of February 23, 2015.This information is current as

Encephalomyelitis DevelopmentExperimental AutoimmuneNegative Regulator of Th17 Response and Acute-Phase Protein Hemopexin Is a

TolosanoSilengo, Fiorella Altruda, Francesco Novelli and Emanuela Simona Rolla, Giada Ingoglia, Valentina Bardina, Lorenzo

http://www.jimmunol.org/content/191/11/5451doi: 10.4049/jimmunol.1203076October 2013;

2013; 191:5451-5459; Prepublished online 23J Immunol 

MaterialSupplementary

6.DC1.htmlhttp://www.jimmunol.org/content/suppl/2013/10/23/jimmunol.120307

Referenceshttp://www.jimmunol.org/content/191/11/5451.full#ref-list-1

, 12 of which you can access for free at: cites 38 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2013 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

at B

ibl Unito on February 23, 2015

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Acute-Phase Protein Hemopexin Is a Negative Regulator ofTh17 Response and Experimental AutoimmuneEncephalomyelitis Development

Simona Rolla,*,†,1 Giada Ingoglia,*,‡,1 Valentina Bardina,*,† Lorenzo Silengo,*,‡

Fiorella Altruda,*,‡ Francesco Novelli,*,†,2 and Emanuela Tolosano*,‡,2

Hemopexin (Hx) is an acute-phase protein synthesized by hepatocytes in response to the proinflammatory cytokines IL-6, IL-1b, and

TNF-a. Hx is the plasma protein with the highest binding affinity to heme and controls heme-iron availability in tissues and also in

T lymphocytes, where it modulates their responsiveness to IFN-g. Recent data have questioned regarding an anti-inflammatory role

of Hx, a role that may be both heme-binding dependent and independent. The aim of this study was to investigate the role of Hx in

the development of a T cell–mediated inflammatory autoimmune response. During experimental autoimmune encephalomyelitis

(EAE), the mouse model of multiple sclerosis, Hx content in serum increased and remained high. When EAE was induced in Hx

knockout (Hx2/2) mice, they developed a clinically earlier and exacerbated EAE compared with wild-type mice, associated to

a higher amount of CD4+-infiltrating T cells. The severe EAE developed by Hx2/2mice could be ascribed to an enhanced expansion

of Th17 cells accounting for both a higher disposition of naive T cells to differentiate toward the Th17 lineage and a higher

production of Th17 differentiating cytokines IL-6 and IL-23 by APCs. When purified human Hx was injected in Hx2/2 mice before

EAE induction, Th17 expansion, as well as disease severity, were comparable with those of wild-type mice. Taken together, these

data indicate that Hx has a negative regulatory role in Th17-mediated inflammation and prospect its pharmacological use to limit

the expansion of this cell subset in inflammatory and autoimmune disease. The Journal of Immunology, 2013, 191: 5451–5459.

Hemopexin (Hx) is an acute-phase protein synthesized byhepatocytes. Systemic concentration of Hx increasesseveral-fold in response to the proinflammatory cyto-

kines IL-6, IL-1b, and TNF-a (1). Other than in plasma, Hx is alsoexpressed in human and mouse brain and in the neural retina(1–5), and it can also be induced in peripheral nerves in responseto injury (6–8). Moreover, Hx has been found in human cere-brospinal fluid both in physiologic and pathologic conditions(9, 10). Hx binds to free heme with high affinity (Kd ,1 pM), andby scavenging heme from the bloodstream, it prevents heme-me-diated endothelial and tissue damage (11). Moreover, recent works

demonstrated an anti-inflammatory role for Hx through its abilityto modulate the expression of proinflammatory cytokines both ina heme-dependent and -independent way (12, 13).In humans, Hx has been chosen as a potential blood biomarker in

pediatric multiple sclerosis (MS) (14) likely because of its role in theacute phase of inflammation. MS is a chronic inflammatory diseaseof the CNS mainly characterized by intervals of remission followedby relapse (15–17). However, acute progressive cases are docu-mented. Onset of this disease initiates outside of the CNS throughactivation of CD4+ T cells by myelin-like antigenic peptides. Thesecells then migrate across the blood–brain barrier initiating focal in-flammation. Encephalitogenic T cells that invade the CNS interactwith APCs, resulting in reactivation of the T cells and activation ofthe APCs (18). Both IFN-g–secreting Th1 and IL-17–secreting Th(Th17) cells have been shown to have a pathogenic role in MS (19)and experimental autoimmune encephalomyelitis (EAE), the mousemodel of MS (20, 21), but an augment of circulating Th17 cells isoften associated with active phases of the disease (22), indicatinga key role of Th17 cells in the exacerbation of the clinical symptoms.To investigate the role of Hx in autoimmune inflammation, we

studied the course of EAE induced by immunization of Hx knockout(Hx2/2) and syngenic wild-type mice with myelin oligodendrocyteglycoprotein peptide (MOG35–55). Our data show that in the absenceof Hx, the clinical symptoms of EAE are more severe and earlierand are associated to a higher number of infiltrating and circulatingTh17 cells. We also found that Hx has a crucial role in the differ-entiation of Th17 cells, and it is able to modulate the expressionlevels of Th17 prodifferentiating cytokines as IL-6 and IL-23.

Materials and MethodsMice

Hx2/2 mice on the 129/Sv background (H-2b) were previously described(23). All mice were housed in our animal facility, with a 12-h dark/light

*Department of Molecular Biotechnology and Health Sciences, University of Turin,10126 Turin, Italy; †Center for Experimental Research and Medical Studies, SanGiovanni Battista Hospital, 10126 Turin, Italy; and ‡Molecular Biotechnology Cen-ter, University of Turin, 10126 Turin, Italy

1S.R. and G.I. equally contributed to this work.

2F.N. and E.T. equally contributed to this work.

Received for publication November 6, 2012. Accepted for publication September 20,2013.

This work was supported by grants from the Federazione Italiana Sclerosi Multipla(2007/R/10 and 2009/R19), Ministero dell’Istruzione, dell’Universita e della Ricerca,Progetti di Rilevante Interesse Nazionale; Regione Piemonte: Ricerca Industriale eSviluppo Precompetitivo, Ricerca Industriale “Converging Technologies”, Pro-getti Strategici su Tematiche di Interesse Regionale o Sovra Regionale, and Pro-getti di Ricerca Sanitaria Finalizzata.

