The transcription factor IRF8 activates integrin-mediated TGF-β signaling and promotes neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

Immunity

Article

The Transcription Factor IRF8 ActivatesIntegrin-Mediated TGF-b Signalingand Promotes NeuroinflammationYuko Yoshida,1,8 Ryusuke Yoshimi,1,8,9 Hiroaki Yoshii,1 Daniel Kim,1 Anup Dey,1 Huabao Xiong,2 Jeeva Munasinghe,3

Itaru Yazawa,4,10 Michael J. O’Donovan,4 Olga A. Maximova,5 Suveena Sharma,6 Jinfang Zhu,6 Hongsheng Wang,7

Herbert C. Morse III,7 and Keiko Ozato1,*1Program in Genomics of Differentiation, NICHD, National Institutes of Health, Bethesda, MD 20892, USA2Immunology Institute, Department of Medicine, Mount Sinai School of Medicine, New York, NY 10029, USA3Laboratory of Functional and Molecular Imaging, NINDS, National Institutes of Health, Bethesda, MD 20892, USA4Laboratory of Neural Control, NINDS, National Institutes of Health, Bethesda, MD 20892, USA5Laboratory of Infectious Diseases, NIAID, National Institutes of Health, Bethesda, MD 20892, USA6Laboratory of Immunology, NIAID, National Institutes of Health, Bethesda, MD 20892, USA7Laboratory of Immunogenetics, NIAID, National Institutes of Health, Rockville, MD 20892, USA8These authors contributed equally to this work9Present address: Department of Internal Medicine and Clinical Immunology, Yokohama City University Graduate School of Medicine,Yokohama 236-0004, Japan10Present address: Department of Neurology, Showa University, Tokyo 142-8555, Japan

*Correspondence: [email protected]

http://dx.doi.org/10.1016/j.immuni.2013.11.022

SUMMARY

Recent epidemiological studies have identifiedinterferon regulatory factor 8 (IRF8) as a susceptibil-ity factor for multiple sclerosis (MS). However, howIRF8 influences the neuroinflammatory diseasehas remained unknown. By studying the role ofIRF8 in experimental autoimmune encephalomyelitis(EAE), a mouse model of MS, we found that Irf8�/�

mice are resistant to EAE. Furthermore, expressionof IRF8 in antigen-presenting cells (APCs, such asmacrophages, dendritic cells, and microglia), butnot in T cells, facilitated disease onset and pro-gression through multiple pathways. IRF8 enhancedavb8 integrin expression in APCs and activatedTGF-b signaling leading to T helper 17 (Th17) celldifferentiation. IRF8 induced a cytokine milieuthat favored growth and maintenance of Th1 andTh17 cells, by stimulating interleukin-12 (IL-12) andIL-23 production, but inhibiting IL-27 during EAE.Finally, IRF8 activated microglia and exacerbatedneuroinflammation. Together, this work providesmechanistic bases by which IRF8 contributes to thepathogenesis of MS.

INTRODUCTION

T helper 17 (Th17) cells promote inflammation and tissue injury

and are associated with autoimmune diseases. However, they

also play a role in pathogen resistance (Weaver et al., 2007).

With the aid of transforming growth factor-b (TGF-b),inter-

leukin-6 (IL-6), and other cytokines, antigen-presenting cells

(APCs), such as dendritic cells (DCs) and macrophages, trigger

development of Th17 cells (Bettelli et al., 2006). Recent studies

revealed that TGF-b signaling is mediated by integrin molecules

on APCs, which enables direct delivery of biologically active

TGF-b into naive T cells (Acharya et al., 2010; Melton et al.,

2010). Integrin-triggered TGF-b signaling results in activation of

ROR family transcription factors that direct Th17 cell differentia-

tion and production of the signature cytokine IL-17 (Ivanov et al.,

2006; Yang et al., 2008). It also activates Treg generation (Travis

et al., 2007).

One of the best-studied autoimmune diseases causally asso-

ciated with Th17 cells is experimental autoimmune encephalo-

myelitis (EAE), a mouse model of multiple sclerosis (MS). MS is

an inflammatory disease of the central nervous system (CNS)

that involves demyelination and neuronal injury (Hauser and

Oksenberg, 2006; Steinman and Zamvil, 2006). Mice lacking

RORa, RORgt, and IL-17 are largely resistant to EAE (Ivanov

et al., 2006; Komiyama et al., 2006). Correlating with results in

the mouse, Th17 cells and IL-17 are present in the CNS of MS

patients (Axtell et al., 2010). In addition to Th17 cells, Th1 cells

play substantive roles in EAE, causing CNS lesions distinct

from those by Th17 cells (Axtell et al., 2010; Kroenke et al.,

2008; Stromnes et al., 2008). Involvement of Th1 cells in MS is

also documented, adding to the similarity of EAE with MS

(Lovett-Racke et al., 2011). Th17 cells, once developed, further

proliferate in lymph nodes and then in the CNS, the processes

dependent on IL-23, a cytokine of the IL-12 family, produced

largely by macrophages and microglia (Becher et al., 2003;

Chen et al., 2006; Cua et al., 2003). The classical IL-12p70 sup-

ports the development and expansion of Th1 cells (Kroenke

et al., 2008). Conversely, development of Th17 cells and EAE is

suppressed by IL-27, another IL-12 family cytokine (Batten

et al., 2006; Bettelli et al., 2006; Stumhofer et al., 2006). Although

infiltrating T cells and APCs initiate CNS inflammation, activation

of the resident microglia worsens the disease by increasing

inflammation and neuronal damage (Sørensen et al., 1999; Star-

ossom et al., 2012).

Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc. 1

Figure 1. Irf8–/– Mice Are Resistant to EAE

(A) Clinical scores of MOG-immunizedWT and Irf8�/�mice. Data represent the

mean clinical scores of 15 mice in three experiments ± SEM (n = 45 mice per

group).

(B) Histopathological analysis of neuroinflammation and demyelination in

spinal cords of WT and Irf8�/� mice 21 days after EAE induction: sections

stained by hematoxylin and eosin (H&E) or Weil’s myelin (see Figure S1A for

detail).

(C) MRI analysis. Images represent quantitative Spin-Spin (T2) relaxation maps

of lumber spinal cords from WT and Irf8�/� mice 21 days after EAE induction.

Greater contrast enhancement in the WT sample is due to higher T2 values

reflecting BBB disruption. Similar differences were seen in all 6 WT or Irf8�/�

spinal cord samples tested.

(D) Average T2 values of selected regions of interest of 16 slices obtained from

all samples. See also Figure S1.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

2 Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc.

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

Despite much progress, the etiology of MS has remained

elusive. This is partly attributable to the multiplicity of path-

ways that affect the disease (Hauser and Oksenberg, 2006).

In this context, genome-wide SNP analyses shed new light

on understanding the onset and progression of MS as they

identify susceptibility factors likely influencing the disease

(De Jager et al., 2009; Disanto et al., 2012; Beecham et al.,

2013). Besides classically known HLA genes, a number of

additional genes have been designated as MS susceptibility

factors, including IRF8 (De Jager et al., 2009; Disanto et al.,

2012). IRF8 is a transcription factor of the IRF family known

to direct development of macrophages and DCs (Tamura

et al., 2005). It drives transcription of IL-12p40 and type I in-

terferons in these cells, thus playing essential roles in defense

against various pathogens (Tailor et al., 2008; Hambleton

et al., 2011; Chang et al., 2012). Additionally, IRF8 regulates

activities in T and B lymphocytes (Feng et al., 2011; Miya-

gawa et al., 2012; Ouyang et al., 2011). The SNP regions

associated with MS susceptibility map to the 30 noncoding

region of the IRF8 gene, suggesting that regulation of IRF8

transcription accounts for MS susceptibility. Nevertheless, lit-

tle information is available as to how IRF8 affects the course

of MS. We show here that Irf8�/� mice are protected from

EAE, and that IRF8 expressed in APCs, rather than T lympho-

cytes, causes the disease by facilitating the onset and expan-

sion of effector T cells and promoting microglia-based

neuroinflammation.

