of April 13, 2017.This information is current as

Autoimmune Encephalomyelitisagainst Acute Fatal ExperimentalCostimulation Protects Rhesus Monkeys Selective Blockade of CD28-Mediated T Cell

Bert A. 't Hart and Bernard VanhoveCaroline Mary, Nicolas Poirier, Paul Baker, Linda Scobie, Krista G. Haanstra, Karin Dijkman, Noun Bashir, Jan Bauer,

http://www.jimmunol.org/content/194/4/1454doi: 10.4049/jimmunol.1402563January 2015;

2015; 194:1454-1466; Prepublished online 14J Immunol 

Referenceshttp://www.jimmunol.org/content/194/4/1454.full#ref-list-1

, 20 of which you can access for free at: cites 49 articlesThis article

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

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

Errata

/content/195/2/749.full.pdfor:

next pageAn erratum has been published regarding this article. Please see

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2015 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Selective Blockade of CD28-Mediated T Cell CostimulationProtects Rhesus Monkeys against Acute Fatal ExperimentalAutoimmune Encephalomyelitis

Krista G. Haanstra,* Karin Dijkman,* Noun Bashir,* Jan Bauer,† Caroline Mary,‡

Nicolas Poirier,‡ Paul Baker,x Linda Scobie,x Bert A. ’t Hart,*,{ and Bernard Vanhove‡,‖

Costimulatory and coinhibitory receptor–ligand pairs on T cells and APC control the immune response. We have investigated

whether selective blockade of CD28–CD80/86 costimulatory interactions, which preserves the coinhibitory CTLA4–CD80/86

interactions and the function of regulatory T (Treg) cells, abrogates the induction of experimental autoimmune encephalomyelitis

(EAE) in rhesus monkeys. EAE was induced by intracutaneous immunization with recombinant human myelin oligodendrocyte

glycoprotein (rhMOG) in CFA on day 0. FR104 is a monovalent, PEGylated-humanized Fab9 Ab fragment against human CD28,

cross-reactive with rhesus monkey CD28. FR104 or placebo was administered on days 0, 7, 14, and 21. FR104 levels remained high

until the end of the study (day 42). Placebo-treated animals all developed clinical EAE between days 12 and 27. FR104-treated

animals did not develop clinical EAE and were sacrificed at the end of the study resulting in a significantly prolonged survival.

FR104 treatment diminished T and B cell responses against rhMOG, significantly reduced CNS inflammation and prevented

demyelination. The inflammatory profile in the cerebrospinal fluid and brain material was also strongly reduced. Recrudescence

of latent virus was investigated in blood, spleen, and brain. No differences between groups were observed for the b-herpesvirus

CMV and the polyomaviruses SV40 and SA12. Cross-sectional measurement of lymphocryptovirus, the rhesus monkey EBV,

demonstrated elevated levels in the blood of FR104-treated animals. Blocking rhesus monkey CD28 with FR104 mitigated

autoreactive T and B cell activation and prevented CNS pathology in the rhMOG/CFA EAE model in rhesus monkeys. The

Journal of Immunology, 2015, 194: 1454–1466.

Tcells have a central pathogenic role in the human neuro-inflammatory disease multiple sclerosis (MS) and itselected animal model experimental autoimmune enceph-

alomyelitis (EAE). The full activation of naive (autoreactive)T cells involves three signals: signal 1 via the TCR recognition ofAg–MHC complexes; signal 2 relayed via one or more costimu-latory pathways; and signal 3 via cytokines released by the acti-vated APC. One of the most important costimulation pathways isthe binding of CD80/86 on APC to CD28 that is constitutivelyexpressed on naive T cells. CD80/86 can also bind CTLA-4 on the

T cell. The pathway leads to inhibition of T cell activation andinduction of T cell anergy. Targeted modulation of this pathway was

found to be therapeutically effective in rheumatoid arthritis as

demonstrated by treatment with the CD80/86 antagonist CTLA4-Ig

(Orencia/abatacept) (1). Abatacept was tested in a Phase I trial in

MS patients and had a positive effect on T cell responses against

myelin basic protein (2). Phase II clinical trials are under way (3).

However, because CTLA-4-Ig fusion proteins are inhibiting co-

stimulatory signaling through CD28 as well as coinhibitory sig-

naling through CTLA-4, it is anticipated that a selective inhibition

of CD28-mediated costimulatory signals might present a better

therapeutic index by maintaining CTLA-4–mediated signals in-

volved in T cell anergy and regulatory T (Treg) cell function.The involvement of CD28 in EAE pathogenesis was established

in mouse models. CD28 knockout mice were resistant to EAEinduction (4–6) and blockade with a CD28-specific Ab amelio-rated the expression of the disease (7, 8).FR104 is a novel human CD28 antagonist that is produced as

a monovalent and PEGylated humanized Fab9 Ab fragment toavoid agonistic effects observed with most anti-CD28 completeIgG mAbs (9, 10). FR104 blocks CD28 binding to CD80/86,thereby favoring CD80/86 interaction with CTLA-4 (11). Thelack of cross-reactivity of FR104 with mouse or rat CD28 pre-cluded efficacy testing in rodent EAE models. However, initialpreclinical efficacy in vitro and in vivo tests indicated a goodreactivity and immunological safety profile in nonhuman primates(i.e., cynomolgus monkeys (11) and baboons (12)). PEGylatedFR104 did not bind marmoset CD28, precluding efficacy testing inthis well-established EAE model (13). Hence, we used an estab-lished EAE model in rhesus macaques induced with recombinanthuman myelin oligodendrocyte glycoprotein (rhMOG) in CFA.

*Biomedical Primate Research Centre, 2280 GH Rijswijk, the Netherlands;†Center for Brain Research, Medical University of Vienna, 1090 Vienna, Austria;‡Effimune SAS, 44035 Nantes, France; xGlasgow Caledonian University, GlasgowG4 0BA, United Kingdom; {University of Groningen, University Medical Center,Department of Neuroscience, 9713 GZ Groningen, the Netherlands; and ‖InstitutNational de la Sante et de la Recherche Medicale Unite Mixte de Recherche1064, 44093 Nantes, France

ORCID: 0000-0002-7654-5096 (K.G.H.).

Received for publication October 7, 2014. Accepted for publication December 11,2014.

This work was supported by the European Union Seventh Framework Programme(FP7/2007-2013) under Grant Agreement 281493: Tolerance Restoration in Autoim-mune Diseases by Selective Manipulation of the CD28 Costimulatory Pathway.

Address correspondence and reprint requests to Dr. Krista G. Haanstra, BiomedicalPrimate Research Centre, P.O. Box 3306, 2280 GH Rijswijk, the Netherlands. E-mailaddress: haanstra@bprc.nl

Abbreviations used in this article: ADA, anti-drug Ab; ADEM, acute disseminatedencephalomyelitis; ALN, axillar lymph node; AU, arbitrary unit; CSF, cerebrospinalfluid; EAE, experimental autoimmune encephalomyelitis; HEV, hepatitis E virus;IL-1RA, IL-1R antagonist; LCV, lymphocryptovirus; LFB, luxol fast blue; MNC,mononuclear cell; MS, multiple sclerosis; PAS, periodic acid–Schiff; rhMOG,recombinant human myelin oligodendrocyte glycoprotein; SI, stimulation index;Treg, regulatory T.

Copyright� 2015 by TheAmericanAssociation of Immunologists, Inc. 0022-1767/15/$25.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1402563

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

This model is characterized by an early disease onset (between 11and 40 d) and a very rapid disease progression from first clinicalsigns until ethical endpoints indicate euthanasia (between a fewhours to 1–2 d) (14–16). The disease incidence is 100%, and themodel has been validated with the anti-a4b1 mAb natalizumab(Tysabri) (15). We report that FR104 has a profound suppressiveeffect on the induction of EAE in six treated animals as comparedwith placebo-treated animals at the level of T and B cell activa-tion, resulting in greatly diminished CNS pathology.Under conditions of immune suppression or indeed altered

immune status, latent viral activation can lead to significantcomplications for the patient, if not treated prophylactically, inparticular for the herpesviruses and polyomaviruses (17–19). Weinvestigated whether reactivation of existing latent endogenousviral infections could be detected in blood and tissues of bothplacebo and FR104-treated animals. We report possible exacer-bation only of macaque lymphocryptovirus (LCV), a gamma-herpesvirus related to EBV.

Materials and MethodsAbs and in vitro evaluation

FR104 is a monovalent, PEGylated, humanized Fab9 Ab fragment thatantagonizes binding of CD28 to CD80 and CD86. Because of its monovalentnature and its target epitope lying outside the C99D loop seen by superagonistanti-CD28 Abs (20), FR104 cannot induce T cell activation (11).

To assess whether FR104 abrogates the function of naturally occurringprimate Treg cells, PBMC were isolated from heparinized venous bloodfrom healthy rhesus monkeys (n = 4) using density gradient centrifugation.CD4+CD252 and CD4+CD25+ T cells were isolated and cultured as pre-viously described (21), with minor modifications. In brief, CD4+CD25+

cells were separated from CD4+CD252 cells using the Treg isolation kitfrom Miltenyi Biotec (Bergisch Gladbach, Germany). CD4+CD252 cellswere stimulated with plate-bound anti-CD3 (1 mg/ml) in the presenceor absence of CD4+CD25+ Treg cells. FR104 (10 mg/ml) was added toevaluate its effect on the inhibitory effect of CD4+CD25+ Tregs on CD4+

CD252 proliferating effector T cells. Proliferation was assessed by[3H]thymidine incorporation during the last 18 h of a 5-d culture.

