Promoting tolerance to proteolipid protein-inducedexperimental autoimmune encephalomyelitisthrough targeting dendritic cellsJoel N. H. Sterna,1,2, Derin B. Keskina,1, Zenichiro Katoa, Hanspeter Waldnerb,3, Sonja Schallenbergc, Ana Andersonb,Harald von Boehmerd,e, Karsten Kretschmerc,d,1,2, and Jack L. Stromingera,2

aDepartment of Molecular and Cellular Biology, Harvard University, Cambridge, MA 02138; bCenter for Neurologic Disease, Brigham and Women’s Hospital,Harvard Medical School, Boston, MA 02115; cCenter for Regenerative Therapies Dresden, Dresden Technical University, 01307 Dresden, Germany;dDepartment of Cancer Immunology and AIDS, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA 02115; and eDepartment of Pathology,Harvard Medical School, Boston, MA 02115

Contributed by Jack L. Strominger, August 9, 2010 (sent for review May 20, 2010)

In T cell-mediated autoimmune diseases, self-reactive T cells withknown antigen specificity appear to be particularly promisingtargets for antigen-specific induction of tolerance without compro-mising desired protective host immune responses. Several lines ofevidence suggest that delivery of antigens to antigen-presentingdendritic cells (DCs) in the steady state (i.e., to immature DCs) mayrepresent a suitable approach to induce antigen-specific T-celltolerance peripherally. Here, we report that anti-DEC205–mediateddelivery of the self-peptide proteolipid protein (PLP)139–151 to DCsameliorated clinical symptoms in the PLP-induced SJL model of ex-perimental autoimmune encephalomyelitis. Splenocytes from trea-ted mice were anergized to PLP139–151, and IL-17 secretion wasmarkedly reduced. Moreover, we show directly, using transgenicCD4+ Vβ6+ TCR T cells specific for PLP139–151, that, under the con-ditions of the present experiments, these cells also became anergic.In addition, evidence for a CD4+ T cell-mediated suppressor mech-anism was obtained.

DEC205 | multiple sclerosis | anergy | monophosphoryl lipid A | T cells

Multiple sclerosis is a T cell-mediated autoimmune diseasecharacterized by immune cell infiltration, inflammatory

demyelination of neuronal axons, and axonal loss in the humancentral nervous system (1, 2). Studies of multiple sclerosis arefacilitated by the animal model experimental autoimmune en-cephalomyelitis (EAE) that recapitulates many aspects of thehuman disease (3). Active induction of EAE is accomplished bystimulation of T cell-mediated immunity to myelin, the insulatingphospholipid layer surrounding the neuronal axons, through im-munization with myelin proteins or synthetic peptide antigensderived from myelin and then emulsified in adjuvant (4). Thistreatment leads to activation of autoreactive myelin-specificCD4+ T cells that circulate in the periphery of naïve animals.Activated autoreactive T cells will cross the blood–brain barrier(5). Within the central nervous system, local and infiltrating an-tigen-presenting cells, such as dendritic cells (DCs) derived frommicroglia, present MHC class II molecule-associated myelinpeptides to infiltrating T cells in the context of costimulation.Myelin-specific CD4+ T cells are reactivated, initiating a cascadeof neuroinflammatory responses that ultimately leads to de-myelination in the central nervous system and neurodegeneration.EAE can also be passively induced by adoptive transfer of pre-activated myelin-specific T cells (6).Although T helper 1 (Th1) cells secreting IFN-γ were consid-

ered to be the primary mediators of EAE, T helper 17 (Th17) cellsrecently were shown to exhibit greater pathogenicity, suggestingthat they play a more decisive role in mediating severe tissuedamage (7, 8). However, both Th1 and Th17 cells, generated withkinetic differences and/or involved at different stages, may beinvolved in development of EAE (9). In fact, the relative contri-bution of both Th subsets was recently suggested to affect the

anatomical location of lesion distribution between brain andspinal cord parenchyma (10).Self-reactive T cells with known antigen specificity, which can be

found in T cell-mediated autoimmune diseases such as multiplesclerosis, appear particularly promising targets for antigen-specifictolerance induction without compromising host immunity to in-fectious insults. Various protocols have been used to interfere withunwanted immunity using peptide-induced tolerance (11), in-cluding the administration of antigens over extended periods oftime via osmotic minipumps (12, 13). In addition, peptide antigenscan also be directly delivered to antigen-presenting cells via tar-geting approaches. In particular, antigens delivered to differentsubsets of DCs after fusion with antibodies to the endocyticreceptors DEC205 (αDEC205) or 33D1 are efficiently processedand presented by MHC class I and class II molecules (14). Thisroute of antigen delivery to murine (15) or human (16) DCs isseveral orders of magnitude more efficient than free peptides andin conjunction with maturation stimuli represents an effectivemethod for inducing strong T-cell responses, i.e., vaccination. Bycontrast, targeting antigen to immatureDCs in the steady state hasbeen described as promoting immunological tolerance but throughdifferent mechanisms in different studies (15, 17–20). It may leadto deletion of antigen-specific T cells with residual cells becomingimmunologically unresponsive, a mechanism that in one study in-creased CD5 expression on activated T cells (17). In addition,delivering minute amounts of peptides via αDEC205 fusion pro-teins to steady-state immature DCs can lead to the de novo gen-eration of antigen-specific Foxp3+ Treg in vivo (18, 21).Previous studies indicated that αDEC205-mediated targeting of

an encephalogenic peptide of the myelin oligodendrocyte glyco-protein (MOG), a minor myelin component, to DCs in vivo pre-vents EAE induction by subsequent injection of the same peptidein complete Freund’s adjuvant (CFA) in C57BL/6 mice (17). Inthis model, pretreatment with large doses of the free peptide inthe absence of adjuvants also leads to protection from subsequentchallenge. Here, we report experiments with αDEC205-mediatedtargeting of the autoantigen of the proteolipid protein peptide(PLP139–151) (derived from a major myelin constituent) in theEAEmodel in SJLmice, which is muchmore prone to disease and

