Pharmaceutical Nanotechnology
Demonstration of the interactions between aromatic
compound-loaded lipid nanocapsules and Acinetobacter baumannii bacterial membrane
A. Montagu
a,b,*, M-L. Joly-Guillou
c, C. Guillet
d, J. Bejaud
a,b, E. Rossines
e, P. Saulnier
a,baLUNAMUniversité,F-49933Angers,France
bINSERMU1066,MicroetNanomédecinesBiomimétiques,IBS-CHU,4RueLarrey,49933Angers,France
cATOMycA,INSERMAtip-AvenirTeam,CRCNA,INSERMU892,6299CNRS,UniversityofAngers,4RueLarrey,49933Angers,France
dServiceCommundeCytometrieetd’analyseNucleotidique(SCCAN),IFR132,IBS–CHU,4RueLarrey,49933Angers,France
eEydopharma,41RueNoelBallay,28000Chartres,France
ARTICLE INFO Articlehistory:
Received12November2015 Accepted19March2016 Availableonline30March2016 Keywords:
Multi-drugresistantbacteria Aromaticcompounds DiO-lipidnanocapsules Hydroxylfunctions Pumpmechanism
ABSTRACT
Acinetobacterbaumanniiisanimportantnosocomialpathogenthatisresistanttomanycommonly-used antibiotics.Onestrategyfortreatmentistheuseofaromaticcompounds(carvacrol,cinnamaldehyde) againstA.baumannii.Theaimofthisstudywastodeterminetheinteractionsbetweenbacteriaandlipid nanocapsules (LNCs) over time based on the fluorescence of 3,30-Dioctadecyloxacarbocyanine Perchlorate-LNCs (DiO-LNCs) and the properties of trypan blue to analyse the physicochemical mechanismsoccurringatthelevelofthebiologicalmembrane.Theresultsdemonstratedthecapacityof carvacrol-loaded LNCstointeract withand penetratethebacterialmembranein comparisonwith cinnamaldehyde-loaded LNCs and unloaded LNCs. Modifications of carvacrol after substitution of hydroxyl functional groups by fatty acids demonstrated the crucial role of hydroxyl functions in antibacterial activity. Finally, after contact with the efflux pump inhibitor, carbonylcyanide-3- chlorophenyl hydrazine(CCCP), the resultsindicated thetotal synergisticantibacterialeffect with Car-LNCs,showingthatCCCPisassociatedwiththeactionmechanismofcarvacrol,especiallyatthelevel oftheeffluxpumpmechanism.
ã2016ElsevierB.V.Allrightsreserved.
1.Introduction
Acinetobacter baumannii is a Gram-negative, non-motile coccobacillus that has become a major cause of nosocomial infections with the overuse of broad-spectrum antibiotics. A.
baumanniicanbeselectedfor theemergenceofmulti-resistant strains.Thisbacteriacausespneumonia,especiallyinmechanically ventilated patients. Infections can occur by urinary catheters, intravenousdevicesoratsurgicalsites(Eveillardetal.,2009).A.
baumanniicandevelopvariousmechanismsofresistance,suchas productionof
b
-lactamases,effluxpumps,lowerpermeabilityof theoutermembrane,mutationsinantibiotictargets(quinolones)and production of enzymes which inactivate aminoglycosides (KarageorgopoulosandFalagas,2008).Thechoiceofanappropri- ateantibacterialtherapyiscomplicatedduetotheresistanceof A.baumanniitomanyantibiotics.Moreover,thesemechanismsare associatedwithgeneticsupportmodifications(mutations,trans- posonacquisitions,plasmids,integrons,andinsertionsequences promoters)(Pelegetal.,2008)thatcantransformthemintohighly multidrugresistantbacteria(MRB).Thebacteriahavetheabilityto acquire multidrugresistance, affectingthe expression ofporins and/orefflux pump(s),which canaffectunrelatedantimicrobial agents(Vilaetal.,2007).