To establish whether the rhesus monkey EAE model was suitable to in-vestigate the effect of FR104, we first tested whether CD28 blocking iseffective to suppress rhMOG-activated cells ex vivo. To test this, we usedstored PBMC, spleen cells, or inguinal lymph node cells from monkeys withEAE, which had an established positive response against rhMOG, asdocumented previously (15, 22). Mononuclear cells (MNC) were culturedin quadruplicate (100,000/well) in culture medium with or without rhMOG(5 mg/ml), in the presence or absence of FR104 (10 mg/ml). Inhibition ofCD28 interaction with CD80/86 by CTLA4-Ig (10 mg/ml; R&D Systems)was used as a positive control.

Animals

Naive, adult rhesus monkeys (Macaca mulatta) were randomly selectedfrom the purpose-bred colony of the Biomedical Primate Research Centreand housed under conventional conditions. Animals were only includedafter a complete physical, hematological, and biochemical checkup hadbeen performed, and monkeys were declared healthy. During the study,monkeys were pair-housed and remained under intensive veterinary care.The daily diet consisted of commercial food pellets for nonhuman primates(Sniff, Soest, Germany), supplemented with rice, raisins, peanuts, andfresh fruit. Drinking water was provided ad libitum.

The experimental design, all study protocols and experimental procedureswere reviewed and approved by the Biomedical Primate Research Centre’sEthics Committee, in accordance with Dutch law on animal experimentation.

Experimental design

The number of animals per group was determined by a power calculationfor graft survival (23). Entered parameters were based on the effectsof natalizumab in this model (15). The power calculation indicated thatstatistically significant results could be obtained with six animals pertreatment group. Twelve animals, all males, were stratified over twogroups, ensuring comparable (i.e., no significant differences in groupmeans) age and weight distribution (Table I). FR104 was dosed at 10 mg/kg(protein equivalent) on days 0, 7, 14, and 21 via the i.v. route and placeboanimals received an equivalent volume of the vehicle for FR104.

Cerebrospinal fluid (CSF) samples were collected via the cisterna magnaprior to euthanasia from sedated monkeys and were processed as describedpreviously (15).

EAE induction and monitoring

EAE was induced with rhMOG, a recombinant protein produced inEscherichia coli that represents the extracellular domain of human MOG,residues 1–125; production and purification was as described previously(24). Animals were immunized on day 0 with 300 mg rhMOG dissolvedin 300 ml PBS and emulsified with an equal volume of CFA (Difco Lab-oratories, Detroit MI). The total volume of 600 ml was injected into thedorsal skin at six sites, targeting both the inguinal and the axillary regions.

Clinical signs were scored daily by observers blinded to the treatmentusing a previously described semiquantitive scale (14, 25, 26): 0, no clinicalsigns; 0.5, loss of appetite and vomiting; 1, substantial reduction of generalcondition; 2, ataxia, sensory loss and/or visual problems; 2.5, incompleteparalysis of one (hemiparesis) or two sides (paraparesis); 3, complete pa-ralysis of one (hemiplegia) or two sides (paraplegia); 4, complete paralysis(quadriplegia); and 5, moribund. The rhMOG-induced EAE model inrhesus monkeys is characterized by acute onset and rapid diseaseprogression, reaching a moribund state within 24 h. To avoid suffering,monkeys were euthanized at EAE score $ 2.5 or at score 2 when theanimal was not expected to survive until the next day.

Pharmacokinetics assessment

Concentrations of FR104 in serum samples from rhesus monkeys werequantified by ELISA. In brief, 96-well microtiter plates were coated withCD28Fc (catalog no. 342-CD-200; R&D Systems) at 2 mg/ml in carbonatebuffer (pH 9) at 50 ml/well overnight at 4˚C. After washing, free bindingsites were blocked using PBS/0.1% Tween/1% BSA) at 100 ml/well for 2 hat 37˚C. Sera were diluted 1/1,000, 1/3,000, 1/9,000, and 1/27,000 andincubated at 50 ml/well for 2 h at 37˚C. After washing with PBS/Tweenbinding of FR104 was detected using a mouse anti-human k mAb (catalogno. NaM76-5F3; Effimune), diluted to 1/10,000 at 50 ml/well for 1 h at37˚C. Detection of bound Abs was enhanced using a peroxidase-labeleddonkey anti-mouse IgG (catalog no. 715-036-151; Jackson Immuno-Research Laboratories) diluted at 1/2000 at 50 ml/well for 1 h at 37˚C.Tetramethylbenzidine substrate was added at 50 ml/well, and color de-velopment was allowed for 10 min at room temperature. The reaction wasstopped with 50 ml/well H2SO4, and absorbance was read at 450 nm.Concentrations were calculated using a two-parameter fit of the linear partof a titration curve with known concentrations of FR104.

Anti-drug Abs monitoring

Anti-drug Abs (ADA) serum levels were evaluated by ELISA. FR104 wascoated on 96-well microtiter plates at 5 mg/ml in borate buffer (pH 9.5)50 ml/well overnight at 4˚C. After washing, plates were blocked with PBS/0.1% Tween/0.25% gelatin at 100 ml/well for 2 h at 37˚C. Sera were di-luted 1/15 at 50 ml/well for 2 h at 37˚C. IgM ADA were detected usingrabbit anti-human IgM (catalog no. F0203; DakoCytomation), and IgGADA were detected using rabbit anti-human IgG (catalog no. F0202;DakoCytomation), both diluted to 1/200 and incubated at 50 ml/well for1 h at 37˚C. Detection was enhanced by incubation with peroxidase-labeled goat anti-rabbit IgG (catalog no. 111-035-144; Jackson Immu-noResearch Laboratories) diluted to 1/2000 and incubated at 50 ml/well for1 h at 37˚C. ABTS substrate was added at 50 ml/well for 15 min at roomtemperature. The reaction was stopped with 50 ml/well H2O and absor-bance was read at 405 nm.

Cellular immune responses against rhMOG

PBMCwere isolated before and once weekly after EAE induction and at thetime of necropsy. At necropsy, MNCwere also isolated from the spleen, andaxillar lymph nodes (ALN). Isolated MNC were dispensed in quadruplicateat 1 3 105 cells/well in 96-well round-bottom microtiter plates with orwithout 0.5 mg/ml rhMOG. MNC proliferation was assayed by the in-corporation of [3H]thymidine (0.5 mCi/well) during the final 18 h of a5-d culture. Cells were harvested for beta scintillation counting (TopcountNXT; Packard, Ramsey, MN). The results obtained from the differentculture conditions were expressed as stimulation index (SI): (cpm of MNCproliferation with rhMOG)/(cpm of MNC proliferation without rhMOG).An SI $ 3 was considered positive.

Ab responses against rhMOG

Serum samples were collected prior to EAE induction, once weeklythereafter, and at necropsy. CSF samples were obtained at necropsy. CSF

The Journal of Immunology 1455

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

and sera were tested for the presence of Abs against rhMOG by ELISA in96-well microtiter plates. Plates were coated with rhMOG (5 mg/ml) andincubated overnight at 4˚C. After washing and blocking with PBS/1%BSA, the wells were incubated in duplicate with 1:100 or 1:1000 dilutedsera or CSF. Bound rhesus monkey Abs were detected with alkalinephosphatase–labeled goat-anti-human IgG (1:2,000; catalog no. AHI1305;Invitrogen) or alkaline phosphatase–labeled goat-anti-human IgM (1:10,000;catalog no. A9794; Sigma-Aldrich, Zwijndrecht, the Netherlands). Conju-gate binding was quantified with p-nitrophenyl phosphate (Sigma-Aldrich).OD values were converted to arbitrary units (AU) using the same positivecontrol on all plates as reference.

Quantification of demyelination and cell infiltration

Left hemispheres were fixed with 4% formaldehyde and subsequentlyprocessed for histopathology as described previously (15, 27). Briefly, threesamples were excised from standardized regions of each brain and em-bedded in paraffin. Sections were deparaffinized and processed for H&Estaining to quantify inflammation and luxol fast blue/periodic acid–Schiff(LFB/PAS) staining for demyelination. In addition, immunohistochemicalstainings for CD3 and MRP14 were performed. LFB/PAS-stained sectionswere scanned and analyzed with ImageJ (version 1.44p, a public domainimage processing and analysis program developed by W. Rasband at theNational Institutes of Health). White and gray matter areas and thedemyelinated areas in the sections were selected with the selection toolsand quantified. Demyelination is expressed relative to total white and graymatter surface. Inflammation in each section (three sections per animal)was scored on an arbitrary scale: 0, no inflammation; 0.5, some inflam-mation (some perivascular infiltrates); 1.0, moderate inflammation (manysmall perivascular infiltrates); and 2.0, strong inflammation (many smalland large perivascular infiltrates, deep infiltration of T cells in the paren-chyma). Afterward, for each animal, the average score of the three sectionswas calculated.

Quantitative RT-PCR

The corpus callosum was sampled for quantitative RT-PCR analysis. Tis-sues were stored at 280˚C. RNA was isolated using the RNeasy LipidTissue Mini Kit (Qiagen, Hilden, Germany), according to the instructionsof the manufacturer. RNA concentrations were determined using theNanoDrop ND-1000 spectrophotometer (Thermo Scientific, Wilmington,DE). cDNA was prepared from mRNA using the RevertAid First-StrandcDNA synthesis kit (Thermo Scientific).