Author contributions: H.W., H.v.B., K.K., and J.L.S. designed research; J.N.H.S., D.B.K., Z.K.,and S.S. performed research; H.W. and A.A. contributed new reagents/analytic tools; andJ.N.H.S. and J.L.S. wrote the paper.

The authors declare no conflict of interest.1J.N.H.S., D.B.K., and K.K. contributed equally to this work.2To whom correspondence may be addressed. E-mail: jstern@fas.harvard.edu, karsten.kretschmer@crt-dresden.de, or jlstrom@fas.harvard.edu.

3Present address: Department of Microbiology and Immunology, Pennsylvania State Uni-versity College of Medicine, Hershey, PA 17033.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1010263107/-/DCSupplemental.

17280–17285 | PNAS | October 5, 2010 | vol. 107 | no. 40 www.pnas.org/cgi/doi/10.1073/pnas.1010263107

Dow

nloa

ded

by g

uest

on

Mar

ch 3

0, 2

020

in which free peptide administration does not lead to protection.Thismodel represents a second example in which targeting ofDCsin the steady state with nanogram amounts of a peptide thatgenerates autoimmunity efficiently ameliorates disease by pro-moting tolerance. In the present case, the amelioration of diseaseresults both from induction of T-cell anergy and by generation ofsuppressor T cells.

ResultsDendritic Cell Targeting of Proteolipid Protein-Derived Peptide viaαDEC205 Fusion Antibodies. To target the encephalogenic antigento DCs, recombinant proteins consisting of amino acids 139–151of proteolipid protein (PLP139–151) fused either to the C ter-minus of the Ig heavy chain of cloned αDEC205 (αDEC205/PLP)or to the GL117 isotype control antibody (GL117/PLP) wereproduced. To confirm that the antigenic peptide delivered by theαDEC205 fusion antibody was properly processed and presented,purified splenic CD11c+ DCs from SJL mice were incubatedfor 3 h with various concentrations of either αDEC205/PLP orGL117/PLP control antibodies. After unbound antibodies wereremoved by extensive washing, DCs were cocultured with antigen-specific CD4+Vβ6+ T cells fromPLP139–151-specific Vβ6+ TCRtransgenic mice (22, 23). 3H-thymidine incorporation at day 4 ofthe culture demonstrated that DCs preincubated with αDEC205/PLP fusion antibody induced vigorous proliferation of thesetransgenic T cells compared with GL117/PLP isotype controlantibody or in the absence of a specific antigen (Fig. S1A).In addition, a PLP139–151-specific T-cell line was established

by immunizing SJL mice with PLP139–151 and restimulatingsplenocytes from the immunizedmice with the same peptide threetimes at 2-wk intervals in vitro. The CD4+ T-cell line obtainedexhibited an activated surface marker phenotype (CD25+,CD69+, CD45+, CD30+, GITR+, CTLA4+, CD71low, orCD62Llow) and secreted high amounts of IL-17 (10,300 pg/mL)along with IL-6 (1,300 pg/mL), IL-5 (772 pg/mL), GM-CSF (2,960pg/mL), and TNF-α (278 pg/mL). Coculture of the PLP139–151T-cell line with CD11c+DCs preincubated with 1 μg of αDEC205/PLP fusion antibody substantially enhanced T-cell proliferation ina dose-dependent manner (Fig. S1B). In contrast, preincubationof DCs with either GL117/PLP isotype control antibody orαDEC205 antibody fused to an irrelevant antigen [peptide 107–119 of hemagglutinin (HA), αDEC205/HA] induced little pro-liferation. Proliferation was accompanied by an ∼10-fold in-crease in IFN-γ secretion only after treatment with αDEC205/PLP (Fig. S1C).

Immunization or Preimmunization with αDEC205/PLP Ameliorates EAEInduced by Either Adoptive Transfer of a PLP139–151-Specific T-CellLine or by Immunization with PLP139–151. The PLP139–151-specificsplenic T-cell line discussed above was adoptively transferredinto naïve SJL mice to passively induce EAE, and then the micewere injected with either 1 μg of αDEC205/PLP or control GL117/PLP fusion antibodies, equivalent to ∼20 ng of PLP139–151. Allrecipients were injected with pertussis toxin (PT) the followingday. As expected, mice that received PLP139–151-specific T cellsand the GL117/PLP isotype control antibody rapidly developedsevere EAE with a maximal mean score of 4 on day 28 of thisexperiment (Fig. 1). In contrast, mice that received PLP139–151-specific T cells followed by αDEC205/PLP exhibited a sub-stantially delayed onset of disease with a low maximal mean scoreof 1 on day 28. Thus, αDEC205-mediated targeting of nanogramamounts of PLP139–151 efficiently interfered with the passiveinduction of EAE by adoptive transfer of highly encephalitogenicT cells with the same antigen specificity.To determine whether preimmunization of SJL mice with