Address correspondence and reprint requests to Prof. Emanuela Tolosano, MolecularBiotechnology Center, Department of Molecular Biotechnology and Health Sciences,University of Turin, Via Nizza 52, 10126 Turin, Italy. E-mail address: emanuela.tolosano@unito.it

The online version of this article contains supplemental material.

Abbreviations used in this article: EAE, experimental autoimmune encephalomyelitis;Hx, hemopexin; Hx2/2, Hx knockout; LN, lymph node; MBP, myelin basic protein;MOG, myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; qRT, quantitativereal-time; SPC, spleen cell; TdR, thymidine.

Copyright� 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1203076

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

cycle and access to standard laboratory chow and tap water ad libitum.Sex- and age-matched wild-type mice on the 129/Sv background wereused as controls. For all experiments, 6- to 8-wk-old male mice were used.Animal studies were approved by the animal ethical committee of theUniversity of Turin, Turin, Italy.

EAE in mice

Hx2/2 and wild-type mice were immunized s.c. with 200 mg MOG35–55

peptide (MEVGWYRSPFSRVVHLYRNGK; Sigma, St. Louis, MO)emulsified in CFA containing 400 mg Mycobacterium tuberculosis H37Ra(Sigma). At days 0 and at day 2 postimmunization, 200 ng pertussis toxin(from Bordetella pertussis; Sigma) was administered i.p. Clinical scorewas assessed daily, in a blinded manner, on a scale of 0–5: 0, no disease; 1,limp tail; 2, hind-limb weakness; 3, hind-limb paralysis; 4, hind-limb andforelimb paralysis; 5, moribund/dead state. The onset of the disease refersto clinical score equal to 1.

In rescue experiments, purified human Hx (Athens Research, Athens,GA) was injected i.p. in Hx2/2 mice at day 0 of EAE at a dose of 2 mg/animal.

Tissue histology and immunohistochemistry

Hx2/2 and wild-type mice were sacrificed at day 28 after induction of EAEand perfused with 35 ml ice-cold PBS through the left ventricle. Spinalcords were isolated and fixed in 4% paraformaldehyde for 24 h, dehydratedin PBS–30% sucrose (Sigma) for 5 d, embedded in optimal cryopreservingtissue compound (OCT; Miles Laboratories, Elkhart, IN), snap-frozen, andsectioned transversally (7 mm/section). Sections were stained with H&E.High-resolution images of sections were captured at 3100 magnificationusing an Olympus BH-2 microscope, coupled to an Olympus camera. Thetotal cell infiltration (inflammatory index) was assessed as follows: eachspinal cord section was subdivided into 16 areas, and each subdivided areadisplaying lymphocyte infiltration was assigned a score of 1; thus, eachanimal had a potential maximum score of 16. Four sections of the lumbarspinal cord per animal were analyzed. The presence of CD4+ cells wasevaluated by immunofluorescence: sections were incubated 1 h at roomtemperature with rat anti-mouse CD4-FITC diluted 1:100 (Biolegend, SanDiego, CA); Hoechst (Invitrogen, Carlsbad, CA) was used to stain the nu-clei. High-resolution images of sections were captured at 2003 magnifi-cation on a Zeiss fluorescence microscope (Zeiss, Thornwood, NY) withZeiss ApoTome image visualization system. The number of CD4+ infiltratedcells in four consecutive sections of the lumbar spinal cord of four wild-typeand four Hx2/2 mice was blindly assessed. Evaluation of demyelizationwas performed by immunohistochemistry. Sections were incubated over-night with a rat anti-mouse myelin basic protein (MBP) Ab (Abcam,Cambridge, U.K.), diluted 1:100, followed by biotin-conjugated secondaryAb. The sections were visualized on an Olympus BH-2microscope, coupledto an Olympus camera. As a control of the primary Ab specificity, serialsections were always processed in parallel with the secondary Ab alone.

Sera analysis

Serumwas collected from the tail vein at day 0 and at days 2, 4, 7, 14, and28 after immunization. One microliter of serum was separated on 10%NaDodSO4-PAGE and analyzed by Western blotting using an anti-Hx Ab(mAb 3D6/E12) (1). An anti-mouse IgG Ab (Abcam) was used as control.Alternatively, red Ponceau staining was used as loading control. Membraneswere then incubated with 1:5000 HRP-conjugated anti-mouse IgG (Abcam)and revealed by ECL on a ChemiDoc system (Bio-Rad Laboratories, Seg-rate, Italy). Quantification of proteins was performed by densitometryanalysis. IL-6 content in the sera was measured by ELISA kit (Biolegend).

Lymphocyte isolation

To isolate spinal cord and brain infiltrates, we sacrificed Hx2/2 and wild-type mice at days 14 or 28 after induction of EAE, and we performedperfusion with 35 ml ice-cold PBS through the left ventricle. Spinal cordsand brains were dissected out and digested with 2.5 mg/ml collagenase D(Roche, Mannheim, Germany) for 45 min at 37˚C. The tissue was ho-mogenized in cold PBS, and the tissue suspensions were passed througha 70-mm cell strainer to obtain a single-cell suspension and then centri-fuged. Cells were resuspended in 37% Percoll (GE Healthcare EuropeGmbH, Milan, Italy) and overlaid with 70% Percoll. After centrifugation at1000 3 g for 25 min, the cell monolayer at the 70–37% interphase wascollected, washed in RPMI 1640 (Invitrogen), and used for flow cytometry.Spleens and lymph nodes (LNs) were collected from mice at day 0 and atdays 7 and 14 of EAE, harvested, and single-cell suspensions were pre-pared by passage through a 70-mm cell strainer. Blood from Hx2/2 andwild-type mice at days 7, 14, 21, and 28 after EAE induction was collected

in heparinized tubes, and PBMCs were isolated by Ficoll (EuroClone, Pero,Italy) gradient.