Consideringthelimitednumberofantibioticsindevelopment, novel treatmentstrategies includetheuseof naturalresources, especiallyessentialoils(EOs).TheantimicrobialactivityofEOsis largelyduetothepresenceoftheirmajorcomponents,aromatic molecules, such as phenols (carvacrol, Car) and aldehydes (cinnamaldehyde,Cin)(Bakkalietal.,2008;Burt,2004).Essential oil components(EOCs)havedemonstratedantibacterialefficacy againstawidevarietyofmicroorganisms,includingMRBs(Yeetal.,
*Correspondingauthorat:INSERMU1066,MicroetNanomédecinesBiomimé- tiques,IBS–CHU,4RueLarrey,49933Angers,France.
E-mailaddresses:[email protected](A.Montagu),
[email protected](M.-L.Joly-Guillou),[email protected] (C. Guillet),[email protected](J. Bejaud),[email protected] (E.Rossines),[email protected](P.Saulnier).
http://dx.doi.org/10.1016/j.ijpharm.2016.03.033 0378-5173/ã2016ElsevierB.V.Allrightsreserved.
ContentslistsavailableatScienceDirect
International Journal of Pharmaceutics
j o u r n al h o m ep a g e: w w w . el s e v i e r . c o m / l o c at e / i j p h a r m
2013). However, EOCs have lipophilic characteristics, reducing theirinvivobioavailabilityandantibacterialactivity.EOCscanbe encapsulatedinsideahydrophilicsystem,suchasthatoflipidic nanocapsules(LNCs),whichhavedemonstratedsuitabilityforthe encapsulationoflipophiliccompounds,enhancingtheantibacte- rialefficacyandpromotingcontactwithbacterialcells.Arecent studyunderlinedtheantibacterialpotentialofEOCs-loadedLNCs (carvacrol,eugenol,cinnamaldehyde)againstA.baumanniiforan invivomodelofsepsisinmice(Montaguetal.,2014).Currently,no specificstudyhasfocusedontheinteractions betweenbacteria and LNCs over time or on the physicochemical mechanisms occurringatthelevelofthebiologicalmembrane.Thisbacterial membrane presents selective permeability controllingthe flow using porins and efflux pumps. The use of nanocarriers could facilitatethecrossingofthesemembranes.
Numerousstudieshavedescribedtheuseoffluorescentdyesto marknanoparticlestostudytheirbio-distribution.Thecarbocya- nine family includes DiO, a lipophilic tracer incorporated into lipids,whichhasbeenusedasamarkerofPLGA(poly(lactide-co- glycolide))nanoparticles(Gaumetetal.,2009).Trypanbluehas already been used in staining methods to quench the green autofluorescenceofcells(Srivastavaetal.,2011).Thepropertiesof trypanbluecouldbeappliedtoquenchDiO-LNCsatthebacterial membrane.
Forthefirsttime,thisresearchfocusesontherealizationand characterizationofDiO-LNCsloadedwithantibacterialactivesand DiO-unloaded LNCs. The antibacterial activity of these loaded nano-carrierswasevaluatedinvitroagainstA.baumanniiandin thepresenceofaneffluxpumpinhibitorCCCP(carbonylcyanide-3- chlorophenyl hydrazine).Finally,we determine theinteractions betweenbacteriaandLNCsovertimebasedonthefluorescenceof DiO-LNCs and the properties of trypan blue to investigate the physicochemical mechanisms occurring at the level of the biologicalmembrane.
2.Materialandmethods
2.1.Chemicalsmaterials
5-Isopropyl-2 methylphenol (Carvacrol, Car) was purchased from Sigma-Aldrich (Saint-Louis, USA). Trans-cinnamaldéhyde (Cin) was purchased from Merck-Millipore (Molsheim,France).
The lipophilic Labrafac1 WL1349 (caprylic-capricacidtriglycer- ides)was purchased from GattefosseS.A. (Saint-Priest,France).
Lipoïd1 S75-3 (soybeanlecithin at 69%of phosphatidylcholine) came from Lipoïd Gmbh (Ludwigshafen, Germany). Kolliphor1 HS15(amixtureoffreepolyethyleneglycol660andpolyethylene glycol 660 (hydroxystearate) was from BASF (Ludwigshafen, Germany) and NaCl was from Prolabo (Fontenay-sous-bois, France).DeionizedwaterwasacquiredfromaMilli-Qplussystem (Millipore, Paris, France) and sterile water was from Cooper (Melun,France).