αDEC205/PLP also ameliorated disease induced in mice immu-nized with unconjugated PLP139–151, SJL mice were either leftuntreated or treated with a single injection of 1 μg of αDEC205/

PLP or GL117/PLP isotype control antibody at day minus 10 orminus 15. EAEwas induced on day 0 by injection of 75 μg PLP139–151 in CFA followed by 200 ng PT (PLP139–151/CFA/PT) thenext day, and mice were monitored daily for 30 d for clinical signsof EAE (Fig. 2 A–C). When EAE was induced in naïve SJL mice(i.e., without pretreatment), all of the mice developed clinicalsymptoms between days 9 and 10 and rapidly progressed to severeEAEwithmeanmaximum scores of 3.8–4.4 by days 16–18 (Fig. 2Aand B). Similarly, pretreatment with the GL117/PLP controlresulted in severe EAE with scores of 3.6–4.4 on days 16–18.Deaths of 40–60% of the mice occurred in these experiments.Thus, pretreatment with GL117/PLP control antibody did not re-sult in an amelioration of disease progression and severity and wascomparable to non-pretreated mice. Other control antibodies,αDEC205 itself (NLDC-145), and recombinant αDEC205/HA107-119 also had no significant effect on the disease course.By contrast, mice pretreated with 1 μg αDEC205/PLP showed

consistently delayed onset of disease by up to 5 d, with maximalscores of 1.4–1.7 on days 16–23 (Fig. 2 A and C) (only twomortalities were observed). This reduction was seen whenαDEC205/PLP was administered 10 or 15 d before induction ofEAE (Fig. 2 A and B) but in one experiment appeared less ef-fective when administered at day 20. Thus, the treatment pre-vented disease when administered 23 d before disease onset incontrols. However, administration of αDEC205/PLP at the sametime as immunization with PLP139–151/CFA/PT did not in-terfere with onset or severity of EAE, possibly due to the rapidconversion of immature to mature DCs by immunization.Moreover, coadministration at day 10 of 1 μg αDEC205/PLP with10 μg monophosphoryl lipid A (MPLA), a low-toxicity derivativeof LPS with potent proinflammatory activity that leads to DCmaturation and activation (24), completely abrogated the bene-ficial effect of αDEC205/PLP alone on PLP139–151/CFA/PT-induced EAE (Fig. 2C).

Effect of αDEC205-Mediated Targeting on Pathogenic IL-17–ProducingT Cells.To determine whether αDEC205/PLP-mediated targetinginterfered with early antigen-specific T-cell induction, SJL micewere either left untreated or treated with a single injection of1 μg of αDEC205/PLP or GL117/PLP control mAb 10 d beforeimmunization with PLP139–151/CFA/PT. Total splenocytes thatcontained both antigen-presenting cells and T cells isolated frommice at day 17, either without pretreatment or pretreated withGL117/PLP control antibody, proliferated vigorously to variousPLP139–151 concentrations in vitro, whereas little proliferationwas seen after pretreatment with αDEC205/PLP even in re-sponse to nonphysiologically high peptide concentrations (Fig.3A). Thus, αDEC205 targeting in vivo reduced either the num-

Fig. 1. αDEC205/PLP ameliorates EAE induced by adoptive transfer ofpathogenic PLP139–151-specific T cells. PLP139–151-specific T-cell lines weregenerated as described in Materials and Methods and 5 × 106 cells wereadoptively transferred into naïve SJL/J mice i.v. into the tail veins. One daylater, mice were immunized i.p. with either 1 μg of αDEC205/PLP mAb (n = 5)or GL117/PLPmAb (n = 5), and then PT (200 ng) was injected i.v. on day 3.Miceweremonitored for 30 d. αDEC205/PLP-treatedmicewere protected, whereasthe GL117/PLP-treated mice developed severe disease (P < 0.02 at 30 d).

Stern et al. PNAS | October 5, 2010 | vol. 107 | no. 40 | 17281

IMMUNOLO

GY

Dow

nloa

ded

by g

uest

on

Mar

ch 3

0, 2

020

bers of antigen-specific T cells or their proliferative capacitytested in vitro.To address this question in more detail, the number of path-

ogenic IL-17–secreting cells in splenocytes from SJL mice thatwere either left untreated or pretreated with a single injection of1 μg of recombinant αDEC205/PLP, GL117/PLP control, or ir-relevant αDEC205/HA fusion mAb followed by PLP139–151/CFA/PT immunization 10 d later, was determined. ELISPOTanalysis at day 17 using total splenocytes and overnight restim-ulation with varying concentrations of PLP139–151 in vitroshowed that αDEC205/PLP resulted in an ∼2- to 3-fold reduc-tion in the number of cells secreting IL-17 compared with micethat were not pretreated (P < 0.02) or were pretreated withαDEC205/HA (P < 0.03) (Fig. 3 B and C). Pretreatment with

GL117/PLP seemed to increase the number of IL-17 secretingcells in the spleen (P < 0.004).