Intron spanning primer-probe pairs were designed with the UniversalProbe Library Assay Design Center (Roche, Basel, Switzerland). Primerswere purchased from Invitrogen Life Technologies (Wilmington, DE). RT-PCRs were performed using iTaq supermix and CFX96 Real-Time system(Bio-Rad, Hercules, CA). Levels of mRNA for genes of interest wereexpressed as relative concentrations compared with ABL or GAPDH, takingPCR efficiencies into account. When mRNA levels of investigated genes inparticular samples were below the detection level, the relative concen-trations to ABL or GAPDH were set to 0. This occurred in three samplesonly (IL-6: animals P6 and F2; CCR7: animal F4).

Cytokine/chemokine analysis of the CSF

The supernatants obtained after centrifugation of the CSF samples wereanalyzed with Luminex technology. A monkey-specific cytokine/chemokine 29-plex panel (Novex; Life Technologies) was used accord-ing to the instructions of the manufacturer. Included cytokines/chemokinesare listed in Table II. Briefly, 50 ml CSF was incubated in duplicate for 2 hwith anticytokine/chemokine Ab–coated beads. All CSF samples wereincubated on the same plate at room temperature on an orbital shaker at750 rpm. The cytokine/chemokine standard curve provided with the kitwas included on the plate. After washing, biotinylated Abs supplied withthe kit were added to the beads and incubated for 1 h and after anotherwashing step, streptavidin–R-PE was added and incubated for 30 min.After removal of the streptavidin–R-PE and washing, the beads wereresuspended in wash solution, after which the plate was read on a Bio-Plex200 system (Bio-Rad, Veenendaal, the Netherlands). Concentrations of thecytokines and chemokines in the CSF were calculated using the Bio-PlexManager Software 4.1 program (Bio-Rad). Results are given as median andrange (minimum – maximum).

Assessment of endogenous viral reactivation

DNAwas isolated from blood, spleen, and brain using the DNeasy blood andtissue kit (Qiagen) and viral RNA from sera using the QIAmp viral RNAminikit (Qiagen), according to the manufacturer’s instructions. Sampleswere taken, at set intervals, before and after treatment with FR104, and

quantitative PCR was used to test the levels of virus. All assays used theTaqMan Gene Expression system (Applied Biosystems) in a final volumeof 25 ml using the Viia 7 real time (Applied Biosystems). For CMV, pri-mers at a concentration of 0.5 mM, 59-GTTAGGGAACCGCCATTCTG-39(forward) and 59-GTATCCGCGTTCCAATGCA-39 (reverse), and probe6Fam-TCCAGCCTCCATAGCCGGGAAGG-TAMRA used at a concen-tration of 0.25 mM, were cycled in a 25 ml volume for 2 min at 50˚C,10 min at 95˚C, followed by 40 cycles of 95˚C for 15 s, 53˚C for 30 s and60˚C for 60 s. For LCV, primers were as follows: 59-GGACGTGCA-CTACAAGGAGA-39 (forward) and 59-GAATCYTGCACGCAGTACAT-39(reverse) and the probe 6Fam-TCGCCTCTTTGCAGAGGGCC-TAMRAwere also used at a concentration of 0.5 and 0.25 mM, respectively, andcycled in a 25-ml volume for 2 min at 50˚C, 10 min at 95˚C, followed by40 cycles of 95˚C for 15 s, 50˚C for 30 s, and 60˚C for 60 s. The hepatitisE virus (HEV) levels were measured using a HepatitisE@ceeram Tools kitby Ceeram (La Chapelle sur Erdre, France), according to the manufacturer’sinstructions. Samples were quantified against the World Health Organi-zation HEV standard (28).

For polyomavirus SA12, amplification was as previously described byCantalupo et al. (29). Standard plasmid control pUC19-SA12 was providedby Prof. P. Cantalupo (University of Pittsburgh, Pittsburgh, PA).

For SV40 amplification, the primers were as follows: 59-GTGTTTCA-GATTCCAACCTATG-39 (forward) and 59-CAGCAGTAGCCTCTTCATC-39and the probe 6 Fam-TGATGAATGGGAGCAGTGGTGGA-TAMRA. Thesewere at a concentration of 0.5 and 0.125 mM, respectively. The total reactionvolume was 25 ml, and the cycling conditions were 2 min at 50˚C, 10 min at95˚C, followed by 40 cycles of 95˚C for 15 s and 60˚C for 1 min. For graphicalpresentation on a log scale, LCV DNA levels below the limit of detectionor quantification were set to 1.

Statistical analysis

Statistical analysis was performed using Prism 6 for Mac OS X (GraphPad,San Diego, CA). Survival curves were compared using the log-rank test(Mantel–Cox). Average group data are given as mean6 SEM or as medianand range (minimum – maximum). Significance of differences betweengroups was calculated using the nonparametric t test (Mann–Whitney Utest). Viral DNA levels are given as mean 6 SD of sample replicates.Significance of difference of viral DNA levels between groups was cal-culated using the parametric t test.

ResultsFR104 blocks an in vitro recall response and does not interferewith Treg cells

We have previously reported that CD28 blockade enhanced Tregfunction in vitro in humans and in vivo in allograft rejection modelsin baboons and cynomolgus macaques (30, 31). To examinewhether the same principle can be observed in rhesus monkeys wetested the effect of FR104 on rhesus monkey Tregs in a previouslypublished assay (21). CD4+CD252 cells stimulated with plate-bound anti-CD3 proliferated vigorously. In this experiment theaverage response was 33,276 6 11,459 cpm. Addition of natu-rally occurring CD4+CD25+ Treg cells reduced the anti-CD3stimulated proliferation to 2.2 6 1.1% of the response withoutadded Tregs (Fig. 1A). The effect of FR104 on the natural Tregswas investigated as well. The response of CD4+CD252 cells initself was inhibited by FR104 down to 43 6 11% of the responsewithout FR104 (data not shown). However, CD4+CD25+ cellsinhibited this proliferation even further, to similar levels as in theabsence of FR104 (Fig. 1A), indicating that suppression by thenatural Treg cells were not affected by FR104. Control culturescontaining the double amount of CD252 cells, thus equaling theamount of cells in the CD4+CD252 and CD4+CD25+ coincuba-tions, was not inhibited (Fig. 1A), indicating that the inhibitionobserved in the CD252/CD25+ coculture was not due to exhaustionof the medium. In conclusion, as previously demonstrated forhuman, baboon, and cynomolgous monkey Treg cells (30, 31),FR104 does not interfere with rhesus monkey natural Treg cellfunction.Next, we tested the effect of FR104 on the ex vivo recall response

against rhMOG of MNC from rhMOG-sensitized rhesus monkeys.

1456 CD28 BLOCKADE PREVENTS EAE IN RHESUS MONKEYS

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

Stored cells from monkeys with EAE, which had an establishedpositive response toward rhMOG, were stimulated with rhMOG exvivo in the presence or absence of FR104 or CTLA4-Ig. Fig. 1Bshows that MNC proliferation was inhibited by FR104 to the sameextent as by CTLA4-Ig (Fig. 1B). These results underline theimportant role of CD28 in secondary T cell responses againstrhMOG.

Pharmacokinetics and immunogenicity of FR104 in rhesusmonkeys

We tested FR104 at a dose of 10 mg/kg FR104 administered on postimmunization days 0, 7, 14, and 21 based on data from extensivepharmacokinetic and pharmacodynamic monitoring in baboons(12) and cynomolgous monkeys (11). FR104 through levels werehomogenous and consistently .50 mg/ml in all animals (Fig. 2A).Only in animal F3 we observed higher levels than in the others.Interestingly, primate IgG ADA levels were detected only in thissame animal after day 11, but this was not observed in any of theother animals (Fig. 2B). IgM ADA levels were below the detectionlimit (data not shown).

FR104 prevents clinical signs of EAE

All placebo-treated animals developed the typical clinical signsof EAE and were euthanized between days 12 and 27 (Fig. 3A,Table I), which is in line with previously reported data in this model(14, 22). None of the FR104-treated animals displayed any signs ofEAE, not even after the last placebo animal had been euthanized;resulting in significantly prolonged survival (p = 0.0005) (seeFig. 3A). It was decided to euthanize the FR104-treated animals ataround day 42, which was 3 wk after the last FR104 injection.Serum samples collected after the last FR104 dosing containedgradually declining levels of FR104, but these were still between 16and 82 mg/ml on the day of sacrifice (see below).

Immune responses to rhMOG

Data obtained in an equivalent nonhuman primate EAE model, therhMOG-induced model in marmosets, show that the core patho-genic T cells are derived from effector memory cells present in thenatural repertoire (32). Thus, we tested cellular and humoral im-mune responses against rhMOG to determine whether blockade ofT cell costimulation via CD28 with FR104 might affect the gen-eration of pathogenic cells.

Mononuclear cells. We monitored anti-rhMOG proliferativeresponses of PBMCs longitudinally at a weekly basis (data notshown) and at necropsy, together with responses by MNC isolatedfrom spleen and axillar lymph nodes. MNC proliferation againstrhMOG was observed in ALN (Fig. 3B) in five of six vehiclecontrol–treated animals. Only in monkey P5 we could not detecta positive anti-rhMOG response. Positive responses in PBMCcollected longitudinally (data not shown) and at necropsy(Fig. 3B) and in splenic MNC (Fig. 3B) were less frequentlyobserved. In FR104-treated animals MNC responses against

FIGURE 1. In vitro assessment of the effects of FR104 on Treg function and on an in vitro anti-rhMOG recall response. (A) CD4+CD252 cells were

stimulated with platebound anti-CD3 (1 mg/ml). This response was set at 100% (first column, ▴). Addition of FR104 (10 mg/ml) reduces the response with

436 11%. This response was also set at 100% (first column, N). Addition of CD4+CD25+ to the CD4+CD252 cells at 1:1 ratio (third column), abrogated the

response completely, as described previously (21). Addition of FR104 to the coculture did not affect the suppression by CD25+ cells. The response in

cultures of CD252 cells containing the equivalent amount of cells as present in cocultures was not affected, precluding exhaustion of medium as an

explanation for the reduced response in the CD252/CD25+ cocultures. (B) Stored MNC (n = 8) from monkeys with EAE and a positive anti-rhMOG

response from previous experiments (15, 22) were thawed and stimulated with rhMOG, resulting in SI with a median of 10, ranging from 3 to 50. Addition

of FR104 (10 mg/ml) or CTLA4-Ig (10 mg/ml) led to reduction in proliferation to ∼25%.

