5), the Il17a promoter (Il17aP), the intergenic enhancer 10 kb downstream of the Il17a TSS (Il17a+10), a region 28 kb downstream of the Il17a TSS (Il17a+28), a region 5 kb upstream of the Il17f TSS (Il17f−5) or the Il17f promoter (Il17fP) (a) and of CNS2 of Foxp3 (b); results are presented relative to input DNA. *P < 0.05 and **P < 0.01 (twotailed unpaired t-test). Data are representative of two independent experiments with three replicates per experiment (mean and s.e.m.).

deletion of Socs3 in developing TH17 cells (Fig. 6). Thus, delivery of Cre recombinase to naive CD4+ T cells from Socs3fl/fl mice via a TatCre fusion protein led to a reduction in RA-induced Foxp3 expression and enhancement of IL-17 that was similar in the presence of IL-23 or IL-1β (Fig. 6). Collectively, these data identified a SOCS3-dependent mechanism by which IL-1-induced NF-κB signaling regulated the amplitude and duration of STAT3 tyrosine-phosphorylation induced by TH17-polarizing cytokines. IL-1b modulates STAT3 and STAT5 binding to consensus sequences Given the dominant effect of IL-1 in overriding RA- and IL-2-induced repression of TH17 cell development, we reasoned that the actions of IL-1 signaling in enhancing the phosphorylation of STAT3 might overcome competition by phosphorylated STAT5 for binding to shared STAT-binding consensus sequences induced by RA-dependent IL-2 signaling15. We performed chromatin-immunoprecipitation analyses to quantify the relative binding of phosphorylated STAT3 and STAT5 to consensus regulatory sites in the Il17a-Il17f and Foxp3 loci under various conditions of the addition RA and IL-1β, in TH17-polarized cells activated by IL-6 alone (STAT3) or IL-6 plus IL-2 (STAT5). As a control for the STAT5-dependent effects of RA, we included TH17-polarized cells deficient in STAT5 (Stat5−/− cells). We found that, with the exception of the intergenic enhancer 10 kilobases (kb) downstream of the Il17a transcription start site, all STAT3-STAT5 consensus sites in the Il17a-Il17f locus of TH17 cells treated with RA alone showed reciprocal decreased binding of STAT3and increased binding of STAT5 compared with their binding in the absence of RA (Fig. 7a). The decrease in STAT3 binding caused by the addition of RA was reversed in STAT5-deficient cells (Fig. 7a), consistent with the actions of RA in enhancing IL-2-dependent signaling via STAT5. The addition of IL-1β reversed both the RA-induced increase in binding of STAT5 and decrease in binding of STAT3 (Fig. 7a), consistent with the augmented phosphorylation of STAT3 induced by IL-1 signaling (Fig. 5b–f). At the intergenic enhancer downstream of Il17a noted above, the binding of both STAT3 and STAT5 was unaffected by addition of RA, although both were significantly affected by IL-1β (Fig. 7a), nature immunology  VOLUME 16  NUMBER 3  MARCH 2015

which indicated that additional regulatory mechanisms might confer enhanced binding of STAT3 at this cis-regulatory element regardless of IL-2-induced phosphorylation of STAT5. Competition between phosphorylated STAT5 and STAT3 in binding to the CNS2 intronic enhancer of Foxp3 has been proposed as regulating the heritable maintenance of Foxp3 expression and, thus, the stability of the Treg cell developmental program29,30. We found that RA, by enhancing the binding of STAT5, also restricted the IL-6-mediated binding of STAT3 at CNS2 in the Foxp3 locus during TH17 differentiation but was outcompeted by increased binding of STAT3 promoted by IL-1 (Fig. 7b). Consistent with those results, while IL-1 alone had no effect on repressing Foxp3 expression during iTreg cell development (Supplementary Fig. 8a), it substantially augmented the limited extinction of Foxp3 expression induced by IL-6 in secondary cultures of sorted Foxp3+ T cells derived from Il17fThy-1.1Foxp3GFP dual-reporter mice (Supplementary Fig. 8b). Indeed, even in the presence of exogenous RA, which promoted stability of Foxp3 expression, greater than 50% of cells extinguished their expression of Foxp3, with a significant population converting into IL-17F-producing cells (Supplementary Fig. 8b). Thus, IL-1 acted cooperatively with IL-6 to subvert the Foxp3–iTreg cell program in favor of TH17 cell development due at least in part to its actions in altering the relative binding of phosphorylated STAT3 and STAT5 at the CNS2 enhancer element in the Foxp3 locus. DISCUSSION The early developmental programs of iTreg cells and TH17 cells are intimately linked, reflected in the coexpression of the lineagespecifying transcription factors Foxp3 and RORγt downstream of TGF-β signaling5,6. The balance between the antagonistic effects of RA and IL-2, which promote iTreg cell development while suppressing TH17 cell development, and those of IL-6 and IL-1, which promote TH17 cell development while suppressing iTreg cell development, is particularly critical in determining homeostatic immunity versus host-protective immunity to microbes in the intestines, where production of RA by resident DCs normally favors antimicrobial tolerance. Here we found that IL-1 signaling overrode the effects of 293

npg

© 2015 Nature America, Inc. All rights reserved.