CD4+ T Cells from αDEC205/PLP-Pretreated Mice Control EAE Induc-tion After Adoptive Transfer. Did αDEC205/PLP-mediated tar-geting also result in induction of regulatory T cells (Treg)? Toaddress the question, SJL mice were either untreated or pre-treated with either 1 μg αDEC205/PLP or GL117/PLP (Fig. 4A).In one of the experiments, as a positive control, an additionalgroup of SJL mice was coimmunized with 500 μg of the syntheticamino acid copolymer poly(F,Y,A,K)n, which has previously beenshown to ameliorate PLP139–151-induced EAE by the genera-tion of IL-10–secreting Tr1-like Tregs (25, 26). Disease was in-duced 10 d later by PLP139–151/CFA/PT administration. Afteran additional 10 d, splenic CD4+ T cells from all four groups werepurified using magnetic beads, and 5 × 106 cells were i.v. trans-ferred into naïve SJL mice. EAE was induced in recipients thefollowing day by PLP139–151/CFA/PT immunization. Recipientsadoptively transferred with 5 × 106 CD4+ T cells from micewithout pretreatment or pretreated with GL117/PLP developedsevere EAE with mean maximum scores of 3.2–3.6 on days 16–18(Fig. 4). As expected, adoptive transfer of CD4+ T cells from poly(F,Y,A,K)n pretreated mice efficiently prevented EAE inductionin recipient SJL mice. Similarly, CD4+ T cells from αDEC205/PLP-treated mice also significantly ameliorated EAE with a meanmaximum score of 2.0 on days 16–18 (P = 0.003 compared withthe control groups). Strikingly, symptoms ameliorated in thetreated groups (but not in the untreated groups) so that, from day23 onward, basically no signs of EAE were detectable (Fig. 4).Thus, the generation of regulatory CD4+ T cells also played a rolein amelioration of EAE after administration of αDEC205/PLP.

Effects of αDEC205/PLP on Pathogenic Vβ6+ TCR Transgenic T Cells.Splenocytes and lymph node cells from Vβ6+ TCR CD4+ T cellsrecognizing PLP139–151 obtained from 5B6 transgenic B10.Smice (22, 23) were adoptively transferred into rag−/− B10.S(I-As)mice. Mice were treated with 1 μg of either αDEC205/PLP orGL117/PLP. Splenocytes and lymph nodes were harvested 10 dlater, and CD4+ T cells were separated using anti-CD4 magneticbeads. Cells from the mice that had been injected with αDEC205/PLP exhibited limited proliferation and reduced IL-17 productionbut unchanged IFN-γ production in response to in vitro restim-ulation, in comparison with PLP139–151-specific CD4+ T cellsfrom GL117/PLP-treated recipients (P < 0.006) (Fig. 5 A–C).Thus, αDEC205/PLP targeting in vivo contributed to ameliora-tion of EAE by reducing the number of antigen-specific patho-genic IL-17–producing T cells and their proliferative capacity invitro. In addition, Foxp3+ cells in the CD4+ T cell populationswere enumerated by FACS (Fig. 5 D). The percentage of Foxp3+

cells among CD4+ cells in αDEC205/PLP and GL117/PLP pre-treated mice was 15% in each case under these conditions. Anti-DEC205/PLP did not result in detectable conversion of CD4+

Foxp3− T cells to Foxp3+ cells. The percentage of these cells innormal B10.S mice that have been shown to express a high level ofCD4+ CD25+ Tregs (27) averaged 6.1%, and in B10.S micebearing the Vβ6 TCR transgene, it averaged 8.3%. Thus, ho-meostatic expansion of Foxp3+ CD4+ T cells (28, 29) in the rag−/−

background likely accounts for the increased numbers found inboth αDEC205/PLP- and αGL117/PLP-treated mice. A smallerspecific conversion to Foxp3+ CD4+ T cells induced by αDEC205/PLP treatment (18) would not have been detected. CD5 was foundto be expressed in a previous study of αDEC205/MOG35–55 treat-ment in anEAEmodel inC57BL/6mice (17).However, noCD5wasexpressed on the isolated anergized Vβ6+ CD4+ T cells fromαDEC205/PLP-treated mice shown here.

A

B

C

Fig. 2. Effect of preadministration of αDEC205/PLP on disease course.(A and B) EAE was induced in SJL/J mice by immunizing with PLP 139–151with or without preadministration of 1 μg of fusion mAbs (αDEC205/PLPmAb, αDEC205/HA mAb, αDEC205 mAb, or GL117/PLP mAb) in sterile PBS oneither (A) day minus 10 or (B) day minus 15. Mice were then immunized with75 μg of PLP139–151 in CFA s.c. on day 0 followed by PT (200 ng) i.v. on day 1.Appearance of clinical signs of EAE was monitored daily, and disease severitywas scored as described inMaterials and Methods. Mean EAE scores for 5–10mice in each group are shown. The majority of mice in groups immunizedwith PLP139–151 that had received αDEC205 mAb, αDEC205/HA mAb, orGL117/PLP mAb were dead by day 12 whereas disease was ameliorated inthose that received αDEC205/PLP. (A) On day 15, αDEC205/PLP mAb (n = 5) vs.GL117/PLP mAb (n = 5) (P < 0.01). (B) On day 15, αDEC205/PLP mAb (n =13) vs. GL117/PLP mAb (n = 9) (P < 0.001). All scoring was performed doubleblind. The data shown are representative of three to six separate experi-ments. (C) Effect of preimmunization with fusion mAbs together with MPLA.MPLA (10 μg) was administered together with either αDEC205/PLP mAb (n =8) or GL117/PLP (n = 5) i.p. 10 d before induction of EAE with PLP139–151 inCFA s.c. and PT (200 ng) i.v. as above. Mice that received MPLA + αDEC205/PLP mAb were not significantly different from controls (P > 0.05). A repre-sentative of two independent experiments is shown. All scoring was per-formed double blind.