Flow cytometry

Spinal cord and brain mononuclear cells, LNs cells, spleen cells (SPCs), andPBMCs were cultured in 48-well plates (1 3 106 cells/well) and activatedwith 50 ng/ml PMA (Sigma), 500 ng/ml Ionomycin (Sigma), and 10 ng/mlbrefeldin A (Sigma) for 18 h. Cells were then harvested, washed in PBS-1%BSA, and labeled on the surface with FITC anti-mouse CD4 mAb (Bio-legend). Cells were then washed and fixed in 4% paraformaldehyde for15 min at room temperature; after washing, cells were permeabilized with5% saponin (Sigma) in PBS and labeled intracellularly with PerCP anti-mouse IL-17 (Biolegend) or PE anti-mouse IFN-g mAb (Biolegend),20 min at 4˚C. Stained cells were acquired on a FACSCalibur and analyzedwith CellQuest (BD Biosciences, Erembodegem, Belgium). Each plotrepresents the results from 50,000 events. To depict cytokine release inresponse to MOG35–55, we cultured SPCs isolated at day 0 and at days 7 and14 postimmunization with or without MOG35–55 (10 mg/ml) in 24-wellplates (43 106 cells/well) for 10 d. After 3 d, IL-2 (1 ng/ml) and IL-23 (10ng/ml) were added to the cultures and replaced every 3 d. At day 9, cellswere activated with PMA, Ionomycin, and brefeldin A for 18 h to performintracellular staining as previously described. The percentage of MOG35–

55–specific Th17 and Th1 cells was measured by flow cytometry subtractingthe percentage of cells cultured in absence of MOG35–55. For the mea-surement of cytokine production, cells were cultured with MOG35–55 pep-tide for 48 h and supernatants tested by using an ELISA kit (Biolegend).

Th17 cell differentiation

Spleens were collected from Hx2/2 and wild-type untreated mice, andsingle-cell suspensions were prepared by passage through a cell strainer(BD Biosciences). CD4+ naive cells were positively isolated by anti-CD4and anti-CD62L microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany),and 2 3 106 CD4+ naive cells were cultured in 24-well plates, coated with 2mg/ml anti-CD3 mAb and 2 mg/ml anti-CD28 (BD Biosciences) in thepresence of Th17 prodifferentiating cytokines, 50 ng/ml IL-6 (Pepro-tech, Rocky Hill, NJ), 5 ng/ml TGF-b (Peprotech), and 10 ng/ml IL-23(R&D Systems, Abingdon, U.K.), with or without IFN-g blocking Ab.Supernatants were collected after 48 h and IL-17 production was measuredby using an ELISA kit (Biolegend). Lymphocytes of the negative fractionwere used as memory cells and cultured with or without 10 mg/ml IL-23plus IFN-g blocking Ab.

Cell proliferation

Spleen cells were harvested and single-cell suspensions were prepared bypassage through a cell strainer (BD Biosciences, Erembodegem, Belgium).Splenocytes were cultured in triplicate in round-bottom 96-well platesin the absence or presence of different concentrations of MOG35–55 pep-tide (0, 1, 10, 100 mg/ml). After 48 h, the cultures were pulsed with 1 mCi[3H]thymidine (TdR) (Amersham, Milan, Italy). After 18 h, the cells wereharvested and [3H]TdR uptake was evaluated by TopCount Microplatereader (Packard Instrument, Wellesley, MA).

Quantitative real-time PCR analysis

Macrophages and dendritic cells were separated from spleens collected atday 0 and at day 7 after EAE induction using anti-F4/80 and anti-CD11cmicrobeads, respectively, following the manufacturer’s instructions (Mil-tenyi Biotec). Total RNA was extracted from cells using RNeasy mini kit(Qiagen, Milan, Italy). cDNAwas prepared from 1 mg total RNA usingPromega M-MLV Retro Transcriptase (Promega Italia, Milan, Italy).Quantitative real-time (qRT)-PCR was performed on a 7300 Real Time PCRSystem (Applied Biosystems, Foster City, CA). Primers and probes weredesigned using the ProbeFinder software (http://www.roche-applied-science.com). All results were normalized to 18S mRNA.

Statistical analyses

Statistical analyses were performed with GraphPad Prism (GraphPadSoftware, San Diego, CA) by using one-way or two-way ANOVA, or thenonparametric two-tailed Mann–Whitney U test. A p value ,0.05 wasconsidered significant.

ResultsHx is upregulated during EAE

To understand whether Hx was modulated during EAE in mice, wefirst analyzed its expression in the liver of wild-type mice before

5452 HEMOPEXIN MODULATES CD4 T CELL DIFFERENTIATION

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

and at day 7 after the immunization with MOG35–55 peptide. HxmRNA levels increased in wild-type liver at day 7 postinjection(Fig. 1A). Moreover, we analyzed Hx in the serum of wild-typemice collected before and at days 2, 4, 7, 14, and 28 after MOG35–55

peptide immunization. As shown in Fig. 1B and 1C, Hx levels in-creased by day 2 after immunization and remained high until day14. In addition, Hx expression was also upregulated in the CNS ofwild-type mice with EAE (Supplemental Fig. 1). Thus, Hx is in-duced during murine EAE.

Loss of Hx exacerbates EAE disease

To investigate whether Hx modulates the autoimmune response, weexamined the effects of Hx deficiency on the development of EAE.Disease progression was analyzed in wild-type and Hx2/2 mice atdifferent time points after immunization with MOG35–55 peptide.In wild-type mice, the initial signs of EAE were observed 11.6 61.9 d after immunization (Fig. 2A, 2B). In contrast, in Hx2/2

mice, the onset of EAE occurred on day 7.7 6 1.3 (Fig. 2A, 2B).Moreover, 28 d after immunization, Hx2/2 mice exhibited ahigher EAE score compared with wild-type mice (3.30 + 0.60versus 2.40 + 0.43; Fig. 2C). Consistent with these observations,Hx2/2 mice displayed a massive inflammatory infiltration in theCNS compared with wild-type mice (Fig. 2D). Zones of demye-lization in CNS, detectable where inflammatory infiltrates occurred,were more extended in Hx2/2 mice than in wild-type mice (Fig.2E). Among inflammatory cells, the number of CD4+ T cells wassignificantly higher in the CNS of Hx2/2 mice than in that of wild-type animals (Fig. 3A). To better characterize CNS-infiltrated CD4T cells, we isolated spinal cord and brain mononuclear cells fromwild-type and Hx2/2 mice at days 14 and 28 postimmunization,activated polyclonally with PMA, Ionomycin, and brefeldin A,and stained with an anti-CD4 Ab and anti–IL-17 or anti–INF-gAbs. As shown in Fig. 3B–D, the number of CD4+ T cells pro-

ducing IL-17 (Th17) in the spinal cord, as well as in the brain, wassignificantly increased in Hx2/2 mice as compared with wild-typeanimals at both days 14 and 28, whereas that of CD4+ T cellsproducing IFN-g (Th1) was comparable. Taken together, theseresults indicate that the loss of Hx results in a more severe EAEassociated to a higher number of Th17-infiltrating cells in theCNS.