Articles the RA–IL-2 axis in promoting iTreg cell differentiation to allow a hostprotective TH17 response to an enteric pathogen, and we defined a mechanism by which IL-1 signaling modulated the TH17 cell–iTreg cell developmental balance. By enhancing the amplitude and duration of phosphorylated STAT3 activity induced by T H17-specifying cytokines (for example, IL-6 and IL-23), IL-1 altered the ratio of phosphorylated STAT3 to phosphorylated STAT5 in developing T cells to enhance TH17 cell development at the expense of iTreg cell development. Indeed, by favoring the binding of phosphorylated STAT3 over that of phosphorylated STAT5 at the CNS2 intronic enhancer in the Foxp3 locus, IL-1-dependent potentiation of STAT3 signaling directly subverted the Treg cell–stabilizing function of IL-2 promoted by RA, thereby contributing to plasticity in the Treg cell developmental program14,15,31,32. These findings extend the role of proinflammatory cytokines in overriding a dominant program of Treg cell development in the intestines to enable the recruitment of TH17 cell–mediated host defense and provide a mechanism by which IL-1 acts together with T H17 cell–specifying cytokines to promote TH17 cell development. Although the contribution of IL-1 to TH17 differentiation is well established2,16, the mechanism of its action has been unclear. Here we identified a link between IL-1 signaling and repression of SOCS3 expression as a mechanism for specifically augmenting signaling by STAT3 but not that of STAT5. Published studies have detailed the structural basis by which SOCS3 interacts with the complex composed of the signal-transducing receptor gp130 and kinase Jak1 following ligand-mediated trans-phosphorylation of the cytoplasmic domain of receptors of the IL-6 family, which explains the specificity of SOCS3 for this family of receptors33,34, and presumably non-gp130 receptors such as the structurally related IL-23R26, but not the IL-2 receptor. The rapid induction of SOCS3 expression downstream of signaling via IL-6 (or IL-23) provides a negative feedback loop that rapidly terminates signaling via the IL-6 receptor or IL-23 receptor by inhibiting receptor-bound Jak1 or Jak2, respectively. Through repression of SOCS3 expression, IL-1 indirectly sustains Jak-STAT interactions that potentiate phosphorylated STAT3 actions. Absent comparable effects on IL-2 receptor–induced phosphorylation of STAT5, IL-1 signaling effectively shifts the ratio of phosphorylated STAT3 to phosphorylated STAT5 in favor of phosphorylated STAT3, providing a competitive advantage for the binding of phosphorylated STAT3 to shared DNA target sequences, as demonstrated in our chromatin-immunoprecipitation studies. Interestingly, IL-1 augmented the IL-21-induced phosphorylation of STAT3 similarly to that induced by IL-6 and IL-23. Although the IL-21 receptor is unlike other common γ-chain cytokine receptors in activating mainly STAT3, not STAT5, as far as we are aware, a direct role for SOCS3 in modulating signaling via the IL-21 receptor has not been defined. Additional studies will be needed to determine whether the IL-21 receptor is unusual among common γ-chain cytokine receptors in recruiting SOCS3 or whether other effects of IL-1 signaling contribute to the increased phosphorylation of STAT3 induced by IL-21. It is notable that TGF-β has been reported to contribute to TH17 cell development through the inhibition of IL-6- and IL-21-induced SOCS3 expression 35, although another study has found that TGF-β also increases STAT5 phosphorylation36, an effect we did not observe here for IL-1 signaling. Thus, although both IL-1 and TGF-β might repress SOCS3 expression in concert with TH17-polarizing cytokines, albeit via fundamentally different signaling pathways, IL-1 would seem to ‘preferentially’ alter the ratio of phosphorylated STAT3 to phosphorylated STAT5 in favor of phosphorylated STAT3. In this context, it is of interest that T H17 cell 294

development can occur in the absence of TGF-β signaling, contingent on co-signaling by IL-1β with either IL-6 or IL-23 (ref. 37). It would therefore be of interest to determine whether this effect is mediated by IL-1-induced inhibition of SOCS3. In addition to positioning SOCS3 for inhibition of the Jak catalytic domain, docking of the SH2 domain of SOCS3 onto phosphorylated Tyr759 of the IL-6-activated gp130 cytoplasmic domain inhibits binding of the cytosolic tyrosine kinase SHP2, which activates signaling via mitogen-activated protein kinases and phosphatidylinositol-3-OH kinase induced by IL-6 and, presumably, IL-23. Thus, while it potentiates the phosphorylation of STAT3, IL-1-mediated repression of SOCS3 might also potentiate STAT3-independent signals from the IL-6 receptor, the effects of which have not been explored. Moreover, deficiency in SOCS3 has been shown to increase activation of STAT1 in myeloid cells38, which raises the possibility that IL-1-induced suppression of SOCS3 might also affect the T H17 cell–iTreg cell developmental balance via modulation of STAT1. It would be of interest to examine these issues in future studies. A published report has found that IL-1 signaling increases expression of the transcription factors IRF4 and RORγt, both of which are required for TH17 differentiation16. Expression of IRF4, which, unlike RORγt39, is central to the development of other effector cells in addition to TH17 cells, appears to be regulated mainly by TCR signals and is STAT independent. Hence, increased expression of IRF trans­ cription factors is probably not mediated by the STAT3-amplifying effects of IL-1 signaling we have described here. However, through cooperative binding with the transcription factor BATF at consensus motifs in multiple genes associated with the TH17 lineage40,41, IRF4 contributes to chromatin accessibility at many sites that are also targeted by STAT3 (ref. 42), as well as sites targeted by both STAT3 and STAT5 (ref. 15). Thus, in addition to any STAT-independent effects that IL-1 signaling might have on enhancing IRF4 expression, it can also indirectly modulate the effects of IRF4 actions by altering the balance of STAT binding at composite consensus sites in target genes. In contrast to IRF4 expression, RORγt expression is STAT3 dependent43. Thus, the effects of IL-1 on augmenting the amplitude and duration of the tyrosine-phosphorylation of STAT3 probably contribute to enhanced expression of RORγt during TH17 differentiation, whether early, downstream of IL-6 signaling, or later, following expression of the IL-23 receptor. iTreg cells differ from thymic Treg cells in their tendency to lose Foxp3 expression during successive rounds of cell division. This reflects differences in the methylation status of CpG islands within CNS2 of the Foxp3 locus30,44,45. CNS2 is highly methylated in newly generated iTreg cells, whereas thymic Treg cells and mature iTreg cells have undergone demethylation of CNS2, which enables binding of Foxp3 and Runx1, among other factors, that confers stable expression of Foxp3 under limiting IL-2 conditions32. Although the mechanism by which CNS2 becomes demethylated during Treg cell maturation remains to be fully elucidated, it has been found that CNS2 is the main site of IL-2-induced binding of STAT5 in the Foxp3 locus32, and, in contrast to other factors that bind CNS2, STAT5 can bind CNS2 regardless of its methylation status36. Although it remains unproven, this suggests that STAT5 might be an important factor in recruiting demethylation complexes to CNS2, in addition to its function in sustaining the competency of CNS2 in immature iTreg cells as they divide32. The data presented here would favor a model wherein IL-1 signaling in intestinal T cells responding to enteropathogenic infection is important for overriding the RA–IL-2–STAT5 axis in favor of enhanced, sustained STAT3 signaling that inhibits binding of STAT5 to Foxp3 CNS2. This would silence expression of Foxp3 VOLUME 16  NUMBER 3  MARCH 2015  nature immunology

