European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200

Contents lists available at ScienceDirect

European Journal of Pharmaceutics and Biopharmaceutics

journal homepage: www.elsevier .com/locate /e jpb

Research paper

DODAB:monoolein liposomes containing Candida albicans cell wallsurface proteins: A novel adjuvant and delivery system

http://dx.doi.org/10.1016/j.ejpb.2014.11.0280939-6411/� 2014 Elsevier B.V. All rights reserved.

Abbreviations: ISCOMS, immune stimulating complexes; DODAB, dioctadecyl-dimethylammonium bromide; PBS, phosphate buffer saline; TDB, trehalose 6,60-dibehenate; MO, 1-monooleoyl-rac-glycerol; CWSP, cell wall surface proteins; DTT,dithiothreitol; HBSS, Hanks’ balanced salt solution; DEMEM, Dulbecco’s ModifiedEagle’s Medium; FBS, foetal bovine serum; NBT, nitroblue tetrazolium; BCIP, 5-bromo-4-chloro-3-indolylphosphate; MW, molecular weight; vDODAB, DODABmolar fraction; DTS, Dispersion Technology Software; Cryo-SEM, cryogenic scan-ning electron microscopy; MTT, 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazo-lium bromide; ELISA, enzyme-linked immunosorbent assay; ADS, antigen deliverysystems; PDI, polydispersivity .⇑ Corresponding author. Centre of Molecular and Environmental Biology (CBMA),

Department of Biology, University of Minho, 4710-057 Braga, Portugal. Tel.: +351253601546; fax: +351253604319.

E-mail address: psampaio@bio.uminho (P. Sampaio).

Catarina Carneiro a, Alexandra Correia b, Tony Collins a, Manuel Vilanova b,c, Célia Pais a,Andreia C. Gomes a,e, M. Elisabete C.D. Real Oliveira d,e, Paula Sampaio a,⇑a Centre of Molecular and Environmental Biology (CBMA), Department of Biology, University of Minho, Braga, Portugalb IBMC – Instituto de Biologia Molecular e Celular, Porto, Portugalc Instituto de Ciências Biomédicas de Abel Salazar (ICBAS), Universidade do Porto, Porto, Portugald Centre of Physics (CFUM), University of Minho, Braga, Portugale NanoDelivery I&D in Biotechnology, Biology Department, Braga, Portugal

a r t i c l e i n f o

Article history:Received 27 July 2014Accepted in revised form 29 November 2014Available online 10 December 2014

Keywords:DODAB:MO liposomesNanoparticlesAdjuvant activityAntigen delivery systemsCandida albicans proteinsLiposome endocytosisHumoral and cellular immune responsesAntigen protection

a b s t r a c t

We describe the preparation and characterization of DODAB:MO-based liposomes and demonstrate theiradjuvant potential and use in antigen delivery. Liposomes loaded with Candida albicans proteins assem-bled as stable negatively charged spherical nanoparticles with a mean size of 280 nm. High adsorptionefficiency (91.0 ± 9.0%) is attained with high lipid concentrations. The nanoparticles were non-toxic,avidly taken up by macrophage cells and accumulated in membrane rich regions with an internalizationtime of 20 min. Immunized mice displayed strong humoral and cell-mediated immune responses,producing antibodies (IgGs) against specific cell wall proteins, Cht3p and Xog1p. DODAB:MO-basedliposomes loaded with C. albicans proteins have an excellent immunogenic potential and can be exploredfor the development of an immunoprotective strategy against Candida infections.

� 2014 Elsevier B.V. All rights reserved.

1. Introduction micro- and nanoparticles, virosomes, immune stimulating com-

Delivery systems and adjuvants are frequently used in vaccineformulations to ensure efficient antigen delivery and induction ofan adequate immune response. In fact, delivery systems have beenapplied in marketed vaccines for over 70 years to protect and carryantigens as well as to act as adjuvants for molecules with lowimmunogenicity [1,2]. Currently, many options for deliverysystems are available, including liposomes, polymer-based

plexes (ISCOMS) and many more [3–6]. Of these, liposomes canalso act as immunological adjuvants [7] and are the most widelyinvestigated delivery system for phagocyte-targeted therapies[8,9]. Liposomes offer a flexible system for manipulation andenable the preparation of vesicles with varying lamellarities,physical characteristics, adsorption efficiencies, and antigen encap-sulation abilities [3]. In addition, liposomes are biodegradable andbiocompatible and may be functionalized to provide cell targetingspecificity and antigen protection [10,11].

Cationic liposomes have been used extensively as drug and vac-cine delivery systems [12,13]. Various types of cationic lipids maybe used in the preparation of liposomes, varying in the charge ofthe hydrophilic head group, the number and/or length of fatty acidchains (tail) and/or their degree of saturation. The cationicsurfactant dioctadecyldimethylammonium bromide (DODAB) is asynthetic amphiphilic lipid composed of a hydrophilic positivelycharged dimethylammonium group (head) attached to two hydro-phobic 18-carbon alkyl chains (tail). In aqueous buffers, DODABmolecules self-assemble into closed liposomal vesicular bilayers[14]. DODAB cationic vesicles have been used as carriers in drugdelivery [15,16] as well as adjuvants for vaccination, displaying

C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200 191

higher colloid stability than alum and better efficacy in inductinghigher cellular immune responses [17–19]. However, cationic lipo-somes alone may be associated with in vivo cytotoxicity [18,20]and, in salt solutions, such as phosphate buffer saline (PBS), caninstantaneously aggregate due to insufficient electrostatic repul-sion forces [21]. To overcome these problems, several stabilizersincluding trehalose 6,60-dibehenate (TDB) [12], and silica [19] havebeen used to increase DODAB liposomes colloidal stability, mini-mize cytotoxicity by dilution effect and increase adjuvant efficacy.In previous studies, we have demonstrated that monoolein,1-monooleoyl-rac-glycerol (MO) could also be used as a stabilizeras it conferred fluidity to the DODAB system by favouring lipidchain mobility [22]. In fact, we have successfully used DODAB:MOas a mammalian cell transfection system and as a system forin vitro gene silencing [16,23]. MO is particularly interestingbecause it exhibits two inverted bicontinuous cubic phases andas a result forms non-lamellar vesicles with negative curvature[24]. This characteristic decreases the structural rigidity of DODABvesicles causing an increase in the lateral mobility of the lipidchain which in turn improves the fusion of the liposomes with cellmembranes [16]. Tuning MO content in DODAB:MO formulationsis also very advantageous in terms of formulation developmentfor drug/protein delivery purposes, since MO has been describedas an enhancer of proteins solubilization [25–28].

Fungal cell wall protein antigens constitute excellentcandidates for vaccine development as the cell wall is a uniquemicrobial feature with an important role in antigen presentationand immunomodulation [29]. Currently, several vaccine strategiesagainst Candida have been investigated, using several cell wallantigens varying from protein carriers conjugated with laminarin[30,31] or mannoprotein conjugates [32] to recombinant cellsurface proteins or peptides, such as Hyr1p [33], or the surfaceadhesins family of proteins (Als proteins) [34–36]. These studiesadministered these antigens with several adjuvants or carrierssuch as complete Freund’s adjuvant (CFA), aluminium hydroxidegel or MPL (lipid A; monophosphoryl) [37–39]. Other studies alsoreported the use of b-mercaptoethanol (b-ME) extracts containingCandida cell wall proteins mixed with Ribi Adjuvant System [40].However, despite the extensive investigation, more studies areneeded to develop a safe and efficient vaccine against Candidainfections.

In the present study, we developed and explored a DODAB:MOlipid based delivery system loaded with dithiothreitol (DTT)extracted Candida albicans cell wall surface proteins (CWSP) anddetermined its immunogenicity for future use against deleteriousfungal infections.

2. Materials and methods

2.1. Materials

Dioctadecyldimethylammonium bromide (DODAB) waspurchased from Tokyo Kasei (Japan). 1-Monooleoyl-rac-glycerol(MO), Hanks’ balanced salt solution (HBSS), glutaraldehyde, propi-dium iodide (PI) and DTT were supplied by Sigma–Aldrich (St.Louis, MO, USA). Wheat Germ Agglutinin Alexa Fluor� 633 Conju-gate (kem = 647 nm) was provided by Alfagene (Carcavelos, Portu-gal), Rhodamine DHPE (Lissamine™ kem = 580 nm) and Tris–HCIBuffer by Invitrogen/Molecular Probes (Eugene, OR, USA) and Eth-anol (high spectral purity) were purchased from Uvasol (Leicester,United Kingdom). Dulbecco’s Modified Eagle’s Medium (DMEM)was supplemented with 10% heat-inactivated foetal bovine serum(FBS), 2 mM L-glutamine, (all from Sigma–Aldrich), HEPES-buffersolution pH = 7.5 supplied by VWR International (Radnor, PA,USA) and 1 mM sodium pyruvate by Merck (Frankfurt, Germany).