Loss of Hx results in an expansion of Th17 cells

Because we observed a difference in the amount of Th17 cellsinfiltrating the CNS between Hx2/2 and wild-type mice, the ex-pansion of Th17 and Th1 cells during EAE in LN cells, SPCs, andPMBCs was assessed. LN cells, SPCs, and PMBCs were isolatedfrom wild-type and Hx2/2 mice at days 7 and 14 postimmuni-zation, activated polyclonally with PMA, Ionomycin, and bre-feldin A, and then stained intracellularly to detect the productionof IL-17 and INF-g. At day 7 of the disease, corresponding to thedisease onset in Hx2/2 mice, Hx2/2 mice showed a higher per-centage of Th17 cells in LNs and spleen compared with wild-typemice (Fig. 4A, 4B). By contrast, at day 14, the percentage of Th17cells was higher in LNs and spleen of the wild-type counterpart. Asignificantly higher percentage of Th17 cells was also observed inPBMCs of Hx2/2 mice compared with wild-type mice (Fig. 4C).No difference in the percentage of Th1 cells in LNs, spleen, andPBMCs between Hx2/2 and wild-type mice was detected (Fig.4A–C).Then the response of lymphocytes after in vitro stimulation

with MOG35–55 was investigated. SPCs were isolated at day 0 andat days 7 and 14 after immunization, cultured for 10 d in thepresence of 10 mg/ml MOG35–55 peptide, and then checked for thecytokine profile. As shown in Fig. 4D, SPCs isolated from Hx2/2

mice at days 7 and 14 showed a higher percentage of MOG35–55–specific Th17 cells compared with wild-type cells. By contrast,SPCs from wild-type mice displayed a higher percentage ofMOG35–55–specific Th1 cells compared with cells from Hx 2/2

mice. Dosage of IL-17 in the supernatants of cells cultured for48 h with MOG35–55 peptide confirmed the difference betweenSPLs isolated from Hx2/2 and wild-type mice (Fig. 4D). Thepercentage of MOG35–55–specific cells producing both IL-17 andIFN-g was very low but significantly higher in Hx2/2 mice than inwild-type controls (Supplemental Fig. 2). Analysis of T cell pro-liferation in the presence of different concentrations of MOG35–55

peptide demonstrated that cells isolated from the spleen of wild-type and Hx2/2 mice had the same proliferative potential (Fig.4E). These results indicate that Hx deficiency results in an un-balance between Th17 and Th1 cell populations, favoring Th17differentiation.

Loss of Hx affects both the differentiative potential of CD4+

T cells and the expression of Th17-polarizing cytokines

Data reported in the previous section demonstrated an expansionof Th17 cell population in Hx2/2 mice subjected to EAE. Thismight be because of a higher predisposition of naive CD4+ T cellsof Hx2/2 mice to differentiate toward Th17 lineage and/or toaltered extrinsic factors promoting Th17 differentiation. To dis-criminate between these two possibilities, we evaluated the Th17differentiative potential of CD4+ naive and memory T cells iso-lated from Hx2/2 mice. As shown in Fig. 5A, naive CD4 T cellsisolated from Hx2/2 mice and activated in vitro with anti-CD3/CD28 mAbs were able to produce an enhanced amount of IL-17compared with that produced by naive CD4+ cells from wild-typemice. This amount of IL-17 was further enhanced in the presenceof Th17 polarizing cytokines alone or admixed with anti–IFN-gneutralizing mAb, and it was higher compared with that produced

FIGURE 1. Hx is upregulated during EAE. (A) qRT-PCR analysis of Hx

mRNA level in the liver of wild-type mice before and at day 7 post-

immunization. Data represent mean 6 SEM. **p , 0.01, n = 5 mice per

group. (B) Representative Western blot of Hx level in the serum of two

wild-type mice during EAE. IgG level was used as loading control. (C)

Serum levels of Hx were quantified by densitometry. The results were

representative of three independent experiments. Data represent the mean

6 SEM. **p , 0.01. AU, Arbitrary units; RQ, relative quantity.

The Journal of Immunology 5453

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

by naive CD4+ from wild-type mice. By contrast, memory CD4T cells from wild-type and Hx2/2 mice produced similar amountsof IL-17 when stimulated with anti-CD3/CD28 mAb alone, whereasmemory CD4 T cells from Hx2/2 mice displayed a striking in-crease of IL-17 production compared with memory CD4 T cellsfrom wild-type mice when activated in the “Th17 polarizingcondition,” namely, IL-23 and anti–IFN-g neutralizing mAb (Fig.5B).Then we evaluated the expression of the cytokines that drive

Th17 differentiation (IL-6 and IL-23) or Th1 differentiation (IL-12)in macrophages (F4/80+ cells) and dendritic cells (CD11c+ cells)of Hx2/2 and wild-type mice with EAE. IL-6 and IL-23 werestrongly induced during disease in macrophages and dendriticcells isolated from both wild-type and Hx2/2 mice, but this in-duction was significant higher in cells from Hx2/2 animals (Fig.

5C, 5D). IL-12 expression did not change during disease and wascomparable between Hx2/2 and wild-type mice (Fig. 5C, 5D).Consistent with these data, the amount of IL-6 in the sera washigher in Hx2/2 mice than in wild-type controls (Fig. 5E). Finally,in the brain of Hx2/2 mice, IL-6 and, to a lesser extent, IL-23

FIGURE 2. Loss of Hx exacerbates EAE disease in vivo. (A) Clinical

scores of EAE in Hx2/2 and wild-type mice, expressed as cumulative

disease score over 28 d. Data represent mean 6 SD of a representative

experiment. ***p , 0.001, n = 4 for each group. (B) Day of the onset

(score = 1) and (C) clinical score at day 28 postimmunization. Data rep-

resent mean 6 SEM of 10 mice per each genotype pooled from 3 inde-

pendent experiments. **p , 0.01, ***p , 0.001. (D) Representative

sections of the spinal cord of a wild-type mouse and an Hx2/2 mouse at

day 28 postimmunization stained with H&E showing the massive inflam-

matory infiltration in Hx2/2 animal. Inflammatory index reported on the

right was calculated as reported in Materials and Methods. Data repre-

sent mean 6 SEM. *p , 0.05, n = 4 mice per each genotype. (E)

Representative sections of the spinal cord of a wild-type mouse and an

Hx2/2 mouse at day 28 postimmunization processed by immunohis-

tochemistry with an anti-MBP Ab. (D and E) Original magnifications

34 (left), 320 (right).