Articles while promoting expression of the gene encoding RORγt and other genes associated with the TH17 lineage that contain consensus sequences regulated by competitive antagonism of STAT5 binding versus STAT3 binding and thereby deviate nascent or immature iTreg cells toward TH17 cell development. These findings extend the role of proinflammatory cytokines in overriding a dominant program of Treg cell development in the intestines to enable the recruitment of TH17 cell–mediated host defense and provide a rationale for the consideration of IL-1 blockade to reestablish homeostasis in the treatment of chronic inflammatory diseases mediated by the TH17 pathway. Methods Methods and any associated references are available in the online version of the paper.

npg

© 2015 Nature America, Inc. All rights reserved.

Note: Any Supplementary Information and Source Data files are available in the online version of the paper. Acknowledgments We thank H. Qin, L. Xu and members of the Weaver laboratory for advice and comments; G. Frankel and S. Wiles (Imperial College, London) for the bioluminescent C. rodentium strain; T. DeSilva (University of Alabama at Birmingham) for Rosa26TdTomato fl/+ mice; D. O’Quinn for assistance with bioluminiscent imaging; B.J. Parsons for technical assistance; G. Gaskins for editorial assistance; the University of Alabama at Birmingham Small Animal Imaging Facility for imaging studies; the University of Alabama at Birmingham Center for AIDS Research Flow Cytometry Core for sorting cells by flow cytometry; and the University of Alabama at Birmingham Epitope Recognition and Immunoreagent Core Facility for antibody preparation. Supported by the US National Institutes of Health (PO1DK71176 and R01DK093015 to C.T.W. and R.D.H., and R01AI047833 to W.S.P.) and the Crohn’s and Colitis Foundation of America (C.T.W. and R.B.). AUTHOR CONTRIBUTIONS R.B., S.K.W., R.D.H. and C.T.W. designed the studies; R.B., S.K.W., S.B. and R.D.H. performed the experiments; C.L.Z. advised on design and execution of figures; T.R.S. performed analyses of pathology; E.N.B. assisted in the design and interpretation of studies involving SOCS3; W.S.P. provided TAT-Cre peptide and provided guidance on its use in naive CD4+ T cells; and R.B., S.K.W. and C.T.W. wrote and edited the manuscript. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. Reprints and permissions information is available online at http://www.nature.com/ reprints/index.html. 1. Basu, R., Hatton, R.D. & Weaver, C.T. The TH17 family: flexibility follows function. Immunol. Rev. 252, 89–103 (2013). 2. Veldhoen, M., Hocking, R.J., Atkins, C.J., Locksley, R.M. & Stockinger, B. TGFβ in the context of an inflammatory cytokine milieu supports de novo differentiation of IL-17-producing T cells. Immunity 24, 179–189 (2006). 3. Mangan, P.R. et al. Transforming growth factor-β induces development of the TH17 lineage. Nature 441, 231–234 (2006). 4. Bettelli, E. et al. Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature 441, 235–238 (2006). 5. Zhou, L. et al. TGF-beta-induced Foxp3 inhibits TH17 cell differentiation by antagonizing RORγt function. Nature 453, 236–240 (2008). 6. Littman, D.R. & Rudensky, A.Y. TH17 and regulatory T cells in mediating and restraining inflammation. Cell 140, 845–858 (2010). 7. Vicente-Suarez, I. et al. Unique lamina propria stromal cells imprint the functional phenotype of mucosal dendritic cells. Mucosal Immunol. 8, 141–151 (2015). 8. Chen, W. et al. Conversion of peripheral CD4+CD25− naive T cells to CD4+CD25+ regulatory T cells by TGF-β induction of transcription factor Foxp3. J. Exp. Med. 198, 1875–1886 (2003). 9. Mucida, D. et al. Reciprocal TH17 and regulatory T cell differentiation mediated by retinoic acid. Science 317, 256–260 (2007). 10. Coombes, J.L. et al. A functionally specialized population of mucosal CD103+ DCs induces Foxp3+ regulatory T cells via a TGF-β and retinoic acid-dependent mechanism. J. Exp. Med. 204, 1757–1764 (2007).

