SMaGYERa

b

vc

vd

e

a

ARRAA

KPTTI

dLiabrg

A

h0

Vaccine 32 (2014) 5681–5689

Contents lists available at ScienceDirect

Vaccine

j our na l ho me page: www.elsev ier .com/ locate /vacc ine

ubcutaneous inverse vaccination with PLGA particles loaded with aOG peptide and IL-10 decreases the severity of experimental

utoimmune encephalomyelitisiuseppe Cappellanoa,∗, Abiy Demeke Woldetsadika, Elisabetta Orilieri a,ogesh Shivakumara, Manuela Rizzi c, Fabio Carniatod, Casimiro Luca Gigliotti a,lena Boggioa, Nausicaa Clementea, Cristoforo Comib, Chiara Dianzanie,enzo Boldorinia, Annalisa Chiocchetti a, Filippo Renòc, Umberto Dianzania

IRCAD and Department of Health Sciences, “A. Avogadro”, University of Eastern Piedmont “A. Avogadro”, via Solaroli 17, Novara, 28100, ItalyIRCAD and Department of Translational Medicine, Section of Neurology, “A. Avogadro”, University of Eastern Piedmont “A. Avogadro”,ia Solaroli 17, Novara, 28100, ItalyInnovative Research Laboratory for Wound Healing, Department of Health Sciences, University of Eastern Piedmont “A. Avogadro”,ia Solaroli 17, 28100 Novara, ItalyDepartment of Science and Technology, University of Eastern Piedmont “A. Avogadro”, viale Teresa Michel 11, 15121 ItalyDepartment of Drug Science and Technology, University of Turin, via Pietro Giuria 9, 10125 Torino, Italy

r t i c l e i n f o

rticle history:eceived 24 February 2014eceived in revised form 19 June 2014ccepted 8 August 2014vailable online 20 August 2014

eywords:LGAolerogenic vaccineREG

nterleukin-10

a b s t r a c t

“Inverse vaccination” refers to antigen-specific tolerogenic immunization treatments that are capableof inhibiting autoimmune responses. In experimental autoimmune encephalomyelitis (EAE), an animalmodel of multiple sclerosis (MS), initial trials using purified myelin antigens required repeated injec-tions because of the rapid clearance of the antigens. This problem has been overcome by DNA-basedvaccines encoding for myelin autoantigens alone or in combination with “adjuvant” molecules, such asinterleukin (IL)-4 or IL-10, that support regulatory immune responses. Phase I and II clinical trials withmyelin basic protein (MBP)-based DNA vaccines showed positive results in reducing magnetic resonanceimaging (MRI)-measured lesions and inducing tolerance to myelin antigens in subsets of MS patients.However, DNA vaccination has potential risks that limit its use in humans. An alternative approach couldbe the use of protein-based inverse vaccines loaded in polymeric biodegradable lactic-glycolic acid (PLGA)nano/microparticles (NP) to obtain the sustained release of antigens and regulatory adjuvants. The aim ofthis work was to test the effectiveness of PLGA-NP loaded with the myelin oligodendrocyte glycoprotein(MOG)35–55 autoantigen and recombinant (r) IL-10 to inverse vaccinate mice with EAE. In vitro experi-ments showed that upon encapsulation in PLGA-NP, both MOG35-55 and rIL-10 were released for severalweeks into the supernatant. PLGA-NP did not display cytotoxic or proinflammatory activity and werepartially endocytosed by phagocytes. In vivo experiments showed that subcutaneous prophylactic and

therapeutic inverse vaccination with PLGA-NP loaded with MOG35-55 and rIL-10 significantly amelioratedthe course of EAE induced with MOG35-55 in C57BL/6 mice. Moreover, they decreased the histopathologiclesions in the central nervous tiin splenic T cells in vitro. Thesebe an effective tool to treat aut

Abbreviations: BCA, bicinchoninic acid; CHOs, Chinese hamster ovary cells; CNS, cenynamic light scattering; EAE, experimental autoimmune encephalomyelitis; FDA, food aPS, lipopolysaccharide; MCF-7, human breast adenocarcinoma cell line; MBP, myelin basing; MOG, myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; MTT, 3-(4,5-dimegarose; NIRF, near-infrared fluorescence; NK, natural killer; NP, nano/microparticles; Ouffered saline; PDI, polydispersity index; PLGA, polymeric biodegradable lactic–glycolic aemitting form; RT, room temperature; s.c., subcutaneous injections; SEM, scanning electrowth factor; TNF, Tumor Necrosis Factor; TREG, regulatory T.∗ Corresponding author at: Interdisciplinary Research Center of Autoimmune Diseasesvogadro”, via Solaroli 17, I-28100 Novara, Italy. Tel.: +39 0321660658; fax: +39 0321620

ttp://dx.doi.org/10.1016/j.vaccine.2014.08.016264-410X/© 2014 Elsevier Ltd. All rights reserved.

ssue and the secretion of IL-17 and interferon (IFN)-� induced by MOG35-55

data suggest that subcutaneous PLGA-NP-based inverse vaccination mayoimmune diseases.

© 2014 Elsevier Ltd. All rights reserved.

tral nervous system; CTL, cytotoxic T lymphocytes; DCM, dichloromethane; DLS,nd drug administration; H&E, hematoxylin&eosin; IFN, interferon; IL, interleukin;

c protein; MFI-R, mean fluorescence intensity ratio; MRI, magnetic resonance imag-thylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; Ni-NTA, nickel-nitrilotriaceticD, optical density; PBMCs, peripheral blood mononuclear cells; PBS, phosphate-cid; PLP, proteolipid protein; r, recombinant; PVA, poly/vinyl alcohol; RR, relapsing

ron microscope; SE, standard error; TH, T helper; TEFF, effector T; TGF, transforming

(IRCAD), and Department of Health Sciences, University of Eastern Piedmont “A.42.E-mail address: giuseppe.cappellano@med.unipmn.it (G. Cappellano).