RESULTS

Irf8–/– Mice Are Resistant to EAETo study the role of IRF8 in EAE pathogenesis, wild-type

(WT) and Irf8�/� mice were injected with MOG35–55 (hereafter

MOG) and clinical signs of the EAE were scored on a daily

basis (Figure 1A). All WT mice exhibited clear signs of

EAE from days 7–9, which peaked on days 17–20, followed

by a slight decline thereafter. In contrast, Irf8�/� mice

were completely resistant to EAE, and showed no clinical

signs. Histopathological analysis of spinal cords of WT mice

revealed typical EAE features with mononuclear cell infiltration,

gliosis, neuronal damage, and demyelination, whereas spinal

cords of Irf8�/� appeared to be normal (Figures 1B; see

Figure S1A available online). EAE in Irf8�/� mice was further

analyzed by quantitative magnetic resonance imaging (MRI)

to detect blood-brain barrier (BBB) disruption and tissue

contrasts (Schellenberg et al., 2007). Tissue contrasts were

examined by probing the inherent spin-spin (T2) relaxation to

obtain a series of T2-weighted images and subsequently to

evaluate T2 maps in the spinal cords and by means of

T1-weighted imaging after injecting the contrast agent Gd-

DTPA (Figure 1C; Figure S1B). All WT spinal cords showed

high contrast enhancement, particularly noticeable in the

peripheral regions, whereas little contrast enhancement was

observed with Irf8�/� samples (see overlay images in Fig-

ure S1B). T2 values obtained from multiple slices in Figure 1D

further reinforced greater enhancement in WT samples than

Irf8�/� samples. These results indicate that unlike WT mice

that had extensive BBB breakdown, Irf8�/� mice retained

intact BBB.

Figure 2. Irf8–/– Mice Fail to Generate Th17,

Th1, and Treg Cells after MOG Immuni-

zation

(A) Relative mRNA expression of Il17a (IL-17A),

Ifng (IFN-g), and Foxp3 in spinal cords, LN and

spleen in WT and Irf8�/� mice on indicated days

after immunization. Samples from spinal cords

were pooled from three mice. Values represent

the means ± SEM. Values for LN and spleen

represent the average of three assays from inde-

pendent sets of mice ± SEM. Similar results were

observed in five independent experiments. *p <

0.01, **p < 0.005.

(B and C) Mononuclear cells from spinal cord

(B) and LN (C) collected on day 21 and day

14 were analyzed for Th17 cells (CD4+

RORgt+IL17+), Th1 cells (CD4+IFN-g+), and

Tregs (CD4+CD25+FOXP3+) after restimulation

with (+) or without (�) MOG (50 mg/ml) in vitro

for 48 hr. Similar results were observed in

five independent experiments. See also Fig-

ure S2.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

Irf8–/– Mice Fail to Generate Th1, Th17, and Treg Cellsafter MOG InjectionMOG immunization stimulates development of autoreactive

Th1, Th17, and regulatory T cells (Tregs) (Bettelli et al., 2006;

Park et al., 2005). To assess whether Irf8�/� mice are capable

of developing these T cells after MOG-injection, we measured

Immunity 40, 1–1

mRNA for Il17a (IL-17A), Ifng

(interferon-g [IFN-g]), the signature

cytokines for Th17 and Th1 cells,

respectively, and Foxp3, the transcription

factor that specifies Tregs. In WT lymph

nodes (LN) and spleen, these transcripts

were sharply increased on day 7 and

peaked on day 14 (Figure 2A). Similar

increase in transcript expression was

seen in spinal cords later, on day 14

and day 21, presumably reflecting

the timing of cellular infiltration into

the CNS. In contrast, expression of

these transcripts was low to undetectable

in the Irf8�/� counterparts during the

entire course of EAE. To ascertain

whether the absence of the mRNA

expression in Irf8�/� mice is attributable

to the lack of corresponding T cells,

we performed flow cytometric analyses

to detect Th1 (IFN-g+, CD4+), Th17

(IL-17+, RORgt+) and Tregs (CD25+,

Foxp3+). As shown in Figures 2B, 2C,

and S2, these effector T cells were

generated to high numbers in WT spinal

cords and LN. However, the percentage

of these T cells was drastically reduced

in Irf8�/� spinal cords and LN. To assess

antigen specificity, we stimulated cells

obtained from MOG-immunized mice

with MOG in vitro for 2 days and analyzed them by flow cytom-

etry. The percentages of these T cells inWTmice were increased

by more than 2-fold but remained very low in Irf8�/� mice (Fig-

ure 2B). These results indicate that Irf8�/� mice fail to sensitize

naive T cells to induce effector T cell differentiation in response

to MOG.

2, February 20, 2014 ª2014 Elsevier Inc. 3

Figure 3. Mice with Monocyte- and Macrophage-Specific, but Not T

Cell-Specific, Irf8 Disruption Are Resistant to EAE

(A) Irf8 mRNA expression was measured in WT mice, conventional Irf8�/�

mice, or mice with monocyte- and macrophage-specific Irf8 disruption (all

unimmunized) for themicroglia andmacrophage population in the spinal cords

(left), BM-derived macrophages (middle), and BM-derived DCs (right).

(B and C) Clinical EAE scores of Irf8f/f -LysM-cre/+ (B) and Irf8f/f-Lck-cre (C)

mice after MOG immunization. Data represent the average disease scores

from three experiments with four or fivemice per group ± SEM (n = 12–15 mice

per group). See also Figure S3.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

Mice with Monocyte- and Macrophage-Specific Irf8

Disruption, but Not with T Cell-Specific Disruption, AreResistant to EAEThe above results revealed the striking absence of effector T cell

development in Irf8�/�mice. Given that IRF8 plays a role in T cell

subset development and functions, it was possible that Irf8�/�

T cells were defective in differentiating into Th1 and Th17 cells

after MOG immunization (Miyagawa et al., 2012; Ouyang et al.,

2011). Conversely, it was possible that Irf8�/� APCs fail to sensi-

tize T cells and or provide a proper cytokine milieu to generate

effector T cells. To assess the cell types through which IRF8

affects EAE, we tested mice with conditional Irf8 disruption.

Irf8f/f mice with LysM-cre should have monocyte- and macro-

phage-specific Irf8 disruption, wheras Irf8f/f mice with Lck-cre

should result in T cell-specific Irf8 disruption (Clausen et al.,

1999; Feng et al., 2011). We confirmed that in Irf8f/f LysM-cre

mice, expression of Irf8 mRNA and the protein was specifically

depleted in macrophages and microglia (Figure 3A, Fig-

ure S3A–S3C). However, noticeable IRF8 expression remained

in the DC populations in these mice (Figure 3A), consistent

with a previous report on the cell types affected by LysM-cre

(Clausen et al., 1999). Irf8 disruption in T cells was confirmed in

our previous report (Ouyang et al., 2011). Mice with monocyte-

and macrophage-specific Irf8 disruption remained resistant to

EAE, although some mice displayed mild clinical signs after

day 20 (Figure 3B). Irf8+/+ LysM-cre mice, tested as a control,

developed EAE, excluding nonspecific effects of Cre recombi-

nase. Conversely, mice with T cell-specific Irf8 disruption devel-

oped EAE, showing similar clinical scores as WT mice or Irf8f/f

mice without Lck-cre (Figure 3C). These results indicate that

IRF8 in the APCs of the monocyte and macrophage lineage is

primarily responsible for causing EAE. IRF8 in T cells, however,

although capable of regulating Th17 cells, does not have a

consequential role in EAE (Ouyang et al., 2011). To investigate

whether IRF8 in DCs plays a role in EAE development, we

performed adoptive-transfer experiments with bone marrow

(BM)-derived DCs. Irf8�/�mice, upon receiving WT BMDCs, dis-

played EAE clinical scores in the first 13 days, although scores

were diminished thereafter (Figure S3F). These data support

the role of IRF8 in macrophages and DCs in promoting EAE.