17282 | www.pnas.org/cgi/doi/10.1073/pnas.1010263107 Stern et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 3

0, 2

020

DiscussionLack or loss of tolerance to several self-molecules that have beenidentified as target antigens in autoimmune diseases is one ofthe key events promoting autoimmunity such as multiple scle-rosis or type I diabetes. Despite many studies in both rodents andhumans to stimulate tolerogenic mechanisms using various pro-tocols of antigen administration with antigens in differentpharmaceutical forms (e.g., peptides or whole antigens) andtesting diverse administration routes, robust data demonstratingclinical benefits are not yet available (18). Recent studies in micehave also indicated that repeated administration of free antigenscan induce fatal autoimmune responses (30). In this context, theability to target minute amounts of antigens to steady-state im-mature DCs in vivo has promise as an approach to obtain antigen-specific immunological tolerance.In earlier studies of immunological tolerance induced by tar-

geting of peptides to immatureDCs by fusion to αDEC205, severaldifferent mechanisms have been reported. In earlier studies usingan artificial system in which HA was the target antigen, the in-duction of immunological unresponsiveness by deletion of autor-eactiveT cells or by anergizationwas emphasized (15, 17, 20). Laterstudies, however, focused on the generation of regulatory T cells asan important mechanism in induction of antigen-specific tolerance(18, 21). In the only previous study using a known autoantigen,MOG35–55-induced EAE in C57BL/6 mice was ameliorated bypretreatment at day minus 7 with αDEC205/MOG35–55 (17). Inthe present experiment, αDEC205/PLP139–151 fusion mAb wassynthesized and used to prevent EAE in themodel in which diseaseis induced by PLP139–151 in SJL mice. Anti-DEC205–mediatedtargeting of low nanogram amounts of the immunodominantPLP139–151 efficiently ameliorated EAE induced either by im-munization with PLP139–151 or by adoptive transfer of PLP139–151-specific T cells (Fig. 2). It is important to note that, in thePLP139–151-induced EAE model in SJL mice, pretreatment withlarge doses of free peptide in the absence of adjuvants does not leadto protection from disease induced by subsequent challenge withpeptide/CFA/PT, in contrast to the MOG35–55-induced EAEmodel in C57BL/6 mice (17). Thus, the fact that αDEC205 tar-geting is several magnitudes more efficient in inducing T-cellresponses compared with free peptide administration does notexplain the tolerogenic effect of small amounts of αDEC205/PLPfusion antibodies in the PLP-induced EAE model.

In an attempt to define the mechanism of PLP139–151-inducedtolerance, we showed that αDEC205-mediated targeting in-terfered with early antigen-specific T-cell induction in peripherallymphoid organs upon active EAE induction, reflected by reducednumbers of pathogenic antigen-specific IL-17–producing T cells(Fig. 3). In addition, and consistent with previous reports (15), theremaining cells exhibited an anergic phenotype upon restim-

A B C

Fig. 3. Effect of αDEC205/PLP on splenocyte proliferation and number of IL-17–producing cells. (A) SJL/J mice were preimmunized with 1 μg of differentfusion antibodies. After 10 d, mice were immunized with PLP139–151 followed by i.v. PT as described in Fig. 1. Seventeen days after disease induction,splenocytes were removed and challenged with a titration of PLP139–151. On day 4 of the proliferation assay, cells were pulsed with 3[H]-thymidine; 16 hlater, proliferative response was measured as cpm. (B) IL-17 ELISPOT analysis of mouse splenocytes isolated on day 17. Splenocytes were plated onto precoatedplates as described in protocols from eBioscience’s IL-17 ELISPOT kit and stimulated with 10 μg/mL PLP139–151. Unstimulated wells were used as controls. Arepresentative of two independent experiments is shown. (C) Quantification of IL-17 ELISPOT. Statistics: none vs. αDEC205/PLP mAb (P < 0.02); αDEC205/PLPmAb vs. GL117/PLP mAb (P < 0.006). Spots per million were calculated by multiplying the average of triplicate wells (2 × 105 cells) by 5-fold.

A

B

Fig. 4. Adoptive transfer (ATx) of CD4+ T cells from anti-DEC205/PLP139–151mAb preimmunized mice ameliorates induction of PLP139–151-induced EAE.Two independent experiments are presented (A and B). (A) The 5 × 106 CD4 +T cells enriched from splenocytes from SJL mice preimmunized on day 10 i.p.with a 1 μg fusion mAbs (αDEC205 /PLP mAb or GL117/PLP mAb) were adop-tively transferred i.v. into the tail veins, followedby immunization onday 1with75 μg of PLP139–151 in CFA and PT i.v. the following day. Controls received PBSinjections. Mean disease scores of five mice/group are shown. At days 20–21,αDEC205 /PLPmAbATx vs. GL117/PLPmAbATxwere compared (P< 0.02). (B) Anadditional identical experiment is shown inwhich SJL micewere preimmunizedat day 10 i.p. with 1 μg of αDEC205/PLP mAb only. Mice were monitored forclinical signs of EAE for 30 d. All scoring was performed double blind.