5 accine

1

bav[a[ic

ttephgbr

rv[p(ateoc

iaoaesevwaptdaD2it[enm

aapc(phPsbu

682 G. Cappellano et al. / V

. Introduction

Autoimmune diseases may be ascribed to a disturbed balanceetween the activated effector T (TEFF) cells that recognize selfntigens and the regulatory T (TREG) cells that suppress TEFF acti-ation and control both the immune and autoimmune reactions1–5]. Several immunoregulatory molecules, such as IL-10, TGF-ß,nd indoleamine 2,3-dioxygenase (IDO), can support TREG function3,4,6,7]. In particular, IL-10 plays a role in the development andmmunosuppressive function of several TREG cell types [2] and is arucial protective cytokine in several autoimmune diseases [3].

Adoptive transfer of several types of TREG cells results in pro-ective activity in animal models of autoimmune diseases, but theranslation of these studies to humans has been hampered by sev-ral issues [8,9]. In particular, the low frequency of these cells, theiroor growth in culture, and their high functional plasticity in vivoave made their use a significant challenge. Antigen-specific strate-ies may potentially circumvent these limitations in vivo [10,11],ut effective approaches to restore the TEFF/TREG balance in vivoemain largely undefined [9].

Multiple sclerosis (MS) is characterized by an autoimmuneesponse against the axons and myelin sheaths of the central ner-ous system (CNS) that results in axonal loss and demyelinization12]. Several myelin proteins, such as myelin basic protein (MBP),roteolipid protein (PLP), and myelin oligodendrocyte glycoproteinMOG), have been implicated as targets of autoreactive T cells in MSnd its animal model, experimental autoimmune encephalomyeli-is (EAE) [12]. According to the “epitope spreading” model, fewpitopes and antigens play a role in the initiation or perpetuationf the disease, but this reactivity tends to spread during the diseaseourse [13].

“Inverse vaccination” refers to antigen-specific tolerogenicmmunization treatments that restore the TEFF/TREG balancend inhibit autoimmune responses. Inverse vaccination may bebtained by the prolonged treatment with high doses of theutoantigens. In EAE, systemic administration of myelin antigensither prevented or treated EAE, but the delivery of repeated mas-ive doses of these proteins were required for the therapeuticffect [14–16]. This problem has been overcome by DNA-basedaccines encoding the myelin antigens alone or in combinationith immunomodulatory molecules, such as IL-4 or IL-10, that act

s “adjuvants” for TREG function [17,18]. The EAE results were soromising that they prompted a first clinical trial in MS patients,hat showed some positive results in reducing the MRI-measuredisease activity and inducing antigen-specific tolerance to myelinntigens in patients with relapse-remitting (RR) MS [19]. However,NA vaccination is still controversial in MS since a second phase

trial showed no significant improvement in magnetic resonancemaging lesion parameters and reduction of new CNS lesions only inhe patients displaying high concentrations of anti-MBP antibodies20]. Moreover while DNA-based vaccines allow for the sustainedxpression of both the antigens and the adjuvant molecules, it isot possible to control their dose or expression kinetics, and thisay be a problem in the case of adverse reactions.An alternative method to obtain the sustained release of

ntigens and adjuvants may be to load them in solid biodegrad-ble particles. Polymeric biodegradable lactic–glycolic acid (PLGA)articles are attractive carriers due to their biodegradability, bio-ompatibility and approval by the Food and Drug administrationFDA) [21]. They maintain effective concentrations of the loadedroteins for prolonged periods of times by trapping them in aydrated polymer network that enables their slow release [21].

LGA allows for the fine-tuning of the degradation rate and sub-equent antigen release from several days to more than one yeary modulating the polymer lactide–glycolide ratio, the molec-lar weight, and the crystal profile [22]. The controlled and/or

32 (2014) 5681–5689

sustained release of antigens and immune modulators from poly-meric materials has been used previously to induce the properimmune effector response [23]. The use of PLGA-NP can enhancetolerance induction in some delivery settings, such as nasal vacci-nation [24]. Moreover, a recent study showed that the intravenousinfusion of either polystyrene or PLGA particles chemically cou-pled with an encephalitogenic peptide can prevent and cure EAE[25]. That strategy stemmed from previous observations showingthat autoantigenic peptides coupled to apoptotic leukocytes caninduce antigen-specific tolerance in models of autoimmune dis-eases, allergy, and transplantation by inducing the depletion andanergy of the antigen-specific lymphocytes. However, a key limi-tation of this approach was that it required systemic (intravenous)injection of the vaccine and was not effective upon local (such assubcutaneous) injection [25]. This implies that the intrinsic risks oflife threatening adverse reactions (such as systemic anaphylaxis)might discourage use of these vaccines in humans.

The aim of the present study was to develop a novel approach forinverse vaccination based on the sustained release of autoantigensand adjuvants from PLGA-NP delivered subcutaneously.

2. Materials and methods

2.1. PLGA production and characterization

PLGA NP were prepared by a modified double solvent evapora-tion method [26]. Nanoparticles containing MOG35-55 (MOG-PLGA),IL-10 (IL10-PLGA), 6-coumarin (C6-PLGA) or IR783 (IR783-PLGA)fluorescent dyes were produced by adding MOG35-55 (200 �g), IL-10 (200 �g), 6-coumarin (C6) (6 mg) or IR783 (6 mg) to the PLGAsolution during the preparation. The release of MOG35-55 and IL-10were evaluated by dialysis in PBS at 37 ◦C, and the proteins releasedwere quantified by the BCA assay (for MOG35-55) and by ELISA (forrIL-10). Details are given in Supplementary information.