FIGURE 2. Serum levels and immunogenicity of FR104. Arrows indi-

cate i.v. injections of FR104. (A) Concentration of FR104 in serum from

individual animals expressed as micrograms per milliliter. (B) IgG ADA

serum levels in FR104-treated animals. IgM ADAwere negative. IgM and

IgG ADA were not found in the placebo-treated monkeys.

The Journal of Immunology 1457

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

rhMOG were below the threshold value (SI . 3) in all analyzedcompartments and all time points, except for monkey F3, whichhad some low positive rhMOG responses in PBMC from day 21onward (data not shown) and in PBMC, spleen, and ALN atnecropsy (Fig. 3B). In the ALN the stimulation indices weresignificantly (p = 0.0108) lower in the FR104 group (2.42 60.595) as compared with the placebo group (10.4 6 2.32).

Abs. High serum levels of anti-rhMOG IgM and IgG were detectedin all placebo animals (Fig. 3C, 3D). Already on day 11 the IgMand IgG serum levels in FR104-treated monkeys were signifi-

cantly (both p = 0.0022) lower (126.1 6 48.85 and 14.02 6 5.799AU, respectively) than in the placebo group (1814 6 396.9 and1773 6 352.1 AU, respectively). Despite a moderate increase,suppression of IgM and IgG Ab production by FR104 persistedthroughout the whole observation period.

Effects on cerebral inflammation and demyelination

Formalin-fixed hemispheres were processed for histological ex-amination. Three tissue blocks were excised from correspondingregions of each hemisphere. Representative sections were prepared

FIGURE 3. Summarized clinical and immunological EAE parameters. Rhesus monkeys were treated with placebo or FR104 just prior to (day 0) and at 7,

14, or 21 d after immunization (arrows). (A) Significantly prolonged (p = 0.0005) disease-free survival during the 42-d observation period. Although all

placebo-treated monkeys were sacrificed before day 28 with severe clinical EAE, monkeys from the FR104-treated group were sacrificed without overt

clinical signs at day 42. (B–D) Treatment with FR104 caused reduced induction of anti-MOG cellular and humoral immune responses. (B) Proliferative anti-

rhMOG responses of PBMC, splenocytes, and ALN from each animal. (C and D) Anti-rhMOG IgM (C) and IgG (D) serum levels in placebo-treated animals

(solid lines) and in FR104-treated animals (dotted lines).

Table I. Overview of individual animal data and EAE scores

Group

Animal

Age (y) Weight (kg) Sex

EAE Onset Euthanasia

ID Code Day EAE Score Day EAE Score

Placebo R05045 P1 8.1 10.74 M 12 0.5 12 5R06052 P2 7.1 7.68 M 12 0.5 13 5R07035 P3 6.1 10.02 M 16 2 16 5R06088 P4 7.0 11.43 M 17 0.5 18 5R08043 P5 5.1 8.05 M 21 2.5 21 5R06030 P6 7.2 12.22 M 26 0.5 27 2.5

Average 6.8 10.0FR104 R06021 F1 7.2 7.76 M — — 41 0

R06087 F2 7.0 9.89 M — — 41 0R04023 F3 9.2 10.75 M — — 42 0R07002 F4 6.3 9.77 M — — 42 0R07091 F5 6.0 9.14 M — — 43 0R08029 F6 5.2 7.13 M — — 43 0

Average 6.8 9.1

1458 CD28 BLOCKADE PREVENTS EAE IN RHESUS MONKEYS

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

FIGURE 4. Effects of FR104 on cerebral inflammation and demyelination. Brain sections from placebo and FR104-treated animals were stained for

myelin (left panels, LFB/PAS), T cells (middle panels, CD3 staining), and recently immigrated macrophages (right panels, MRP14). From both groups, two

animals are presented, one with the least and one with the most severe pathology. (A–C) White matter area of animal P1 (least (Figure legend continues)

The Journal of Immunology 1459

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

and examined for the presence and intensity of demyelination(Fig. 4, LFB/PAS panels) and inflammation.In the placebo group, all animals showed inflammation in all

areas investigated. Inflammation ranged from moderate numbersof perivascular infiltrates to many large perivascular infiltrates

with additional infiltration deep in the parenchyma. These lesionsconsisted mainly of T cells and infiltrating macrophages (Fig. 4,CD3 and MRP14 panels). In the FR104 group, five of sixanimals did not reveal inflammatory infiltrates. The other ani-mal of this group only showed some small perivascular infil-

affected). LFB/PAS staining shows some perivascular demyelination but no confluent plaque. Staining for CD3 and MRP14 show moderate numbers of

T cells and macrophages. Rectangles indicate areas enlarged in the insert. (D–F) White matter area of animal P5 (most severely affected). The depicted

section contains large demyelinating (necrotic) lesions. Stainings for CD3 show moderate numbers of T cells and large numbers of MRP14+ myeloid

cells representing macrophages and granulocytes that are abundant in the lesions. The inset shows an area (blood vessel) at the border of the lesion with

MRP14+ macrophages. (G–I) White matter of animal F3 (least affected). This animal showed no demyelination and no infiltration of T cells and

macrophages. The inserts shows a small area next to a blood vessel without any infiltrating cells. (J–L) White matter of animal F1 (most severely

affected). This animal showed some infiltration of T cells but absence of demyelination and infiltrating macrophages. The inserts show some infiltrating

T cells but the absence of infiltrating macrophages. (A, D, G, and J) Scale bar, 500 mm. Arbitrary scores for infiltration (M) and percentage demye-

lination of the total gray and white matter areas (N) per group. Only infiltration is significantly less in the FR104 group (p = 0.0022) whereas demyelination is

not (p = 0.0606). Arrows marked with an asterisk (A and D) indicate demyelinated areas. Areas marked with GM (A, B, G, and H) indicated gray matter areas in

the white matter of the lower part of the corpus callosum. In rhesus monkeys, unlike in rodents, gray matter areas are embedded within the white matter tracts of

the tempolateral corpus callosum.

FIGURE 5. Treatment with FR104 modulates the expression in the brain of mRNA for inflammation markers, as analyzed with quantitative PCR.

Expression levels of mRNA was normalized against the household gene ABL. Relative concentrations of mRNA levels of investigated genes that were

below the detection limit (IL-6: animals P6 and F2; CCR7: animal F4) were set to 0.

1460 CD28 BLOCKADE PREVENTS EAE IN RHESUS MONKEYS

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

trates. The severity of inflammation was significantly less inFR104-treated monkeys than in the placebo-treated group(p = 0.0022; Fig. 5M). Demyelination was observed in four ofsix animals from the placebo groups. Demyelination in thisplacebo group ranged from small to moderate demyelinatingrims around perivascular infiltrates (three animals) to the pres-ence of large demyelinating (necrotic) plaques as seen in oneanimal. These large demyelinating/necrotic plaques besides

many macrophages also contained large numbers of polymor-phonuclear granulocytes. No demyelination was observed in anyof the FR104-treated animals, but because of the absence ofdemyelination in the analyzed sections of the placebo animalsP2 and P6, this is only trending toward significance (p = 0.0606;Fig. 4N).Only small-sized perivascular cuffs of inflammatory cells were

detected in the spinal cords of only two animals (P2 and P2), and in

FIGURE 6. Parallel analysis of CSF and serum collected at necropsy. (A) Absolute leukocyte counts in the CSF. In the same CSF samples, anti-rhMOG IgM (B)

and IgG (C) levels were recorded. All parameters illustrate the profound suppressive effect of FR104. (D) Cytokine and chemokine levels were measured in the

serum (CSF see Table II). Serum of placebo animals contained significantly less IL-1b, CCL2, CCL11, and CCL22 as compared with FR104-treated animals. For

CCL2, this difference is due to a significant decrease in necropsy sera in placebo animals. For CCL11, this is due to a significant increase in FR104-treated animals.

The Journal of Immunology 1461

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

none of the animals spinal cord demyelination was found (data notshown).