Chemical syntheses of carvacrol derivatives were achieved by esterification by the condensation of carboxylic acid with the hydroxyl group of carvacrol. The reaction occurred in dichloromethane in the presence of dicyclohexylcarbodiimide (Sigma–Aldrich) and dimethylaminopyridine (Sigma–Aldrich).
Twoderivativesofcarvacrolwereproducedafterpurificationon asilicagelcolumn.Tworadicalswereaddedattheestergroup:the firstoneincludingaceticacidandthesecondpalmiticacid(PA).
Thecarbocyanine-basedfluorophore,DiO(3,30-Dioctadecylox- acarbocyaninePerchlorateem.=484;exc.=501nm)wassupplied byInvitrogen(Eugene,OR,USA).Phosphate-bufferedsaline(PBS) wasobtainedfromLonza(Verviers,Belgium).
CCCP (carbonylcyanide-3-chlorophenyl hydrazone), an efflux pumpinhibitor(EPI),waspurchasedfromSigma–Aldrich.
2.2.PreparationofblankLNCs,CarCin-LNCs
LNCs were prepared according tothe method described by Heurtaultetal.(2002).Briefly,theformulationconsistedofmixing all of the components, (Kolliphor1 HS15, Labrafac1 WL1349, Lipoid1S75-3,NaClanddeionizedwater)undermagneticstirring and heating from room temperature to 90C. Two cycles of progressivecoolingandheatingbetween90Cand60Cwerethen performedtohomogenize themixture.Then,activecompounds wereaddedtothemixtureatdifferentcyclesaccordingtotheir phase inversiontemperature(PIT) whentheyare introducedin emulsion systems. Finally, an irreversible shock induced by a suddendilutionofthemixturewithcoldwater(69.36%w/w)was performedaccordingtothePITofthefinalmixture.Slowmagnetic stirringwasthenappliedtothesuspensionfor5min.TheCarCin- LNCs(carvacrolandcinnamaldehydeloadedLNCs)wereobtained atafinalconcentrationof29mgofactivespergofLNCssuspension (17mg/gofcarvacroland12mg/gofcinnamaldehyde).Unloaded- LNCs (blank-LNCs), Car-LNCs, modified Car-LNCs and Cin-LNCs werepreparedasdescribedpreviously,withthesameproportion ofeachactiveintheCarCin-LNCs(17mg/gofcarvacroland12mg/g ofcinnamaldehyde)(Table1).Theconcentrationofexcipientswas 131mg/gto165mg/gforthefiveformulations.FortheDiO-loaded LNCs,DiOwasaddedtomixtureofthecomponentstoobtainafinal concentrationof1mgpargofLabrafac1.
2.3.Characterizationofblank-LNCs,Car-LNCs,Cin-LNCs,CarCin-LNCs The average hydrodynamic diameter and the polydispersity index (PdI) of the nanocapsules were determined at 25C, in triplicate,usingaMalvernZetasizerNanoZS(MalvernInstruments S.A., Worcestershire, UK). For the measurements, the LNC suspensionswerediluted1:60(v/v)indeionizedwater.
2.4.Determinationofantibacterialactivity
The antibacterial effects of Car, Cin and CarCin actives and blank-LNCs,Car-LNCs,Cin-LNCsandCarCin-LNCswereevaluated bytheminimalinhibitoryconcentrations(MICs)andkillkinetic studies. We used A.baumannii (AYE), a cephalosporinase-over- producingandexpandedspectrum
b
-lactamaseproducing(BLSE)strainresistanttomost
b
-lactams, aminoglycosidesandfluoro- quinolones. For the determination of the MICs, a bacterial suspension with a turbidity equivalent to the 0.5 McFarland standardwasprepared(108CFU/ml)inbrothliquidmedium.The solution was diluted in brain heart infusion (BHI, bioMerieux, Marcy l’Etoile, France). Briefly, in each wellof a 96-well plate, serially dilutedactivesoractives loaded-LNCsatconcentrations between0.045mg/mland5mg/mlwereusedinthepresenceof bacterial suspensions. The control corresponded to bacteria withoutactives.Aftera24-hincubationat37C,theMICvalues weredeterminedasthelowestconcentrationoftheantimicrobialTable1
Ratiooftheweight/weight(percentage)ofeachcomponentofblank-LNCs,CarCin- LNCs,Car-LNCs,CarAA-LNCs,CarPA-LNCsandCin-LNCs.