2.2. In vitro biological analyses

Cell toxicity was assessed in human breast adenocarcinomacell line (MCF-7) using the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay.

Proinflammatory activity was assessed on human peripheralblood mononuclear cells (PBMCs) that were cultured with PLGAparticles and/or bacterial lipopolysaccharide (LPS; 1 �g/ml) for48 h. Then, Tumor Necrosis Factor (TNF)-� secretion in the super-natant was measured by ELISA.

The cellular uptake of PLGA-NP was investigated in MCF-7 cellsincubated with C6-PLGA-NP at 37 ◦C for 4 h. Then, the cells wereanalyzed by confocal laser microscopy and flow cytometry.

Lymphocyte activation was assessed in EAE mouse spleen cellscultured in the presence and absence of 10 �g/ml MOG35-55 for 5d. Then, the supernatants were collected, and the levels of IFN-�,IL-10, and IL-17 were evaluated by ELISA.

Histologic analysis of the EAE lesions was performed in thespinal cords fixed in 10% formalin. Sections were stained withhematoxylin&eosin (H&E) and with Luxol fast blue dye for myelinstaining. Details of these methods are given in Supplementaryinformation.

2.3. In vivo analyses

Female C57BL/6 mice were purchased from Harlan (Horst, TheNetherlands) and were housed under specific pathogen-free con-

ditions at the animal facility at the University of Eastern Piedmont“A. Avogadro” (Novara). The experiments were performed in accor-dance with the guidelines of the Institutional Ethical Committee forAnimal Experimentation.

G. Cappellano et al. / Vaccine 32 (2014) 5681–5689 5683

Fig. 1. Physical and chemical analysis of PLGA-NP. (A) SEM micrographs of blank-PLGA and (B) the particle size distribution of blank-PLGA, IL10-PLGA and MOG-PLGAassessed by DLS analysis; (C) In vitro cumulative release [%] of MOG35-55 and human IL-10 from PLGA-NP in PBS (pH 7.4) at 37 ◦C. The data are shown as the mean ± SE from2

ada

f[PNflfi5t

2

M(

3

3

t

independent experiments.

The biodistribution of PLGA-NP was assessed in C57BL/6 micefter a subcutaneous (s.c.) injection of IR783-PLGA or PBS. After 7, the mice were euthanized, and their organs were analyzed withn IVIS® imaging system (Perkin Elmer).

Prophylactic and therapeutic vaccination was performed inemale C57BL/6 mice with EAE induced as previously described27]. Mice (4–8 weeks old) were vaccinated by s.c. injection ofLGA-NP loaded with either 2 �g MOG35-55 or mouse IL-10. TheP were dissolved in PBS, and the injections were given in oneank either (i) on days-30 and -15 (prophylactic vaccination)

rom EAE induction (day 0) or (ii) on days 8 and 22 after EAEnduction (therapeutic vaccination). Each experiment involved–10 mice/group. Details are given in Supplementary informa-ion.

.4. Statistical analysis

The statistical analyses were performed using theann–Whitney U-test calculated with GraphPad Instat software

GraphPad Software, San Diego, CA).

. Results

.1. Physical and chemical characterization of PLGA-NP

We produced PLGA 65:35 lyophilized nanoparticleshat were unloaded (blank-PLGA) or loaded with either

MOG35-55 (MOG-PLGA) or rIL-10 (IL10-PLGA). PLGA-NP weredispersed in distilled water and analyzed for size distribution andzeta potential; the particles were also imaged using a scanningelectron microscope (SEM). Similar results were obtained usingblank-PLGA particles and MOG-PLGA or IL10-PLGA particles. TheSEM showed a spherical shape with a smooth surface lacking anyevident irregularities (Fig. 1A). The size distribution was evaluatedby dynamic light scattering (DLS) and ranged 100–1000 nm witha peak approximately 200 nm and a polydispersity index (PDI) ofapproximately 0.2 for all formulations (Fig. 1B). This behavior canbe related to the negative charge density (�-potential = −19 mVfor PLGA; −21.1 for IL-10 and -−24.9 mV for MOG35-55) exposedon the surface of the particles as a consequence of the presence ofcarboxylates moieties in the structure.

The loading efficiency of MOG35-55 and IL-10 was evaluatedby calculating the amount of protein recovered after dissolvingMOG-PLGA and IL10-PLGA in NaOH (1 M) and neutralizing by HCl(1N). The results showed that the loading efficiency was approxi-mately 23% for rIL-10 and 90% for MOG35-55. This difference may beascribed to the different sizes of these molecules (19 kDa for rIL-10and 2.5 kDa for MOG35-55).

The in vitro release of human IL-10 or MOG35-55 from theirrespective PLGA preparations was assessed by incubating these

particles in PBS at 37 ◦C and sampling the supernatant weekly.The in vitro release profiles showed that PLGA provided sustainedrelease of each protein for several weeks with similar kinetics foreach protein (Fig. 1C).

5 accine 32 (2014) 5681–5689

3

pas

wswrsE�oi

iccbtdC

aw7bw

Fig. 2. TNF-� production was assessed by ELISA in PBMCs cultured with blank-PLGA and IL-10 (human) in the presence or absence of lipopolysaccharide (LPS) at

FbI(

684 G. Cappellano et al. / V

.2. Biological characterization of the PLGA-NP

The cell toxicity of the PLGA preparations was assessed byerforming an MTT assay on MCF-7 cells incubated with titratedmounts (0.6, 6, 60, and 600 �g/ml) of blank-PLGA. The resultshowed no significant toxicity at any dose (data not shown).