Quantitative PCR of brain specimens

The corpus callosum was sampled for RT-PCR analysis as in therhesus monkey EAE model lesions are commonly detected in thisregion. Expression levels of mRNA for CD3 (T cells), CD19 (B cells),CD63 (neutrophils) and a series of T cell homing and activation

markers (CD28, CTLA4, CD25, andCCR7) and cytokines (IL-6, IL-8,and IFN-g) were assessed and expressed relative to the expression ofABL mRNA for all animals (Fig. 5). Expression levels relative toGAPDH gave similar results (data not shown). Similar expressionlevels of CD3 and CD19 mRNA were detected in both groups,but, consistent with histological data, the relative expression of CD63as a marker for neutrophils was significantly higher in placebo-treated animals. Expression levels of mRNA for a number of

Table II. Luminex analysis of cytokines and chemokines in necropsy CSF samples

Placebo Group Median (Min – Max) FR104 group Median (Min – Max) p Value

Median Placebo – Median FR104 $ 25 pg/ml; p # 0.05

IL-2 228 (79.3–299) 0.00 (0.00–0.00) 0.0022IL-6 .15,140* (n = 6) 5.61 (4.85–6.25) 0.0022IL-12 117 (21.6–215) 0.00 (0.00–0.00) 0.0022IL-15 29.7 (15.2–76.5) 0.00 (0.00–0.00) 0.0022IFN-g 37.7 (2.42–257) 0.00 (0.00–0.00) 0.0022IL-1RA 6,875 (625–15663) 8.75 (0.00–46.9) 0.0022G-CSF 68.7 (21.3–99.0) 12.0 (2.73–36.3) 0.0130MIF 122 (71.1–200) 34.2 (18.0–99.9) 0.0152CCL2 (MCP-1) 2,542 (674–6720) 123 (89.9–206) 0.0022CXCL10 (IP-10) 675 (145–3657) 2.79 (1.03–3.85) 0.0022CXCL11 (I-TAC) 43.7 (14.7–526) 3.36 (195–4.78) 0.0022

Median Placebo – Median FR104 , 25 pg/ml; p # 0.05

IL-1b 2.68 (1.38–5.51) 0.56 (0.52–0.62) 0.0022IL-4 15.0 (14.1–15.9) 13.1 (12.2 -13.1) 0.0022IL-5 8.21 (5.77–9.80) 5.12 (4.48–5.33) 0.0022IL-10 9.78 (6.29–17.2) 3.19 (2.94–3.44) 0.0022FGF-basic 18.6 (13.7–33.7) 0.00 (0.00–0.00) 0.0022GM-CSF 1.25 (0.78–1.63) 0.60 (0.55–0.64) 0.0022VEGF 13.0 (5.10–19.0) 0.00 (0.00–0.00) 0.0022CXCL8 (IL-8) 6.94 (5.56–12.7) 2.56 (1.08–4.55) 0.0022

Not Significantly Different (p . 0.05)

IL-17 0.00 (0.00–76.5) 0.00 (0.00–0.00) 0.4545TNF-a 0.00 (0.00–1.36) 0.00 (0.00–0.00) .0.9999EGF 0.00 (0.00–15.8) 0.00 (0.00–0.00) 0.4545HGF 12.1 (0.00–79.1) 0.00 (0.00–0.00) 0.0606CCL3 (MIP-1a) 0.00 (0.00–0.00) 0.00 (0.00–0.00) .0.9999CCL4 (MIP-1b) 0.00 (0.00–1.36) 0.00 (0.00–0.00) .0.9999CCL5 (RANTES) 14.0 (10.0–21.0) 14.8 (6.32–48.3) 0.9740CCL11 (Eotaxin) 4.84 (1.41–9.77) 2.15 (1.83–2.40) 0.0649CCL22 (MDC) 176 (0.00–473) 0.00 (0.00–0.00) 0.1818CXCL9 (MIG) 0.00 (0.00–42.5) 0.00 (0.00–0.00) 0.4545

*Above detection range.

Table III. Detection of virus in the spleen and brain tissue at necropsy

Animal

Spleen Brain

CMV LCV SA12 SA12 LCV SV40

P1 6,100 (6 678) 22,270 (6 1274) 1,019 (6 228) BLD BLD BLDP2 1,312 (6 60) 9,293 (6 79) 418 (6 44) BLD BLD BLDP3 1,678 (6 202) 9,810 (6 983) 854 (6 216) BLD BLD 93 (6 23)P4 BLD 708 (6 98) BLD BLD BLD 58 (6 10)P5 BLD 5,712 (6 163) BLD BLD BLD 344 (6 20)P6 BLD 1,368 (6 110) BLD BLD BLD BLDF1 369 (6 43) 7,898 (6 1863) BLD BLD BLD BLDF2 2,807 (6 902) 16,728 (6 3483) 398 (6 7) BLD BLD 392 (6 25)F3 83,093 (6 1465) 256,777 (6 7408) 28,265 (6 1243) BLD BLD BLDF4 603 (6 63) 4,844 (6 48) BLD BLD BLD 49 (6 5)F5 1,334 (6 104) 11,668 (6 701) 504 (6 37) BLD BLD 252 (6 37)F6 3,111 (6 239) 26,469 (6 3720) 847 (6 87) BLD BLD BLD

Viral copy number values are reported as copies per microgram DNA (6 SD) in each column. Assays were carried out in triplicate over a minimum of three experiments.BLD, below limit of detection or limit of quantification of the assay.

1462 CD28 BLOCKADE PREVENTS EAE IN RHESUS MONKEYS

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

T cell–associated genes (CTLA-4, CD25, and CCR7) as well asthe neutrophil-chemoattractant CXCL8 (IL-8) and the proinflam-matory cytokine IL-6 were significantly elevated in placebo-treatedanimals.

FR104 mitigates the inflammatory profile in the CSF

CSF samples from rhesus monkeys developing EAE contain ele-vated leukocyte numbers (15). The absence of leukocytes in theCSF of animals treated with FR104 supports the conclusion thatneuroinflammation was suppressed in these animals (Fig. 6A).After centrifugation for removal of cells, CSF samples were

analyzed for the presence of anti-rhMOG IgM and IgG Abs. Thesewere clearly detectable in the placebo-treated monkeys, but theywere significantly reduced in monkeys treated with FR104(Fig. 6B, 6C). IgG Abs were found in all placebo animals (range,31.803–249.0 AU), whereas these were absent in normal CSF(data not shown). IgG Abs were undetectable in the FR104-treatedanimals (range, 0.000–2.100 AU; p = 0.0022). IgM titres are,because of some background staining, not always negative innormal CSF, but elevated levels were found in some but not allplacebo animals (range, 8.500–310.9 AU). In the FR104 group,IgM levels were significantly (p = 0.0238) reduced (range, 0.000–20.90 AU).The CSF was also analyzed by Luminex multiplex technology

for the presence of cytokines and chemokines. An inflammatoryprofile was evident in all placebo animals. Of the 29 testedcytokines and chemokines, 19 were significantly elevated inplacebo-treated animals as compared with monkeys treated withFR104 (Table II). Of these 19 significantly elevated cytokines/chemokines, some had only a low difference between medians

and the biological relevance of these significant, but low-leveldifferences is questionable. Using an arbitrarily chosen cutoffdifference between medians of $25 pg/ml, we found higher levelsin placebo-treated than FR104-treated monkeys for 11 cytokinesand chemokines: IL-2, IL-6, IL-12, IL-15, IFN-g, IL-1R antago-nist (IL-1RA), G-CSF, MIF, CCL2, CXCL10, and CXCL11. In-terestingly, IL-17 was detected in only two of the six placebo-treated animals, and TNF-a was not detected in any of them norwere IL-17 and TNF-a detected in any of the FR104-treatedanimals (Table II). CSF samples taken prior to EAE inductionwere not available from the monkeys in the current study. How-ever, in four pre-EAE induction CSF samples available froma previously reported study (15), the two most prominentlyexpressed cytokines/chemokines IL-6 and CCL2 could not bedetected (data not shown), further indicating that the high levels ofcytokines and chemokines found in the CSF of the animals withEAE, were not already present before EAE induction, but are trulya result of the inflammation in the brain.

FR104 affects cytokine and chemokine levels in the blood

Cytokines and chemokines were also determined in sera collectedprior to EAE induction and at necropsy. This enabled us to calculateserum/CSF ratios. From the 11 cytokines and chemokines in theCSF that were higher (difference between medians of$ 25 pg/ml)in placebo animals, IL-2, IL-6, IL-12, IL-15, CCL2, and CXCL10were significantly higher in the CSF than in the serum (notshown), indicating that these factors might be locally produced inthe brain and had not leaked into the CSF from the blood. IL-10,FGF-basic and VEGF were also significantly higher in the CSF, ascompared with the serum.A comparison of the cytokine and chemokine levels in sera

collected at necropsy from all animals, revealed significant dif-ferences between the two groups for IL-1b, CCL2, CCL11, andCCL22 (Fig. 6D). Interestingly, the levels of all these cytokineswere lower in the placebo animals as compared with the FR104-treated animals. Comparison with pre-EAE induction sera indi-cated that for CCL2 this was due to decreased levels in the nec-ropsy sera of EAE animals compared with preinduction levels. ForCCL11, this was due to an increase in necropsy sera of FR104-treated animals as compared with pre-EAE induction sera.

FR104 and reactivation of existing endogenous viral infectionsin treated animals

A potential adverse effect of strong immunosuppression is therecrudescence of latent virus infection. To test whether the treat-ment with FR104 impairs the capacity of the animals to controlchronic latent virus infection, we examined by quantitative PCRblood, spleen, and brain for expression of a number of virusesrelevant to human infections/complications to see whether reac-tivation was induced by the FR104 treatment. CMV, LCV, SA12,SV40, and HEV were all assayed in blood at day 27 and at days 7and 14 and up to day 21 post-EAE induction and at euthanasia aswell as in spleen and brain tissue sampled at euthanasia.The polyomavirus SA12 was detected in the blood of F5 only at

euthanasia, albeit at very low levels (871 6 302 mean copies/mgDNA). SA12 was also detected in the spleen tissue at necropsy butnot in brain (Table III). HEV was not detected in blood or feces inany of the animals in either group (data not shown).CMV virus levels in the blood were at the limit of detection and

therefore considered insignificant in both the treated and controlgroups. However, virus was detected in the spleens of animals inboth groups confirming presence of the virus and that lack ofreactivation was not due to the lack of prior exposure (Table III).No CMV DNA was detected in the brain tissue.

FIGURE 7. LCV reactivation response in whole blood in placebo (A)

and FR104-treated (B) animals. Arrows indicate points at which FR104

was administered at days 0, 7, 14, and 21.