2.2. Extraction of yeast cell wall surface proteins

C. albicans strain sc5314, kindly provided by Prof. JoachimMorschhauser (Wurzburg, Germany), was obtained from a 2 dayYPD agar (2% D-glucose, 1% Difco yeast extract, 2% peptone and2% agar) (w/v) culture incubated at 26 �C. The following proceduresused for CWSP extraction were all performed in a sterile environ-ment using apyrogenic solutions. The CWSP were released fromwhole intact cells by DTT treatment as described previously [41]with some modifications. Briefly, yeast cells were grown in YPDmedium at 26 �C, 180 rpm, to the start of the stationary growthphase. Cells were then harvested by centrifugation and washedtwice with 50 mM-Tris–HCl pH 7.5 buffer. Subsequently, cells wereresuspended in the same buffer supplemented with 2 mM dithio-threitol (DTT), shaken vigorously for 2 h at 4 �C and centrifugedat 7943g for 15 min. The yeast pellet obtained was used for theassessment of membrane integrity while the yeast CWSP in thesupernatant were concentrated and retained. Membrane integritywas assessed by incubating cells with PI (6 lg/ml) and analysisby flow cytometry (EPICS XL-MCL, Beckman-Coulter Corporation).Supernatant proteins were concentrated with a 3 kD Amicon cen-trifugal filter supplied by Millipore (Madrid, Spain) at 4 �C, washedwith 25 mM HEPES-buffer solution pH = 7.5 and quantified bymeans of the Bradford assay [42]. The concentrated proteinsobtained were stored at �80 �C in aliquots of 100 lg/ml.

2.2.1. Western blotExtracted proteins were subjected to 12% SDS–polyacrylamide

gel electrophoresis and transferred to polyvinylidene fluoridemembranes (hybond-P; Amersham). The membranes were blockedwith 3% BSA in Tris-saline buffer (w/v) containing 0.05% Tween (v/v) for 2 h at room temperature. Membranes were then incubatedovernight at 4 �C with a pool of antisera (1:250) from immunizedmice. Bound antibodies were detected with goat anti-mouse IgsAP conjugate and membranes were developed using the conven-tional method with nitroblue tetrazolium (NBT)/5-bromo-4-chloro-3-indolylphosphate (BCIP) (Roche) in 100 mM Tris,100 mM NaCl, and 5 mM MgCl2, pH 9.5. A broad range SDS–PAGEmolecular weight (MW) marker (15–250 kDa) was used (Bio-Rad).

2.2.2. Identification of the proteinsThe protein bands, excised from the gel, were destained,

reduced, alkylated and digested with trypsin (Promega, 6.7 ng/ll)overnight at 37 �C. The tryptic peptides were desalted and concen-trated using POROS R2 (Applied Biosystems) and eluted directlyonto the MALDI plate using 0.6 ll of 5 mg/ml CHCA (alpha-cyano-4-hydroxycinnamic acid, Sigma) in 50% (v/v) acetonitrileand 5% (v/v) formic acid.

The data were acquired in positive reflector MS and MS/MSmodes using a 4800plus MALDI-TOF/TOF (AB Sciex) mass spec-trometer and using 4000 Series Explorer Software v.3.5.3 (AppliedBiosystems). External calibration was performed using Pepmix1(Laser BioLabs).

The twenty fifth most intense precursor ions from the MS spec-tra were selected for MS/MS analysis.

The MS/MS data were analysed using Protein Pilot Software v.4.5 (ABSciex) with the Mascot search engine (MOWSE algorithm).The search parameters were as follows: monoisotopic peptidemass values were considered, maximum precursor mass tolerance(MS) of 50 ppm and a maximum fragment mass tolerance (MS/MS)of 0.3 Da. The searches were performed against SwissProt databasewith or without taxonomic restriction for Fungi. A maximum oftwo missed cleavage was allowed. Carboxyamidomethylation ofcysteines and oxidation of methionines, were set as fixed andvariable modifications, respectively.

192 C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200

Protein identification was only accepted when significant pro-tein homology scores were obtained (p < 0.05) and at least onepeptide was fragmented with a significant individual ion score(p < 0.05). These data were obtained at the Mass Spectrometry Unit(UniMS), at ITQB/iBET, Oeiras, Portugal.

2.3. Preparation and characterization of CWSP-loaded liposomes

DODAB:MO based liposomes were prepared using the lipid-filmhydration method as previously described [43]. DODAB and MO, ata DODAB molar fraction (vDODAB) of 0.33, were dissolved in eth-anol and mixed in a round-bottom flask. The vDODAB is given bythe following equation [25]:

vDODAB ¼ ½DODAB�½DODAB� þ ½MO�

The solvent was removed by rotary evaporation at 55 �C andliposomes formed after hydration of the lipid film with 25 mMHEPES pH 7.5 at 55 �C. The liposomal stock dispersion was pre-pared at a total lipid concentration of 1774 lg/ml. Using 5 lg/mlof CWSP, seven different formulations containing different pro-tein/lipid ratio (w/w), 0.006, 0.011, 0.014, 0.028, 0.056 and 0.113were prepared (Table 1). The formulations were then incubatedfor 1 h at 55 �C. The z-average diameter of all formulations wasdetermined by dynamic light scattering at 25 �C with a MalvernZetaSizer Nano ZS particle analyser. The Malvern Dispersion Tech-nology Software (DTS) was used with multiple narrow mode (high-resolution) data processing, and mean size (nm), polydispersityindex (PDI) and error values were determined. The nanoparticlesurface charge was measured indirectly by zeta potential analysisat 25 �C using DTS with mono-modal mode data processing todetermine average f-potential (mV) and error values. For severalstudies, an up-scale of some formulations, using 50 lg/ml of pro-tein, was necessary but in all cases the protein-to-lipid weightratios were maintained as described in Table 1.

2.3.1. Quantification of protein adsorption to liposomesThe prepared formulations were pelleted by ultracentrifugation

(100,000g for 1 h), the pellet submitted to TCA protein precipita-tion (Thermo Scientific Pierce), and the concentration of CWSPdetermined using the BCA Protein Assay Kit (Thermo ScientificPierce) according to the manufacturer’s instructions. The emptyliposomes were used as negative control in order to exclude lipidinterference in the protein quantification method.

2.3.2. Cryo-SEM analysis of liposomesThe morphology of the liposomes was evaluated by cryogenic

scanning electron microscopy (Cryo-SEM) using a High resolutionscanning electron microscope with X-ray Microanalysis (JEOLJSM 6301F/Oxford INCA Energy 350/Gatan Alto 2500). Aliquots of4 ll of empty liposomes (total lipid concentration of 888 lg/ml)

Table 1Total lipid and of antigen concentrations for the different protein to lipid weightratios. Equal volumes of CWSP were added to the liposomes dispersions and the finalconcentrations of the formulations are presented. For some experiments, theconcentration of total lipid and of CWSP was increased but protein to lipid ratiowas maintained.

CWSP:total lipid (weight ratio) Total lipid (lg/ml) CWSP (lg/ml)

0.006 833 50.011 454.5 50.014 357.1 50.028 178.5 50.056 88.8 50.113 44.2 50.376 13.3 5

or 4 ll of formulation ADS1 (50 lg/ml of CWSP and 888 lg/ml oflipids, CWSP:Lipid weight ratio of 0.056) were placed into carbonfilm grids, rapidly cooled (plunging into sub-cooled nitrogen–slushnitrogen) and transferred under vacuum to the cold stage of thepreparation chamber. The samples were fractured, sublimated(‘etched’) for 120 s at �90 �C and coated with Au/Pd by sputteringfor 40 s. Samples were transferred into the SEM chamber and stud-ied at a temperature of �150 �C. Micrographs were captured at anacceleration voltage of 15 kV and a working distance of 15 mmwith a secondary electron in lens detector.

2.4. Macrophage studies

The in vitro studies were performed using the murine macro-phage cell line J774A.1 as previously described [44]. This cell linewas cultured in complete DMEM at 37 �C in a 5% CO2 atmosphere.After confluent growth, macrophage cells were recovered, washedand re-suspended in complete DMEM to the desired finalconcentration.