The pro-inflammatory activity of blank-PLGA and IL10-PLGAas evaluated by assessing the ability of the particles to induce the

ecretion of TNF-� in human PBMCs. Human PBMCs were incubatedith titrated amounts (0, 6, 60, 600 �g/ml) of each PLGA prepa-

ation in the presence and absence of LPS (1 �g/ml), and TNF-�ecretion into the culture supernatant was assessed after 48 h byLISA. At concentrations ≥60 �g/ml, the blank-PLGA induced TNF-

secretion and increased the response to LPS. By contrast, nonef the IL10-PLGA doses induced TNF-� secretion, and each dosenhibited that induced by LPS (Fig. 2).

The cellular uptake of the PLGA nanoparticles was assessed byncubating MCF-7 cells for 4 h with either coumarin 6 (C6), a fluores-ent lipophilic dye that emits in the FL-1 channel, or C6-PLGA. Theellular uptake of the fluorescent substance was then assessed byoth confocal microscopy and flow cytometry. The results showedhat C6-PLGA was partially taken up by phagocytes in a dose-ependent manner, and no substantial uptake was detected with6 alone (Fig. 3A and B).

The in vivo biodistribution of the PLGA preparations wasssessed by injecting mice subcutaneously with PLGA-NP loaded

ith the IR783 infrared dye (IR783-PLGA) or PBS (control). After

d, mice were euthanized, and their organs were analyzedy fluorescence imaging. As shown in Fig. 3C, strong signalsere detected in the brain, heart, kidneys and, especially, the

ig. 3. Cellular uptake and biodistribution of PLGA-NP. (A) and (B) MCF-7 cells were incuby confocal laser scanning microscopy (A) and flow cytometry (B); representative resu

R783-PLGA, and the heart (H), liver (L), spleen (S), lung (L), kidney (K), brain (B) and lymNIRF).

1 �g/ml. Statistical analysis was performed with the non-parametric ANOVA test;the data are the mean ± SE from 5 experiments. *Significantly different (p < 0.05)from blank-PLGA; significantly different (p < 0.001) from blank-PLGA plus LPS.

lungs. Mild signals were detected in the liver, spleen and lymphnodes.

3.3. Inverse vaccination of mice with EAE

We evaluated the in vivo effect of treatments with MOG35-55-based PLGA inverse vaccines in EAE induced in C57BL/6 mice.

ated with 120 �g/ml of C6-PLGA-NP for 4 h, and fluorescence uptake was assessedlts from three independent experiments are shown. (C) Mice were injected withph nodes (Ly) were collected after 7 d and analyzed by near-infrared fluorescence

G. Cappellano et al. / Vaccine 32 (2014) 5681–5689 5685

Fig. 4. Prophylactic inverse vaccination of EAE mice with different PLGA preparations. Individual (upper panels) and cumulative (lower panel) results from two independentexperiments. Daily clinical scores of EAE in mice treated with blank-PLGA, MOG-PLGA, IL10-PLGA or MOG-PLGA plus IL10-PLGA on days-30 and -15 prior to EAE induction.Experiment 1: 5–8 mice/group; experiment 2: 8–10 mice/group. Results are shown as mean ± SE. The Mann–Whitney U-test was used to compare the clinical scores:I

MbipeMmafwiPMbnsc

lcMacMpstiwa

L10-PLGA + MOG-PLGA vs PLGA (*), IL10-PLGA (§), MOG-PLGA (#); *,§,# p < 0.05.

ice (n = 5–10/group) were vaccinated via s.c. injection of eitherlank-PLGA, MOG-PLGA, IL10-PLGA or MOG-PLGA plus IL10-PLGA

n one flank on days-31 and -15 prior to EAE induction (pro-hylactic inverse vaccination). The results from two independentxperiments showed that, at several time points, treatment withOG-PLGA plus IL10-PLGA significantly inhibited EAE develop-ent compared to the other treatments (Fig. 4). Moreover, the

verage daily disease score calculated on the cumulative datarom both experiments was significantly lower in mice treatedith MOG-PLGA plus IL10-PLGA (mean ± SE, 1.3 ± 0.17) than

n those treated with blank-PLGA (1.8 ± 0.23, p < 0.05) or IL10-LGA (1.8 ± 0.22, p < 0.05). A small effect was displayed also byOG-PLGA alone (1.5 ± 0.19) which inhibited EAE compared to

lank-PLGA and IL-10-PLGA alone at some time points, but no sig-ificant differences were detected at the level of average diseasecores (Fig. 4). By contrast, IL10-PLGA alone had no effect. No vac-ination delayed the disease onset.

The effect of the vaccination on the immune response and CNSesions was evaluated by comparing the anti-MOG spleen lympho-yte response and the histopathologic lesions in mice treated withOG-PLGA plus IL10-PLGA and those treated with blank-PLGA. The

nti-MOG response was assessed by incubating spleen lympho-ytes collected at day 21 (the disease peak) after EAE induction withOG35-55 and assessing cell proliferation (by 3H-thymidine incor-

oration) and the secretion of IFN-�, IL-17, and IL-10 in the cultureupernatants (by ELISA) after a 5 d culture. The results showed

hat vaccination with MOG-PLGA plus IL10-PLGA significantlynhibited the secretion of IL-17 and IFN-� compared to vaccination

ith blank-PLGA, whereas no significant effect on cell prolifer-tion was detected (Fig. 5A). The secretion of IL-10 was always

undetectable (data not shown); this observation is in accordancewith data reported by other authors [28].