Percentageofw/w Blank-LNCs CarCin-LNCs Car-LNCs CarAA-LNCs CarPA-LNCs
Cin-LNCs
Carvacrol – 1.73 1,71 –
Cinnamaldehyde – 1.15 – 1.23
Kolliphor1HS15 4.83 4.70 4.76 4.75
NaCl 0.51 0.50 0.51 0.50
Lipoid1S75-3 0.43 0.42 0.44 0.42
Labrafac1WL1349 5.87 5.70 5.78 5.8
Deionizedwater 88.36 85.8 86.8 87.30
active that inhibited the visible growth of the microorganism tested.MBCs(minimalbactericidalkillconcentrations)werebased onthelowestconcentrationoftheextractsrequiredtokill99.9%of bacteriafromtheinitialinoculum asdetermined byplating on agar.TheMBCsofeachtestedproductinthepresenceofCCCPwere determinedintheBHIwherenogrowthwasdetectedafter48h.
TheFBC(fractionalkillconcentration)valueswerecalculatedby equation1.TheFBCindexwasinterpretedasfollows:totalsynergy, 0.5;partialsynergy,0.5–1;additiveeffect,1.0;andantagonism,
>1.0(Berenbaum,1978).
XFBC¼MBCcombination:a
MBCalone:a þMBCcombination:b
MBCalone:b ð1Þ
ThekillkineticsstudieswereperformedatonceandtwicetheMIC ofCarCin-LNCsinBHIwithabacterialsuspensionat105CFU/ml.
Thebacterialcountsweremeasuredat0h,1h,3h,6hand24h afterincubationat37Cinahotwaterbath.Ateachtime,bacterial suspensions were plated on Columbia agar with sheep blood (GmbH, Wesel, Germany). Then, colonies were counted. Each bacterialcountwasperformedintriplicate.Allexperimentswere performedatleastthreetimesinindependentconditions.
To explore the efflux pump mechanism, the antibacterial activityofcarvacrolandCar-LNCswasstudiedbydilutioninthe presence of CCCP. A plate containing CCCP alone was used as thecontrol. CCCP was added toplates containing 8.6
m
g/ml to550
m
g/ml Car-LNCsand 2.5m
g/ml to160m
g/mlCar alone.ThefinalconcentrationofCCCPintheplatewas12.5
m
g/ml.Allinvitroexperimentswereperformedatleastthreetimesinindependent conditions.
2.5.Flowcytometry
This bacterial suspension was prepared at 105CFU/ml in BHI. The solution was incubated with blank-LNCs, Car-LNCs, CarAA-LNCs,CarPA-LNCs,Cin-LNCs,CarCin-LNCsinBHIat37Cin ahotwaterbathinthedark(concentrationscorrespondingtothe MICofCarCin-LNCstolimitthedeadbacteriainthesamplewhich canaltertheinterpretationoftheresults)for10min,1h,2hand
3h. These studies were conducted over short times with no modificationsoftheLNCsphysicochemicalcharacteristicsinthe presenceofBHIandbacteria(datanotshown).Ateach timeof incubation,thebacterialsuspensionswerewashedthreetimesin PBS, and the solutions were treated with trypan blue (Lonza) (0.2mg/ml)for1minfollowedbytwoPBSrinses.Analyseswere performed witha MASCquant flowcytometer(Miltenyi). Allin vitro experiments were performed at least three times in independentconditions.
3.Results
3.1.Studyofthebehaviorofblank-LNCs,CarCin-LNCs,Car-LNCs,Cin- LNCs
3.1.1.PhysicochemicalpropertiesoftheLNCsuspensions
Themaindiameteroftheblank-LNCswas48.31.5nmwitha polydispersity index (PdI) of less than 0.2, indicating the monodispersity of the prepara tion (Table 2(A)). CarCin-LNCs showedanincreasedsizecomparedtoblanks-LNCs,dependingon theconcentrationsofactives.TheCarCin-LNCsweretwiceaslarge asblank-LNCs.Thisphenomenonis explainedbythestericand amphiphilic propertiesof carvacrol increasing the sizeof LNCs comparedtoCin-LNCs(893nmvs600.5nm)withthesame proportionsofeachactive(Table2(A)).Concurrently,anincreasein thezetapotential(ZP)wasobservedforCarCin-LNCscomparedto Car-LNCs,Cin-LNCsandblank-LNCs.