nature immunology  VOLUME 16  NUMBER 3  MARCH 2015

11. Benson, M.J., Pino-Lagos, K., Rosemblatt, M. & Noelle, R.J. All-trans retinoic acid mediates enhanced T reg cell growth, differentiation, and gut homing in the face of high levels of co-stimulation. J. Exp. Med. 204, 1765–1774 (2007). 12. Sun, C.M. et al. Small intestine lamina propria dendritic cells promote de novo generation of Foxp3 T reg cells via retinoic acid. J. Exp. Med. 204, 1775–1785 (2007). 13. Davidson, T.S., DiPaolo, R.J., Andersson, J. & Shevach, E.M. IL-2 is essential for TGF-β-mediated induction of Foxp3+ T regulatory cells. J. Immunol. 178, 4022–4026 (2007). 14. Laurence, A. et al. Interleukin-2 signaling via STAT5 constrains T helper 17 cell generation. Immunity 26, 371–381 (2007). 15. Yang, X.P. et al. Opposing regulation of the locus encoding IL-17 through direct, reciprocal actions of STAT3 and STAT5. Nat. Immunol. 12, 247–254 (2011). 16. Chung, Y. et al. Critical regulation of early Th17 cell differentiation by interleukin-1 signaling. Immunity 30, 576–587 (2009). 17. Shaw, M.H., Kamada, N., Kim, Y.G. & Nunez, G. Microbiota-induced IL-1β, but not IL-6, is critical for the development of steady-state TH17 cells in the intestine. J. Exp. Med. 209, 251–258 (2012). 18. Atarashi, K. et al. ATP drives lamina propria TH17 cell differentiation. Nature 455, 808–812 (2008). 19. Ivanov, I.I. et al. Induction of intestinal Th17 cells by segmented filamentous bacteria. Cell 139, 485–498 (2009). 20. Ishigame, H. et al. Differential roles of interleukin-17A and -17F in host defense against mucoepithelial bacterial infection and allergic responses. Immunity 30, 108–119 (2009). 21. Basu, R. et al. Th22 cells are an important source of IL-22 for host protection against enteropathogenic bacteria. Immunity 37, 1061–1075 (2012). 22. Zheng, Y. et al. Interleukin-22 mediates early host defense against attaching and effacing bacterial pathogens. Nat. Med. 14, 282–289 (2008). 23. Hirota, K. et al. Fate mapping of IL-17-producing T cells in inflammatory responses. Nat. Immunol. 12, 255–263 (2011). 24. Lee, Y.K. et al. Late developmental plasticity in the T helper 17 lineage. Immunity 30, 92–107 (2009). 25. Harrington, L.E., Janowski, K.M., Oliver, J.R., Zajac, A.J. & Weaver, C.T. Memory CD4 T cells emerge from effector T-cell progenitors. Nature 452, 356–360 (2008). 26. Chen, Z. et al. Selective regulatory function of Socs3 in the formation of IL-17secreting T cells. Proc. Natl. Acad. Sci. USA 103, 8137–8142 (2006). 27. Yang, X.-P. et al. Dual function of interleukin-1beta for the regulation of interleukin6-induced suppressor of cytokine signaling 3 expression. J. Biol. Chem. 279, 45279–45289 (2004). 28. Wen, Z., Zhong, Z. & Darnell, J.E. Jr. Maximal activation of transcription by Stat1 and Stat3 requires both tyrosine and serine phosphorylation. Cell 82, 241–250 (1995). 29. Laurence, A. et al. STAT3 transcription factor promotes instability of nTreg cells and limits generation of iTreg cells during acute murine graft-versus-host disease. Immunity 37, 209–222 (2012). 30. Zheng, Y. et al. Role of conserved non-coding DNA elements in the Foxp3 gene in regulatory T-cell fate. Nature 463, 808–812 (2010). 31. Murphy, K.M. & Stockinger, B. Effector T cell plasticity: flexibility in the face of changing circumstances. Nat. Immunol. 11, 674–680 (2010). 32. Feng, Y. et al. Control of the inheritance of regulatory T cell Identity by a cis element in the Foxp3 locus. Cell 158, 749–763 (2014). 33. Babon, J.J., Varghese, L.N. & Nicola, N.A. Inhibition of IL-6 family cytokines by SOCS3. Semin. Immunol. 26, 13–19 (2014). 34. Kershaw, N.J. et al. SOCS3 binds specific receptor-JAK complexes to control cytokine signaling by direct kinase inhibition. Nat. Struct. Mol. Biol. 20, 469–476 (2013). 35. Qin, H. et al. TGF-β promotes Th17 cell development through inhibition of SOCS3. J. Immunol. 183, 97–105 (2009). 36. Ogawa, C. et al. TGF-β-mediated Foxp3 gene expression is cooperatively regulated by Stat5, Creb, and AP-1 through CNS2. J. Immunol. 192, 475–483 (2014). 37. Ghoreschi, K. et al. Generation of pathogenic TH17 cells in the absence of TGF-β signalling. Nature 467, 967–971 (2010). 38. Lang, R. et al. SOCS3 regulates the plasticity of gp130 signaling. Nat. Immunol. 4, 546–550 (2003). 39. Ivanov, I.I. et al. The orphan nuclear receptor RORγt directs the differentiation program of proinflammatory IL-17+ T helper cells. Cell 126, 1121–1133 (2006). 40. Glasmacher, E. et al. A genomic regulatory element that directs assembly and function of immune-specific AP-1-IRF complexes. Science 338, 975–980 (2012). 41. Li, P. et al. BATF-JUN is critical for IRF4-mediated transcription in T cells. Nature 490, 543–546 (2012). 42. Ciofani, M. et al. A validated regulatory network for Th17 cell specification. Cell 151, 289–303 (2012). 43. Yang, X.O. et al. STAT3 regulates cytokine-mediated generation of inflammatory helper T cells. J. Biol. Chem. 282, 9358–9363 (2007). 44. Miyao, T. et al. Plasticity of Foxp3+ T cells reflects promiscuous Foxp3 expression in conventional T cells but not reprogramming of regulatory T cells. Immunity 36, 262–275 (2012). 45. Ohkura, N. et al. T cell receptor stimulation-induced epigenetic changes and Foxp3 expression are independent and complementary events required for Treg cell development. Immunity 37, 785–799 (2012).

295

ONLINE METHODS

npg

© 2015 Nature America, Inc. All rights reserved.