FIGURE 3. EAE in Hx2/2 mice is associated with an increased infil-

tration of Th17 cells. (A) Representative sections of the spinal cord of a

wild-type mouse and an Hx2/2 mouse at day 28 postimmunization pro-

cessed by IFL with an anti–CD4-FITC Ab. The number of CD4+ cells,

counted as reported in Materials and Methods (right panel). Original

magnification 3200. Data represent mean 6 SEM. *p , 0.05, n = 4 mice

per each genotype. (B) Th17 and Th1 cells infiltrating the spinal cord and

the brain of wild-type and Hx2/2 mice at day 14 of the disease, detected by

flow cytometry. Data represent mean 6 SEM. *p , 0.05, ***p , 0.001,

n = 4 mice per genotype. (C) Th17 and Th1 cells infiltrating the spinal

cord and the brain of wild-type and Hx2/2 mice at day 28 of the disease,

detected by flow cytometry. Data represent mean 6 SEM. *p , 0.05,

**p , 0.01, n = 6 mice per genotype. (D) Representative dot plots are

shown: numbers in the quadrants indicate the percentage of cytokine-

secreting cells gated on CD4+ cells. The mean clinical score of mice used

in the analyses is reported in (B) and (C).

5454 HEMOPEXIN MODULATES CD4 T CELL DIFFERENTIATION

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

mRNAwere significantly higher than in that of wild-type animalsat day 28 postimmunization (Fig. 5F). These results indicate thatboth intrinsic and extrinsic factors may contribute to the expansionof Th17 cells in Hx2/2 mice.

Injection of purified Hx in Hx2/2 mice rescues the diseaseseverity and the expansion of Th17 cells

To further assess the potential protective role of Hx in the de-velopment of EAE, we injected a single dose of purified human Hxin Hx2/2 mice at day 0 of EAE and evaluated disease progressionand Th17 expansion. As shown in Fig. 6A, exogenous Hx wasdetectable by Western blotting in the serum of Hx2/2 mice onlyuntil day 2 of EAE.A single dose of Hx was able to rescue the EAE severity in Hx2/2

mice (Fig. 6B). The amount of circulating Th17 cells (Fig. 6C)

during disease progression as well as the amount of infiltratedTh17 cells in the spinal cord and in the brain (Fig. 6D) at day 28 ofEAE were significantly reduced in Hx-injected Hx2/2 mice andwere comparable with those of wild-type animals. As expected,the amount of circulating (Fig. 6C) and infiltrated (Fig. 6D) Th1cells were not affected by Hx injection.As shown in Fig. 6E, the expression of IL-6 and IL-23 in the CNS

at day 28 of EAE was reduced in Hx-injected Hx2/2mice comparedwith untreated Hx2/2 animals and was similar to that of wild-types.These data indicate that a single dose of Hx in Hx2/2 mice was

able to limit Th17 cell expansion, thus controlling EAE severity.

DiscussionIn this study, we demonstrated that Hx affects the severity of EAEby limiting the expansion of the Th17 cell population. This

FIGURE 4. Loss of Hx results in an expansion of

Th17 cell population. Ex vivo analysis of Th17 and

Th1 cells in LN cells (A) and SPCs (B) of wild-type

and Hx2/2 mice at days 7 and 14 postimmunization

and in PMBCs during the disease (C). Cells from

wild-type (n = 5) and Hx2/2 (n = 5) mice were

stimulated in vitro with PMA and Ionomycin for

18 h, and IL-17 and IFN-g quantified by flow cy-

tometry analysis. Data represent mean 6 SEM. *p ,0.05, ***p , 0.001. (D) Flow cytometry analysis of

Th17 and Th1 cells and ELISA assay for IL-17 and

INF-g after in vitro stimulation with MOG35–55 peptide

of SPCs isolated from wild-type and Hx2/2 mice

before and at days 7 and 14 postimmunization. Data

represent mean 6 SEM. *p , 0.05, **p , 0.01. (E)

[3H]TdR uptake of SPCs isolated from wild-type

and Hx2/2 mice, and cultured in the absence or

presence of different concentrations of MOG35–55

peptide. Results are expressed as the Dcpm.

The Journal of Immunology 5455

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

conclusion comes from the observation that Hx2/2mice developeda more severe and earlier EAE than wild-type animals, associatedwith a higher amount of circulating and infiltrating Th17 cellsautoreactive against myelin. A direct effect of Hx in controllingEAE development is further demonstrated by the “rescue experi-ment” in which Hx administration in Hx2/2 mice before EAEinduction has been shown to limit Th17 cell expansion and at-tenuate disease severity.Both Th1 and Th17 cells have been defined as effector T cell

subsets involved in the pathogenesis of the EAE, but Th17 cellsare considered to be more pathogenic than Th1 cells (24). Auto-antigen-specific Th17 cells have been shown to break and trans-migrate across the blood–brain barrier, and in the CNS induce

inflammation by neutrophil recruitment (20) and killing of neu-rons (25). Thus, the more severe EAE developed by Hx2/2 micemay be ascribed to the higher amount of Th17 cells infiltrating theCNS and circulating in the periphery of Hx2/2mice comparedwith wild-type controls.The Th17-driven EAE developed by Hx2/2 mice is due to both

intrinsic and extrinsic factors. We showed that CD4+ T cells iso-lated from Hx2/2 mice polarize more efficiently toward Th17lineage than cells isolated from wild-types controls. The absenceof Hx seems to predispose naive CD4 cells to differentiate intoTh17 lineage as CD4+ naive cells from Hx2/2mice produce anenhanced amount of IL-17 compared with the same cells fromwild-type mice. By contrast, activated memory CD4 cells produce

FIGURE 5. Loss of Hx results in enhanced dif-

ferentiative potential toward Th17 lineage, as well

as in an enhanced production of Th17-polarizing

cytokines. (A) IL-17 production in cultures of CD4

naive T cells isolated from untreated wild-type and

Hx2/2 mice, and stimulated with polarizing cyto-

kines (IL-6, TGF-b, and IL-23) alone or in com-

bination with anti–IFN-g mAb. Data represent

mean 6 SEM. *p , 0.05, **p , 0.01. (B) IL-17

production in cultures of CD4 memory T cells

isolated from untreated wild-type and Hx2/2 mice,

and stimulated with IL-23 plus anti–IFN-g mAb.