IRF8 Expression Increases in Macrophages, DCs, andMicroglia during EAETo identify the cells expressing IRF8 during EAE, we analyzed

Irf8-EGFP gene-targeted mice, in which the IRF8-GFP fusion

protein is expressed from the endogenous Irf8 promoter (details

of these mice will be described elsewhere). When immunized

with MOG, both homozygous and heterozygous Irf8-EGFP

mice succumbed to EAE, exhibiting essentially the same clinical

symptoms asWTmice (Figure S4A). This result is consistent with

our previous report showing that the IRF8-GFP fusion protein

functions in a manner identical to IRF8 (Laricchia-Robbio et al.,

2005). FACS analysis of GFP+ cells in spleen from unimmunized

mice showed that IRF8-GFP is expressed at high levels in DCs

and macrophages, slightly lower levels in B cells, but at back-

ground levels in T cells and granulocytes (neutrophils) (Fig-

ure S4B). IRF8-GFP was also expressed in microglia (CD11b+,

CD45lo), a population distinct from macrophages (CD11b+,

CD45hi) in the spinal cord (Figure S4D). IRF8-GFP+ macro-

4 Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc.

phages, DCs, and microglia were examined during EAE (Fig-

ure 4A). The number of IRF8-GFP+ microglia increased greatly

during EAE along with a substantial increase in IRF8-EGFP

expression, consistent with microglia activation during EAE

Figure 4. IRF8-Expressing APCs Increase in Number during EAE

(A) Flow cytometric detection of GFP+ cells in microglia (CD45loCD11b+), DCs

(CD11c+MHC II+), macrophages (CD45hiCD11b+ F4/80+) in spinal cords, LN,

and spleen in MOG-injected Irf8-EGFP gene-targeted mice. Similar results

were observed in five independent experiments.

(B) The total number of DCs, macrophages, and granulocytes (CD11bhiGr-1hi)

(left panel) and IRF8-GFP+ cells (right panel) in LN and spleen in MOG-

immunized Irf8-GFP gene-targeted mice. Values represent the average of

three independent experiments ± SEM. *p < 0.01, **p < 0.005. See also Fig-

ure S4.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

(Starossom et al., 2012). Also in line with the inability of microglia

activation in Irf8�/� mice, mRNA expression of Cxcl9 (Mig1) and

Nos2 was much lower in Irf8�/� spinal cords than in WT spinal

cords, factors likely involved in neuronal inflammation and

damage (Figure S6B) (Hauser and Oksenberg, 2006; Masuda

et al., 2012). IRF8-GFP+ macrophages also increased in spinal

cords during EAE, although numbers were lower than in micro-

glia (Figure 4A). DCs in LN and spleen of unimmunized mice

displayed two populations, IRF8-GFPhi and IRF8-GFPint, corre-

sponding to plasmacytoid DCs/CD8a+ DCs and myeloid DCs

(Tamura et al., 2005).The IRF8-GFPhi population increased in

number as well as GFP intensity during EAE. IRF8-GFP expres-

sion in macrophages in LNs and spleens was also higher in EAE

mice than in naive mice. Data in Figure 4B (left panel) showed

that the number of total DCs and macrophages sharply

increased on days 7 and 14, followed by a significant decline

by day 21. As expected, the majority of these cells expressed

IRF8-GFP (Figure 4B, right panel). Together, the total number

and IRF8-GFP+ APCs markedly increase in peripheral lymphoid

organs and in spinal cords during EAE. Granulocytes, however,

did not express IRF8 to a substantial degree throughout the

course of EAE, suggesting that IRF8 in these cells does not

play a major role in the disease. IRF8-GFP+ T cells were very

low in unimmunized mice, and only a small fraction of CD4+ or

CD8+ T cells showed increased IRF8-GFP expression during

EAE (Figure S4C).

Aberrant Production of IL-12 Family Cytokines in Irf8–/–

MiceCytokines of the IL-12 family, IL-12 (p70), IL-23, and IL-27 are

produced in APCs and influence the course of EAE (Vignali and

Kuchroo, 2012). IL-23, composed of IL-12p40 and p19, pro-

motes Th17 cell expansion. Mice lacking either subunit or

IL-23R are resistant to EAE (Becher et al., 2003; Chen et al.,

2006; Cua et al., 2003). IL-12(p70), a heterodimer of IL-12p40

and p35, affects Th1 cell development (Vignali and Kuchroo,

2012). Alternatively, IL-27, composed of p28 and Ebi3, represses

differentiation of Th17 and other T cells and reduces EAE patho-

genesis (Batten et al., 2006; Stumhofer et al., 2006). Because we

found that IRF8 in APCs, not T cells, accounts for EAE develop-

ment, and because IRF8 has been shown to activate Il12b

(IL-12p40) transcription in response to pathogens and TLR stim-

uli, it was possible that IRF8 promotes EAE by regulating these

cytokines (Chang et al., 2012; Tailor et al., 2008). We therefore

tested mRNA expression of each subunit of the cytokines. As

seen in Figure 5A, Il12b expression greatly increased in WT

mice, but not in Irf8�/� mice. In keeping with these data, IL-

12p40 protein expression in sera markedly increased in WT

mice during EAE, but were near background in Irf8�/� mice (Fig-

ure 5B). Il23a (IL-23p19) and Il12a (IL-12p35) did not change

greatly and was similar in WT and Irf8�/� mice (Figure 5A; Fig-

ure S5A). Given that the biological activity of IL-12 (p70) and

IL-23 depends on IL-12p40, Irf8�/� APCs fail to create an EAE

promoting cytokine environment. Ebi3 mRNA levels also

increased during EAE and were consistently higher in Irf8�/�

mice than WT mice (Figure 5A). Il27 (IL-27p28) mRNA levels,

also increased during EAE, were variable in WT and Irf8�/�

mice (Figure S5A). This might be due to the variable expression

in non-APCs, because Il27 levels were higher in Irf8�/� adherent

cells than WT cells (Figure S5B). In accordance, IL-27 producing

macrophages and microglia were much higher in Irf8�/� mice

compared to those in WT mice during EAE (Figures 5C–5E).

IL-27 protein expression in sera was also higher in Irf8�/�

compared to WT mice during EAE (Figure S5C). The increase

in IL-27 production might contribute to defective Th17 genera-

tion in Irf8�/� mice, as evidenced by the following in vitro cocul-

ture assays: when CD4+ T cells were cultured with DCs obtained

from day 21 EAE mice, addition of anti-IL-27 Ab increased Th17

cell generation and IL-17 induction with Irf8�/� DCs to a greater

extent than with WT DCs (see Figures S7F and S7G). These

Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc. 5

Figure 5. Irf8–/– Mice Do Not Produce IL-12p40, but Overproduce IL-27 during EAE

(A) mRNA expression of Il12b (IL-12p40), Il23a (IL-23p19), and Ebi3 in WT and Irf8�/� mice on indicated days of EAE. Values for LN and spleen represent the

average of three independent experiments ± SEM. *p < 0.01, **p < 0.005. Samples from spinal cords were pooled from three mice.

(B) ELISA analysis of IL-12p40 protein levels in sera from WT and Irf8�/� mice on indicated days of EAE. Values represent the average of data from three mice ±

SEM. **p < 0.005.

(C) The number of IL-27+ macrophages in LN and spleen fromWT and Irf8�/�mice on day 14 of EAE (the average data from three independent mice). **p < 0.005.

(D) Flow cytometric detection of microglia (CD45loCD11b+) and macrophages (CD45hiCD11b+) in spinal cords from WT and Irf8�/� mice on day 21 of EAE.