Mice. The following mouse strains used were from the Jackson Laboratory and/or were bred at the University of Alabama at Birmingham (UAB) animal facility: C57BL/6J (B6); B6.129P2-Tcrbtm1Mom/J (Tcrb−/−); B6.129S2-Il1btm1Dch (Il1b−/−); B6.Cg-Tg-Il17ftm1(Thy1)Weav (Il17f Thy–1.1); B6.Cg-Tg(TcraTcrb)425Cbn/J (OT-II); B6N.129(Cg)-Foxp3tm3Ayr/J (Foxp3GFP); B6.129S7-Il1r1tm1Imx/J (Il1r1−/−); B6.129S6-Stat5a-Stat5btm2Mam/Mmjax (Stat5fl/fl); B6.129S1-Relafl/fl (Relafl/fl); B6.Cg-Gt(ROSA)26Sortm14(CAG–tdTomato)Hze/J (Rosa26TdTomato fl/+); and B6.129S4-Socs3tm1Ayos/J (Socs3fl/fl). B6.Cg-Tg(Cd4-Cre)1Cwi (Cd4-Cre) mice were from Taconic, and B6.CD45.1 mice were from Frederick Cancer Center. All strains were on a C57BL/6 (B6) background. All intercrosses to generate additional strains, such as wild-type and IL-1R-deficient dual reporter (Il17fThy–1.1Foxp3GFP) mice, Il17fThy–1.1 CD45.1+CD45.2+ mice and Stat5bfl/fl Cd4-Cre mice, were generated by crosses in UAB’s breeding facility. Animals were bred and maintained under specific pathogen–free condition in accordance with institutional animal care and use committee regulations. C. rodentium infection and mAb treatments. C. rodentium strain DBS100 (American Type Culture Collection) was used for all inoculations, with the exception of experiments in which whole-body imaging was performed. For imaging experiments, the bioluminescent C. rodentium strain ICC180 (derived from DBS100) was used46. This strain has a constitutive luxCDABE operon that encodes luciferase of the nematode Photorhabdus luminescens. C. rodentium strain DBS100 was grown at 37 °C in LB broth, and C. rodentium strain ICC180 was grown in LB broth containing kanamycin (100 µg/ml). Mice were inoculated with 2 × 109 colony-forming units of C. rodentium (a high dose) in a total volume of 200 ml of PBS via gastric gavage. For depletion of Il17fThy–1.1 cells in vivo, mice were given intraperitoneal injection of 400 µg anti-Thy-1.1 (19E1.2; University of Alabama-Birmingham Epitope Recognition and Immunoreagent Core Facility) on days 3 and 7 after infection. In all studies, age- and sex-matched mice 6–12 weeks of age were used. Bioluminescence imaging. Mice were anesthetized with isoflurane and were placed in the supine position in a custom-built chamber for imaging with the IVIS-100 system and Living Image Software (Xenogen). Baseline images were obtained before gavage with 2 × 109 colony-forming units of C. rodentium strain ICC180, and whole-body images were obtained at a binning of 4 over 1 for 3 min at the appropriate times after infection with C. rodentium ICC180. Luminescence emitted from the same gate in individual mice was quantified as counts per second, and pseudocolor images represent light intensity generated as a measure of colonization of the luminescent bacterial strain. Cell isolation, in vitro T cell differentiation, protein transduction and inhibitor treatment. Naive CD4+ T cells from spleens or lymph nodes of 8- to 10-week-old mice were purified by sorting on a FACSAria II (BD Bioscience) with gating on the CD4+CD25−CD62LhiCD44lo fraction. Isolated CD4+ T cells were cultured in R-10 medium (RPMI medium containing 10% FBS, 100 IU/ml penicillin, 100 µg/ml streptomycin, 1 mM sodium pyruvate, nonessential amino acids, 50 µM β-mercaptoethanol and 2 mM l-glutamine). For stimulation with plate-bound polyclonal antibodies, sorted CD4 + T cells were stimulated with 5 µg of plate-bound anti-CD3 (145-2C11; BD Pharmingen) and 1 µg/ml of soluble anti-CD28 (37.51; BD Pharmingen). For polyclonal feeder-based stimulation, sorted CD4+ T cells were stimulated with 1 µg/ml of soluble anti-CD3 (identified above) in presence of in presence of populations of feeder cells obtained from the spleen or MLNs of Il1b−/− mice, then irradiated and depleted of T cells. For antigen-specific stimulation, sorted naive OT-II CD4+ T cells (which have transgenic expression of an ovalbuminspecific TCR) were activated with 5 µg/ml ovalbumin peptide in presence of populations of splenic feeder cells obtained from Il1b−/− mice, then irradiated and depleted of T cells. In all cases in which feeder cells were used for TH17 polarization, T cells and feeder cells were added at a ratio of 1:1. All cultures were done in a volume of 200 µl in 96-well U-bottomed plates. For T H17 or iTreg polarization, the following exogenous cytokines or antibodies were added (concentrations in parentheses), unless stated otherwise: IL-6 (20 ng/ml; R&D Systems); TGF-β (2 ng/ml; R&D Systems); IL-21 (10 ng/ml; R&D Systems); IL-23 (5 ng/ml; R&D Systems); IL-1β (20 ng/ml; R&D Systems),