Data represent mean 6 SEM. ***p , 0.001. (C

and D) qRT-PCR analysis of IL-6, IL-23, and IL-12

mRNA level in macrophages (C) and dendritic cells

(D) isolated from the spleen of wild-type and Hx2/2

mice before and at day 7 postimmunization. Results

were representative of three independent experi-

ments. Data represent mean 6 SEM. *p , 0.05,

**p , 0.01, n = 4 mice/genotype for each experi-

mental point. (E). IL-6 serum levels in wild-type

and Hx2/2 mice before and at days 7 and 14 post-

immunization. IL-6 (pg/ml) was measured by ELISA

assay. Data represent the mean6 SEM. (F) qRT-PCR

analysis of IL-6 and IL-23 mRNA level in the brain

of wild-type and Hx2/2 mice at day 28 postimmuni-

zation. Data represent mean6 SEM. *p, 0.05, n = 4

mice per genotype.

5456 HEMOPEXIN MODULATES CD4 T CELL DIFFERENTIATION

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

an equal amount of IL-17 in Hx2/2 and wild-type mice, and this isincreased in memory CD4 T cells from Hx2/2 mice only whenactivated in the presence of IL-23 and in the absence of IFN-g.These data suggest that the absence of Hx favors the differentia-tion of naive CD4+ cells toward Th17 lineage and enhances thestabilization and expansion of memory Th17 T cells by IL-23,particularly when IFN-g is neutralized.The enhanced potential of CD4+ naive T cells to differentiate

toward Th17 lineage might be because of the reduced sensitivityto IFN-g of lymphocytes of Hx2/2 mice as we previously dem-

onstrated that, in the absence of Hx, T cells displayed a reducedIFN-g–STAT1–dependent phosphorylation (26). IFN-g signalingplays a key role in Th17 differentiation and function (27). It wasalready demonstrated that the neutralization of IFN-g effect bya specific mAb to IFN-g or to the IFN-gR1 chain was required forthe massive development of activated T lymphocytes cultured inthe presence of IL-23 into Th17 cells (28). The Hx-mediatedcontrol of CD4+ T cell responsiveness to IFN-g might be relatedto the Hx ability to modulate iron availability within cells (1), thusaffecting the expression of Transferrin receptor that has already

FIGURE 6. Injection of purified Hx in Hx2/2

mice rescues the disease severity and the expansion

of Th17 cells. (A) Representative Western blot of

Hx in the serum of three Hx2/2 mice at different

times of EAE. Ponceau red staining was used as

loading control. (B) Clinical scores of EAE in wild-

type, Hx2/2, and Hx-injected Hx2/2 mice, ex-

pressed as cumulative disease score over 28 d. Data

represent mean 6 SD of a representative experi-

ment. *p , 0.05, ***p , 0.001, #p , 0.05, ###p ,0.001; n = 4 for each group. (C) Ex vivo analysis of

Th17 and Th1 cells in PMBCs of wild-type, Hx2/2,

and Hx-injected Hx2/2 mice at days 0, 7, 14, 21, and

28 postimmunization. Data represent mean 6 SEM.

*p , 0.05, **p , 0.01, ***p , 0.001, ##p , 0.01.

(D) Th17 and Th1 cells infiltrating the spinal cord

and the brain of wild-type, Hx2/2, and Hx-injected

Hx2/2 mice at day 28 of the disease, detected by

flow cytometry. Data represent mean6 SEM. *p,0.05, **p , 0.01, ***p , 0.001, n = 4 mice per

genotype. (E) qRT-PCR analysis of IL-6 and IL-23

mRNA level in the brain of wild-type, Hx2/2, and

Hx-injected Hx2/2 mice at day 28 postimmuni-

zation. Data represent mean 6 SEM. *p , 0.05,

**p , 0.01, n = 4 mice per genotype.

The Journal of Immunology 5457

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

been shown to regulate INFg-R2 expression at the plasma mem-brane (29, 30). Because Transferrin receptor expression was lowerin T lymphocytes from Hx2/2 mice both in normal condition (26)and after EAE induction (data not shown), developing T cells fromMOG35–55–immunized Hx2/2 mice showed refractoriness to IFN-gand were more prone to polarize into Th17 cells.Other than the intrinsic higher capability of CD4 T cells from

Hx2/2 mice to differentiate toward Th17 lineage, our data indicatethat also extrinsic factors may contribute to Th17 cell expansion inHx2/2 mice. We showed that macrophages and dendritic cellsisolated from Hx2/2 mice with EAE produced higher amounts ofIL-6 and IL-23 than cells isolated from wild-type animals. IL-6 isa key stimulator for Th17 differentiation and, accordingly, IL-6–deficient mice are protected from EAE (31). IL-23 produced byAPCs stabilizes the Th17 phenotype by promoting secretion ofIL-17A, IL-17F, and IL-22, and by helping Th17 cells to acquireeffector function (32). Of note, mice deficient for IL-23 arecompletely resistant to EAE induction (33). The increase of IL-6and IL-23 in Hx2/2 mice could be related to the Hx ability tomodulate cytokine production (12). It has already been shown thatHx was able to downregulate the secretion of IL-6 and TNF-afrom macrophages in response to LPS in a manner that is distinctfrom its function as a heme scavenger (12). This might occur bylimiting TLR4- and TLR2-induced cytokine production. TLRsignals, induced by pathogen-associated molecular patterns con-tained in the CFA, are necessary for the induction of EAE andinfluence the CD4 Th cell–dependent inflammation that causesEAE (34). Absence of TLR4 exacerbates EAE and is associatedwith increased expression of IL-6 and IL-23 in splenic dendriticcells and increased frequency of Th17 cells (35). By contrast,deficiency of TLR2 in CD4+ T cells substantially impaired Th17response in vivo and the pathogenesis of EAE (36). In this article,we show that Hx is able to modulate not only the secretion of IL-6,but also that of IL-23, an effect similar to that obtained by TLR4blockade (35), confirming the emerging hypothesis that Hx maymodulate the response to TLR4 and TLR2 ligands, thus playinga role in acute and chronic inflammatory diseases (12, 13). Theelucidation of the mechanisms by which Hx could modulate TLRactivity needs more studies.Hx was reported to be produced in the CNS by ependymal cells,