2.4.1. Cytotoxicity analysisMacrophages were plated onto 96-well tissue culture plates

(Falcon) at 3 � 105 cells/well and incubated to adhere overnightat 37 �C in a humidified atmosphere of 5% CO2. Different concentra-tions of empty liposomes, CWSP alone and two formulations ADS1(CWSP:Lipid weight ratio of 0.056) and ADS2 (CWSP:Lipid weightratio of 0.376), were added to the cells. Cell viability was analysedby the 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bro-mide (MTT) assay [45] following 24 h and 48 h incubation with thedifferent formulations, at different concentrations. MTT formazanwas solubilized by adding DMSO–Ethanol (1:1) solution and absor-bance measured at 570 nm. Untreated cells were used as a controlof viability (100%) and cells treated with DMEM:DMSO (4:1) as acontrol of cytotoxicity (100%). Results were expressed as percent-age of viability according to the following equation:

Viability ð%Þ ¼ Experimental value ðaverageÞ � Control of cytotoxicity ðaverageÞControl of viability ðaverageÞ � Control of cytotoxicity ðaverageÞ �100

2.4.2. Microscopy analysis of internalized liposomesFor confocal microscopy, Rhodamine DHPE was incorporated

into the DODAB:MO liposome (at a molar ratio of 1:200) duringthe preparation phase, as described in Section 2.3. Labelledliposomes were then pelleted by ultracentrifugation (100,000gfor 1 h) to remove unbound dye and resuspended in the same buf-fer volume. This labelled liposome stock dispersion was used toprepare the various protein to lipids weight ratio formulations, asdescribed in Section 2.3. Macrophages were plated in a microscopychamber plate (Ibidi) at 3 � 105 cells/well and left to adhere over-night at 37 �C in a humidified atmosphere of 5% CO2. Immediatelyprior to incubation with liposomes, the macrophages were labelledwith Wheat Germ Agglutinin Alexa Fluor� 633 Conjugate (2.5 lg/ml) for 10 min and then washed twice with HBSS. The micros-copy chamber plate was then placed in the integrated chamber(37 �C, 5% CO2) of the LSM 780 Carl Zeiss Microscope and 8 ll ofDODAB:MO (1:2) empty labelled liposomes (total lipid concentra-tion of 888 lg/ml) and 8 ll of labelled ADS1 or ADS2 were addedto the adhered macrophages. A time-lapse imaging video and az-stack image after 75 min were obtained and images were ana-lysed using ZEN 2012 lite software (ZEISS).

For scanning electron microscopy (SEM), macrophages wereplated at 5 � 105 cells/well onto 24-well tissue culture plates(Falcon) containing clean sterile glass coverslips (Ø13 mm) and leftto adhere overnight at 37 �C in a humidified atmosphere of 5% CO2.8 ll of DODAB:MO (1:2) empty liposomes (total lipid concentra-tion of 888 lg/ml), 8 ll of ADS1 and ADS2 were added to the

C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200 193

adhered macrophages. After 30 min, cells were washed with 1�PBS and fixed according to Hoess et al. [46] with some modifica-tions. Briefly, cells were fixed with 2% glutaraldehyde in PBS (v/v)for 2 h at 4 �C and dehydrated through a series of 10 min incuba-tions in increasing ethanol concentrations (10%, 30%, 50%, 70%,90% and 100%). Desiccation was carried out with two mixtures ofhexamethyldisilazane:ethanol (1:2, 2:1) (v/v) and pure hexameth-yldisilazane for 3 min each followed by air-drying. Glass coverslipswere mounted onto aluminium stubs, sputter-coated with goldand observed with an S-360 scanning electron microscope (Leo,Cambridge, USA).

2.4.3. Cytokine productionMacrophages were plated onto 96-well tissue culture plates

(Falcon) at 3 � 105 cells/well, incubated with the prepared lipidformulations and the release of pro-inflammatory cytokine TNF-awas quantified after 2, 6 and 24 h of incubation. The Mouse TNFaElisa Kit (Thermo Scientific) (Sensitivity: 9 pg/mL) was usedaccording to the manufacturer’s instructions. Macrophagesincubated with 1 lg/ml LPS were used as a positive control andmacrophages alone as a negative control.

2.5. Immunization procedures

Female BALB/c mice, 8–10 weeks old, were purchased fromCharles River (Barcelona, Spain) and kept under specific-patho-gen-free conditions at the Animal Facility of the Instituto deCiências Biomédicas Abel Salazar, Porto, Portugal. All proceduresinvolving mice were performed according to the European Conven-tion for the Protection of Vertebrate Animals used for Experimentaland other Scientific Purposes (ETS 123), the 86/609/EEC directiveand Portuguese rules (DL 129/92).

2.5.1. Immunization protocolEight BALB/c mice per group, total of 4 groups, were injected

subcutaneously three times with a 2-week intervening period,with 200 ll of one of the following preparations: CWSP alone(50 lg/ml); DODAB:MO (1:2) empty liposomes (total lipid concen-tration of 888 lg/ml), ADS1 or ADS2. Following the last immuniza-tion, serum was collected by retro-orbital bleeding and micehumanely terminated by placing them in a closed chamber filledwith CO2. Spleens were aseptically removed and total splenocytesobtained by homogenization of the tissue in HBSS buffer.

2.5.2. Quantification of CWSP specific antibody isotypesSpecific anti-CWSP immunoglobulins in the collected serum

were quantified by an enzyme-linked immunosorbent assay(ELISA) according to Ferreirinha et al. [47]. Briefly, polystyrenemicrotitre plates (Nunc, Roskilde, Denmark) were coated with5 lg/ml of CWSP and incubated overnight. Wells were then satu-rated for 1 h at room temperature with 1% BSA in PBS (w/v) andserial dilutions of the serum samples were plated and incubatedfor 2 h at room temperature. After washing, alkaline phospha-tase-coupled monoclonal goat anti-mouse IgG1 and IgG2a (South-ern Biotechnology Associates, Birmingham, AL) were added andincubated for 30 min at room temperature. After washing, thebound antibodies were detected by development with a substratesolution containing p-nitrophenyl phosphate (Sigma) during30 min, and the reaction stopped by addition of 0.1 M EDTA, pH8. The absorbance was measured at 405 nm, subtracting for eachwell the value of the absorbance at 570 nm. The antibody titreswere expressed as the reciprocal of the highest dilution with anabsorbance 2� higher than the value of the control (no serumadded).

2.5.3. Analysis of cytokine productionSplenocytes isolated from mice were washed, re-suspended in

RPMI 1640-FCS and seeded at 1 � 106 cells/ml in 96-well flat-bot-tom plates (Nunc). Spleen cells were stimulated ex vivo for 6 dayswith CWSP (5 lg/ml) at 37 �C in 5% CO2. The culture supernatantswere collected and IFN-c (Sensitivity: 2 pg/ml), IL-17 (Sensitivity:5 pg/ml), IL-10 (Sensitivity: 5.22 pg/ml) and IL-4 (Sensitivity:2 pg/ml) levels quantified with Duo-Set ELISA kits (all from R&DSystems, Minneapolis, MN) according to the manufacturer’sinstructions.

2.5.4. Proliferation of splenocytes ex vivoAfter ex vivo stimulation with 5 lg/ml CWSP, splenocytes were

incubated with polystyrene monodisperse microparticles (Fluka)and the number of splenocytes determined by Flow Cytometry(Beckman-Coulter Corporation, Hialeah, FL, USA) according to themanufacturer’s instructions. CD3 was used as a positive controlfor proliferation.

2.6. Statistical analyses

Data were analysed using analysis of variance (ANOVA)followed by the Bonferroni test to compare the mean values of dif-ferent groups. Unless otherwise stated, results shown are fromthree independent experiments with three replicates. Differenceswere considered significant when the p value was less than 0.05.

3. Results

3.1. Characterization of CWSP-loaded liposomes

In previous studies, we observed that DODAB:MO formulationswith an enriched MO content assembled mainly as vesicles withlamellar structures containing inverted non-lamellar structures.This morphology is observed probably because MO is preferentiallylocalized in the interior of the lipidic structures, due to its naturalnegative curvature [16,23]. These structures will offer a highsurface area that may be advantageous for entrapping/adsorbingantigens for vaccine development.

In the present study, we used liposomes based on DODAB andMO with a vDODAB of 0.33 to prepare antigen delivery systems(ADS) and evaluated the adjuvant potential of this lipid formula-tion. C. albicans CWSP obtained by using the reducing agent DTTwere selected as antigens for the DODAB:MO delivery systems.Flow cytometry analysis of C. albicans cells after the extractiontreatment confirmed the membrane integrity of yeast cells (>99%PI negative) (Fig. 1), thereby ensuring that the majority of theextracted proteins are derived from the surface of the yeast celland that cytoplasmic protein contamination is low. This methodyielded 3 ± 0.04 lg of proteins/g of cell dry weight and DLS andf-potential measurements suggested polydisperse (PDI)(0.63 ± 0.01), negatively charged (�14 ± 0.70 mV) proteins with amean size of 88.76 ± 5.1. The DLS size distributions of the extractedCWSP can be found in Supplementary information (Fig. S1).