(E) Flow cytometric detection of IL-27+ microglia and macrophages in spinal cords of WT and Irf8�/� mice on day 21. Data were obtained from a pool of three

spinal cords. Similar results were observed in three independent experiments. See also Figure S5.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

results indicate that IRF8 supports production of cytokines

favoring EAE, while repressing those that antagonize the

disease.

avb8 Integrin Transcription in APCs Depends on IRF8Whereas the above results pointed to the contribution of IL-12

family cytokines to EAE development, the complete absence

of effector T cell differentiation in Irf8�/� mice remained un-

explained, suggesting an additional defect at an earlier, sensiti-

zation stage. APC activation of TGFb signaling is a key event in

initiating Th17-cell development leading to EAE (Acharya et al.,

2010; Melton et al., 2010). This activity is mediated by integrin

molecules of the avb8 subtype expressed on APCs. The avb8

integrin liberates biologically active TGF-b at the APC-T cell junc-

tion to achieve effective delivery of TGF-b into naive T cells,

which in turn triggers Th17 cell differentiation (Acharya et al.,

6 Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc.

2010; Melton et al., 2010). We postulated that IRF8 acts on

this step by regulating transcription of Itgb8, the gene encoding

the avb8 integrin. To test this hypothesis, we tested Itgb8mRNA

expression in macrophages and DCs sorted fromWT and Irf8�/�

mice. Itgb8 was highly expressed in WT DCs and macrophages,

but expression was striking lower in the Irf8�/� counterparts

(Figure 6A). Additionally, Itgb8 expression was induced by

IFN-g and LPS stimulation both in WT BMDCs and BM-derived

macrophages, but Itgb8 induction was meager in Irf8�/� cells

(Figure 6B). Luciferase reporter assays were performed to test

IRF8 regulation of Itgb8 transcription in 293T cells transfected

with an Irf8 expression vector (Figure 6C). The reporter con-

taining the 1945 bp Itgb8 upstream fragment gave a robust

luciferase activity, only when cotransfected with Irf8 vector (Fig-

ure 6D). However, reporter activity by the 1081 bp fragment and

the basal (null) promoter was low and not enhanced by Irf8,

Figure 6. IRF8 Is Required for Itgb8 Expression in DCs and Macrophages

(A) Expression of integrin Itgb8mRNA in macrophages (CD11b+F4/80+) and DCs (CD11c+MHC II+) sorted from spleen of WT and Irf8�/� mice. Values represent

data from three independent sorts ± SEM. *p < 0.01.

(B) Left shows expression of Itgb8mRNA in BM-derived macrophages stimulated with IFN-g (100 U/ml) for 10 hr, followed by stimulation with LPS and CpG for

indicated hr. (B) Right shows expression of Itgb8 mRNA in BMDCs stimulated with LPS and CpG (1 mg/ml each). The values represent the average of three

independent experiments ± SEM. *p < 0.01, **p < 0.005.

(C) Diagram of Itgb8 reporter constructs. Indicated lengths of Itgb8 upstream fragments were tested for reporter activity. The �1,945 mut reporter contained

indicated mutations in the consensus ISRE motif.

(D) 293T cells in 12-well plates were transfected with the reporter constructs (1 mg), pcDNA-IRF8 or empty vector (0.5 mg), and pRL-TK vector (50 ng) for 24 hr.

Data represent the average of three independent assays ± SEM. **p < 0.005.

(E) ChIP detection of IRF8 binding to the Itgb8 promoter. IRF8 binding was normalized by control IgG binding, which was uniformly at background levels. Data

represent the average of three independent experiments ± SEM.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

suggesting that IRF8 enhances promoter activity through the

region between �1,081 and �1,945. The region contains puta-

tive IRF8 binding sites resembling the IFN-stimulated response

element (ISRE) (Figure 6C). When a mutation was placed in

one of these sites (�1,945 mut, Site 1), IRF8 failed to enhance

reporter activity, supporting the involvement of this site in

Itgb8 transcription (Figure 6D). Chromatin immunoprecipi-

tation (ChIP) assays were next performed for BMDCs from Irf8-

EGFP gene-targeted mice with antibody against GFP and

IRF8 (Figure 6E) for Site1 and two additional ISRE like sites,

Site 2 and Site 3. Substantial IRF8 binding was detected on

Site 1 by both antibodies, but IRF8 binding on other sites was

meager to background. These results show that Itgb8 is a direct

target of IRF8, which activates its transcription in APCs through

Site 1.

IRF8 Activates Latent TGF-b and Triggers Developmentof Th17 and Treg cellsTo test the possibility that IRF8 is involved in activation of TGF-b

signaling in APCs to prime naive T cells, we examined develop-

ment of Th17 cells in the presence of latent TGF-b from naive

CD4+ T cells that were cultured along with CD11c+ DCs

in vitro. Flow cytometric analyses in Figure 7A (upper panel)

and Figure S7B showed that with latent TGF-b, WT DCs, but

not Irf8�/� DCs, consistently stimulated Th17 cell development.

When cultured with active TGF-b, however, both WT and Irf8�/�

Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc. 7

Figure 7. IRF8 Is Required for Activation of

TGF-b Signaling and Initiation of Th17 Cell

Development

(A) Naive CD4+ T cells were cultured with DCs from

WT or Irf8�/� mice for 5 days in the absence or

presence of active TGF-b1, or latent TGF-b1 for

5 days with fresh media added on day 3. Cells

expressing IL-17A (upper panel) or P-SMAD2/3

(lower panel) were detected after PMA/iomomycin

treatment for 4 hr. Similar results were observed in

three independent experiments.

(B) Naive CD4+ T cells were cultured with DCs from

WT or Irf8�/� mice and analyzed for expression of

FOXP3. Similar results were observed in three in-

dependent experiments.

(C) P-SMAD2/3 expressing CD4+ T cells from WT

and Irf8�/� mice were analyzed 3 days after MOG

immunization by FACs. The table below represents

mean MFI values of five assays performed with in-

dependent mice ± SEM. *p < 0.01.

(D) Itgb8 and Irf8 mRNA levels were measured in LN

from WT and Irf8�/� mice on indicated days after

MOG immunization. Values represent the average of

three assays ± SEM. Similar results were observed in

five independent experiments. *p < 0.01, **p < 0.005.

(E) ELISA analysis of active and total TGF-b1 protein

levels in sera from WT and Irf8�/� mice on indicated

days after MOG immunization. Values represent the

average of data from five mice ± SEM. *p < 0.01,

**p < 0.005. See also Figures S6 and S7.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

8 Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc.

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

DCs supported Th17 cell generation, indicating defective TGF-b

activation in Irf8�/�DCs. As expected, no Th17 cells were gener-

ated without TGF-b or without DCs (Figure 7A; Figure S7A).

Consequently, the amount of IL-17 secreted in supernatants

was much lower when cultured with Irf8�/� DCs and latent

TGF-b, but comparable when cultured with active TGF-b

(Figure S7D). Moreover, Irf8�/� DCs did not stimulate Tregs

generation when cultured with latent TGF-b, while WT DCs did,

further supporting the idea that deficient TGF-b signaling

accounts for the priming defects in Irf8�/� DCs (Figure 7B). To

verify the role of IRF8 in TGF-b signaling, we tested SMAD

activation in APCs: among SMAD family of transcription factors

activated by TGF-b signaling, SMAD2 is shown to be critical for

Th17 cell activation (Malhotra et al., 2010). As shown in Figure 7A

(lower panel) and Figure S7C, WT DCs, but not Irf8�/� DCs led to

enhanced SMAD2 phosphorylation in T cells when incubated

with latent TGF-b. When incubated with active TGF-b, both WT

and Irf8�/� DCs enhanced SMAD2 phosphorylation. It is of

note that LPS induction of Il6 and Tnfa was comparable in WT

DCs and Irf8�/� DCs, verifying specificity of TGF-b signal

deficiency in Irf8�/� DCs (Figure S7E). Extending the above

in vitro observations to in vivo, we found that SMAD2 phosphor-

ylation was increased in CD4+ T cells upon EAE induction in WT

mice, but not Irf8�/�mice (Figure 7C). Moreover, Itgb8 and Irf8

transcript expression was enhanced in WT DCs during EAE,

but negligible in Irf8�/�DCs (Figure 7D).We also show that active

TGF-b protein levels in sera were lower in Irf8�/� mice than WT

mice during EAE, whereas total TGF-b protein levels varied,

although they tended to be lower in Irf8�/� mice (Figure 7E). In

addition, expression of Tgfb and Il6 mRNA was similar in WT

and Irf8�/� mice during EAE, indicating selective inability of

Irf8�/� mice in producing biologically active TGF-b (Figure S6A).