neurons, and glial cells (1, 2, 4, 5). In the CNS, Hx controls ironaccumulation in oligodendrocytes and affects their differentiation(4, 37). Accordingly, the expression of the MBP along with thedensity of myelinated fibers in the basal ganglia and in the motorand somatosensory cortex of Hx2/2 mice was strongly reducedcompared with wild-type controls (37). Thus, it is possible that theexacerbated demyelination we observed in Hx2/2 mice with EAEwas due not only to massive Th17 cell infiltration, but also to theimpairment of Hx-deficient oligodendrocytes to respond to ex-ogenous insults. This conclusion is consistent with data showingthat iron deposition in the brain contributes to MS pathogenesis(38). Thus, Hx would control, on one hand, systemic signalsdriving Th17 differentiation and, on the other hand, local signalsin CNS affecting oligodendrocyte function.Hx has been reported to be induced in pediatric MS (14). Our

data on Hx2/2 mice with EAE may give an explanation to suchinduction. Hx upregulation is expected to contribute to the in-flammatory response thus controlling the severity of the disease.Moreover, by controlling T cell responsiveness to IFN-g, Hx in-duction might limit Th17 cell differentiation. Results obtained byinjecting purified Hx in Hx2/2 mice strongly support these con-clusions. Finally, the previously reported observation that Hx ex-pression in CNS controls oligodendrocyte function suggests thatall of these factors may contribute to disease establishment and

progression. Thus, Hx administration may be beneficial in clinicalsettings, such as autoimmune and inflammatory diseases, char-acterized by dysregulation of T cell differentiation.

DisclosuresThe authors have no financial conflicts of interest.

References1. Tolosano, E., S. Fagoonee, N. Morello, F. Vinchi, and V. Fiorito. 2010. Heme scav-

enging and the other facets of hemopexin. Antioxid. Redox Signal. 12: 305–320.2. Morris, C. M., J. M. Candy, J. A. Edwardson, C. A. Bloxham, and A. Smith.

1993. Evidence for the localization of haemopexin immunoreactivity in neuronesin the human brain. Neurosci. Lett. 149: 141–144.

3. Hunt, R. C., D. M. Hunt, N. Gaur, and A. Smith. 1996. Hemopexin in the humanretina: protection of the retina against heme-mediated toxicity. J. Cell. Physiol.168: 71–80.

4. Morello, N., E. Tonoli, F. Logrand, V. Fiorito, S. Fagoonee, E. Turco, L. Silengo,A. Vercelli, F. Altruda, and E. Tolosano. 2009. Haemopexin affects iron distri-bution and ferritin expression in mouse brain. J. Cell. Mol. Med. 13: 4192–4204.

5. Li, R. C., S. Saleem, G. Zhen, W. Cao, H. Zhuang, J. Lee, A. Smith, F. Altruda,E. Tolosano, and S. Dore. 2009. Heme-hemopexin complex attenuates neuronalcell death and stroke damage. J. Cereb. Blood Flow Metab. 29: 953–964.

6. Madore, N., L. Camborieux, N. Bertrand, and J. P. Swerts. 1999. Regulationof hemopexin synthesis in degenerating and regenerating rat sciatic nerve.J. Neurochem. 72: 708–715.

7. Camborieux, L., V. Julia, B. Pipy, and J. P. Swerts. 2000. Respective roles ofinflammation and axonal breakdown in the regulation of peripheral nervehemopexin: an analysis in rats and in C57BL/Wlds mice. J. Neuroimmunol. 107:29–41.

8. Swerts, J. P., C. Soula, Y. Sagot, M. J. Guinaudy, J. C. Guillemot, P. Ferrara,A. M. Duprat, and P. Cochard. 1992. Hemopexin is synthesized in peripheralnerves but not in central nervous system and accumulates after axotomy. J. Biol.Chem. 267: 10596–10600.

9. Davidsson, P., S. Folkesson, M. Christiansson, M. Lindbjer, B. Dellheden,K. Blennow, and A. Westman-Brinkmalm. 2002. Identification of proteins inhuman cerebrospinal fluid using liquid-phase isoelectric focusing as a pre-fractionation step followed by two-dimensional gel electrophoresis and matrix-assisted laser desorption/ionisation mass spectrometry. Rapid Commun. MassSpectrom. 16: 2083–2088.

10. Castano, E. M., A. E. Roher, C. L. Esh, T. A. Kokjohn, and T. Beach. 2006.Comparative proteomics of cerebrospinal fluid in neuropathologically-confirmedAlzheimer’s disease and non-demented elderly subjects. Neurol. Res. 28: 155–163.

11. Vinchi, F., S. Gastaldi, L. Silengo, F. Altruda, and E. Tolosano. 2008. Hemopexinprevents endothelial damage and liver congestion in a mouse model of hemeoverload. Am. J. Pathol. 173: 289–299.

12. Liang, X., T. Lin, G. Sun, L. Beasley-Topliffe, J. M. Cavaillon, and H. S. Warren.2009. Hemopexin down-regulates LPS-induced proinflammatory cytokines frommacrophages. J. Leukoc. Biol. 86: 229–235.

13. Lin, T., F. Sammy, H. Yang, S. Thundivalappil, J. Hellman, K. J. Tracey, andH. S. Warren. 2012. Identification of hemopexin as an anti-inflammatory factorthat inhibits synergy of hemoglobin with HMGB1 in sterile and infectious in-flammation. J. Immunol. 189: 2017–2022.

14. Rithidech, K. N., L. Honikel, M. Milazzo, D. Madigan, R. Troxell, andL. B. Krupp. 2009. Protein expression profiles in pediatric multiple sclerosis:potential biomarkers. Mult. Scler. 15: 455–464.

15. Noseworthy, J. H., C. Lucchinetti, M. Rodriguez, and B. G. Weinshenker. 2000.Multiple sclerosis. N. Engl. J. Med. 343: 938–952.

16. Frohman, E. M., M. K. Racke, and C. S. Raine. 2006. Multiple sclerosis—theplaque and its pathogenesis. N. Engl. J. Med. 354: 942–955.