Stern et al. PNAS | October 5, 2010 | vol. 107 | no. 40 | 17283

IMMUNOLO

GY

Dow

nloa

ded

by g

uest

on

Mar

ch 3

0, 2

020

ulation in vitro. It is likely that both deletion and induction of ananergic phenotype in pathogenic T cells contributed to αDEC205/PLP-mediated amelioration of EAE.In addition, however, adoptively transferred CD4+ T cells from

αDEC205/PLP-treated mice efficiently prevented EAE inductionin recipients (Fig. 4 A and B). These data point toward an addi-tional dominant T-cell suppressive mechanism of immunologicaltolerance promoted by αDEC205/PLP-mediated targeting.However, this experiment does not make clear to what extent denovo generation or expansion of preexisting Foxp3− expressingCD4+ Tregs or IL-10 secreting T cells, or conversion of patho-

genic CD4+ Foxp3− cells mediated by αDEC205/PLP, contributesto disease amelioration.To approach the latter possibility, pathogenic CD4+ Vβ6+ T

cells were adoptively transferred to B10.S rag−/− mice. Aftertreatment with αDEC205/PLP, splenocytes or lymph node cellswere markedly anergic to PLP139–151 and had severely reducedIL-17 production but little or no change in IFNγ secretion. Thisexperiment may reinforce the relative importance of IL-17 in thepathogenesis of EAE in this model system (31). A high level ofFoxp3+ CD4+ Vβ6+ T cells was seen after treatment with controlGL117 mAb, and no further increase was found after treatmentwith αDEC205/PLP. Thus, no evidence of specific conversioncould be detected under the conditions of the present experiment.These experiments demonstrate that αDEC205/PLP139–151

ameliorates EAE inductionmainly by inducing anergy in PLP139–151-specific T cells. In addition, evidence of T-cell suppressionwas obtained, although induction of neither IL-10 secretionnor Foxp3+ T cells was seen. In a previous study (17), MOG35–55induced EAE was ameliorated by αDEC205/MOG35-55. In ad-dition to these two autoantigens, MBP85–99 has also been shownto induce EAE, and all have been shown to be potentially im-portant in multiple sclerosis (32, 33). Conceivably, a combinationof these three αDEC205 fusion proteins could represent a thera-peutic modality for this disease.

Materials and MethodsMice. Six- to 12-wk-old female SJL/J (H-2s) mice were purchased from theJackson Laboratory. Vβ6+ PLP139–151-specific 5B6 TCR transgenic mice onthe rag−/− B10.S (B10/I-As) background along with nontransgenic rag−/−

B10.S mice were previously described (22). All animals were maintained atthe animal facilities of Harvard University according to the animal protocolguidelines of Harvard University.

Recombinant Fusion Antibody Production. Double-stranded DNA fragmentscoding for PLP139–151 with spacer residues on both sides were constructedusing synthetic oligonucleotides as described previously (34) using the fol-lowing oligonucleotides: PLP-1 forward, 5′-cta gcg aca tgg cca aga agg agacag tct gga ggc tcg agg agt tcg gta ggt tca caa aca ggC AT; PLP-1 reverse, 5′-CAG GC Tat gcc tgt ttg tga acc tac cga act cct cga gcc tcc aga ctg tct cct tct tggcca tgt cg; PLP-2 forward, 5′- AGC CTG GGC AAA TGG CTG GGC CAT CCG GATAAA TTT tat tat gac ggt agg aca tga tag gc; PLP-2 reverse, 5′-ggc cgc cta tca tgtcct acc gtc ata ata AAA TTT ATC CGGATGGCC CAG CCA TTT GCC (the PLP139–151 peptide-encoding nucleotide sequence split between the two sets ofoligonucleotides is shown in uppercase letters). DNA fragments were addedin-frame to the C terminus of the heavy chains of cloned NLDC-145 (αDEC205)and III/10 isotype control constant regions. To ensure the specificity of antigentargeting the rat IgG2a, constant regions of the original NLDC-145 and iso-type control antibodies were replaced with mouse IgG1 constant regions,which carry point mutations interfering with Fc receptor binding (35). Theplasmid vectors of the IgH chain cDNA of the cloned NLDC-145 (pDEC-IgH)and GL117 (GL117/10-IgH) and their respective IgL-k light chain cDNA (pDEC-IgL-k and pGL117/10-IgL-k) were kindly provided by M. C. Nussenzweig (TheRockefeller University, New York, NY). The plasmid vectors containing thecDNA of amino acids 107–119 of HA (HA107-119) added to the C terminus ofcloned αDEC205 and III/10 control antibody have been described previously(18). Hybrid antibodies were produced using the FreeStyle MAX 293 expres-sion system (Invitrogen) according to the manufacturer’s recommendations.In brief, suspension cultures of FreeStyle 293-F cells were maintained inserum-free FreeStyle 293 expression medium and transiently transfectedwith plasmid vectors of the respective IgH chain and Igk chain cDNA usingFreeStyle MAX reagent. The original anti-DEC205 antibody NLDC-145 (with-out peptide tag), which was included in some experiments as a control, wasproduced by hybridoma cells in serum-free Hybridoma medium (Invitrogen).All antibodies were purified on prepacked HiTrapTM Protein G HP columns(Amersham Biosciences). Protein concentrations were determined specto-photometrically bymeasuring the absorbance at 280 nm. The amount and thepresence of full-length recombinant fusion protein were verified by SDS/PAGE with an IgG1/IgLκ antibody as a reference.