3.1.2.AntibacterialactivityofLNCsuspension
Theantibacterialactivities ofCa,Cin,CarCinandtheactives- loaded LNCs were evaluated by CMI and kill kinetics against A.baumannii(Table2(C)).TheresultsshowedthatalloftheMICs werelessthanorequalto0.55mg/ml,exceptfortheblank-LNCs, whichpresentedaMICabove5mg/ml.Nosignificantdifference wasnotedbetweentheMICsoftheCinandCarCinactivesandthe Cin-LNCsandCarCin-LNCs.TheMICsfortheCinactiveandtheCin- LNCswere0.31mg/mland0.36mg/ml,respectively.TheMICsfor the CarCin actives and the CarCin-LNCs were 0.16mg/ml and
Table2
(A)Physicochemicalpropertiesoftheunloaded-LNCs,CarCin-LNCs,Cin-LNCsandCar-LNCslabelledwithDiO.(B)MICofCar,Cin,CarCinandtheseactivesnano-encapsulated andlabelledwithDiOagainstA.baumannii.(C)MBCandFBCindexofCCCP,carvacrol,Car-LNCsandmixturesofCCCPandcarvacrolandCar-LNCsagainstA.baumannii(data areexpressedasthemeanSD).
A Meanparticlesize(nm) Polydispersity Zetapotential(mV)
Unloaded-LNCs 48.31.5 0.030.005 8.51.8
CarCin-LNCs 107.30.6 0.130.03 13.17
Meanparticlesize(nm) Polydispersity Zetapotential(mV)
Cin-LNCs 600.5 0.080.01 8.72.4
Car-LNCs 893 0.150.03 8.83.9
B
Carvacrol Cinnamaldehyde CarCin CarCin-LNCs Blank-LNCs
MIC(mg/ml) 0.16 0.31 0.16 0.31 5
CarCin-LNCs Cin-LNCs Car-LNCs
MIC(mg/ml) 0.31 0.36 0.55
C
MBC(mg/ml) Alone +CCCP FBCindex Effect
CCCP 25 6.25 12.5 6.25 12.5 6.25 12.5
– – – – – –
Carvacrol 160 80 40 0.75 0.75 Partialsynergistic Partialsynergistic
0.31mg/ml, respectively. We noted a small difference of MIC between the Car active and the Car-LNCs, which ranged from 0.16mg/mlto0.55mg/ml.
CCCPhadaMICof25
m
g/mlagainstA.baumannii(Table2(C)).WeevaluatedtheMBCsofcarvacrolandCar-LNCsinthepresence of 6.25
m
g/ml and 12.5m
g/ml CCCP (under the MIC/MBC) andcomparedtheMBCswithandwithoutCCCP.Theresultsshowed thatthestrainbecamelessresistanttocarvacrol(2–4folds)and Car-LNCs(4–8folds)inthepresenceofthisEPI(Table2(C)).
Forthe killkinetics, theconcentrationsused wereonceand twice the MIC for CarCin-LNCs and the concentrations corre- sponding tothe proportion of each encapsulatedactive in the CarCinmixture(Fig.2(A/B)).Nobactericidaleffectwasobserved for the four formulations, except for CarCin-LNCs after 6h of contact(Fig.1(A)).Whenthekillkineticswereanalyzedattwice theMICofCarCin-LNCs,abactericidaleffectofCarCin-LNCswas observedfrom3hto24h(Fig.1(B)).