DODAB:MO liposomes were relatively polydisperse,PDI � 0.39 ± 0.01, with a main (95 ± 6%) population size of around245 ± 18 nm and were highly positively charged (+52 ± 2.4 mV).Since the protein-to-lipid weight ratio is highly decisive for theaggregation behaviour of the liposomes [48], seven differentformulations were prepared and characterized. Fig. 2 includes aschematic representation of the possible protein-to-lipid interac-tions observed at different protein–lipid weight ratios. Analysesof the mean size and f-potential of these formulations indicatedthat at protein/lipid weight ratios below 0.014, an increase in thenanoparticle size and heterogeneity is observed in comparison

Fig. 1. (A) Flow cytometry evaluation of membrane integrity of C. albicans. Cells were stained with the propidium iodide (PI), a membrane impermeant dye, before (I) andafter (II) treatment with DTT used for CWSP extraction. Membrane disrupted, by heat, C. albicans cells were used as a negative control (III). In the histograms, PI stained cellspercentage is indicated and gated in H-2. Histograms are a representative example from one of three independent experiments (mean ± SD). No less than 20000 cells wereanalysed per condition.

Fig. 2. Representative mean size, PDI and zeta potential of DODAB:MO liposomes atdifferent protein-to-lipid weight ratios. At the bottom, a schematic representationof the possible interactions between proteins and liposomes: (I) empty liposomes,(II) at low protein to lipid weight ratios, liposomes and proteins interact formingunstable aggregates, (III) at intermediate protein to lipid weight ratios, liposomesadsorb more proteins, forming a protein corona around the liposomal surface thatstabilizes liposomes and reduces liposome-to-liposome interactions, and (IV) athigh, protein to lipid weight ratios, liposomes with a protein corona are formed andfree proteins remain in solution. Significant differences between the differentformulations and the empty liposomes are: ⁄P < 0.05, and ⁄⁄⁄P < 0.001. All valuesshown are the average values from three independent experiments ± SEM.

194 C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200

with the naked liposomes (Fig. 2). The size of the vesicles at thelowest protein/lipid ratio (0.006) was approximately 200 nm, sim-ilar to the empty liposomes, what confirms a small adsorption of

the protein at this point. The PDI of this vesicles revealed howevera more polydisperse population in comparison with the naked ves-icles, which can be explained by the presence of a high amount ofliposomes competing for the proteins. However, at protein/lipidratio from 0.006 to 0.014, the size of the nanoparticles increasedsignificantly, from 207.2 ± 49.8 nm to 4143 ± 1903 nm, and thisincrease was characterized by aggregation at room temperature(data not shown). It is also important to mention that the increasein size and PDI was accompanied by a decrease in f-potential from+51.1 ± 1.02 mV (naked liposomes) to +22.7 ± 1.1 mV (at weightratio 0.014).

In contrast, a nanoparticle size reduction is observed at protein/lipid weight ratios of 0.028 and above, ranging from1157 ± 143.5 nm to 223.1 ± 23.1 nm. In addition, at these protein/lipid ratios the f-potential became negative, ranging from�17.95 ± 0.07 mV to �24.03 ± 0.14 mV. This is a clear indicationthat, by decreasing the concentration of lipids the positive chargeof this vesicles is shielded by the negative residues of the adsorbedproteins and the vesicles become negatively charged. The DLS sizedistributions of DODAB:MO liposomes at the different protein/lipidweight ratios can be found in Supplementary information (Fig. S2).

The morphology of the liposomal structures was determined byCryo-SEM (Supplementary information, Fig. S3) and demonstratedthat the DODAB:MO microstructure is clearly dominated by spher-ical vesicles of variable sizes, as already observed by DLS assays.Large vesicles (P500 nm) were also detected, in some of which itis possible to observe smaller vesicles inside. It was also observedthat the vesicles with a protein-to-lipid weight ratio of 0.056 had amean particle size of around 350 nm, showing that the addition ofproteins did not changed the main morphology; it only reduced themean size of the nanoparticles.

Based on physicochemical properties (Fig. 2), negativelycharged particles with sizes around 200 nm, and acceptable PDIvalues were selected. The formulation with a protein/lipidweight ratio of 0.056, designated as ADS1 (antigen delivery sys-tem 1), and the formulation with a protein/lipid ratio of 0.376,designated as ADS2 (antigen delivery system 2) were selected.These were the formulations with the highest (833 lg/ml) andlowest (125 lg/ml) lipid concentrations, respectively, that meetthe criteria defined.

After selecting the formulations, the protein adsorption wasquantified. ADS1 and ADS2 nanoparticles presented differentdegrees of antigen adsorption efficiency: 91.0 ± 9.0% for the formerand 16.5 ± 4.5% for the latter, showing that an increase in the pro-tein/lipid ratio reduced the protein adsorption. In order to confirmthis trend the adsorption efficiency of the protein/lipid weight ratio

C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200 195

0.113 was quantified and, as expected, an intermediate value of45.2 ± 4.9% was obtained.

3.2. In vitro activation of phagocytic cells

Before evaluating the immunostimulatory potential of the for-mulations their associated cytotoxicity was considered. Fig. 3shows cell viability values above 81% for all formulations and lipidconcentrations upon 24 and 48 h of co-incubation, indicating thatthe CWSP and ADSs systems are not cytotoxic to these cells. Wealso characterized macrophage morphology after one hour of incu-bation with the different formulations by scanning electronmicroscopy (SEM) and found these to be similar to that of the con-trol cells (Supplementary Fig. S4), in agreement with the toxicityresults.

ADS1 liposomes uptake by J447 cells were monitored for 70 minby confocal microscopy in order to determine uptake by macro-phages (Fig. 4A). A gradual intracellular accumulation of rhoda-mine labelled liposomes was observed with time, showing thatADS1 is efficiently taken up by macrophages (Supplementary dataV1). Interestingly, an image viewed at the mid-point of cell thick-ness shows that most of the rhodamine fluorescence did not spreaduniformly and seemed to concentrate intra-cellularly in discretesites (Fig. 4B). Rhodamine fluorescence appears to be distributedthroughout the cytoplasm, the transport endosome region and inthe Golgi area next to the nucleus, which itself does not revealany fluorescence signal. By analysing co-localization with AlexaFluor 633 the fluorescence was observed to be within mem-brane-rich regions, i.e., in the cytoplasmic area and in the cellularextensions (Fig. 4A – arrow head), resembling the pattern seenafter endocytosis described previously [49]. The internalization ofa labelled ADS1 was estimated to be around 20 min, in agreementwith other studies that assessed endocytosis of a liposome to occurwithin 30 min (Supplementary Fig. S5) [49,50].

Secretion of pro-inflammatory cytokines, namely TNF-a, by themacrophages was quantified by ELISA after co-incubation with thedifferent liposomes (Fig. 5). Untreated cells and cells stimulatedwith liposomes alone produced residual or no TNF-a. CWSP signif-icantly (P < 0.01) potentiated the production of TNF-a in a timedependent manner in comparison with lipids alone. ADS2 alsoinduced TNF-a production as expected since free CWSP are alsopresent in this formulation, but the levels were significantly lowerthan for CWSP (P < 0.05). ADS1 however failed to enhance the basallevels of TNF-a production.

Fig. 3. Effects of different concentrations of empty liposomes, CWSP, ADS1 and ADS2 onand 48 h incubation. The liposomes, CWSP and the ADSs were diluted in the cells mmean ± SEM of three independent experiments. Significant differences between the two

3.3. Immunological characterization of CWSP-loaded liposomes

The potential of the ADSs to act as adjuvant/antigen deliveryvehicles in vivo, was analysed by determining both humoral andcell mediated immune responses elicited by the loaded CWSP.The humoral response was analysed by quantification of CWSP-specific IgG1 and IgG2a antibody production, after immunizationwith ADS1, ADS2, CWSP alone or empty liposomes (Fig. 6). Immu-nization with ADS2 induced similar levels of IgGs as immunizationwith the proteins alone (P > 0.05). On the contrary, the IgGresponse of mice immunized with ADS1 was significantly higher(P < 0.05) than the response observed with proteins alone. Compar-ing the two systems, immunization with ADS1 resulted in signifi-cantly higher levels of IgG1 and IgG2a than immunization withADS2 (P < 0.05). IgG1 was the predominantly produced isotype asthe ratio of IgG1:IgG2a in mice immunized with the CWSP was21.4. However, in mice immunized with ADS1, the IgG1:IgG2aratio was three times lower as compared to the one detected whenusing free-proteins, indicating that immunization with ADS1shifted the immune response to a Th1-type (P < 0.05) in compari-son with mice immunized with the CWSP alone.