These results provide compelling evidence that IRF8 is essential

for TGF-b activation in APCs and hence sensitization of naive

CD4+ T to differentiate into Th17 cells and Tregs.

DISCUSSION

We show that Irf8�/� mice are resistant to EAE: Th17, Th1, and

Treg cells were not generated in Irf8�/� mice upon MOG

immunization. IRF8 in APCs, but not in T cells, promoted EAE

pathogenesis in a multifaceted manner. We show several key

mechanisms by which IRF8 promotes EAE, which likely con-

stitutes a basis of the MS risk factor (De Jager et al., 2009; Dis-

anto et al., 2012; Beecham et al., 2013).

IRF8 was required for TGF-b signaling that takes place at the

interface between APCs and naive T cells. The bulk of secreted

TGF-b is biologically inactive, as it is bound to the latency asso-

ciated protein (LAP). However, the avb8 integrin molecule on the

APC cell surface dissociates LAP to liberate functional TGF-b,

allowing direct delivery of TGF-b signal to T cells (Acharya

et al., 2010; Melton et al., 2010; Henderson and Sheppard,

2013). This step, coupled with TCR engagement drives T cells

to differentiate into Th17 cells (Acharya et al., 2010; Melton

et al., 2010). Our data show that IRF8 plays a role in this priming

step by enhancing Itgb8 transcription in APCs. The conclusion

that integrin-mediated TGF-b signaling depends on IRF8 is sup-

ported by the following observations: (1) Itgb8 transcript levels

were markedly lower in Irf8�/� APCs than WT cells, (2) IRF8

occupied the Itgb8 promoter and stimulated Itgb8 promoter

activity, (3) Irf8�/� DCs did not activate latent TGF-b and did

not initiate SMAD-based TGF-b signaling in T cells in vivo and

in vitro, and (4) Irf8�/� DCs failed to generate Th17 and Treg cells

in the presence of latent TGF-b. Consistent with these defects,

active TGF-b levels were lower in Irf8�/� mice during EAE.

Because Tregs are reported to promote Th17 cell development

under some conditions, defective Treg generation in Irf8�/�

mice might further contribute to the absence of EAE (Chen

et al., 2006; Pandiyan et al., 2011). Together, in view of the noted

requirement of TGF-b signaling for the EAE and for MS, IRF8 is

likely to exert a decisive influence on the onset of the disease

(Veldhoen et al., 2006).

Not only Th17 cells, but Th1 cells facilitate pathogenesis of

EAE and MS (Axtell et al., 2010; Kroenke et al., 2008; Lovett-

Racke et al., 2011). Th1 cells prompt macrophage infiltration

and NOS2 induction, whereas Th17 cells attract neutrophils

and stimulate GM-CSF production (Kroenke et al., 2008).

Although Th1 cell development might or might not depend on

TGF-b signaling, initial triggering of Th1 differentiation is likely

to be compromised in Irf8�/� mice as a result of the reduced

expression of MHC II and costimulatory molecules on Irf8�/�

APCs (Figure S6C).

Our results show that in addition to priming, IRF8 exerts a

major influence on subsequent stages of the disease, by facili-

tating expansion andmaintenance of Th17 and Th1 cells through

the regulation of IL-12 family cytokines. IL-12(p70) is critical for

development of Th1 cells in EAE, whereas IL-23 is necessary

for maintenance of Th17 cells (Becher et al., 2003; Cua et al.,

2003; Kroenke et al., 2008). IL-23 likely contributes not only to

EAE but to MS, because it is produced by microglia of MS

patients (Li et al., 2007). Due to the absence of the shared

IL-12p40 subunit, Irf8�/� mice are defective in producing both

IL-12p70 and IL-23, which would create an adverse environment

for the survival of Th1 and Th17 cells. Interestingly, Irf8�/� APCs

expressed greater levels of IL-27 thanWT cells, the cytokine that

negatively regulates development of Th17 and other T cells

(Stumhofer et al., 2006). Expression of Ebi3 mRNA was sig-

nificantly increased in Irf8�/� APCs relative to WT cells during

EAE. Irf8�/� mice had greater number of IL-27p28+ macro-

phages and microglia than WT mice. Together, IRF8 appears

to facilitate pro-Th1/Th17 cytokines and inhibits the anti-Th17

cytokine, generating a condition favoring inflammation and

exacerbation of EAE. The increase in IL-27 expression in Irf8�/�

mice was somewhat unexpected, considering the previous

report that IRF8 enhances IL-27p28 transcription in macro-

phages after TLR stimulation in vitro, while not affecting Ebi3

expression (Zhang et al., 2010). The different results obtained

in the two studies might be due to the differing cytokine environ-

ments in which IL-27 expression was examined, one in vitro after

TLR stimulation, the other in vivo after MOG immunization,

because it was previously found that IRF8 can act as a transcrip-

tional activator or a repressor in macrophages, depending on

stimulation (Chang et al., 2012). In addition, binding partners

for IRF8, such as IRF1, might also affect target gene regulation

(Horiuchi et al., 2012). In this respect, it is interesting to note

that Irf1�/� mice are partially resistant to EAE, suggesting the

possibility that IRF1 might cooperate with IRF8 to exacerbate

EAE (Tada et al., 1997).

Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc. 9

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

Microglia, the CNS-resident APCs, are activated in EAE and

play a role in disease progression (Starossom et al., 2012).

Analysis of Irf8-EGFP knockin mice found that the number of

IRF8-GFP+ microglia and levels of IRF8-GFP increased during

EAE. Although microglia are present in the CNS of Irf8�/� mice,

they express reduced levels of Iba1 and other microglia markers

and are deficient in IL-12p40 induction in vitro (Horiuchi et al.,

2012; Minten et al., 2012). Moreover, IRF8 is shown to be im-

portant for transforming quiescent microglia to a reactive

phenotype that produces proinflammatory cytokines after

peripheral nerve injury (Masuda et al., 2012). In agreement, our

results show that IRF8 plays a key role in microglial activation

during EAE by stimulating IL-12p40 expression, which in turn

increases intra-CNS amplification of Th1 and Th17 cells. We

also noted that Irf8�/� microglia produce more IL-27 than WT

mice. Thus, the ability to activate microglia might be another

attribute of IRF8 that boosts EAE disease progression. IRF8

might reinforce disease by enhancing expression of Ripk2 in

CNS-infiltrating DCs (Shaw et al., 2011). In light of the recent

report that IRF8 promotes Ly6C+ monocyte differentiation

(Kurotaki et al., 2013), which supports proinflammatory M1-

type macrophages, IRF8 appears to have a strong propensity

to steer APCs toward proinflammatory pathways and advance

EAE progression.

Lastly, it should be noted that IRF8 SNPs linked to the MS

susceptibility map to the 30 noncoding region of the IRF8 gene,

away from the transcription end sites (De Jager et al., 2009;

Disanto et al., 2012). This suggests that the SNPs are in a down-

stream regulatory sequence(s) yet to be fully characterized,

which influences IRF8 transcription in APCs. It will be of interest

to elucidate how this region affects IRF8 transcription and

whether drugs currently used or proposed for MS treatment

affect IRF8 expression.