nature immunology

human IL-2 (1–100 U; Roche), anti-IL-4 (10 µg/ml; XMG1.2; prepared in-house); anti-IFN-γ (10 µg/ml; 11B11; prepared in-house); and anti-IL-2 (10 µg/ml; JES6-1A12; eBioscience). All-trans RA (Sigma) was added at various concentrations (1–100 nM), and the RA antagonist LE540 (Wako) was added at a concentration of 1 µM. For the purification of Foxp3+CD4+ T cells, Foxp3-GFP+ cells were sorted by flow cytometry and subjected to various conditions. For the purification of CD4+ single-positive thymocytes, CD8+ T cells were initially removed by depletion via magnetic-activated cell sorting (Mouse CD8α MicroBeads; Miltenyi Biotec), and CD4+ single-positive cells were further purified by isolation of the flow-through by magnetic-activated cell sorting (Mouse CD4+ T Cell Isolation Kit II; Miltenyi Biotec). In experiments in which cells were treated with cell-signaling inhibitors, recovered cells were washed and then were deprived of serum in complete medium for 1 h before addition of the following reagents (concentrations in parentheses): the Jak2 inhibitor AG490 (50 µM), the Jnk inhibitor SP600125 (1 µM), and the NF-κB inhibitor PDTC (1 µM). With the exception of PDTC (Sigma), all inhibitors were from Calbiochem. In experiments in which naive CD4+ T cells were transduced by TAT-Cre protein, sorted naive CD4+ T cells from Rosa26TdTomato fl/+ reporter mice with a loxP-flanked STOP cassette were initially standardized by transduction of TAT-Cre. Cells were incubated for 15–20 min at 37 °C in 1 ml of 100 µg/ml TAT-Cre peptide in serum-free RPMI medium, then were washed and suspended in RPMI medium containing IL-7 (10 ng/ml) plus 10% FCS, followed by incubation for 6, 12, 24 or 48 h; maximum expression of the reporter was observed at 24 h. Subsequently, sorted naive CD4+ T cells from Ai14 td-Tomato reporter and Socs3fl/fl mice were treated in tandem with TAT-Cre or control peptide and were cultured for 24 h in IL-7-containing medium, followed by polarization under TH17 conditions, with the additional culture modifications indicated in figure legends. Flow cytometry, including flow cytometry with phosphorylationspecific antibodies. The following mAbs were used for sorting cells by flow cytometry: anti CD4-eFlour 780 (eBioscience, clone RM4-5); anti-CD25-FITC (BD Pharmingen, clone 7D4); anti-CD62L-PE (eBioscience, clone MEL-14); and anti-CD44-APC (eBioscience, clone IM-7). For flow cytometry and intracellular staining, the following mAbs were used: fluorescein isothiocyanate– or peridinin chlorophyll protein–cyanine 5.5–conjugated anti-CD3 (145-2C11; eBioscience); phycoerythrin-indotricarbocyanine–conjugated anti-CD4 (RM4-5; eBioscience); allophycocyanin-conjugated anti-IL-17A (17B7; eBioscience); peridinin chlorophyll protein–cyanine 5.5–conjugated anti-CD25 (PC61.5; eBioscience), peridinin chlorophyll protein–cyanine 5.5–conjugated anti-CD45.1 (A20; eBioscience) and phycoerythrin-conjugated anti-RORγt (AFKJS-9; eBioscience). Differentiated TH17 cells were directly stained with anti-Thy-1.1 (OX-7; BD Pharmingen) without secondary stimulation. Only in cases in which intracellular IL-17A was detected, cells were stimulated for 4 h with PMA (phorbol 12-myristate 13-acetate; 50 ng/ml; Sigma) and ionomycin (750 ng/ml; Calbiochem) in the presence of Golgi Plug (BD Pharmingen). For detection of intracellular IL-17A and Foxp3, cells were resuspended in Foxp3 staining Buffer (eBioscience). In all cases, LIVE/DEAD Fixable Near-IR Dead Cell Stain (Invitrogen) was included before surface staining to exclude dead cells in flow cytometry. For flow cytometry with phosphorylation-specific antibodies, sorted naive CD4+ T cells were differentiated with plate-bound anti-CD3 and soluble antiCD28 (both identified above) under TH17-polarizing conditions alone or in presence of RA or RA plus IL-1β. After 4 d, cells were allowed to ‘rest’ briefly in medium at a neutral pH before being stimulated for 20 min with IL-6 alone (50 ng/ml; R&D Systems) or with IL-2 (500 U/ml). The cells were then washed and suspended in prewarmed PBS before being fixed for 10 min, in an incubator at 37 °C, with 4% paraformaldehyde. The cells were washed again and resuspended in ice-cold phospho-wash buffer (PBS containing 1× HALT protease inhibitor (78437; Thermo Scientific)) before the addition of chilled 100% methanol in drops and incubation for 30 min, followed by two washes in PBS containing 0.05% Tween. Cells were preincubated for 15 min in PBS-Tween containing 3% goat serum and 1 µl Fc block before incubation for 1 h with phycoerythrin-conjugated antibody to STAT3 phosphorylated at Tyr705 (612569; BD Biosciences) or Alexa Fluor 647–conjugated

doi:10.1038/ni.3099

antibody to STAT5 phosphorylated at Tyr694 (612599; BD Pharmingen). Samples were acquired on an LSRII (BD Biosciences), and data were analyzed with FlowJo software (TreeStar).

npg

© 2015 Nature America, Inc. All rights reserved.