The Journal of Immunology 1463

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

DNA from the polyomavirus SV40 was detectable in brain tissuefrom three of six placebo-treated animals (P3, P4, and P5) and three ofsix FR104-treated animals (F2, F4, and F5) but themean copy numberbetween the groups was found not to be significant (Table III).The gammaherpesvirus macaque LCV, the rhesus monkey repre-

sentative of EBV, was detected in the blood of animals F4 andP3 at day 27 (Fig. 7). In the placebo-treated group, low levels ofviral DNAwere detected continuously, until euthanasia (Fig. 7A).Virus levels for the FR104-treated group appeared to increase afterday 7 peaking at day 14 then decreasing at day 21. In animals F1,F2, F3, and F4, levels increased again after the fourth injection ofFR104 at day 21 post-EAE induction. Levels of virus remainedlow in animals F5 and F6. LCV was detectable in the spleen also(Table III). Interestingly, animal F3 had elevated levels in thetissues for CMV, LCV, and SA12.

DiscussionDespite demonstrated effects in RA and promising results in MS,abatacept has the theoretical disadvantage to block both inter-actions of CD80/CD86 with CD28 and with CTLA4, therebypotentially abrogating downmodulatory effects of CTLA-4. Incontrast, CD28 blockade by specific mAbs will abrogate immuneactivation by interfering with the interaction of CD80/CD86 withCD28 but will not affect the interaction of CD80/CD86 withCTLA4. However, anti-CD28 mAbs have been notorious for theirpotential (super)agonistic effects on T cells (for review, see Ref. 9).Modified versions without cross-linking activity may be usefulantagonists of T cell–APC interaction without T cell stimulation(9). In this paper, we report the results of a preclinical efficacystudy of FR104, a novel monovalent, PEGylated mAb againsthuman CD28, which presents no agonistic and superagonisticproperties in vitro (12) and in vivo (11).PEGylated FR104 did not bind marmoset CD28 but bound

rhesus monkey CD28 with equivalent affinity as human CD28 (datanot shown). We report in this paper that FR104 did not affect thefunction of rhesus monkey naturally occurring Tregs (Fig. 1A)and that it inhibited a rhMOG-induced ex vivo recall response(Fig. 1B). These data formed a solid basis for subsequent efficacytesting of FR104 in vivo in the rhesus monkey EAE model. Thismodel recapitulates aspects of the human neuroinflammatorydisease MS but is also used as a generic autoimmune diseasemodel. This implies that promising results obtained in the EAEmodel may have broader relevance than only for MS.The validity of CD28 signaling as target of immunotherapy in the

EAE model has been firmly established in the mouse EAE modelbecause CD28 knockout mice are either resistant to EAE inductionor the disease is less severe (4–6, 33). Moreover, blocking of CD28signaling by antagonists of CD80/CD86 (7, 33) or CD28 directly (8)prevented EAE induction. Our studies in the nonhuman primateEAE model revealed an essential difference between the EAE modelin inbred/SPF laboratory mice and outbred nonspecific pathogen-free primates, namely that the encephalitogenic T cells in the latterseem to be effector memory cells (34). Activation of these effectormemory T cells is considered less sensitive to costimulation blockade(35, 36). These considerations warranted us to test the efficacy ofFR104 in the nonhuman primate rhMOG/CFA EAE model.B cells also seem to have a crucial role in the nonhuman primate

EAE model. We recently reported that in rhesus monkeys im-munized with rhMOG formulated with IFA, instead of the CFAused in the current study, clinical EAE is only observed in caseswith the highest anti-rhMOG Ab levels (22). More detailedanalysis has been performed in another primate EAE model,i.e., in the common marmoset. In that model we found a dual roleof B cells, which depended on the disease stage (37). B cells

contribute to the early EAE initiation phase by producing anti-rhMOG Abs. In late-stage disease, they act as APC for a patho-genic subset of cytotoxic T cells (37).EAE was induced with rhMOG in CFA, resulting in acute onset

of clinical EAE in all placebo animals necessitating euthanasiabecause of fast progression to severe EAE symptoms between days12 and 27 (Fig. 3A). This is in line with previous observations inthis model (14, 15), and it confirms that the EAE model in rhesusmonkeys is characterized by an early onset and a very rapid dis-ease progression.It was demonstrated in rodents that CD28 plays a role in the

induction of EAE but that primed cells can promote EAE in theabsence of CD28-mediated signaling (5). Therefore, treatmentwith FR104 was initiated on day 0 to avoid that treatment wasinitiated at the stage where the pathogenic process could notbe reverted by FR104 and thereby missing a possible effect ofFR104. None of the FR104-treated animals developed EAE,leading to significantly prolonged disease-free survival in thisgroup (p = 0.0005; Fig. 3A). Treatment with FR104 was stoppedwell after the last animal in the placebo group had been eutha-nized. We chose to euthanize FR104-treated animals beforeFR104 levels had waned completely (Fig. 2A) because we wantedto examine the effect of FR104 treatment on (the absence of) CNSpathology, which may have disappeared in the case of rebounddisease.FR104 treatment suppressed peripheral T cell activation against

rhMOG in vivo (Fig. 3B). Although anti-rhMOG IgG serum levelswere detectable, these were substantially lower than those ob-served in placebo animals. Anti-rhMOG IgM Abs were notdetected at all. Furthermore, levels of IgM-ADA were negative,and IgG-ADA were very low in five of six animals. Only monkeyF3 had high IgG-ADA. These combined observations hint ata possible inhibitory effect of FR104 on CD28-mediated T cellhelp for Ig isotype switching, as observed in mouse models (38,39) and as class-switching was also affected by abatacept (40).In comparison with our previous study (15), we found less

demyelination. The reason for this is not an on average lowerpercentage demyelinated area but merely results from inclusionof both white and gray matter in this study as compared withfocusing on white matter only in our previous study. Probablybecause of the absence of demyelination in two of the placebo-treated animals, the difference between the groups for demyelin-ation shows only a trend toward significance (p = 0.0606).Interestingly, immunohistology analysis revealed presence of

CD3+ cell infiltration in all placebo animals but only in onemonkey of the FR104-treated group (F1, Fig. 4K, 4M). However,CD3 mRNA expression in the brain was found in all FR104 ani-mals, and the relative levels did not differ between the groups(Fig. 5, upper left panel). The cause of this discrepancy is notknown but could be due to the sampling procedure because tissuefor mRNA isolation was collected from the right hemisphere andtissue for histopathology from the left hemisphere, or it could bedue to regulation on the translational level. However, more con-sistent with the clinical data, we observed that mRNA levels formarkers for T cells (CD28) and of T cell activation (CTLA4 andCD25) for the T cell secondary lymphoid organ homing markerCCR7 and the Th1 cytokine IFN-g were significantly lower inFR104-treated animals as compared with the placebo group (Fig. 5).Expression of CD19 mRNA was, like CD3 mRNA, also not

different between the groups. B cell numbers were not quantified byhistopathology because these were previously not found in lesionsof untreated controls (15), which seems in contrast with thepresence of CD19 mRNA in the brains of placebo and FR104animals. The strikingly elevated expression of mRNA for CD6,

1464 CD28 BLOCKADE PREVENTS EAE IN RHESUS MONKEYS

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

a marker of neutrophils, and of mRNA for CXCL8 (IL-8), a neu-trophil chemoattractant, is consistent with the abundant presenceof this cell type in brain lesions in the rhesus monkey rhMOG/CFA EAE model. Because FR104 prevented inflammation in thebrain, mRNA for these markers remained significantly lower inthe treated animals as compared with placebo-treated animals.Interestingly, although IL-6 mRNA levels were significantlyhigher in the placebo-treated group, the levels were very low, toeven being undetectable in animals P6 and F2. However, this wasthe cytokine most abundantly present in the CSF.The mRNA levels of indicated immune markers were also

assayed in the lymphoid compartment (i.e., spleen, ALN, andblood). The significantly higher CD63 mRNA level in blood ofplacebo animals is consistent with the observed elevated bloodneutrophil counts at necropsy in animals with EAE in this study(data not shown) and unpublished results from other rhesus monkeyEAE studies by our group (15, 22) as well as published resultsin the myelin/CFA EAE model in rhesus monkeys (41). CD63mRNA was also elevated in ALN, although only trending towardsignificance, but not in the spleen. Neutrophils play a role incertain rodent EAE models as well (42–44) but their role in MShas not yet been thoroughly investigated.Elevated MNC numbers were detected in the CSF of placebo-

treated animals. This is reminiscent of observations in patientswith acute disseminated encephalomyelitis (ADEM) (45, 46).Analysis of the cytokine and chemokine levels in CSF revealedvery high amounts of IL-6, IL-1RA, CCL2, and CXCL10 of theplacebo-treated animals, whereas levels were very low or absentin the CSF of the FR104-treated animals. The presence of thesecytokines and chemokines has been observed in MS patients, aswell as in patients with ADEM (47, 48).Biologics have been documented to activate viruses of concern

(17–19), and viral reactivation after immunosuppression has alsobeen documented in nonhuman primates (49–51). To determinewhether FR104 led to reactivation of latent viral pathogens, animalswere assayed for the presence of herpesviruses, polyomaviruses,and HEV. Overall, no significant reactivation was observed forrhesus CMV or the polyomaviruses, although the viruses weredetected in the spleen. HEV was undetectable. However, someevidence of LCV reactivation was seen in animals of both groupswith a higher viremia observed in those treated with FR104. EBV,the human representative within the genus LCV, is a known pre-dictor of malignancy development (52) and was shown to be as-sociated with acute encephalitis during treatment with abatacept(19). In this study, even though viral levels increased in the blood,there was no evidence of virus in the brain tissue, and high levels ofLCV in spleen tissue were only observed in one animal. Given theshort duration of these experiments and because the levels in bloodwere extremely variable between animals, it is not possible toconvincingly demonstrate whether LCV is indeed truly reactivatedin this model and poses a risk for use of FR104.In conclusion, we report that i.v. administration of FR104 is

well tolerated and shows promising clinical effects in the rhMOG-induced rhesus monkey EAE model. The results confirm the centralpathogenic role of T cells in the model. Moreover, the preclinical dataare promising in view of a future therapy for patients with autoimmuneneuroinflammatory disease. Regarding the clinical perspectives, theseobservations are encouraging for future treatment of ADEM patientsand possibly also for MS, although the effect on on-going chronicinflammation needs to be established. However, the observation thatFR104 has a similar suppressive effect as CTLA4-Ig on a recallimmune response against rhMOG in vitro (Fig. 1B) offers an en-couraging perspective for further exploring the immunosuppressiveproperties of FR104 in MS patients. A point of caution is the possible

exacerbation of macaque LCV, which indicates that it may be wise tomonitor patients treated with FR104 for recrudescence of latent EBV.