The histopathologic analysis was performed on the spinal cordsof mice collected at day 21 after EAE induction. Light microscopyanalysis revealed that the mice treated with blank-PLGA dis-played a higher inflammatory cell infiltrate than those treatedwith IL10-PLGA plus MOG-PLGA, but this difference was not sta-tistically significant (Fig. 5B). However, detection of CD3+ T cellsby immunofluorescence showed that they were strikingly moreabundant in the former than in the latter group. Moreover, myelinstaining showed a significantly dramatic loss of myelin in the micetreated with blank-PLGA, whereas mice treated with MOG-PLGAplus IL10-PLGA showed mild myelin damage only in the subpialmyelin (Fig. 5B).

Finally, we assessed the effect of treatment with the differ-ent PLGA preparations delivered after EAE induction (therapeuticinverse vaccination). Mice (n = 8–10/group) were injected s.c. witheither blank-PLGA, MOG-PLGA, IL10-PLGA or MOG-PLGA plusIL10-PLGA on days 8 and 22 after EAE induction. The resultsfrom two independent experiments showed that, at most timepoints, treatment with MOG-PLGA plus IL10-PLGA significantlyinhibited EAE development compared to treatment with blank-PLGA or IL-10-PLGA (Fig. 6). Moreover, the average daily diseasescore calculated on the cumulative data from both experimentswas significantly lower in the mice treated with MOG-PLGA plusIL10-PLGA (mean ± SE, 1.2 ± 0.15) compared to those treated with

blank-PLGA (1.8 ± 0.24, p < 0.05) or IL10-PLGA (1.8 ± 0.23, p < 0.05)or MOG-PLGA (1.6 ± 0.20, p < 0.05) (Fig. 6). By contrast, MOG-PLGAand IL-10-PLGA had no substantial effect. No vaccination delayedthe disease onset.

5686 G. Cappellano et al. / Vaccine 32 (2014) 5681–5689

Fig. 5. T cell response and histopathologic lesions in EAE mice vaccinated with blank-PLGA or MOG-PLGA plus IL10-PLGA. (A) Spleen lymphocytes were collected 21 d afterEAE induction and cultured with MOG35-55; cell proliferation and the secretion of IFN-� and IL-17 in the culture supernatants were assessed after a 5 d culture period. (B)Histopathologic analysis of the spinal cords collected from mice 21 d after EAE induction; data are shown as representative stainings and quantitative analysis performed inmultiple experiments (bar graphs). H&E staining: (a) blank-PLGA, (b) MOG-PLGA plus IL10-PLGA (arrows mark the inflammatory infiltrate). Immunofluorescence staining ofC lank-P2 was p

4

od

itigbbm

cva

D3+ T cells: (c) blank-PLGA, (d) MOG-10 plus IL-10 PLGA; KB (myelin) staining: (e) b0×. The data are the mean ± SE of the results from 6 to 8 mice. Statistical analysis

. Discussion

This work shows that s.c. inverse vaccination with a mixturef PLGA loaded with MOG35-55 and IL-10 is effective at decreasingisease severity in mouse EAE.

Polymeric materials have been investigated for their ability tonduce the proper immune effector responses, and in some cases,o promote tolerance. Because the induction and maintenance ofmmune tolerance is influenced by the persistence of the anti-en and can be supported by tolerogenic cytokines, the use ofiodegradable micro-nanoparticles is ideal for inverse vaccinationecause they can support a sustained release of several entrappedolecules [23].

An ideal nanoparticle for inverse vaccination should display low

ell toxicity, low proinflammatory activity, sustained release of theaccine active components, and slow elimination by phagocytes tollow their persistence in vivo to promote regulatory rather than

LGA, (f) MOG-PLGA plus IL10-PLGA; arrows indicate myelin vacuoles. Magnificationerformed with the Mann–Whitney U-test (*p < 0.05).

effector immune functions. The MTT assay on MCF-7 cells showedthat our particles were not toxic, which is in accordance with datareported by other groups [29].

Regarding the proinflammatory activity of PLGA, the datareported in the literature are controversial. Nicolete et al. showedthat PLGA nano/microparticles per se can activate the inflammatoryresponse; in particular, microparticles (used at 1 mg/ml) with a sizeranging from 5 to 7 �m induced the production of higher amountsof TNF-� in murine macrophages than those induced by nanopar-ticles [30,31]. In our experimental setting, blank-PLGA inducedTNF-� production at doses ≥60 �g/ml, but this proinflammatoryeffect was not detected when using IL10-PLGA; instead, these parti-cles inhibited the TNF-� secretion induced by LPS. However, it must

be noted that the in vivo experiments showed that the EAE onsetwas earlier in the prophylactic vaccination than in the therapeuticvaccination which suggests that pretreatment with the PLGA-NPmay precondition the induction of EAE by stimulating the innate

G. Cappellano et al. / Vaccine 32 (2014) 5681–5689 5687

Fig. 6. Therapeutic inverse vaccination of EAE mice with different PLGA-NP preparations. Individual (upper panels) and cumulative (lower panel) results from two independentexperiments. Daily clinical scores of EAE in mice treated with blank-PLGA, MOG-PLGA, IL10-PLGA or MOG-PLGA plus IL10-PLGA on day 8 and 22 after EAE induction.E wn asI

im

tTfi

mtwppi

pdtntMpefprbmwit

xperiment 1: 5–8 mice/group; experiment 2: 8–10 mice/group. Results are shoL10-PLGA + MOG-PLGA vs PLGA (*), vs MOG-PLGA (#), IL10-PLGA (§), #,*,§, p < 0.05.

mmunity. This is in line with works showing that PLGA particlesay support immunization in classical vaccination settings [32].The release of MOG35-55 and IL-10 from the PLGA-NP was sus-

ained, and both molecules were released for several weeks in vitro.he particles were partially endocytosed, an effect that is beneficialor antigen presentation, but this did not substantially affect theirn vivo biodistribution.