Isolation and staining of LPLs. For isolation of LPLs, the large intestine was removed and cleared of luminal contents and fat, then was cut into small pieces and washed in chilled Hank’s balanced-salt solution without Ca2+ or Mg2+. Minced tissue pieces were incubated for 30 min in presence of EDTA and were vortexed thoroughly to remove epithelial cells, then were incubated in RPMI medium containing collagenase IV (1 mg/ml; Sigma-Aldrich), dispase (0.5 mg/ml; Gibco, Invitrogen) and DNase I (0.25 mg/ml; Sigma-Aldrich). LP lymphocytes were collected by centrifugation. In all experiments with Il17fThy–1.1Foxp3GFP reporter mice, unmanipulated LPLs were analyzed after undergoing surface staining with anti-CD3, anti-CD4, anti-CD25, antiCD45.1, anti-TCRβ and anti-Thy-1.1 (all antibodies identified above). For analyses of Il1r1+/+Il17fThy–1.1 (CD45.1+CD45.2+) and Il1r1−/−Il17fThy–1.1 (CD45.2+) recipient Tcrb−/− mice, unmanipulated LPLs were analyzed after undergoing surface staining with anti-TCRβ, anti-CD25, anti-CD45.1 and anti-CD90.1 (all antibodies identified above) and were subsequently fixed in fixation buffer (Foxp3-staining buffer; eBioscience) before intracellular staining with mAb to Foxp3 (identified above). Adoptive transfer and in vivo treatment with LE135. For co-transfer of CD4+ T cells from congenically marked (donor) Il1r1+/+Il17fThy–1.1 (CD45.1+CD45.2+) and Il1r1−/−Il17fThy–1.1 (CD45.2+) reporter mice into Tcrb−/− recipient mice, CD4+ T cells were isolated directly ex vivo by depletion of cells via magneticactivated cell sorting with phycoerythrin-labeled anti-CD8α (53-6.7; eBioscience), anti-CD11b (M1/70; eBioscience), anti-CD11c (N418; eBioscience), anti-CD19 (1D3, eBioscience), anti-CD45R (B220) (RA3-6B2; eBioscience), anti-CD49b (DX5) (DX5; eBioscience), anti-CD105 (MJ7/18; eBioscience), anti–MHC class II (M5/114.15.2; Miltenyi Biotec), anti-γδ TCR (eBioGL3; eBioscience) and anti-Ter-119 (TER-119; eBioscience), captured by antiphycoerythrin magnetic microbeads (Miltenyi Biotec). For co-transfer of wild-type and IL-1R1-deficient Foxp3+ T cells, CD4+Foxp3+GFP+IL-17F. Thy-1.1− T cells were sorted by flow cytometry from C. rodentium–infected donor mice. In all cases, recovered T cells were transferred by intraperitoneal injection at the appropriate dose and time point. For in vivo treatment of Il1r1−/− reporter mice with the retinoic acid inhibitor LE135 (Torcis Bioscience), 100 µg of LE135 dissolved in dimethyl sulfoxide (10 mg/ml) in a volume of 200 µl (DMSO:PBS, 1:20) was introduced by gavage from day 3 through day 7 after infection, followed by analysis at day 8 after infection by flow cytometry or on day 10 by histopathology. Control mice were given vehicle alone by gavage on the same schedule. ELISA. For determination of cytokine concentrations in vivo, supernatants of differentiated T cells were collected on day 3 of culture and analyzed by capture ELISA. For determination of cytokine concentrations ex vivo, colonic tissues from uninfected or C. rodentium–infected mice were collected at the appropriate time after infection, then were finely minced and were cultured for 24 h in 24-well plates in R-10 medium supplemented with 10 µg/ml gentamicin. Supernatants were collected after centrifugation and were analyzed by capture ELISA. ELISA of IL-2, IFN-γ, IL-17A and IL-22 used mouse quantikine ELISA kits according to the manufacturer’s recommendations (R&D Systems). Immunoblot analysis. B6 CD4+ T cells were cultured for 5 d under TH17-polarizing conditions in presence of anti-CD3 (identified above) and splenocyte populations that were irradiated and depleted of T cells. Following Ficoll separation to isolate viable cells, T cells were allowed to ‘rest’ in medium and were restimulated for the appropriate time with IL-6 (20 ng/ml), IL-1β (20 ng/ml) or IL-21 (20 ng/ml). Cell lysates were prepared in lysis buffer (RIPA buffer: 50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate and 0.1% SDS) containing a protease- and phosphataseinhibitor mixture (Pierce). Proteins were quantified by Bradford Assay before equivalent amounts were separated by SDS-PAGE and transferred to a PVDF membrane (Millipore). Primary antibodies were as follows: anti-STAT3 (9132; Cell Signaling), antibody to STAT3 phosphorylated at Tyr705 (9145;