Cell mediated immune responses to CWSP delivered by theprepared liposomes were evaluated by splenocytes proliferationanalysis and subsequent quantification of cytokine secretion afterantigen re-stimulation. Results showed that, after re-stimulation,splenocytes from all groups of mice, except from mice immunizedwith empty liposomes, exhibited a higher (P < 0.05) proliferativeresponse in comparison with the respective non-restimulated con-trols (Fig. 7). Splenocytes from mice immunized with ADS1 showedthe highest levels of proliferation and only these presented a sig-nificant proliferation (P < 0.05) in comparison with cells fromCWSP immunized mice. Quantification of IFN-c and IL-17 was per-formed after re-stimulation with CWSP as indicators of a Th1 orTh17-type polarized response, respectively, while IL-4 and IL-10were used as indicators of a Th2 pathway. Re-stimulated spleno-cytes derived from mice immunized with ADS1 or ADS2 secretedIFN-c and IL-17A, while splenocytes from mice immunized withCWSP or liposomes alone failed to do so (Fig. 8). Importantly, thelevels of these cytokines were much higher in mice immunizedwith ADS1 (380.0 pg/ml of IFN-c and 124.0 pg/ml of IL-17A) thanwith ADS2 (41.6 pg/ml of IFN-c and 2.0 pg/ml of IL-17A) (P < 0.05).

The production of IL-4 and IL-10 was investigated as anindicator of Th2 polarization. Low levels of IL-4 were detected insplenocytes of mice immunized with ADS1 (31 pg/ml) and ADS2

the viability of J774A.1 cells. Viability was assessed using the MTT assay after 24 hedium and the final concentration is indicated in the x-axis. Results indicate thetime points, 24 h and 48 h of incubation, are indicated: (⁄⁄P < 0.01, ⁄⁄⁄P < 0.001).

Fig. 4. Representative confocal microscopy images of ADS1 uptake by J774 cells. (A) Uptake as a function of incubation time over 75 min. Rhodamine labelled ADS1 (red)superimposed with Alexa Fluor� 633 Conjugate labelled macrophages (blue). (B) A mid-point thickness view (22.8 lm � 20.7 lm � 0.7 lm) of the cell, indicated by arectangle in (A), shows rhodamine fluorescence superimposed with Alexa Fluor� 633 Conjugate fluorescence (z). Z-axis rotations of a single transverse slice through twosections of the cell: view in the x–0–z plane (x) and view in the y–0–z plane (y) are also presented.

Fig. 5. Production of TNF-a by J774A.1 cells after incubation with CWSP (2 lg/ml),liposomes alone (36 lg/ml), ADS1 (2:36 lg/ml; CWSP:Lipid), ADS2 (2:6 lg/ml;CWSP:Lipid), or LPS (1 lg/ml) for 2, 6 and 24 h. Results indicate the mean ± SEM ofthree measurements and represent three independent experiments. Significantdifferences between ADSs and the CWSP are shown above the symbol / andsignificant differences between the ADSs are shown below /: ⁄P < 0.05, ⁄⁄P < 0.01,and ⁄⁄⁄⁄P < 0.0001.

Fig. 6. Anti-CWSP specific IgG1 and IgG2a responses in serum of mice immunizedwith CWSP alone (10 lg/ml), ASD1 (10:178 lg/ml; CWSP:Lipid), ASD2 (10:27 lg/ml; CWSP:Lipid), or empty liposomes (178 lg/ml). The antibody titres areexpressed as the reciprocal highest dilution with an absorbance 2� higher thanthe value of the control (no serum added). Each point represents an individualmouse. Horizontal lines correspond to the mean value in each group. Results arerepresentative of two independent experiments. Significant differences betweenthe ADSs and the CWSP are shown above the symbol / and significant differencesbetween the ADSs are shown below /: ⁄P < 0.05, ⁄⁄P < 0.01 and ⁄⁄⁄P < 0.001.

196 C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200

(11.6 pg/ml) while residual or no production was observed in miceimmunized with CWSP or liposomes alone. IL-10 levels were sim-ilar in all groups (between 55.1 and 99.6 pg/ml) except for thegroup of mice injected with lipids alone where only residual levelswere observed (1.7 pg/ml). As expected, no significant differences

were observed regarding levels of IL-4 or IL-10 between miceimmunized with CWSP and ADS2. Comparing both ADSs systems,the levels of IL-4 and IL-10 secreted by splenocytes of mice immu-nized with ADS1 were higher than of mice immunized with ADS2(P < 0.01).

Fig. 7. Splenocytes proliferation after re-stimulation with CWSP (5 lg/ml) after6 days incubation. Controls cells are incubated in the same conditions without re-stimulation and CD3 was used as a positive control for proliferation. Each pointrepresents an individual mouse. Results are representative of two independentexperiments. Significant differences between control and re-stimulated cells areindicated above the bars and differences between the groups are indicated abovethe horizontal line: ⁄P < 0.05, ⁄⁄P < 0.01.

Fig. 8. Cytokine production from CWSP re-stimulated splenocytes derived frommice immunized with CWSP alone (10 lg/ml), ADS1 (10:178 lg/ml; CWSP:Lipid),ADS2 (10:27 lg/ml; CWSP:Lipid), or empty liposomes (178 lg/ml). Bars denotemean cytokine levels, and error bars denote SD of 8 mice per group. Results arerepresentative of two independent experiments. Significant differences betweenthe systems and the CWSP are shown above the symbol / and significant differencesbetween systems are shown below the symbol /: ⁄P < 0.05, ⁄⁄P < 0.01 and⁄⁄⁄P < 0.001.

Fig. 9. Western blot analysis of the C. albicans cell wall surface proteins (CWSP)used in the preparation of the ADSs. Proteins were separated on 12% SDS–PAGE(between the black arrows a profile with increased resolution of the low molecularweight proteins is shown) (A), and transferred onto a PVDF membrane to be probedwith serum from mice immunized with CWSP, ASD1, ASD2 or liposomes alone (B).

C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200 197

3.4. Analysis of antibody specificities by Western blotting

In order to determine whether the different ADS formulationswere similarly capable of delivering proteins, immunoblottingassays were performed by combining the CWSP proteins as anti-gens with the serum from immunized mice. Reactivity of serumobtained from mice immunized with CWSP revealed the presenceof antibodies specific for protein bands located at a molecularweight of circa 60 kDa, for a protein band of approximately40 kDa and for others less intensely, between 10 and 15 kDa(Fig. 9). Curiously, serum from mice immunized with ADS1 con-tained antibodies mainly reactive with a protein (or proteins) witha molecular weight of approximately 200 kDa, while serum frommice immunized with ADS2 mainly contained antibodies specificfor this protein but also for proteins with a molecular weight of�60 kDa, as observed for the serum of mice immunized withCWSP. The reactivity detected with serum of mice immunized withADS2 is in agreement with the fact that, with this lipid formulation,only 16% of the proteins are entrapped and most is free, as in miceimmunized with CWSP alone. Protein sequence analysis revealedthat the protein that was mainly identified by antibodies induced

by ADS1 was Cht3p (60.0 kDa) (Fig. 9, band 1), the major chitinasethat has been reported to be present at C. albicans cell wall [51]. Incontrast, the protein identified by serum of mice immunized withCWSP with a molecular weight circa 40 kDa was Xog1p (50 kDa)identified as exo-1,3-beta-glucanase (Fig. 9, band 3), that has alsobeen identified at C. albicans cell wall [52]. Unfortunately, the otherprotein identified by the serum of CWSP immunized mice with amolecular weight �60 kDa (Fig. 9, band 2) could not be identifiedby the method applied due to the presence of several proteins inthat molecular weight range. Nevertheless it is clear that theADS1 favours the induction of antibodies against Cht3p, a proteinwith none or very low reactivity to be delivered alone.

4. Discussion

In this study, we evaluated for the first time the ability ofDODAB:MO liposomes to act simultaneously as protein deliveryvehicles and immunoadjuvants. In previous studies, we observedthat the introduction of MO in DODAB based liposomes at a molarfraction lower than 0.5, allowed the liposomes to assemble mainlyinto spherical densely packed nanoparticles [25]. TheseDODAB:MO based liposomes offer a high surface area that maybe advantageous for adsorbing antigens for vaccine development.Therefore in the present study, we used these liposomes to prepareantigen delivery systems (ADS) with C. albicans cell wall surfaceproteins (CWSP) as antigens and evaluated their adjuvantpotential. Indeed, the surface of C. albicans is a significant sourceof candidal antigens [53] and previous studies have shown thatthe major cell wall components that elicit a response from the hostimmune system are proteins and mannoproteins, in which boththe carbohydrate and protein moieties are able to trigger theresponse [54–56].