The in vivo experiments in EAE showed that vaccination with aixture of MOG-PLGA and IL10-PLGA displayed a significant pro-

ective effect and decreased the severity of the disease. This effectas detected not only when vaccination was performed in a pro-hylactic setting, i.e., before EAE induction, but also when it waserformed in a therapeutic setting, i.e., after the disease onset. Thus,

t may be a potential treatment for patients with MS.The protective effect of IL10-PLGA and MOG-PLGA was accom-

anied by decreased histopathologic lesions in the spinal cord andecreased lymphocytes production of IFN-� and IL-17-in responseo MOG35-55 in vitro. These results suggest that this inverse vacci-ation inhibits the autoantigen-specific TH1 and TH17 responseshat are crucial in the pathogenesis of MS and EAE. By contrast,

OG-PLGA had a minimal effect that was detectable only in therophylactic treatment. Moreover, IL10-PLGA had no protectiveffect in any protocol, which is in accordance with previous datarom Cannella et al. showing that treatment with IL-10 does notrotect from EAE [33]. These results are in line with those recentlyeported by Peine et al. showing effectiveness of EAE treatmenty codelivery of MOG and dexamethasone in acetalated dextran

icroparticles [34]. Therefore, delivery of the autoantigen togetherith immunosuppressive agents, such as IL-10 or dexamethasone,

s crucial to suppress the autoimmune response. Data obtained witholerogenic DNA vaccines indicate that these immunosuppressive

mean ± SE. The Mann–Whitney U-test was used to compare the clinical scores:

agents may act as “inverse adjuvants” supporting the induction oftolerance. Lack of these “inverse adjuvants” may explain the fail-ure of clinical trials using autoantigens or “altered peptide ligands”alone which were terminated because of lack of substantial clinicaleffects and development of hypersensitivity reactions [35].

It is noteworthy that our vaccination was effective by the s.c.route. This marks a key difference from the inverse vaccinationperformed by Getts et al. with MOG35-55 covalently linked to thesurface of either polystyrene beads or PLGA particles [25]. Theseauthors obtained a striking protective effect using both prophylac-tic and therapeutic treatments, but this effect strictly depended onthe intravenous delivery of the vaccine, which may restrict its usein humans because of the potential risk of anaphylactic reactions.The effect of this intravenous vaccine seemed to be mediated bythe induction of T cell anergy, as indicated by a strong inhibitionof lymphocyte proliferation; this was possibly driven by tolero-genic macrophages taking up the antigen after scavenger receptorsrecognized the nanoparticles covalently linked to the self peptide.By contrast, our vaccine uses particles that slowly release boththe autoantigen and IL-10. It does not inhibit T cell proliferation,but instead inhibits the secretion of IFN-� and IL-17. Therefore, itseems to induce immune deviation instead of anergy and opensthe way to further studies aiming to ameliorate the mixture ofautoantigens and immunosuppressive adjuvants. Our data do notsupport the possibility that the protective effect was mediated byCD4+CD25+FoxP3+ TREG cells since their numbers were similar inthe spleens of mice treated with blank-PLGA and those treated

with IL-10-PLGA plus MOG-PLGA (data not shown). This is in linewith other reports [34], but it does not rule out the involvementof other populations of regulatory cells or CD4+CD25+FoxP3+ TREGcells compartimentalized in other tissues [36,37].

5 accine

bwwt[ifspicoanttt

imritttim

caeaelpabtab

C

A

FAd

A

i2

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

688 G. Cappellano et al. / V

The in vivo biodistribution of the PLGA-NP was intriguingecause substantial amounts were detectable in several organs oneeek after the injection. This biodistribution is in accordance withorks showing that particles < 50 nm rapidly diffuse from the injec-

ion site, whereas those > 1–5 �m tend to stay at the injection site38]. Because the sizes of our particles ranged from 100 nm to 1 �m,t is conceivable that the smallest of them were readily able to dif-use. It is noteworthy that among the other organs, the strongestignal was detected in the lungs. The lungs have been shown tolay a key role in licensing T cells to enter CNS [39]. The signal

n the brain is also intriguing because it suggests that the vac-ine might exert an immunoregulatory effect directly at the sitef the autoimmune aggression. This effect was not simply ascrib-ble to an aspecific immunosuppressive activity of IL-10 becauseo effect was detected using IL10-PLGA alone. Because changinghe particle size would profoundly affect the biodistribution, fur-her work is needed to assess the role of the PLGA particle size inhe immunosuppressive effect.

A key problem for translating this approach to humans is that,n MS, the target autoantigens are not fully known and may involve

olecules different from MBP, MOG and PLP known to play aole in EAE and MS. Moreover, the autoreactivity is heterogeneousn different patients and changes during the disease course dueo epitope spreading [12]. However, several works showed thatolerogenic vaccination, too, may induce an epitope spreading ofhe tolerance and may gradually involve also autoantigens notncluded in the vaccine. Moreover, amelioration of the vaccines

ight be obtained using mixtures of autoantigens [40,41].In conclusion, this work showed that PLGA-based inverse vac-

ination delivered subcutaneously may be effective in treatingutoimmune diseases. Current treatments for autoimmune dis-ases (i.e., MS) are based on immunosuppressive therapies thatffect the global immune response and display significant toxicity,specially when chronic therapy is needed. Even novel biotechno-ogical drugs do not solve this problem but instead add the furtherroblem of a dramatic increase in the economic costs of the ther-py. Effective therapies with low cost and low toxicity are neededy patients and by the public health systems. Inverse vaccina-ion would meet these criteria because it is capable of inducingn autoantigen-specific immunosuppression with the potential ofeing low cost and long lasting.