AcknowledgmentsWe thank S. Anwar Jagessar for the production of the rhMOG, Nikki van Driel

for technical assistance, and Henk van Westbroek for help with the artwork.

DisclosuresC.M., N.P., and B.V. are current employees of and own stock or equity

from Effimmune.

References1. Tanaka, Y., S. Kubo, H. Yamanaka, K. Amano, S. Hirata, E. Tanaka,

H. Nagasawa, H. Yasuoka, and T. Takeuchi. 2014. Efficacy and safety of aba-tacept in routine care of patients with rheumatoid arthritis: Orencia(�) as Bio-logical Intensive Treatment for RA (ORBIT) study. Mod. Rheumatol. 24: 754–762.

2. Viglietta, V., K. Bourcier, G. J. Buckle, B. Healy, H. L. Weiner, D. A. Hafler,S. Egorova, C. R. Guttmann, J. R. Rusche, and S. J. Khoury. 2008. CTLA4Igtreatment in patients with multiple sclerosis: an open-label, phase 1 clinical trial.Neurology 71: 917–924.

3. National Institutes of Health. Clinical Trials database. Available at: http://www.clinicaltrials.gov/ct2/show/study/NCT01116427. Accessed: December 29, 2014.

4. Chang, T. T., C. Jabs, R. A. Sobel, V. K. Kuchroo, and A. H. Sharpe. 1999.Studies in B7-deficient mice reveal a critical role for B7 costimulation in bothinduction and effector phases of experimental autoimmune encephalomyelitis. J.Exp. Med. 190: 733–740.

5. Chitnis, T., N. Najafian, K. A. Abdallah, V. Dong, H. Yagita, M. H. Sayegh, andS. J. Khoury. 2001. CD28-independent induction of experimental autoimmuneencephalomyelitis. J. Clin. Invest. 107: 575–583.

6. Oliveira-dos-Santos, A. J., A. Ho, Y. Tada, J. J. Lafaille, S. Tonegawa,T. W. Mak, and J. M. Penninger. 1999. CD28 costimulation is crucial for thedevelopment of spontaneous autoimmune encephalomyelitis. J. Immunol. 162:4490–4495.

7. Khoury, S. J., E. Akalin, A. Chandraker, L. A. Turka, P. S. Linsley,M. H. Sayegh, and W. W. Hancock. 1995. CD28-B7 costimulatory blockade byCTLA4Ig prevents actively induced experimental autoimmune encephalomy-elitis and inhibits Th1 but spares Th2 cytokines in the central nervous system.J. Immunol. 155: 4521–4524.

8. Perrin, P. J., C. H. June, J. H. Maldonado, R. B. Ratts, and M. K. Racke. 1999.Blockade of CD28 during in vitro activation of encephalitogenic T cells orafter disease onset ameliorates experimental autoimmune encephalomyelitis.J. Immunol. 163: 1704–1710.

9. Poirier, N., G. Blancho, and B. Vanhove. 2012. CD28-specific immunomodu-lating antibodies: what can be learned from experimental models? Am.J. Transplant. 12: 1682–1690.

10. Suntharalingam, G., M. R. Perry, S. Ward, S. J. Brett, A. Castello-Cortes,M. D. Brunner, and N. Panoskaltsis. 2006. Cytokine storm in a phase 1 trial ofthe anti-CD28 monoclonal antibody TGN1412. N. Engl. J. Med. 355: 1018–1028.

11. Poirier, N., C. Mary, N. Dilek, J. Hervouet, D. Minault, G. Blancho, and B. Vanhove.2012. Preclinical efficacy and immunological safety of FR104, an antagonist anti-CD28 monovalent Fab9 antibody. Am. J. Transplant. 12: 2630–2640.

12. Poirier, N., C. Mary, S. Le Bas-Bernardet, V. Daguin, L. Belarif, M. Chevalier,J. Hervouet, D. Minault, S. Ville, V. Charpy, et al. 2014. Advantages of Papioanubis for preclinical testing of immunotoxicity of candidate therapeutic an-tagonist antibodies targeting CD28. MAbs 6: 697–707.

13. ’t Hart, B. A., and L. Massacesi. 2009. Clinical, pathological, and immunologicaspects of the multiple sclerosis model in common marmosets (Callithrix jac-chus). J. Neuropathol. Exp. Neurol. 68: 341–355.

14. Kerlero de Rosbo, N., H. P. Brok, J. Bauer, J. F. Kaye, B. A. ’t Hart, and A. Ben-Nun. 2000. Rhesus monkeys are highly susceptible to experimental autoimmuneencephalomyelitis induced by myelin oligodendrocyte glycoprotein: character-isation of immunodominant T- and B-cell epitopes. J. Neuroimmunol. 110: 83–96.

15. Haanstra, K. G., S. O. Hofman, D. M. Lopes Estevao, E. L. Blezer, J. Bauer,L. L. Yang, T. Wyant, V. Csizmadia, B. A. ’t Hart, and E. R. Fedyk. 2013.Antagonizing the a4b1 integrin, but not a4b7, inhibits leukocytic infiltration ofthe central nervous system in rhesus monkey experimental autoimmune en-cephalomyelitis. J. Immunol. 190: 1961–1973.

16. ’t Hart, B. A., J. Bauer, H. P. Brok, and S. Amor. 2005. Non-human primatemodels of experimental autoimmune encephalomyelitis: variations on a theme.J. Neuroimmunol. 168: 1–12.

17. Bellizzi, A., C. Nardis, E. Anzivino, D. Rodıo, D. Fioriti, M. Mischitelli,F. Chiarini, and V. Pietropaolo. 2012. Human polyomavirus JC reactivation andpathogenetic mechanisms of progressive multifocal leukoencephalopathy andcancer in the era of monoclonal antibody therapies. J. Neurovirol. 18: 1–11.

18. Green, M., and M. G. Michaels. 2013. Epstein-Barr virus infection and post-transplant lymphoproliferative disorder. Am. J. Transplant. 13(Suppl. 3): 41–54,quiz 54.

19. Nakajima, H., A. Takayama, T. Ito, and T. Yoshikawa. 2013. Acute encepha-lomyelitis with multiple herpes viral reactivations during abatacept therapy. BMJCase Rep. pii: bcr2013009731.

The Journal of Immunology 1465

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

20. L€uhder, F., Y. Huang, K. M. Dennehy, C. Guntermann, I. M€uller, E. Winkler,T. Kerkau, S. Ikemizu, S. J. Davis, T. Hanke, and T. H€unig. 2003. Topologicalrequirements and signaling properties of T cell-activating, anti-CD28 antibodysuperagonists. J. Exp. Med. 197: 955–966.

21. Haanstra, K. G., M. J. van der Maas, B. A. ’t Hart, and M. Jonker. 2008.Characterization of naturally occurring CD4+CD25+ regulatory T cells in rhesusmonkeys. Transplantation 85: 1185–1192.

22. Haanstra, K. G., S. A. Jagessar, A. L. Bauchet, M. Doussau, C. M. Fovet,N. Heijmans, S. O. Hofman, J. van Lubeek-Veth, J. J. Bajramovic, Y. S. Kap,et al. 2013. Induction of experimental autoimmune encephalomyelitis withrecombinant human myelin oligodendrocyte glycoprotein in incompleteFreund’s adjuvant in three non-human primate species. J. Neuroimmune Phar-macol. 8: 1251–1264.

23. Quesgen Systems, Inc. Data management for clinical research. Available at:http://www.quesgen.com/SSSurvival.php. Accessed December 29, 2014.

24. Kerlero de Rosbo, N., M. Hoffman, I. Mendel, I. Yust, J. Kaye, R. Bakimer,S. Flechter, O. Abramsky, R. Milo, A. Karni, and A. Ben-Nun. 1997. Predom-inance of the autoimmune response to myelin oligodendrocyte glycoprotein(MOG) in multiple sclerosis: reactivity to the extracellular domain of MOG isdirected against three main regions. Eur. J. Immunol. 27: 3059–3069.

25. Brok, H. P., L. Boven, M. van Meurs, N. Kerlero de Rosbo, L. Celebi-Paul,Y. S. Kap, A. Jagessar, R. Q. Hintzen, G. Keir, J. Bajramovic, et al. 2007. Thehuman CMV-UL86 peptide 981-1003 shares a crossreactive T-cell epitope withthe encephalitogenic MOG peptide 34-56, but lacks the capacity to induce EAEin rhesus monkeys. J. Neuroimmunol. 182: 135–152.