The characterization of the liposomes prepared revealedinteresting features for the nanoparticle liposomal formulationsused. At low protein/lipid weight ratios (<0.014; higher lipid con-centrations) an increase in the nanoparticle size and heterogeneityis observed, in comparison with the naked liposomes, accompaniedby a decrease in f-potential. This can be explained stepwise consid-ering the decreasing effect of liposomes concentration for a fixed

198 C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200

amount of protein. For the lowest protein/lipid weight ratio (0.006)there is the highest amount of cationic vesicles, and hence the zetapotential presents positive values although slightly smaller com-paring to the naked vesicles of DODAB:MO, indicating the begin-ning of some charge shielding made by the protein negativelycharged residues. This charge neutralization may reduce the repul-sion of the vesicles that tend to aggregate and thus the increase insize and PDI is due to the formation of aggregates of different sizes.In addition, the protein adsorbed at the surface might behave aslinker between some positively charged vesicles promoting theaggregation, as represented in the schematic drawing proposed(Fig. 2II). At ratio of 0.028 and above (lower lipid concentrations)the nanoparticles are now negatively charged, the size is reducingstabilizing at a protein/lipid ratio of 0.056 with lower PDI and col-loidal stability. This stabilization of liposomes against aggregationmay be due to the adsorption of a protein layer around the surfaceof the liposomes and could result in a progressive coating of thecationic head groups of the liposomes with the negatively chargedCWSP and thereby forming a protein corona as already suggestedby other authors [48,57,58]. In fact, one of the main advantagesof using positively charged liposomes lies on its ability to adsorboppositely charged antigens. Furthermore, complexation of lipo-somes with the CWSP might also occur by the hydrophobic effectsince some of these proteins might exhibit a substantial proportionof hydrophobic alpha-helix domains as secondary structures whichwill be prone to interact with hydrophobic borders of theDODAB:MO layers. In fact, the main protein recognized by theserum of mice immunized with ADS1, the Cht3, is indicated as aprobable unstable protein, considering the method of Guruprasadet al. [59].

In a previous work we characterized the DODAB:MO liposomesystem used in this study by Cryo-TEM and showed the presenceof larger and smaller vesicles with some internal structures [25].Indeed, Cryo-SEM performed in this work confirmed spherical ves-icles of variable sizes, and in some, smaller vesicles inside, in accor-dance with Cryo-TEM observations. In addition, Cryo-SEM showedthat the addition of proteins does not change the main morphologyof the liposomes but instead stabilizes the nanoparticles to a moreuniform size.

One of the advantages of using liposomes is the possibility ofquantitative evaluation of CWSP adsorption by means of ultracen-trifugation to separate free from adsorbed proteins. ADS1 showedan adsorption efficiency of more than 90% and ADS2 around 17%with ADS1 being thus an excellent vehicle for CWSP adsorption.We also observed that from the 50 lg/ml of CWSP initially added,the amount of protein effectively adsorbed into ADS1 and ADS2equals 45.5 and 8.25 lg/ml, respectively, resulting in an effectiveratio of protein/lipid of 0.05 for ADS1 and 0.06 for ADS2, when tak-ing into account the amount of lipid used to prepare the nanopar-ticles. Since ADS1 and ADS2 present similar liposomal sizes itappears that the stabilizing protein corona is achieved at a pro-tein/lipid ratio of around 0.05–0.06 and that addition of more lip-osomes does not significantly increased the proteins absorbed,instead, the formulations aggregate and precipitation ensues. Asexpected, at an intermediate weight ratio of 0.113 approximately45% of CWSP added were adsorbed confirming the optimal ratiofor protein corona formation. A schematic representation of thepossible liposome structures based on electrostatic interactionswith the CWSP is presented in Fig. 2.

The immunostimulatory potential of these DODAB:MO basednanoparticles was evaluated but firstly their cytotoxic effect wasexamined. Cationic liposomes are well known for their toxicity incomparison with neutral or anionic liposomes and DODABliposomes were previously shown to be cytotoxic to macrophagecultures at concentrations P50 lg/ml [60]. However, since theconcentration of DODAB needed for the preparation of the liposo-

mal systems used in this study is reduced by the introduction ofMO, the cytotoxic effects associated with high doses of DODABwere avoided and all formulations showed viabilities above 80%.Indeed, cationic liposomes cytotoxic effects are well-described[61] and that is why the interest in neutral and anionic liposomesis increasing [62]. However, the adsorption of CWSP masked thecationic charge of liposomes and that might be the reason whypronounced toxicity was avoided when ADS where incubated withthe cells. Moreover, the negative charge of ADSs did not interferewith cell uptake. In fact, ADS1 was avidly internalized by macro-phages as observed by the gradually intracellular accumulationof rhodamine labelled ADS1, stained with Alexa Fluor 633, in dis-crete sites membrane-rich regions. This pattern of internalizationis similar to endocytosis pattern observed in other studies[49,50] and was estimated to occur within 20 min, in agreementwith other studies that assessed endocytosis of a liposome within30 min [49,50]. The mean size of the two ADSs selected may alsocontribute for adequate cell uptake since size range for optimaluptake of antigens and elicitation of a cellular response by den-dritic cells have been determined to be below 500 nm [18,63].

After confirmation of internalization, it was important to evalu-ate whether the liposomes were able to activate macrophages. Ourstudies indicated that only, ADS2 and the CWSP were able toinduce the production of the pro-inflammatory cytokine, TNF-a,in J774 cells. Only residual TNF-a was obtained upon incubationof these cells with ADS1 or liposomes alone. Several studies havealready reported that the cell wall components of C. albicans havethe capacity to induce the release of pro-inflammatory cytokinesupon incubation with phagocytic cells [64–67]. The fact thatADS1, contrary to ADS2, fails to stimulate these cells is probablydue to all CWSP being entrapped/protected and unavailable to acti-vate whereas free non-encapsulated CWSP (approx. 83%) arebelieved to be present in ADS2. Therefore, as discussed previously,the phagocytic cells rapidly internalize ADS1, but fail to activateTNF-a secretion. Similarly to these observations, it was previouslydescribed that a C. albicans recombinant mannoprotein was trans-ported into the compartments of endosomes and lysosomes,degraded and loaded on major histocompatibility molecules with-out stimulating pro-inflammatory cytokines production [68]. Thelack of evident TNF-a production could be an interesting featureof that delivery system by preventing deleterious side effects thatcould result from excessive inflammation still allowing antigen-delivery and the generation of an immune response [69].

Due to the results of the studies on the activation of macro-phages it was thus important to evaluate whether the liposomeswere able to activate the immune system in vivo and generatehumoral and cellular responses. As expected, immunization withCWSP and ADS2 induced similar levels of IgGs, while immuniza-tion with ADS1 induced significantly higher levels of IgGs thanimmunization with ADS2 (P < 0.05). The ratio of IgG1:IgG2a inmice immunized with the CWSP was 21.4, suggesting a Th2-typeimmune response was generated. However, in mice immunizedwith ADS1, this ratio was three times lower in comparison withthe one detected when using free-proteins, indicating that immu-nization with ADS1, promoted a shifted in the Th1/Th2 balancetowards the Th1 type (P < 0.05), as observed in previous studies[12]. These results show that increasing the amount of DODAB:MOliposomes used to incorporate the CWSP (ADS1) is effective inenhancing not only the levels of CWSP-IgGs but also the balancetowards a Th1-type response, which is known to be protective inC. albicans infections [70]. In accordance with this, splenocytes ofADS1 immunized mice produced higher levels of IFN-c and Il-17than splenocytes from mice immunized with the other for-mulations after re-stimulation with the proteins. These are impor-tant indicators of T-cell activation and protective immunity andnumerous vaccine studies have shown their importance in

C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200 199

C. albicans vaccine efficacy [71–74]. Finally, activation of a cell-mediated immune response was also shown by the observationof splenocyte proliferation after specific re-stimulation (P < 0.05).

An important feature of using different formulations to deliverproteins is the evaluation of whether the antibodies induced bythe different liposomes/CWSP are reactive against the same pro-teins. In our work, immunoblotting using CWSP as antigensrevealed major differences in the serum CWSP-specific antibodiesproduced by the different groups of immunized mice. Serum fromADS1 immunized mice showed a pattern of CWSP-specific antibod-ies that hybridized mainly with a high molecular weight proteinidentified as Cht3p (C. albicans cell wall chitinase). Although themolecular weight estimated in our analysis was of around 150–200 kDa, the molecular weight of this protein has been estimatedto be around 60.0 kDa. This discrepancy is probably due to glyco-sylation of the protein, indeed it has been described that CHT3has four N-glycosylation sites (Asn-Xaa-Ser/Thr) and several sitesfor 0-mannosylation [75]. In contrast, serum from mice immunizedwith CWSP alone presented antibodies against proteins around60 kDa (not identified) and the 50 kDa Xog1p (C. albicans exo-1,3-beta-glucanase). Curiously, ADS2 showed a pattern of hybrid-ization that appears to include the range of proteins that hybridizewith serum from mice immunized with CWSP alone and withADS1. The differences in the hybridization patterns between theADS1 and ADS2 are in agreement with the fact that ADS2 entrapsonly 16% of the CWSP while most is free, as in mice immunizedwith CWSP alone and ADS1 entraps 91% of the CWSP, presentinga different pattern. The major differences between Xog1p andCht3p, besides the molecular weight, is the fact that Cht3p isunstable in vitro. These results reveal interesting features for theseliposomal formulations, showing that these liposomes not onlyenhance the immune response against proteins from the patho-genic yeast C. albicans but also adsorb and efficiently deliver a pro-tein considered to be unstable in vitro, that otherwise was notefficiently delivered.