In conclusion, IRF8, acting in APCs, activates TGF-b signaling

and primes naive T cells to initiate EAE. During the course of

disease, IRF8 fortifies inflammatory pathways to exacerbate

EAE pathogenesis. Together, this work highlights the multiplicity

of mechanisms by which IRF8 functions as a risk factor for MS.

EXPERIMENTAL PROCEDURES

Mice and EAE

Irf8�/�, Irf8-EGFP, and Irf8f/f mice of the C57BL/6 background and WT mice

were maintained in the NICHD animal facility (Feng et al., 2011) (H.W., data

not shown). LysM-cre (Lyz2tm1(cre)Ifo) and Lck-cre (Tg Lck-cre) mice (Jackson

Laboratories) were crossed with Irf8f/f mice to generate Irf8f/f-LysM-cre and

Irf8f/f Lck-cre mice (Clausen et al., 1999; Jakubzick et al., 2008). All animal

experiments were performed according to the animal study (ASP# 11-044)

approved by the Animal Care and Use Committees of NICHD, NIH. Female

mice (8–12 weeks old) were injected with 200 mg MOG35–55 in CFA containing

400 mg of killedM. tuberculosis subcutaneously at two sites and intraperitone-

ally with 100 ng of pertussis toxin, followed by injection of 100 ng pertussis

toxin 24 hr later (Hooke Laboratories). EAE symptoms were scored daily

according to the EAE scoring system.

Magnetic Resonance Imaging

MRI experiments were performed on a 7-T (Bruker Avance), 21 cm horizontal

scanner. A contiguous set of 1 mm thick, T1 weighted images (field-of-view

[FOV] = 25.6 3 12.8 mm, in-plane resolution of 100 mm, TE/TR = 6/100 ms),

encompassing the region between the Th9-L5 vertebrae (15 slices), were

acquired before and 5 min after intravenous administration of Gd-DTPA

(Berlex Laboratories). Quantitative T2 images (TE/TR = 20/2500 ms, in-plane

10 Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc.

resolution of 200 mm, number of echoes = 16) were acquired with the same

number of slices and slice locations as the T1-weighted images. T2 data

were analyzed with software by MATLAB (Mathworks Inc). The resultant T2maps were used to evaluate T2 values for four regions of interests (area

0.016 cm2) per slice.

Cell Preparations, Culture, and Flow Cytometry

Microglia-macrophage populations were prepared from spinal cord single-cell

suspensions by a 37%/70% percoll density gradient centrifugation, followed

by incubation on tissue culture plates for 1 hr. CD11c+ DCs were isolated

from draining LN and spleen (90%–95%) as described Acharya et al., 2010;

Melton et al., 2010). Naive CD4+ T cells were isolated from LN of WT mice

with biotin-conjugated anti-CD8a, CD11b, CD11c, CD19, CD45R, NK1.1,

Gr-1 Abs (Biolegend), and streptavidin microbeads (Miltenyi Biotec). For

Th17 development, 2.5 3 105 T cells were cultured with 105 DCs for 5 days

with anti-IL4 (10 mg/ml, Bio X cell), anti-IL-12 (10 mg/ml, Bio X Cell), anti-

IFN-g (10 mg/ml, Bio X Cell), anti-CD3 (1 mg/ml, eBiosciences), IL-6

(10 ng/ml, Pepro Tec), IL-1b (10 ng/ml, Pepro Tec), and TGF-b1 (5 ng/ml,

R&D Systems) or latent TGF-b1 (5 or 20 ng/ml, R&D Systems), supplemented

with freshmedia on day 3. For Treg development, cells were cultured with anti-

IL-4 (10 mg/ml, Bio X cell), anti-IL12 (10 mg/ml, Bio X Cell), anti-IFN-g (10 mg/ml,

Bio X Cell), anti-CD3 (1 mg/ml, eBiosciences), anti-CD28 (3 mg/ml,

eBiosciences), IL-2 (100 U/ml, Pepro Tec), and TGF-b1 (5 ng/ml, R&D

Systems) or latent TGF-b1 (5 or 20 ng/ml, R&D Systems). Transcripts were

measured by quantitative RT-PCR and normalized by Hprt as described

(Chang et al., 2012). The concentrations of IL-12p40, IL-27, and TGF-b1 in

sera from MOG-immunized mice were measured by ELISA using commercial

kits (eBioscience and R&D Systems). Flow cytometry cells were blocked with

anti-mouse FcgR Ab (CD16/32, clone 2.4G2), stained with Abs below, and

analyzed on FACSCalibur (BD Biosciences). Data were analyzed by FlowJo

software. The following antibodies were used: PerCP-conjugated anti-CD4,

allophycocyanin-conjugated anti-CD25 Abs (BD Biosciences), allophycocya-

nin-conjugated anti-F4/80, PE-conjugated anti-CD11b and PerCP-conjugated

anti-CD45 Abs, BD Biosciences), PE-conjugated anti-IL17, FITC-conjugated

anti-IFN-g (BD Biosciences) and FITC-conjugated anti-IL27 Abs (R&D

Systems) with Cytyofix/Cytoperm kits (BD Biosciences), or allophycocyanin-

conjugated anti-RORgt, PE-conjugated anti-FOXP3 Abs (eBiosciences), and

PE-conjugated anti-Phospho SMAD3/2 Abs (BD Biosciences). To purify DCs

and macrophages, we treated spleen cell suspensions with the Fc receptor

blocker 2.4G2 mAb followed by staining with antibodies against Gr-1,

CD11c, MHC II, CD11b, and F4/80 and sorted with a FACSAria-sorter (BD

Biosciences). Dead cells were excluded by gating on 7AAD� cells. DCs

were sorted as MHC II+CD11c+ and macrophages as CD11b+F4/80+. Purity

of the sorted populations was greater than 95%.

Itgb8 Promoter and ChIP Assays

Itgb8 upstream fragments (�1,952 to +3, �1,081 to +3) were amplified from

mouse genomic DNA with Phusion high-fidelity DNA polymerase (New

England Biolab) with the forward primer containing KpnI, 50-CGCAGGTAC

CAGGAAGAAATTTGAAAAACC-30,or 50-CGCAGGTACCATCTGACAGTAGAA

TACATC-30 and the reverse primer containing BamHI, 50-TGTGGGATCCTA

GAAGGCCAAGG-3. The amplicons were inserted into the KpnI-BglII site of

pGL4.20 luciferase reporter (Promega). 293T cells were transfected with the

reporter, pcDNA-IRF8, and pRL-TK for 24 hr, and luciferase activity was

measured with Dual-Glo Luciferase System (Promega). Chromatin from 106

BMDCs was immunoprecipitated with 0.5 to 1 mg of normal immunoglobulin

G (IgG), affinity purified anti-IRF8 antibody, or anti-GFP antibody bound to

Dynabeads (Chang et al., 2012). The following primers were used for site 1:

50-GTGGCTCTAACCATATCTTC-30 (forward), and 50-TGAAGCTAAGGGGG

CAATCC-30 (reverse), site 2; 50-CAAATGATCCTCCTTAACAG-30 (forward),

and 50-GTAGGAGGTCCCACCAACCC-30 (reverse), site 3; 50-ATCTGCA

ACTTCTTGTTTCC-30 (forward), and 50-CAGACTTTGGAGATTTAGAC-30

(reverse). Input DNA (2%) was used for normalization.

SUPPLEMENTAL INFORMATION

Supplemental Information includes seven figures and can be found with this

article online at http://dx.doi.org/10.1016/j.immuni.2013.11.022.

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

ACKNOWLEDGMENTS

We thank Alan Koretsky (NINDS) for his support and critical reading of the

manuscript, Takayuki Ito and Makoto Horiuchi (UC, Davis), Ulrich Siebenlist

(NIAID) for useful advice on experiments and reagents, and Mehrnoosh

Abshari for assistance in cell sorting and in vitro experiments. This work was

supported by the Intramural Program of NICHD, NIAID, and NINDS, National

Institutes of Health.