17. Stys, P. K., G. W. Zamponi, J. van Minnen, and J. J. Geurts. 2012. Will the realmultiple sclerosis please stand up? Nat. Rev. Neurosci. 13: 507–514.

18. Shrikant, P., and E. N. Benveniste. 1996. The central nervous system as animmunocompetent organ: role of glial cells in antigen presentation. J. Immunol.157: 1819–1822.

19. Hedegaard, C. J., M. Krakauer, K. Bendtzen, H. Lund, F. Sellebjerg, andC. H. Nielsen. 2008. T helper cell type 1 (Th1), Th2 and Th17 responses tomyelin basic protein and disease activity in multiple sclerosis. Immunology 125:161–169.

20. Kroenke, M. A., T. J. Carlson, A. V. Andjelkovic, and B. M. Segal. 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.

21. Stromnes, I. M., L. M. Cerretti, D. Liggitt, R. A. Harris, and J. M. Goverman.2008. Differential regulation of central nervous system autoimmunity by T(H)1and T(H)17 cells. Nat. Med. 14: 337–342.

22. Durelli, L., L. Conti, M. Clerico, D. Boselli, G. Contessa, P. Ripellino,B. Ferrero, P. Eid, and F. Novelli. 2009. T-helper 17 cells expand in multiplesclerosis and are inhibited by interferon-beta. Ann. Neurol. 65: 499–509.

23. Tolosano, E., E. Hirsch, E. Patrucco, C. Camaschella, R. Navone, L. Silengo, andF. Altruda. 1999. Defective recovery and severe renal damage after acute he-molysis in hemopexin-deficient mice. Blood 94: 3906–3914.

24. Becher, B., and B. M. Segal. 2011. T(H)17 cytokines in autoimmune neuro-inflammation. Curr. Opin. Immunol. 23: 707–712.

5458 HEMOPEXIN MODULATES CD4 T CELL DIFFERENTIATION

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from

25. Kebir, H., K. Kreymborg, I. Ifergan, A. Dodelet-Devillers, R. Cayrol,M. Bernard, F. Giuliani, N. Arbour, B. Becher, and A. Prat. 2007. HumanTH17 lymphocytes promote blood-brain barrier disruption and central nervoussystem inflammation. Nat. Med. 13: 1173–1175.

26. Fagoonee, S., C. Caorsi, M. Giovarelli, M. Stoltenberg, L. Silengo, F. Altruda,G. Camussi, E. Tolosano, and B. Bussolati. 2008. Lack of plasma proteinhemopexin dampens mercury-induced autoimmune response in mice. J. Immu-nol. 181: 1937–1947.

27. Harrington, L. E., R. D. Hatton, P. R. Mangan, H. Turner, T. L. Murphy,K. M. Murphy, and C. T. Weaver. 2005. Interleukin 17-producing CD4+ effectorT cells develop via a lineage distinct from the T helper type 1 and 2 lineages.Nat. Immunol. 6: 1123–1132.

28. Conti, L., R. De Palma, S. Rolla, D. Boselli, G. Rodolico, S. Kaur,O. Silvennoinen, E. Niccolai, A. Amedei, F. Ivaldi, et al. 2012. Th17 cells inmultiple sclerosis express higher levels of JAK2, which increases their surfaceexpression of IFN-gR2. J. Immunol. 188: 1011–1018.

29. Regis, G., L. Conti, D. Boselli, and F. Novelli. 2006. IFNgammaR2 traffickingtunes IFNgamma-STAT1 signaling in T lymphocytes. Trends Immunol. 27:96–101.

30. Regis, G., M. Bosticardo, L. Conti, S. De Angelis, D. Boselli, B. Tomaino,P. Bernabei, M. Giovarelli, and F. Novelli. 2005. Iron regulates T-lymphocytesensitivity to the IFN-gamma/STAT1 signaling pathway in vitro and in vivo.Blood 105: 3214–3221.

31. Samoilova, E. B., J. L. Horton, B. Hilliard, T. S. Liu, and Y. Chen. 1998. IL-6-deficient mice are resistant to experimental autoimmune encephalomyelitis: roles

of IL-6 in the activation and differentiation of autoreactive T cells. J. Immunol.161: 6480–6486.

32. Takatori, H., Y. Kanno, W. T. Watford, C. M. Tato, G. Weiss, I. I. Ivanov,D. R. Littman, and J. J. O’Shea. 2009. Lymphoid tissue inducer-like cells are aninnate source of IL-17 and IL-22. J. Exp. Med. 206: 35–41.

33. Cua, D. J., J. Sherlock, Y. Chen, C. A. Murphy, B. Joyce, B. Seymour, L. Lucian,W. To, S. Kwan, T. Churakova, et al. 2003. Interleukin-23 rather thaninterleukin-12 is the critical cytokine for autoimmune inflammation of the brain.Nature 421: 744–748.

34. Marta, M., U. C. Meier, and A. Lobell. 2009. Regulation of autoimmune en-cephalomyelitis by toll-like receptors. Autoimmun. Rev. 8: 506–509.

35. Marta, M., A. Andersson, M. Isaksson, O. Kampe, and A. Lobell. 2008. Un-expected regulatory roles of TLR4 and TLR9 in experimental autoimmune en-cephalomyelitis. Eur. J. Immunol. 38: 565–575.

36. Reynolds, J. M., B. P. Pappu, J. Peng, G. J. Martinez, Y. Zhang, Y. Chung, L. Ma,X. O. Yang, R. I. Nurieva, Q. Tian, and C. Dong. 2010. Toll-like receptor 2signaling in CD4(+) T lymphocytes promotes T helper 17 responses and reg-ulates the pathogenesis of autoimmune disease. Immunity 32: 692–702.

37. Morello, N., F. T. Bianchi, P. Marmiroli, E. Tonoli, V. Rodriguez Menendez,L. Silengo, G. Cavaletti, A. Vercelli, F. Altruda, and E. Tolosano. 2011. A rolefor hemopexin in oligodendrocyte differentiation and myelin formation. PLoSONE 6: e20173.

38. van Rensburg, S. J., M. J. Kotze, and R. van Toorn. 2012. The conundrum of ironin multiple sclerosis—time for an individualised approach.Metab. Brain Dis. 27:239–253.

The Journal of Immunology 5459

at Bibl U

nito on February 23, 2015http://w

ww

.jimm

unol.org/D

ownloaded from