5. Conclusions

In this study, we prepared and evaluated for the first time theability of DODAB:MO liposomes to act simultaneously as proteindelivery vehicles and immunoadjuvants. The selected liposomeformulations were stable, non-cytotoxic, avidly taken up by macro-phages and showed a good adjuvant activity, inducing both strongantibody responses and cell-mediated immunity. In contrast toimmunization with CWSP alone, these systems induced high levelsof IL-17 and IFN-c but low levels of IL-10, indicating polarizationtowards a Th1-type immune response and confirming their immu-noadjuvant potential. Additionally, these systems favoured theinduction of antibodies against proteins different to those inducedby CWSP alone, with mainly a protein(s), probably glycosylated, ofhigh molecular weight being observed. The advantage of theseliposomes may be associated with the fact that, as particulate car-riers, they entrap and therefore protect the specific protein(s)against degradation.

Conflict of interest

The authors declare that they have no conflict of interest.

Acknowledgements

We thank Professor Eduardo F. Marques, Faculdade de Ciências,Universidade do Porto, and Professor Claudia Botelho, Centre ofBiological Engineering, Minho University, for technical supportwith the Cryo-SEM and confocal microscopy analysis, respectively.

We thank also Deborah Penque from the Laboratório de Proteómi-ca, Instituto Nacional de Saúde Dr. Ricardo Jorge INSA I.P., Lisbon,Portugal, and Patrícia Alves from the Mass Spectrometry Unit(UniMS), at ITQB/iBET, Oeiras, Portugal, for their help with theproteins identification.

This work was supported by FCT/MEC through Portuguesefunds (PIDDAC) - PEst-OE/BIA/UI4050/2014, PEst-C/FIS/UI0607/2013 (CFUM) and PTDC/QUI/69795/2006, while Catarina Carneiroholds scholarship SFRH/BD/69068/2010. We acknowledge Nano-Delivery-I&D em Bionanotecnologia, Lda. for access to theirequipment.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.ejpb.2014.11.028.

References

[1] M. Henriksen-Lacey et al., Liposomal vaccine delivery systems, Expert Opin.Drug Deliv. 8 (4) (2011) 505–519.

[2] F. Azmi et al., Recent progress in adjuvant discovery for peptide-based subunitvaccines, Hum. Vaccines Immunother. 10 (3) (2013).

[3] A.R. Mohammed et al., Increased potential of a cationic liposome-baseddelivery system: enhancing stability and sustained immunological activity inpre-clinical development, Eur. J. Pharm. Biopharm. 76 (3) (2010) 404–412.

[4] A.E. Gregory, R. Titball, D. Williamson, Vaccine delivery using nanoparticles,Front. Cell. Infect. Microbiol. 3 (2013) 13.

[5] V.B. Joshi, S.M. Geary, A.K. Salem, Biodegradable particles as vaccine deliverysystems: size matters, AAPS J. 15 (1) (2013) 85–94.

[6] A. Garcia, J.B. De Sanctis, An overview of adjuvant formulations and deliverysystems, APMIS 122 (4) (2014) 257–267.

[7] A.C. Allison, Immunological adjuvants and their modes of action, Arch.Immunol. Ther. Exp. (Warsz) 45 (2–3) (1997) 141–147.

[8] P.M. Daftarian et al., A targeted and adjuvanted nanocarrier lowers theeffective dose of liposomal amphotericin B and enhances adaptive immunity inmurine cutaneous leishmaniasis, J. Infect. Dis. 208 (11) (2013) 1914–1922.

[9] N.K. Jain, V. Mishra, N.K. Mehra, Targeted drug delivery to macrophages, ExpertOpin. Drug Deliv. 10 (3) (2013) 353–367.

[10] K.A. Ghaffar et al., Liposomes as nanovaccine delivery systems, Curr. Top. Med.Chem. (2014).

[11] P. Vabbilisetty, X.L. Sun, Liposome surface functionalization based on differentanchoring lipids via Staudinger ligation, Org. Biomol. Chem. 12 (8) (2014)1237–1244.

[12] J. Davidsen et al., Characterization of cationic liposomes based ondimethyldioctadecylammonium and synthetic cord factor from M.tuberculosis (trehalose 6,60-dibehenate) – a novel adjuvant inducing bothstrong CMI and antibody responses, Biochim. Biophys. Acta 1718 (1–2) (2005)22–31.

[13] P.T. Ingvarsson et al., Designing CAF-adjuvanted dry powder vaccines: spraydrying preserves the adjuvant activity of CAF01, J. Control. Release 167 (3)(2013) 256–264.

[14] J.P. Neves Silva, P.J. Coutinho, M.E. Real Oliveira, Characterization ofmonoolein-based lipoplexes using fluorescence spectroscopy, J. Fluoresc. 18(2) (2008) 555–562.

[15] M.J. Hussain et al., Th1 immune responses can be modulated by varyingdimethyldioctadecylammonium and distearoyl-sn-glycero-3-phosphocholinecontent in liposomal adjuvants, J. Pharm. Pharmacol. 66 (3) (2014) 358–366.

[16] A.C. Oliveira et al., Dioctadecyldimethylammonium:monoolein nanocarriersfor efficient in vitro gene silencing, ACS Appl. Mater. Interfaces (2014).

[17] A.M. Carmona-Ribeiro, Biomimetic particles in drug and vaccine delivery, J.Liposome Res. 17 (3–4) (2007) 165–172.

[18] N. Lincopan et al., Novel immunoadjuvants based on cationic lipid:preparation, characterization and activity in vivo, Vaccine 27 (42) (2009)5760–5771.

[19] N. Lincopan et al., Silica-based cationic bilayers as immunoadjuvants, BMCBiotechnol. 9 (2009) 5.

[20] A.M. Carmona-Ribeiro, Bilayer-forming synthetic lipids: drugs or carriers?,Curr Med. Chem. 10 (22) (2003) 2425–2446.

[21] M. Ogris et al., The size of DNA/transferrin-PEI complexes is an importantfactor for gene expression in cultured cells, Gene Ther. 5 (10) (1998) 1425–1433.

[22] J.P. Silva et al., Structural dynamics and physicochemical properties of pDNA/DODAB:MO lipoplexes: effect of pH and anionic lipids in inverted non-lamellarphases versus lamellar phases, Biochim. Biophys. Acta 1838 (10) (2014) 2555–2567.

[23] J.P. Silva et al., DODAB:monoolein-based lipoplexes as non-viral vectors fortransfection of mammalian cells, Biochim. Biophys. Acta 1808 (10) (2011)2440–2449.

200 C. Carneiro et al. / European Journal of Pharmaceutics and Biopharmaceutics 89 (2015) 190–200

[24] L. Sagalowicz, R. Mezzenga, M.E. Leser, Investigating reversed liquid crystallinemesophases, Curr. Opin. Colloid Interface Sci. 11 (2006) 224–229.

[25] I.M. Oliveira et al., Aggregation behavior of aqueousdioctadecyldimethylammonium bromide/monoolein mixtures: amultitechnique investigation on the influence of composition andtemperature, J. Colloid Interface Sci. 374 (1) (2012) 206–217.

[26] M.L. Lynch et al., Enhanced loading of water-soluble actives into bicontinuouscubic phase liquid crystals using cationic surfactants, J. Colloid Interface Sci.260 (2) (2003) 404–413.

[27] M. Caffrey, A lipid’s eye view of membrane protein crystallization inmesophases, Curr. Opin. Struct. Biol. 10 (4) (2000) 486–497.

[28] X. Ai, M. Caffrey, Membrane protein crystallization in lipidic mesophases:detergent effects, Biophys. J. 79 (1) (2000) 394–405.

[29] S. El-Kirat-Chatel et al., Single-molecule analysis of the major glycopolymers ofpathogenic and non-pathogenic yeast cells, Nanoscale 5 (11) (2013) 4855–4863.

[30] D. Pietrella et al., A beta-glucan-conjugate vaccine and anti-beta-glucanantibodies are effective against murine vaginal candidiasis as assessed by anovel in vivo imaging technique, Vaccine 28 (7) (2010) 1717–1725.

[31] A. Torosantucci et al., A novel glyco-conjugate vaccine against fungalpathogens, J. Exp. Med. 202 (5) (2005) 597–606.