26. Brok, H. P., J. Bauer, M. Jonker, E. Blezer, S. Amor, R. E. Bontrop, J. D. Laman,and B. A. ’t Hart. 2001. Non-human primate models of multiple sclerosis.Immunol. Rev. 183: 173–185.

27. Jagessar, S. A., P. A. Smith, E. Blezer, C. Delarasse, D. Pham-Dinh, J. D. Laman,J. Bauer, S. Amor, and B. ’t Hart. 2008. Autoimmunity against myelin oligo-dendrocyte glycoprotein is dispensable for the initiation although essential forthe progression of chronic encephalomyelitis in common marmosets. J. Neuro-pathol. Exp. Neurol. 67: 326–340.

28. World Health Organization Hepatitis E virus standard for RNA assay. PaulEhrlich Institute. Available at: http://www.pei.de/EN/information/license-applicants/standard-and-referencematerials/who/hepatitis-e-rna-nat/referencematerial-standard-hepatitis-e-rna-nat-node.html. Accessed December 29, 2014.

29. Cantalupo, P., A. Doering, C. S. Sullivan, A. Pal, K. W. Peden, A. M. Lewis, andJ. M. Pipas. 2005. Complete nucleotide sequence of polyomavirus SA12.J. Virol. 79: 13094–13104.

30. Dilek, N., N. Poirier, P. Hulin, F. Coulon, C. Mary, S. Ville, H. Vie,B. Clemenceau, G. Blancho, and B. Vanhove. 2013. Targeting CD28, CTLA-4and PD-L1 costimulation differentially controls immune synapses and functionof human regulatory and conventional T-cells. PLoS One 8: e83139.

31. Poirier, N., A. M. Azimzadeh, T. Zhang, N. Dilek, C. Mary, B. Nguyen,X. Tillou, G. Wu, K. Reneaudin, J. Hervouet, et al. 2010. Inducing CTLA-4-dependent immune regulation by selective CD28 blockade promotes regulatoryT cells in organ transplantation. Sci. Transl. Med. 2: 17ra10.

32. Jagessar, S. A., N. Heijmans, E. L. Blezer, J. Bauer, J. H. Blokhuis,J. A. Wubben, J. W. Drijfhout, P. J. van den Elsen, J. D. Laman, and B. A. Hart.2012. Unravelling the T-cell‑mediated autoimmune attack on CNS myelin ina new primate EAE model induced with MOG34-56 peptide in incomplete ad-juvant. Eur. J. Immunol. 42: 217–227.

33. Girvin, A. M., M. C. Dal Canto, L. Rhee, B. Salomon, A. Sharpe,J. A. Bluestone, and S. D. Miller. 2000. A critical role for B7/CD28 costimu-lation in experimental autoimmune encephalomyelitis: a comparative study us-ing costimulatory molecule-deficient mice and monoclonal antibody blockade.J. Immunol. 164: 136–143.

34. ’t Hart, B. A., B. Gran, and R. Weissert. 2011. EAE: imperfect but useful modelsof multiple sclerosis. Trends Mol. Med. 17: 119–125.

35. Adams, A. B., M. A. Williams, T. R. Jones, N. Shirasugi, M. M. Durham,S. M. Kaech, E. J. Wherry, T. Onami, J. G. Lanier, K. E. Kokko, et al. 2003.Heterologous immunity provides a potent barrier to transplantation tolerance.J. Clin. Invest. 111: 1887–1895.

36. Williams, M. A., T. M. Onami, A. B. Adams, M. M. Durham, T. C. Pearson,R. Ahmed, and C. P. Larsen. 2002. Cutting edge: persistent viral infectionprevents tolerance induction and escapes immune control following CD28/CD40blockade-based regimen. J. Immunol. 169: 5387–5391.

37. ’t Hart, B. A., S. A. Jagessar, K. Haanstra, E. Verschoor, J. D. Laman, andY. S. Kap. 2013. The primate EAE model points at EBV-infected B cells asa preferential therapy target in multiple sclerosis. Front. Immunol. 4: 145.

38. Lumsden, J. M., J. A. Williams, and R. J. Hodes. 2003. Differential requirementsfor expression of CD80/86 and CD40 on B cells for T-dependent antibodyresponses in vivo. J. Immunol. 170: 781–787.

39. Rau, F. C., J. Dieter, Z. Luo, S. O. Priest, and N. Baumgarth. 2009. B7-1/2(CD80/CD86) direct signaling to B cells enhances IgG secretion. J. Immunol.183: 7661–7671.

40. Scarsi, M., L. Paolini, D. Ricotta, A. Pedrini, S. Piantoni, L. Caimi, A. Tincani,and P. Airo. 2014. Abatacept reduces levels of switched memory B cells,autoantibodies, and immunoglobulins in patients with rheumatoid arthritis.J. Rheumatol. 41: 666–672.

41. van Lambalgen, R., and M. Jonker. 1987. Experimental allergic encephalomy-elitis in rhesus monkeys: I. Immunological parameters in EAE resistant andsusceptible rhesus monkeys. Clin. Exp. Immunol. 68: 100–107.

42. Aube, B., S. A. Levesque, A. Pare, E. Chamma, H. Kebir, R. Gorina,M. A. Lecuyer, J. I. Alvarez, Y. De Koninck, B. Engelhardt, et al. 2014. Neu-trophils mediate blood-spinal cord barrier disruption in demyelinating neuro-inflammatory diseases. J. Immunol. 193: 2438–2454.

43. Stoolman, J. S., P. C. Duncker, A. K. Huber, and B. M. Segal. 2014. Site-specificchemokine expression regulates central nervous system inflammation anddetermines clinical phenotype in autoimmune encephalomyelitis. J. Immunol.193: 564–570.

44. Simmons, S. B., D. Liggitt, and J. M. Goverman. 2014. Cytokine-regulatedneutrophil recruitment is required for brain but not spinal cord inflammationduring experimental autoimmune encephalomyelitis. J. Immunol. 193: 555–563.

45. Dale, R. C., C. de Sousa, W. K. Chong, T. C. Cox, B. Harding, and B. G. Neville.2000. Acute disseminated encephalomyelitis, multiphasic disseminated en-cephalomyelitis and multiple sclerosis in children. Brain 123: 2407–2422.

46. Schwarz, S., A. Mohr, M. Knauth, B. Wildemann, and B. Storch-Hagenlocher.2001. Acute disseminated encephalomyelitis: a follow-up study of 40 adultpatients. Neurology 56: 1313–1318.

47. Franciotta, D., E. Zardini, S. Ravaglia, G. Piccolo, L. Andreoni, R. Bergamaschi,A. Romani, E. Tavazzi, P. Naldi, M. Ceroni, and E. Marchioni. 2006. Cytokinesand chemokines in cerebrospinal fluid and serum of adult patients with acutedisseminated encephalomyelitis. J. Neurol. Sci. 247: 202–207.

48. Ishizu, T., M. Minohara, T. Ichiyama, R. Kira, M. Tanaka, M. Osoegawa,T. Hara, S. Furukawa, and J. Kira. 2006. CSF cytokine and chemokine profiles inacute disseminated encephalomyelitis. J. Neuroimmunol. 175: 52–58.

49. Maki, T., A. Carville, I. E. Stillman, K. Sato, T. Kodaka, K. Minamimura,N. Ogawa, A. Kanamoto, R. Gottschalk, A. P. Monaco, et al. 2008. SV40 in-fection associated with rituximab treatment after kidney transplantation innonhuman primates. Transplantation 85: 893–902.

50. Simon, M. A. 2008. Polyomaviruses of nonhuman primates: implications forresearch. Comp. Med. 58: 51–56.

51. Simmons, J. H. 2010. Herpesvirus infections of laboratory macaques.J. Immunotoxicol. 7: 102–113.

52. Gulley, M. L., and W. Tang. 2010. Using Epstein-Barr viral load assays to di-agnose, monitor, and prevent posttransplant lymphoproliferative disorder. Clin.Microbiol. Rev. 23: 350–366.

1466 CD28 BLOCKADE PREVENTS EAE IN RHESUS MONKEYS

by guest on April 13, 2017

http://ww

w.jim

munol.org/

Dow

nloaded from

Corrections

Haanstra, K. G., K. Dijkman, N. Bashir, J. Bauer, C. Mary, N. Poirier, P. Baker, L. Scobie, B. A. ’t Hart, and B. Vanhove. 2015. Selectiveblockade of CD28-mediated T cell costimulation protects rhesus monkeys against acute fatal experimental autoimmune encephalo-myelitis. J Immunol. 194: 1454–1466.

The eighth author’s name was omitted from the article. The corrected author and affiliation lines are shown below.

Krista G. Haanstra,* Karin Dijkman,* Noun Bashir,* Jan Bauer,† Caroline Mary,‡ Nicolas Poirier,‡ Paul Baker,x Claire L. Crossan,x

Linda Scobie,x Bert A. ’t Hart,*,{ and Bernard Vanhove‡,‖

*Biomedical Primate Research Centre, 2280 GH Rijswijk, the Netherlands; †Center for Brain Research, Medical University of Vienna,1090 Vienna, Austria; ‡Effimune SAS, 44035 Nantes, France; xGlasgow Caledonian University, Glasgow G4 0BA, United Kingdom;{Department of Neuroscience, University Medical Center, University of Groningen, 9713 GZ Groningen, the Netherlands; and ‖InstitutNational de la Sante et de la Recherche Medicale, Unite Mixte de Recherche 1064, 44093 Nantes, France

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1501150

Copyright � 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00

The Journal of Immunology