doi:10.1038/ni.3099 

Cell Signaling Technology), or antibody to STAT3 phosphorylated at Ser727 (9134; Cell Signaling Technology). Horseradish peroxidase–cojugated donkey anti-rabbit (SA1200; Affinity Bioreagents) or anti-mouse (SA1200; Affinity Bioreagents) was used for detection of target proteins by ECL detection kit (GE Healthcare or Pierce SuperSignal Dura). Real-time PCR. Total RNA was isolated using RNeasy Isolation Kit according to the manufacturer’s instructions (Qiagen) and was treated with DNAse (Qiagen). cDNA was synthesized with the Superscript III FirstStrand Synthesis system (Invitrogen), and real-time PCR was performed on a Bio-Rad iCycler with iQTM SYBR Green Supermix (Bio-Rad) and primer pairs specific for cDNA of Il17f, Il1r1, Il1r2, Il1rap, Rara, Rarb, Rarg, Rxra, Rxrb, Rxrc, Rorc and 18S transcripts. The primer sets were as follows: Il17f, 5′-CGCCATTCAGCAAGAAATCC-3′ (forward) and 5′-CTCCAACCTGA AGGAATTAGAACAG-3′; Il1r1, 5′-CCGGGCACGCCCAGGAGAA-3′ (forward) and 5′-TCCAGCGACAGCAGAGGCAC-3′ (reverse); Il1r2, 5′-TCGGAGAAGCCCACAGTCCA-3′ (forward) and 5′-GGAGTGAGGTGCC AAGGGGC-3′ (reverse); Il1rap, 5′-TTGCCACCCCAGATCTATTC-3′ (forward) and 5′-CCAGACCTCATTGTGGGAGT-3′ (reverse); Rara, 5′-TCC GACGAAGCATCCAGAA-3′ (forward) and 5′-GGTTCCGGGTCACCTTG TT-3′ (reverse); Rarb, 5′-CAGTGAGCTGGCCACCAAGT-3′ (forward) and 5′-GCGATGGTCAGACCTGTGAA-3′ (reverse); Rarg, 5′-GTCAG CTCCTGTGAAGGCTGC-3′ (forward) and 5′-GTTTCGATCGTTCCTTA CAG-3′ (reverse); Rxra, 5′-TAGTCGCAGACATGGACACC-3′ (forward) and 5′-GTTGGAGAGTTGAGGGACGA-3′ (reverse); Rxrb, 5′-GCACAGA AACTCAGCCCATT-3′ (forward) and 5′-CATCCTCATGTCACGCATTT-3′ (reverse); Rxrg, 5′-GCCTGGGATTGGAAATATGA-3′ (forward) and 5′-AC ACCGTAGTGCTTCCCTGA-3′ (reverse); Rorc, 5′-CAGCCAACATGTGGA AAAGCT-3′ (forward) and 5′-GGGAAGGCGGCTTGGA-3′ (reverse); Cd25, 5′-TACAAGAACGGCACCATCCTAA-3′ (forward) and 5′-TTGCT GCTCCAGGAGTTTCC-3′ (reverse); Cd122, 5′-GTCCATGCCAAGTCG AACCT-3′ (forward) and 5′-GGATGCCTGCCTCACAAGAG-3′ (reverse); Socs3, 5′-AGTGCAGAGTAGTGACTAAACATTACAAGA-3′ (forward) and 5′-AGCAGGCGAGTGTAGAGTCAGAGT-3′ (reverse); and 18S rRNA, 5′-GCC GCTAGAGGTGAAATTCTTG-3′ (forward) and 5′-CATTCTTGGCAAA TGCTTTCG-3′ (reverse). Reactions were run in duplicate and samples were normalized to 18S rRNA as a induction relative to expression in control samples unless stated otherwise. Chromatin-immunoprecipitation assay. CD4+ single-positive thymocytes from wild-type or Stat5fl/flCd4-Cre mice were polarized for 3–4 d under TH0 conditions (no cytokine) or TH17-polarizing conditions (IL-6 plus TGF-β), with or without RA alone or RA plus IL-1β, followed by 20 min of stimulation in CO2 incubator by the addition of IL-6 (50 ng/ml; R&D Systems) or IL-2 (500 U/ml). The cells were crosslinked for 5 min with 1% (vol/vol) formaldehyde, quenched with 125 mM glycine and then suspended in HALT protease inhibitor ‘cocktail’ (78437; Thermo Scientific) containing PBS. Chromatin immunoprecipitation was performed with an ExactaChIP Kit in accordance with the manufacturer’s instruction (R&D Systems). Cells were suspended overnight at 4 °C in lysis buffer containing protease inhibitor and were immunoprecipitated with biotinylated anti-STAT3 (ECP1799; R&D Systems) or anti-STAT5 (ECP2168; R&D Systems). Bound DNA was collected through the use of agarose-streptavidin beads and was purified with a QIAprep Spin Miniprep Kit (Qiagen). Immunoprecipitated DNA was quantified by realtime PCR with iQTM SYBR Green Supermix (Bio-Rad) on a Bio-Rad iCycler with the following primer pairs specific for elements in the Il17a-Il17f locus47 and Foxp3 locus: Il17a–5, 5′-CAGGTATTATTCTCAGGGCTTTGG-3′ (forward) and 5′-TGGCAATGGTGTCTTTTCTTTG-3′ (reverse); Il17a promoter, 5′-CACCTCACACGAGGCACAAG-3′ (forward) and 5′-ATG TTTGCGCGTCCTGATC-3′ (reverse); Il17a+10, 5′-GGATTAAGGGCACA CGTGTTG-3′ (forward) and 5′-TTTCCCCACTCTGTCTTTCCA-3′ (reverse); Il17a+28, 5′-TCATCGGCTCCCACACAGA-3′ (forward) and 5′-GGCAGTA CCGAAGCTGTTTCA-3′ (reverse); Il17f promoter, 5′-CCCACAAAGCAA CACTCTTGTC-3′ (forward) and 5′-ACTGCATGACCCGAAAGCA-3′ (reverse); Il17f–5, 5′-GCATCGCATCTTTCAAACCA-3′ (forward) and 5′-TT AGGATAAGCGCCCAGTGAAT-3′ (reverse); and Foxp3 CNS2 for, 5′-GT TGCCGATGAAGCCCAAT-3′ (forward) and 5′-ATCTGGGCCCTGTTG

nature immunology

TCACA-3′. Data are presented as relative values normalized to the input DNA samples (percentage of input).

46. Wiles, S., Pickard, K.M., Peng, K., MacDonald, T.T. & Frankel, G. In vivo bioluminescence imaging of the murine pathogen Citrobacter rodentium. Infect. Immun. 74, 5391–5396 (2006). 47. Yang, X.P. et al. Opposing regulation of the locus encoding IL-17 through direct, reciprocal actions of STAT3 and STAT5. Nat. Immunol. 12, 247–254 (2011). 48. Bleich, A. et al. Refined histopathologic scoring system improves power to detect colitis QTL in mice. Mamm. Genome 15, 865–871 (2004).

npg

© 2015 Nature America, Inc. All rights reserved.

Histolopathological evaluation. Tissue samples obtained from proximal, middle and distal portions of the large intestine were fixed in 10% neutral buffered formalin and embedded in paraffin for the preparation of sections 5 µm in thickness that were stained with hematoxylin and eosin. The tissue sections were examined and assigned scores for tissue pathology as described48. In all score assigments, the identity of specimens was concealed from the pathologist.

Statistical analysis. For statistical analyses, P values were calculated by a paired or unpaired Student’s t-test, unless stated otherwise in figure legends. A P value of