Received: March 14, 2013

Accepted: November 25, 2013

Published: January 30, 2014

REFERENCES

Acharya, M., Mukhopadhyay, S., Paıdassi, H., Jamil, T., Chow, C., Kissler, S.,

Stuart, L.M., Hynes, R.O., and Lacy-Hulbert, A. (2010). av Integrin expression

by DCs is required for Th17 cell differentiation and development of experi-

mental autoimmune encephalomyelitis in mice. J. Clin. Invest. 120, 4445–

4452.

Axtell, R.C., de Jong, B.A., Boniface, K., van der Voort, L.F., Bhat, R., De

Sarno, P., Naves, R., Han,M., Zhong, F., Castellanos, J.G., et al. (2010). T help-

er type 1 and 17 cells determine efficacy of interferon-beta in multiple sclerosis

and experimental encephalomyelitis. Nat. Med. 16, 406–412.

Batten, M., Li, J., Yi, S., Kljavin, N.M., Danilenko, D.M., Lucas, S., Lee, J., de

Sauvage, F.J., and Ghilardi, N. (2006). Interleukin 27 limits autoimmune

encephalomyelitis by suppressing the development of interleukin 17-produc-

ing T cells. Nat. Immunol. 7, 929–936.

Becher, B., Durell, B.G., and Noelle, R.J. (2003). IL-23 produced by CNS-resi-

dent cells controls T cell encephalitogenicity during the effector phase of

experimental autoimmune encephalomyelitis. J. Clin. Invest. 112, 1186–1191.

Beecham, A.H., Patsopoulos, N.A., Xifara, D.K., Davis, M.F., Kemppinen, A.,

Cotsapas, C., Shah, T.S., Spencer, C., Booth, D., Goris, A., et al.;

International Multiple Sclerosis Genetics Consortium (IMSGC); Wellcome

Trust Case Control Consortium 2 (WTCCC2); International IBD Genetics

Consortium (IIBDGC) (2013). Analysis of immune-related loci identifies 48

new susceptibility variants for multiple sclerosis. Nat. Genet. 45, 1353–1360.

Bettelli, E., Carrier, Y., Gao, W., Korn, T., Strom, T.B., Oukka, M., Weiner, H.L.,

and Kuchroo, V.K. (2006). Reciprocal developmental pathways for the gener-

ation of pathogenic effector TH17 and regulatory T cells. Nature 441, 235–238.

Chang, T.H., Xu, S., Tailor, P., Kanno, T., and Ozato, K. (2012). The small ubiq-

uitin-like modifier-deconjugating enzyme sentrin-specific peptidase 1

switches IFN regulatory factor 8 from a repressor to an activator during macro-

phage activation. J. Immunol. 189, 3548–3556.

Chen, Y., Langrish, C.L., McKenzie, B., Joyce-Shaikh, B., Stumhofer, J.S.,

McClanahan, T., Blumenschein, W., Churakovsa, T., Low, J., Presta, L.,

et al. (2006). Anti-IL-23 therapy inhibits multiple inflammatory pathways and

ameliorates autoimmune encephalomyelitis. J. Clin. Invest. 116, 1317–1326.

Clausen, B.E., Burkhardt, C., Reith, W., Renkawitz, R., and Forster, I. (1999).

Conditional gene targeting in macrophages and granulocytes using LysMcre

mice. Transgenic Res. 8, 265–277.

Cua, D.J., Sherlock, J., Chen, Y., Murphy, C.A., Joyce, B., Seymour, B.,

Lucian, L., To, W., Kwan, S., Churakova, T., et al. (2003). Interleukin-23 rather

than interleukin-12 is the critical cytokine for autoimmune inflammation of the

brain. Nature 421, 744–748.

De Jager, P.L., Jia, X., Wang, J., de Bakker, P.I.W., Ottoboni, L., Aggarwal,

N.T., Piccio, L., Raychaudhuri, S., Tran, D., Aubin, C., et al.; International MS

Genetics Consortium (2009). Meta-analysis of genome scans and replication

identify CD6, IRF8 and TNFRSF1A as new multiple sclerosis susceptibility

loci. Nat. Genet. 41, 776–782.

Disanto, G., Sandve, G.K., Berlanga-Taylor, A.J., Morahan, J.M., Dobson, R.,

Giovannoni, G., and Ramagopalan, S.V. (2012). Genomic regions associated

with multiple sclerosis are active in B cells. PLoS ONE 7, e32281.

Feng, J., Wang, H., Shin, D.M., Masiuk, M., Qi, C.F., and Morse, H.C., 3rd.

(2011). IFN regulatory factor 8 restricts the size of the marginal zone and follic-

ular B cell pools. J. Immunol. 186, 1458–1466.

Hambleton, S., Salem, S., Bustamante, J., Bigley, V., Boisson-Dupuis, S.,

Azevedo, J., Fortin, A., Haniffa, M., Ceron-Gutierrez, L., Bacon, C.M., et al.

(2011). IRF8 Mutations and Human Dendritic-Cell Immunodeficiency. N Engl

J Med.

Hauser, S.L., and Oksenberg, J.R. (2006). The neurobiology of multiple scle-

rosis: genes, inflammation, and neurodegeneration. Neuron 52, 61–76.

Henderson, N.C., and Sheppard, D. (2013). Integrin-mediated regulation of

TGFb in fibrosis. Biochim. Biophys. Acta 1832, 891–896.

Horiuchi, M., Wakayama, K., Itoh, A., Kawai, K., Pleasure, D., Ozato, K., and

Itoh, T. (2012). Interferon regulatory factor 8/interferon consensus sequence

binding protein is a critical transcription factor for the physiological phenotype

of microglia. J. Neuroinflammation 9, 227.

Ivanov, I.I., McKenzie, B.S., Zhou, L., Tadokoro, C.E., Lepelley, A., Lafaille,

J.J., Cua, D.J., and Littman, D.R. (2006). The orphan nuclear receptor

RORgammat directs the differentiation program of proinflammatory IL-17+ T

helper cells. Cell 126, 1121–1133.

Jakubzick, C., Bogunovic, M., Bonito, A.J., Kuan, E.L., Merad, M., and

Randolph, G.J. (2008). Lymph-migrating, tissue-derived dendritic cells are

minor constituents within steady-state lymph nodes. J. Exp. Med. 205,

2839–2850.

Komiyama, Y., Nakae, S., Matsuki, T., Nambu, A., Ishigame, H., Kakuta, S.,

Sudo, K., and Iwakura, Y. (2006). IL-17 plays an important role in the develop-

ment of experimental autoimmune encephalomyelitis. J. Immunol. 177,

566–573.

Kroenke, M.A., Carlson, T.J., Andjelkovic, A.V., and Segal, B.M. (2008). IL-12-

and IL-23-modulated T cells induce distinct types of EAE based on histology,

CNS chemokine profile, and response to cytokine inhibition. J. Exp. Med. 205,

1535–1541.

Kurotaki, D., Osato, N., Nishiyama, A., Yamamoto, M., Ban, T., Sato, H.,

Nakabayashi, J., Umehara, M., Miyake, N., Matsumoto, N., et al. (2013).

Essential role of the IRF8-KLF4 transcription factor cascade in murine mono-

cyte differentiation. Blood 121, 1839–1849.

Laricchia-Robbio, L., Tamura, T., Karpova, T., Sprague, B.L., McNally, J.G.,

and Ozato, K. (2005). Partner-regulated interaction of IFN regulatory factor 8

with chromatin visualized in live macrophages. Proc. Natl. Acad. Sci. USA

102, 14368–14373.

Li, Y., Chu, N., Hu, A., Gran, B., Rostami, A., and Zhang, G.-X. (2007).