3.1.3.Flowcytometrywithblank-LNCs,CarCin-LNCs,Car-LNCsand Cin-LNCslabelledwithDiO
3.1.3.1.Blank-LNCsandCarCin-LNCs. Thebehaviorofblank-LNCs andCarCin-LNCslabelledwithDiOwasevaluatedafteraddition totheA.baumanniistrain.After10minofincubation,therewas a difference in the DiO-bacteria fluorescence between the blank-LNCsand CarCin-LNCs (Fig. 2(1)). Therewas 32%4% of DiO-bacteriaaftercontactwithblank-LNCsand94%2%ofDiO- bacteria after contact with CarCin-LNCs. After fluorescence extinction at the bacterial membrane due to trypan blue exposure, the percentage of DiO-bacteria fluorescence
(internalized DiO-LNCs fluorescence) was 17%4% for blank- LNCs and84%7%forCarCin-LNCs.Withblank-LNCstherewas adecreaseofthebacteriafluorescence.Oneexperimentshowed a decrease of the internalized blank-LNCs fluorescence. In addition, a significant amount of adsorbed blank-LNCs (approximately half) was observedat 10minof incubation.For CarCin-LNCs, the fluorescence of bacteria remained constant over time (between 85%-95%) (Fig. 2(2)). It is important, to determine which active(s) is(are) responsible for these loaded LNC-bacteria interactions.
3.1.3.2. Car-LNCs and Cin-LNCs. The behavior of Car-LNCs and Cin-LNCs labelled DiO was evaluated with comparable concentrationsofeachactivetotheCarCin-LNCsintheprevious study(Fig.2(2)).
After10minofincubation,therewasadifferenceintheDiO- bacteriafluorescencebetweentheCin-LNCsandCar-LNCs.These results werecomparablewiththeblank-LNCsand CarCin-LNCs.
Therewere43%3%ofDiO-bacteriaaftercontactwithCin-LNCs and 91%2% of DiO-bacteria after contact with Car-LNCs (Fig.2(2)).Afterfluorescenceextinctionatthebacterialmembrane due to trypan blue exposure, the percentage of DiO-bacteria fluorescenceresultingininternalizedDiO-LNCsfluorescencewas 25%2%forCin-LNCsand86%4%Car-LNCs.Aftercontactwith Cin-LNCs and Car-LNCs, the fluorescence of bacteria remained constantovertime(between35–45%and85–95%,respectively).In the sameway, theinternalized Cin-LNCs/Car-LNCs fluorescence was constant (18–25% and 85–90%, respectively). A significant amountofadsorbedCin-LNCs(approximatelyhalf)wasobserved foralltimescomparedtotheadsorbedCar-LNCs.
Fig.1.Killkineticsoftheactivesandactivesmixtures-loadedLNCslabelledwithDiOagainstA.baumanniiattheMIC(A)andtwicetheMIC(B)(dataareexpressedasthe meanSD).
3.2.Studyofthebehaviorofthemodifiedcarvacrol-loadedLNCs 3.2.1.PhysicochemicalpropertiesoftheLNCsuspensions
CarPAandCarAA(Fig.3(A))wereencapsulatedbyLNCsusing the same process as that for carvacrol. Three formulations presentedaverynarrowsizepolydispersity(PdI<0.15),indicating verylowheterogeneityinthesuspension(Fig.3(B)).Thesizesof
themodifiedcarvacrol-loadedLNCsweresmallerthanCar-LNCs:
733nm for CarPA-LNCs and 580.2nm for CarAA-LNCs (Fig.3(B)).
3.2.2.AntibacterialactivityoftheLNCsuspension
The antibacterial activities of CarPA,CarAA and theactives- loadedLNCswereevaluatedagainstA.baumannii(Fig.3(B)).The Fig.2. (1)Representationofbacteriaeventsafter10minofexposuretoblank-LNCsorCarCin-LNCs(0,3mg/mlofactives)labelledwithDiObefore(A,B)andafterexposureto trypanblue(C,D)byflowcytometry.(2)HistogramrepresentingtheDiO-labelledbacteriafluorescence(%)aftercontactwithblank-LNCs,CarCin-LNCs,Car-LNCsandCin- LNCs(dataareexpressedasthemeanSEM).
resultsshowedthattheseactivesdidnotexhibitMICssimilarto CarPA-LNCsandCarAA-LNCsatthetestedconcentrations(5mg/
ml).