Effect of Fusion Antibodies on the Induction of EAE. For preimmunization, SJL/Jmice were immunized with 1 μg i.p. of fusion antibodies (αDEC205/PLP mAb,αDEC205/HA mAb, αDEC205 mAb alone, or GL117/PLP mAb) either 10 or 15

A

B

C

D

Fig. 5. Effect of αDEC205/PLP on adoptively transferred CD4+ Vβ6+ TCR 5B6transgenic (tg) T cells. Splenocytes were isolated from B10.S mice that carrya transgenic TCR 5B6 recognizing PLP139–151 presented on I-As. Splenocyteswere enriched for Vβ6+ CD4+ tg T cells using Miltenyi CD4-positive selectionkits (purity ∼89%). (A and B) The 10 × 106 T cells were injected i.v. into naïveB10.S rag−/− mice along with 1 μg i.p. of fusion antibodies (αDEC205/PLPmAb or GL117/PLP mAb). Splenocytes (A) and axillary lymph nodes (B) wereremoved 10 d later. Single cell suspensions were stimulated with PLP139–151for 4 d, and 3H-thymidine incorporation was measured. The αDEC205/PLPmAb-treated Vβ6+ CD4+ tg T cells did not proliferate in response to PLP139–151 peptide, whereas Vβ6+ CD4+ tg T cells treated with GL117/PLP mAbproliferated (P < 0.03). (C) Vβ6+ TCR 5B6 tg CD4+ T cells were stimulated bycross-linking using plate-bound CD3 and CD28 mAb coated overnight todetect cytokine production. Supernatants from the proliferation assay wereremoved 3 d after stimulation, and cytokines were measured by Luminexassay as described in Materials and Methods. IL-17 was significantly reducedupon administration of 1 μg of αDEC205/PLP mAb compared with a controlgroup treated with 1 μg of GL117/PLP mAb (P < 0.005). (D) Splenocytes usedwere obtained in A. FACS analysis of gated CD4+ cells stained for in-tracellular Foxp3 was carried out using CD4-FITC and Foxp3-PE.

17284 | www.pnas.org/cgi/doi/10.1073/pnas.1010263107 Stern et al.

Dow

nloa

ded

by g

uest

on

Mar

ch 3

0, 2

020

d before inducing EAE. Six- to 10-wk-old female mice were immunized s.c.with 75 μg of PLP139–151 emulsified in CFA; 200 ng PT (List BiologicalLaboratories) was given i.v. on the day after immunization. The mice weremonitored for clinical signs of EAE, and they were scored from 0 to 5: 1, limptail; 2, hind limb paralysis; 3, complete hind limb paralysis; 4, four limbsparalyzed; 5, moribund. All scoring was performed double blind.

Details of the proliferation assay, cytokine measurements, and adoptivetransfer experiments are included in SI Materials and Methods.

ACKNOWLEDGMENTS. We thank M. Nussenzweig (The Rockefeller Univer-sity, New York) for providing the plasmid vectors of the IgH and respectiveIgκ light chain cDNA of cloned anti-DEC205 NLDC-145 and III/10 isotypecontrol recombinant antibodies and T. Koenig (Kretschmer Laboratory) forexcellent technical assistance in recombinant antibody production. This workwas supported by grants from the National Institute of Allergy and Infec-tious Diseases (AI049524) and from the National Multiple Sclerosis Society(RG 3796A3/1).

1. Hafler DA, Weiner HL (1989) MS: a CNS and systemic autoimmune disease. ImmunolToday 10:104–107.

2. Zamvil SS, Steinman L (1990) The T lymphocyte in experimental allergic enceph-alomyelitis. Annu Rev Immunol 8:579–621.

3. Wekerle H (1993) Experimental autoimmune encephalomyelitis as a model ofimmune-mediated CNS disease. Curr Opin Neurobiol 3:779–784.

4. Tuohy VK, Lu Z, Sobel RA, Laursen RA, Lees MB (1989) Identification of anencephalitogenic determinant of myelin proteolipid protein for SJL mice. J Immunol142:1523–1527.

5. Risau W, Engelhardt B, Wekerle H (1990) Immune function of the blood-brain barrier:Incomplete presentation of protein (auto-)antigens by rat brain microvascularendothelium in vitro. J Cell Biol 110:1757–1766.

6. Stromnes IM, Goverman JM (2006) Active induction of experimental allergicencephalomyelitis. Nat Protoc 1:1810–1819.

7. McKenzie BS, Kastelein RA, Cua DJ (2006) Understanding the IL-23-IL-17 immunepathway. Trends Immunol 27:17–23.

8. Bettelli E, Oukka M, Kuchroo VK (2007) T(H)-17 cells in the circle of immunity andautoimmunity. Nat Immunol 8:345–350.

9. Steinman L (2007) A brief history of T(H)17, the first major revision in the T(H)1/T(H)2hypothesis of T cell-mediated tissue damage. Nat Med 13:139–145.

10. Stromnes IM, Cerretti LM, Liggitt D, Harris RA, Goverman JM (2008) Differentialregulation of central nervous system autoimmunity by T(H)1 and T(H)17 cells. NatMed 14:337–342.