Increased IL-23p19 expression in multiple sclerosis lesions and its induction

in microglia. Brain 130, 490–501.

Lovett-Racke, A.E., Yang, Y., and Racke, M.K. (2011). Th1 versus Th17: are

T cell cytokines relevant in multiple sclerosis? Biochim. Biophys. Acta 1812,

246–251.

Malhotra, N., Robertson, E., and Kang, J. (2010). SMAD2 is essential for TGF

beta-mediated Th17 cell generation. J. Biol. Chem. 285, 29044–29048.

Masuda, T., Tsuda, M., Yoshinaga, R., Tozaki-Saitoh, H., Ozato, K., Tamura,

T., and Inoue, K. (2012). IRF8 is a critical transcription factor for transforming

microglia into a reactive phenotype. Cell Rep 1, 334–340.

Melton, A.C., Bailey-Bucktrout, S.L., Travis, M.A., Fife, B.T., Bluestone, J.A.,

and Sheppard, D. (2010). Expression of avb8 integrin on dendritic cells regu-

lates Th17 cell development and experimental autoimmune encephalomyelitis

in mice. J. Clin. Invest. 120, 4436–4444.

Minten, C., Terry, R., Deffrasnes, C., King, N.J., and Campbell, I.L. (2012). IFN

regulatory factor 8 is a key constitutive determinant of the morphological and

molecular properties of microglia in the CNS. PLoS ONE 7, e49851.

Miyagawa, F., Zhang, H., Terunuma, A., Ozato, K., Tagaya, Y., and Katz, S.I.

(2012). Interferon regulatory factor 8 integrates T-cell receptor and cytokine-

signaling pathways and drives effector differentiation of CD8 T cells. Proc.

Natl. Acad. Sci. USA 109, 12123–12128.

Ouyang, X., Zhang, R., Yang, J., Li, Q., Qin, L., Zhu, C., Liu, J., Ning, H., Shin,

M.S., Gupta, M., et al. (2011). Transcription factor IRF8 directs a silencing

programme for TH17 cell differentiation. Nat Commun 2, 314.

Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc. 11

Immunity

Multifaceted Role of IRF8 in Neuroinflammation

Please cite this article in press as: Yoshida et al., The Transcription Factor IRF8 Activates Integrin-Mediated TGF-b Signaling and Promotes Neuro-inflammation, Immunity (2014), http://dx.doi.org/10.1016/j.immuni.2013.11.022

Pandiyan, P., Conti, H.R., Zheng, L., Peterson, A.C., Mathern, D.R.,

Hernandez-Santos, N., Edgerton, M., Gaffen, S.L., and Lenardo, M.J. (2011).

CD4(+)CD25(+)Foxp3(+) regulatory T cells promote Th17 cells in vitro and

enhance host resistance in mouseCandida albicans Th17 cell infection model.

Immunity 34, 422–434.

Park, H., Li, Z., Yang, X.O., Chang, S.H., Nurieva, R., Wang, Y.H., Wang, Y.,

Hood, L., Zhu, Z., Tian, Q., and Dong, C. (2005). A distinct lineage of CD4

T cells regulates tissue inflammation by producing interleukin 17. Nat.

Immunol. 6, 1133–1141.

Schellenberg, A.E., Buist, R., Yong, V.W., Del Bigio, M.R., and Peeling, J.

(2007). Magnetic resonance imaging of blood-spinal cord barrier disruption

in mice with experimental autoimmune encephalomyelitis. Magn. Reson.

Med. 58, 298–305.

Shaw, P.J., Barr, M.J., Lukens, J.R., McGargill, M.A., Chi, H., Mak, T.W., and

Kanneganti, T.D. (2011). Signaling via the RIP2 adaptor protein in central

nervous system-infiltrating dendritic cells promotes inflammation and autoim-

munity. Immunity 34, 75–84.

Sørensen, T.L., Tani, M., Jensen, J., Pierce, V., Lucchinetti, C., Folcik, V.A.,

Qin, S., Rottman, J., Sellebjerg, F., Strieter, R.M., et al. (1999). Expression of

specific chemokines and chemokine receptors in the central nervous system

of multiple sclerosis patients. J. Clin. Invest. 103, 807–815.

Starossom, S.C., Mascanfroni, I.D., Imitola, J., Cao, L., Raddassi, K.,

Hernandez, S.F., Bassil, R., Croci, D.O., Cerliani, J.P., Delacour, D., et al.

(2012). Galectin-1 deactivates classically activated microglia and protects

from inflammation-induced neurodegeneration. Immunity 37, 249–263.

Steinman, L., and Zamvil, S.S. (2006). How to successfully apply animal

studies in experimental allergic encephalomyelitis to research onmultiple scle-

rosis. Ann. Neurol. 60, 12–21.

Stromnes, I.M., Cerretti, L.M., Liggitt, D., Harris, R.A., and Goverman, J.M.

(2008). Differential regulation of central nervous system autoimmunity by

T(H)1 and T(H)17 cells. Nat. Med. 14, 337–342.

Stumhofer, J.S., Laurence, A., Wilson, E.H., Huang, E., Tato, C.M., Johnson,

L.M., Villarino, A.V., Huang, Q., Yoshimura, A., Sehy, D., et al. (2006).

Interleukin 27 negatively regulates the development of interleukin 17-produc-

12 Immunity 40, 1–12, February 20, 2014 ª2014 Elsevier Inc.

ing T helper cells during chronic inflammation of the central nervous system.

Nat. Immunol. 7, 937–945.

Tada, Y., Ho, A., Matsuyama, T., and Mak, T.W. (1997). Reduced incidence

and severity of antigen-induced autoimmune diseases in mice lacking inter-

feron regulatory factor-1. J. Exp. Med. 185, 231–238.

Tailor, P., Tamura, T., Morse, H.C., 3rd, and Ozato, K. (2008). The BXH2

mutation in IRF8 differentially impairs dendritic cell subset development in

the mouse. Blood 111, 1942–1945.

Tamura, T., Tailor, P., Yamaoka, K., Kong, H.J., Tsujimura, H., O’Shea, J.J.,

Singh, H., and Ozato, K. (2005). IFN regulatory factor-4 and -8 govern dendritic

cell subset development and their functional diversity. J. Immunol. 174, 2573–

2581.

Travis, M.A., Reizis, B., Melton, A.C., Masteller, E., Tang, Q., Proctor, J.M.,

Wang, Y., Bernstein, X., Huang, X., Reichardt, L.F., et al. (2007). Loss of integ-

rin alpha(v)beta8 on dendritic cells causes autoimmunity and colitis in mice.

Nature 449, 361–365.

Veldhoen, M., Hocking, R.J., Flavell, R.A., and Stockinger, B. (2006). Signals

mediated by transforming growth factor-beta initiate autoimmune encephalo-

myelitis, but chronic inflammation is needed to sustain disease. Nat. Immunol.

7, 1151–1156.

Vignali, D.A., and Kuchroo, V.K. (2012). IL-12 family cytokines: immunological

playmakers. Nat. Immunol. 13, 722–728.

Weaver, C.T., Hatton, R.D., Mangan, P.R., and Harrington, L.E. (2007). IL-17

family cytokines and the expanding diversity of effector T cell lineages.

Annu. Rev. Immunol. 25, 821–852.

Yang, X.O., Pappu, B.P., Nurieva, R., Akimzhanov, A., Kang, H.S., Chung, Y.,

Ma, L., Shah, B., Panopoulos, A.D., Schluns, K.S., et al. (2008). T helper 17

lineage differentiation is programmed by orphan nuclear receptors ROR alpha

and ROR gamma. Immunity 28, 29–39.

Zhang, J., Qian, X., Ning, H., Yang, J., Xiong, H., and Liu, J. (2010). Activation of

IL-27 p28 gene transcription by interferon regulatory factor 8 in cooperation

with interferon regulatory factor 1. J. Biol. Chem. 285, 21269–21281.