onflict of interest

The authors state that there is no conflict of interest.

cknowledgments

This work was supported by Compagnia di San Paolo (Torino),ondazione Cariplo (Milano), Fondazione Amici di Jean (Torino),ssociazione Italiana Ricerca sul Cancro (Milano), Fondazione Cassai Risparmio di Cuneo.

ppendix A. Supplementary data

Supplementary material related to this article can be found,n the online version, at http://dx.doi.org/10.1016/j.vaccine.014.08.016.

eferences

[1] Platten M, Steinman L. Multiple sclerosis: trapped in deadly glue. Nat Med2005;11:252–3.

[2] Hsieh CS, Zheng Y, Liang Y, Fontenot JD, Rudensky AY. An intersection betweenthe self-reactive regulatory and non regulatory T cell receptor repertoires. NatImmunol 2006;7:401–10.

[

32 (2014) 5681–5689

[3] Zhang X, Koldzic DN, Izikson L, Reddy J, Nazareno RF, Sakaguchi S, et al. IL-10 isinvolved in the suppression of experimental autoimmune encephalomyelitisby CD25+CD4+ regulatory T cells. Int Immunol 2004;16:249–56.

[4] Huber S, Schramm C, Lehr HA, Mann A, Schmitt S, Becker C, et al. Cuttingedge: TGF-beta signaling is required for the in vivo expansion and immunosup-pressive capacity of regulatory CD4+CD25+ T cells. J Immunol 2004;173:6526–31.

[5] Chang CC, Ciubotariu R, Manavalan JS, Yuan J, Colovai AI, Piazza F, et al. Toler-ization of dendritic cells by T(S) cells: the crucial role of inhibitory receptorsILT3 and ILT4. Nat Immunol 2002;3:237–43.

[6] Huss DJ, Winger RC, Peng H, Yang Y, Racke MK, Lovett-Racke AE. TGF-betaenhances effector Th1 cell activation but promotes self-regulation via IL-10. JImmunol 2010;184:5628–36.

[7] Yan Y, Zhang GX, Gran B, Fallarino F, Yu S, Li H, et al. IDO upregulates reg-ulatory T cells via tryptophan catabolite and suppresses encephalitogenic Tcell responses in experimental autoimmune encephalomyelitis. J Immunol2010;185:5953–61.

[8] Tang Q, Bluestone JA. Regulatory T cell physiology and application to treatautoimmunity. Immunol Rev 2005;212:217–37.

[9] Ozdemir C, Kucuksezer UC, Akdis M, Akdis CA. Specific immunotherapy andturning off the T cell: how does it work? Ann Allergy Asthma Immunol2011;107:381–92.

10] Lutterotti A, Sospedra M, Martin R. Antigen-specific therapies in MS – currentconcepts and novel approaches. J Neurol Sci 2008;274:18–22.

11] Badawi AH, Siahaan TJ. Immune modulating peptides for the treatment andsuppression of multiple sclerosis. Clin Immunol 2012;144:127–38.

12] Johnson D, Hafler DA, Fallis RJ, Lees MB, Brady RO, Quarles RH, et al.Cell-mediated immunity to myelin-associated glycoprotein, proteolipid pro-tein, and myelin basic protein in multiple sclerosis. J Neuroimmunol1986;13:99–108.

13] Smith CE, Miller SD. Multi-peptide coupled-cell tolerance ameliorates ongoingrelapsing EAE associated with multiple pathogenic autoreactivities. J Autoim-mun 2006;27:218–31.

14] Tsai CY, Chow NH, Ho TS, Lei HY. Intracerebral injection of myelin basic protein(MBP) induces inflammation in brain and causes paraplegia in MBP-sensitizedB6 mice. Clin Exp Immunol 1997;109:127–33.

15] Karpus WJ, Kennedy KJ, Smith WS, Miller SD. Inhibition of relaps-ing experimental autoimmune encephalomyelitis in SJL mice by feedingthe immunodominant PLP139-151 peptide. J Neurosci Res 1996;45:410–23.

16] Devaux B, Enderlin F, Wallner B, Smilek DE. Induction of EAE in mice withrecombinant human MOG, and treatment of EAE with a MOG peptide. J Neu-roimmunol 1997;75:169–73.

17] Ho PP, Fontoura P, Platten M, Sobel RA, DeVoss JJ, Lee LY, et al. Asuppressive oligodeoxynucleotide enhances the efficacy of myelin cocktail/IL-4-tolerizing DNA vaccination and treats autoimmune disease. J Immunol2005;175:6226–34.

18] Steinman L. Inverse vaccination, the opposite of Jenner’s concept, for therapyof autoimmunity. J Intern Med 2010;267:441–51.

19] Bar-Or A, Vollmer T, Antel J, Arnold DL, Bodner CA, Campagnolo D, et al. Induc-tion of antigen-specific tolerance in multiple sclerosis after immunization withDNA encoding myelin basic protein in a randomized, placebo-controlled phase1/2 trial. Arch Neurol 2007;64:1407–15.

20] Garren H, Robinson WH, Krasulová E, Havrdová E, Nadj C, Selmaj K, et al. Phase2 trial of a DNA vaccine encoding myelin basic protein for multiple sclerosis.Ann Neurol 2008;63:611–20.

21] Houchin ML, Topp EM. Chemical degradation of peptides and proteins in PLGA:a review of reactions and mechanisms. J Pharm Sci 2008;97:2395–404.