11. Miller SD, Turley DM, Podojil JR (2007) Antigen-specific tolerance strategies for theprevention and treatment of autoimmune disease. Nat Rev Immunol 7:665–677.

12. Verginis P, McLaughlin KA, Wucherpfennig KW, von Boehmer H, Apostolou I (2008)Induction of antigen-specific regulatory T cells in wild-type mice: Visualization andtargets of suppression. Proc Natl Acad Sci USA 105:3479–3484.

13. Apostolou I, von Boehmer H (2004) In vivo instruction of suppressor commitment innaive T cells. J Exp Med 199:1401–1408.

14. Dudziak D, et al. (2007) Differential antigen processing by dendritic cell subsets invivo. Science 315:107–111.

15. Hawiger D, et al. (2001) Dendritic cells induce peripheral T cell unresponsivenessunder steady state conditions in vivo. J Exp Med 194:769–779.

16. Bozzacco L, et al. (2007) DEC-205 receptor on dendritic cells mediates presentation ofHIV gag protein to CD8+ T cells in a spectrum of human MHC I haplotypes. Proc NatlAcad Sci USA 104:1289–1294.

17. Hawiger D, Masilamani RF, Bettelli E, Kuchroo VK, Nussenzweig MC (2004)Immunological unresponsiveness characterized by increased expression of CD5 onperipheral T cells induced by dendritic cells in vivo. Immunity 20:695–705.

18. Kretschmer K, et al. (2005) Inducing and expanding regulatory T cell populations byforeign antigen. Nat Immunol 6:1219–1227.

19. Mukhopadhaya A, et al. (2008) Selective delivery of beta cell antigen to dendritic cellsin vivo leads to deletion and tolerance of autoreactive CD8+ T cells in NOD mice. ProcNatl Acad Sci USA 105:6374–6379.

20. Bruder D, et al. (2005) On the edge of autoimmunity: T-cell stimulation by steady-state dendritic cells prevents autoimmune diabetes. Diabetes 54:3395–3401.

21. Yamazaki S, et al. (2008) CD8+ CD205+ splenic dendritic cells are specialized to induceFoxp3+ regulatory T cells. J Immunol 181:6923–6933.

22. Waldner H, Collins M, Kuchroo VK (2004) Activation of antigen-presenting cells bymicrobial products breaks self tolerance and induces autoimmune disease. J ClinInvest 113:990–997.

23. Kuchroo VK, et al. (2002) T cell response in experimental autoimmuneencephalomyelitis (EAE): Role of self and cross-reactive antigens in shaping, tuning,and regulating the autopathogenic T cell repertoire. Annu Rev Immunol 20:101–123.

24. Mata-Haro V, et al. (2007) The vaccine adjuvant monophosphoryl lipid A as a TRIF-biased agonist of TLR4. Science 316:1628–1632.

25. Stern JN, et al. (2008) Amino acid copolymer-specific IL-10-secreting regulatory T cellsthat ameliorate autoimmune diseases in mice. Proc Natl Acad Sci USA 105:5172–5176.

26. Stern JN, et al. (2004) Amelioration of proteolipid protein 139-151-inducedencephalomyelitis in SJL mice by modified amino acid copolymers and theirmechanisms. Proc Natl Acad Sci USA 101:11743–11748.

27. Reddy J, et al. (2004) Myelin proteolipid protein-specific CD4+CD25+ regulatory cellsmediate genetic resistance to experimental autoimmune encephalomyelitis. Proc NatlAcad Sci USA 101:15434–15439.

28. Jameson SC (2002) Maintaining the norm: T-cell homeostasis. Nat Rev Immunol 2:547–556.

29. Nishio J, Feuerer M, Wong J, Mathis D, Benoist C (2010) Anti-CD3 therapy permitsregulatory T cells to surmount T cell receptor-specified peripheral niche constraints.J Exp Med, in press.

30. Pugliese A, et al. (2001) Self-antigen-presenting cells expressing diabetes-associatedautoantigens exist in both thymus and peripheral lymphoid organs. J Clin Invest 107:555–564.

31. Axtell RC, et al. (2010) T helper type 1 and 17 cells determine efficacy of interferon-βin multiple sclerosis and experimental encephalomyelitis. Nat Med 16:406–412.

32. Zhang J, et al. (1994) Increased frequency of interleukin 2-responsive T cells specificfor myelin basic protein and proteolipid protein in peripheral blood andcerebrospinal fluid of patients with multiple sclerosis. J Exp Med 179:3973–3984.

33. Bettelli E, Baeten D, Jäger A, Sobel RA, Kuchroo VK (2006) Myelin oligodendrocyteglycoprotein-specific T and B cells cooperate to induce a Devic-like disease in mice.J Clin Invest 116:2393–2402.

34. Kretschmer K, Heng TS, von Boehmer H (2006) De novo production of antigen-specificsuppressor cells in vivo. Nat Protoc 1:653–661.

35. Clynes RA, Towers TL, Presta LG, Ravetch JV (2000) Inhibitory Fc receptors modulate invivo cytoxicity against tumor targets. Nat Med 6:443–446.

Stern et al. PNAS | October 5, 2010 | vol. 107 | no. 40 | 17285

IMMUNOLO

GY

Dow

nloa

ded

by g

uest

on

Mar

ch 3

0, 2

020