[32] L. Paulovicova et al., Humoral and cell-mediated immunity followingvaccination with synthetic Candida cell wall mannan derivedheptamannoside-protein conjugate: immunomodulatory properties ofheptamannoside-BSA conjugate, Int. Immunopharmacol. 14 (2) (2012) 179–187.

[33] G. Luo et al., Candida albicans Hyr1p confers resistance to neutrophil killingand is a potential vaccine target, J. Infect. Dis. 201 (11) (2010) 1718–1728.

[34] A.S. Ibrahim et al., Vaccination with recombinant N-terminal domain of Als1pimproves survival during murine disseminated candidiasis by enhancing cell-mediated, not humoral, immunity, Infect. Immun. 73 (2) (2005) 999–1005.

[35] B. Spellberg et al., The antifungal vaccine derived from the recombinant Nterminus of Als3p protects mice against the bacterium Staphylococcus aureus,Infect. Immun. 76 (10) (2008) 4574–4580.

[36] G. Luo et al., Active and passive immunization with rHyr1p-N protects miceagainst hematogenously disseminated candidiasis, PLoS One 6 (10) (2011)e25909.

[37] H. Xin et al., Synthetic glycopeptide vaccines combining beta-mannan andpeptide epitopes induce protection against candidiasis, Proc. Natl. Acad. Sci.USA 105 (36) (2008) 13526–13531.

[38] H. Xin, J.E. Cutler, Vaccine and monoclonal antibody that enhance mouseresistance to candidiasis, Clin. Vaccine Immunol. 18 (10) (2011) 1656–1667.

[39] E. Paulovicova et al., Synthetically prepared glycooligosaccharides mimickingCandida albicans cell wall glycan antigens – novel tools to study host–pathogeninteractions, FEMS Yeast Res. (2013).

[40] D.P. Thomas et al., A proteomic-based approach for the identification ofCandida albicans protein components present in a subunit vaccine thatprotects against disseminated candidiasis, Proteomics 6 (22) (2006) 6033–6041.

[41] M.R. Insenser et al., Gel and gel-free proteomics to identify Saccharomycescerevisiae cell surface proteins, J. Proteomics 73 (6) (2010) 1183–1195.

[42] M.M. Bradford, A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein–dyebinding, Anal. Biochem. 72 (1976) 248–254.

[43] A.D. Bangham, M.M. Standish, J.C. Watkins, Diffusion of univalent ions acrossthe lamellae of swollen phospholipids, J. Mol. Biol. 13 (1) (1965) 238–252.

[44] R. Sabino et al., Isolates from hospital environments are the most virulent ofthe Candida parapsilosis complex, BMC Microbiol. 11 (2011) 180.

[45] M. Madesh, K.A. Balasubramanian, A microtiter plate assay for superoxideusing MTT reduction method, Indian J. Biochem. Biophys. 34 (6) (1997) 535–539.

[46] A. Hoess et al., Cultivation of hepatoma cell line HepG2 on nanoporousaluminum oxide membranes, Acta Biomater. 3 (1) (2007) 43–50.

[47] P. Ferreirinha et al., Protective effect of intranasal immunization with Neosporacaninum membrane antigens against murine neosporosis established throughthe gastrointestinal tract, Immunology 141 (2) (2014) 256–267.

[48] Y. Perrie et al., A case-study investigating the physicochemical characteristicsthat dictate the function of a liposomal adjuvant, Hum. Vaccines Immunother.9 (6) (2013).

[49] M.S. Martina et al., The in vitro kinetics of the interactions between PEG-ylatedmagnetic-fluid-loaded liposomes and macrophages, Biomaterials 28 (28)(2007) 4143–4153.

[50] D.L. Daleke, K. Hong, D. Papahadjopoulos, Endocytosis of liposomes bymacrophages: binding, acidification and leakage of liposomes monitored bya new fluorescence assay, Biochim. Biophys. Acta 1024 (2) (1990) 352–366.

[51] M.L. Hernaez et al., Identification of Candida albicans exposed surface proteinsin vivo by a rapid proteomic approach, J. Proteomics 73 (7) (2010) 1404–1409.

[52] P.W. Tsai et al., Characterizing the role of cell-wall beta-1,3-exoglucanaseXog1p in Candida albicans adhesion by the human antimicrobial peptide LL-37,PLoS One 6 (6) (2011) e21394.

[53] G.G. Gauglitz et al., Host defence against Candida albicans and the role ofpattern-recognition receptors, Acta Derm. Venereol. 92 (3) (2012) 291–298.

[54] P.W. de Groot et al., Adhesins in human fungal pathogens: glue with plenty ofstick, Eukaryot. Cell 12 (4) (2013) 470–481.

[55] J.L. Lopez-Ribot et al., Antibody response to Candida albicans cell wall antigens,FEMS Immunol. Med. Microbiol. 41 (3) (2004) 187–196.

[56] D.W. Lowman et al., Novel structural features in Candida albicans hyphalglucan provide a basis for differential innate immune recognition of hyphaeversus yeast, J. Biol. Chem. 289 (6) (2014) 3432–3443.

[57] A. Hirano et al., Adsorption and disruption of lipid bilayers by nanoscaleprotein aggregates, Langmuir 28 (8) (2012) 3887–3895.

[58] M. Hamborg et al., Protein antigen adsorption to the DDA/TDB liposomaladjuvant: effect on protein structure, stability, and liposome physicochemicalcharacteristics, Pharm. Res. 30 (1) (2013) 140–155.

[59] K. Guruprasad, B.V. Reddy, M.W. Pandit, Correlation between stability of aprotein and its dipeptide composition: a novel approach for predicting in vivostability of a protein from its primary sequence, Protein Eng. 4 (2) (1990) 155–161.

[60] K.S. Korsholm et al., The adjuvant mechanism of cationicdimethyldioctadecylammonium liposomes, Immunology 121 (2) (2007)216–226.

[61] A.M. Carmona-Ribeiro, Biomimetic nanoparticles: preparation,characterization and biomedical applications, Int. J. Nanomed. 5 (2010) 249–259.

[62] C. Kelly, C. Jefferies, S.A. Cryan, Targeted liposomal drug delivery to monocytesand macrophages, J. Drug Deliv. 2011 (2011) 727241.

[63] S.D. Xiang et al., Pathogen recognition and development of particulatevaccines: does size matter?, Methods 40 (1) (2006) 1–9

[64] A.D. de Boer et al., The Candida albicans cell wall protein Rhd3/Pga29 isabundant in the yeast form and contributes to virulence, Yeast 27 (8) (2010)611–624.

[65] D. Pietrella et al., Candida albicans mannoprotein influences the biologicalfunction of dendritic cells, Cell. Microbiol. 8 (4) (2006) 602–612.

[66] K. Ueno et al., The mannan of Candida albicans lacking beta-1,2-linkedoligomannosides increases the production of inflammatory cytokines bydendritic cells, Med. Mycol. 51 (4) (2013) 385–395.

[67] M. Martinez-Esparza et al., Glycoconjugate expression on the cell wall of tps1/tps1 trehalose-deficient Candida albicans strain and implications for itsinteraction with macrophages, Glycobiology 21 (6) (2011) 796–805.

[68] D. Pietrella et al., A Candida albicans mannoprotein deprived of its mannanmoiety is efficiently taken up and processed by human dendritic cells andinduces T-cell activation without stimulating proinflammatory cytokineproduction, Infect. Immun. 76 (9) (2008) 4359–4367.

[69] M. Inoue et al., T cells down-regulate macrophage TNF production by IRAK1-mediated IL-10 expression and control innate hyperinflammation, Proc. Natl.Acad. Sci. USA 111 (14) (2014) 5295–5300.

[70] B. Spellberg, Vaccines for invasive fungal infections, F1000 Med. Rep. 3 (2011)13.

[71] N. Hernandez-Santos, S.L. Gaffen, Th17 cells in immunity to Candida albicans,Cell Host Microbe 11 (5) (2012) 425–435.

[72] L. Romani, Immunity to fungal infections, Nat. Rev. Immunol. 11 (4) (2011)275–288.

[73] L. Lin et al., Th1–Th17 cells mediate protective adaptive immunity againstStaphylococcus aureus and Candida albicans infection in mice, PLoS Pathog. 5(12) (2009) e1000703.

[74] B. Spellberg et al., Antibody titer threshold predicts anti-candidal vaccineefficacy even though the mechanism of protection is induction of cell-mediated immunity, J. Infect. Dis. 197 (7) (2008) 967–971.

[75] K.J. McCreath, C.A. Specht, P.W. Robbins, Molecular cloning andcharacterization of chitinase genes from Candida albicans, Proc. Natl. Acad.Sci. USA 92 (7) (1995) 2544–2548.