22] Jiang W, Gupta RK, Deshpande MC, Schwendeman SP. Biodegradablepoly(lactic-co-glycolic acid) microparticles for injectable delivery of vaccineantigens. Adv Drug Deliv Rev 2005;57:391–410.

23] Leleux J, Roy K. Micro and nanoparticle-based delivery systems for vaccineimmunotherapy: an immunological and materials perspective. Adv HealthcMater 2013;2:72–94.

24] Keijzer C, Slütter B, van der Zee R, Jiskoot W, van Eden W, Broere F. PLGA, PLGA-TMC and TMC-TPP nanoparticles differentially modulate the outcome of nasalvaccination by inducing tolerance or enhancing humoral immunity. PLoS ONE2011;6:e26684.

25] Getts DR, Martin AJ, McCarthy DP, Terry RL, Hunter ZN, Yap WT, et al.Microparticles bearing encephalitogenic peptides induce T-cell toleranceand ameliorate experimental autoimmune encephalomyelitis. Nat Biotechnol2012;30:1217–24.

26] Zhou X, Liu B, Yu X, Zha X, Zhang X, Wang X, et al. Controlled release of PEI/DNAcomplexes from PLGA microspheres as a potent delivery system to enhanceimmune response to HIV vaccine DNA prime/MVA boost regime. Eur J PharmBiopharm 2008;68:589–95.

27] Rojo JM, Pini E, Ojeda G, Bello R, Dong C, Flavell RA, et al. CD4+ICOS+ T lympho-cytes inhibit T cell activation ‘in vitro’ and attenuate autoimmune encephalitis‘in vivo’. Int Immunol 2008;20:577–89.

28] Fissolo N, Costa C, Nurtdinov RN, Bustamante MF, Llombart V, Mansilla MJ, et al.Treatment with MOG-DNA vaccines induces CD4+CD25+FoxP3+ regulatory T

cells and up-regulates genes with neuroprotective functions in experimentalautoimmune encephalomyelitis. J Neuroinflamm 2012;9:139.

29] Semete B, Booysen L, Lemmer Y, Kalombo L, Katata L, Verschoor J, et al. In vivoevaluation of the biodistribution and safety of PLGA nanoparticles as drugdelivery systems. Nanomedicine 2010;6:662–71.

ccine

[

[

[

[

[

[

[

[

[

[

[

1 trial in multiple sclerosis. Sci Transl Med 2013;5:188ra75.

G. Cappellano et al. / Va

30] Nicolete R, dos Santos DF, Faccioli LH. The uptake of PLGA micro or nanopar-ticles by macrophages provokes distinct in vitro inflammatory response. IntImmunopharmacol 2011;11:1557–63.

31] Look M, Saltzman WM, Craft J, Fahmy TM. The nanomaterial-dependentmodulation of dendritic cells and its potential influence on therapeuticimmunosuppression in lupus. Biomaterials 2014;35:1089–95.

32] Sharp FA, Ruane D, Claass B, Creagh E, Harris J, Malyala P, et al. Uptake of partic-ulate vaccine adjuvants by dendritic cells activates the NALP3 inflammasome.Proc Natl Acad Sci U S A 2009;106:870–5.

33] Cannella B, Gao YL, Brosnan C, Raine CS. IL-10 fails to abrogate exper-imental autoimmune encephalomyelitis. J Neurosci Res 1996;45:735–46.

34] Peine KJ, Guerau-de-Arellano M, Lee P, Kanthamneni N, Severin M, ProbstGD, et al. Treatment of experimental autoimmune encephalomyeli-tis by codelivery of disease associated peptide and dexamethasonein acetalated dextran microparticles. Mol Pharm 2014;11:828–35,

http://dx.doi.org/10.1021/mp4005172.

35] Bielekova B, Goodwin B, Richert N, Cortese I, Kondo T, Afshar G, et al. Encephal-itogenic potential of the myelin basic protein peptide (aa 83–99) in multiplesclerosis: results of a phase II clinical trial with an altered peptide ligand. NatMed 2000;6:1167–75.

[

32 (2014) 5681–5689 5689

36] Liu Y, Lan Q, Lu L, Chen M, Xia Z, Ma J, et al. Phenotypic and functional charac-teristic of a newly identified CD8+Foxp3−CD103+ regulatory T cells. J Mol CellBiol 2014;6:81–92.

37] Chen ML, Yan BS, Kozoriz D, Weiner HL. Novel CD8+ Treg suppress EAEby TGF-beta- and IFN-gamma-dependent mechanisms. Eur J Immunol2009;39:3423–35.

38] Blank F, Stumbles PA, Seydoux E, Holt PG, Fink A, Rothen-Rutishauser B,et al. Size-dependent uptake of particles by pulmonary antigen-presenting cellpopulations and trafficking to regional lymph nodes. Am J Respir Cell Mol Biol2013;49:67–77.

39] Odoardi F, Sie C, Streyl K, Ulaganathan VK, Schläger C, Lodygin D, et al. Tcells become licensed in the lung to enter the central nervous system. Nature2012;488:675–9.

40] Lutterotti A, Yousef S, Sputtek A, Stürner KH, Stellmann JP, Breiden P, et al.Antigen-specific tolerance by autologous myelin peptide-coupled cells: a phase

41] Kaushansky N, Kerlero de Rosbo N, Zilkha-Falb R, Yosef-Hemo R, Cohen L, et al.‘Multi-epitope targeted’ immune-specific therapy for a multiple sclerosis-likedisease via engineered multi-epitope protein is superior to peptides. PLoS ONE2011;6:e27860.