3.2.3.FlowcytometrywithCarAA-LNCsandCarPA-LNCslabelledwith DiO
ToconfirmthehypothesisthatOHfunctionalgroupspromote the interactions with bacterial membranes and improve the intrusionofCar-LNCs,andthuscarvacrol,acrossthemembrane, CarPA-LNCsandCarAA-LNCSlabelledwithDiOwerestudiedwith bacteriabyflowcytometry.After10minofincubation,theresults showedamassivedifferenceinDiO-bacteriafluorescencebetween the modified carvacrol-loaded LNCs and Ca-LNCs. There was 40%16% of DiO-bacteria after contact with CarPA-LNCs and
21%5%ofDiO-bacteriaforCarAA-LNCs(91%ofDiO-bacteriafor Car-LNCs)(Fig.3(C)).Aftertrypanblueexposure,thepercentageof DiO-bacteria fluorescence resulting in internalized DiO-LNCs fluorescence was 29%4% for CarPA-LNCs and 10%1.5%
CarAA-LNCs.Atthesametimeandformodifiedcarvacrol-loaded LNCs, there was a decrease of bacteria fluorescence, which vanished after 2h of contact. In addition,a significantamount ofadsorbedCarPA-LNCsand CarAA-LNCs(approximatelya third andahalf,respectively)wasobservedafter10minofincubation.
4.Discussion
A.baumannii hasemergedasmultidrugresistantand highly pathogenicforintensivecareunitpatients.Currently,thelackof Fig.3.(A)Structureofcarvacrolandcarvacrolderivatives,CarAAandCarPA.(B)MICsofCarPAandCarAAnano-encapsulatedandlabelledwithDiOagainstA.baumannii.(C) HistogramrepresentingtheDiOlabelledbacteriafluorescence(%)aftercontactwithCar-LNCs,CarPAandCarAA-LNCs(dataareexpressedasthemeanSEM).
thedevelopmentofnewantibacterialmoleculesbythepharma- ceuticalindustryisamajorproblem.Itisthereforeimportantto findnewtherapeuticapproachesthatareeffectiveandlimitthe emergenceofbacterialresistance.Weselectedtwocomponents fromOriganumandcinnamonEOs,carvacrolandcinnamaldehyde, respectively,duetotheirantibacterialactivitywithdistinctaction mechanismswiththeaimofobtaininganoptimizedantibacterial actionandalimitedemergenceofresistance(Michielsetal.,2007;
Peietal.,2009).
Carvacrolisaphenoliccompoundwithahydroxylgrouponan aromaticring.Thelipophilicpropertiesofthiscomponentallowits accumulation in the cytoplasmic membrane and consequently modify of the membrane permeability. The penetration of
carvacrolcausestheformationofporesinthebacterialmembrane, resulting in ion and cellular component leakage, leading to bacterialdeath(Ulteeetal.,2002).
Cinnamaldehyde is an aldehyde compound with a carboxyl group on its aromatic ring. This compound is not able to disorganizetheoutercellmembraneordepletetheintracellular pool.Thebiologicalactivityismainlyduetothecarboxylgroup, whichbindstoproteinsatthebacterialmembrane(Burt,2004).
Phenoliccompoundspossessthegreatestantibacterialpoten- tial, followedbyaldehydes, ketones, alcoholsandhydrocarbons (Bajpaietal.,2012).
However,thesemoleculespossessvariouslipophiliccharacter- isticsthatreducetheirbioavailability.Toovercomethesechallenges,
Fig.4.MechanismrepresentationofCar-LNCsatthebacterialmembraneofA.baumannii.(A)ModeofantibacterialactionofCarCin-LNCsandCar-LNCsviatheOHfunctional groupsofcarvacrolaccessibleattheLNCsurface,whichalloweddirectcontactwiththebacterialmembrane,crossingthemembraneandcausingantibacterialeffects.(B) AftercontactwithCCCP,therewasasynergisticeffectbasedontheCar-LNCsaccumulationinthebacteriacausingagreaterantibacterialeffect.(C)Modeofactionofblank- LNCsandmodifiedCar-LNCs.TheseLNCscouldcrossthebacterialmembraneandreleaseintothemediumbyeffluxpumpswithoutantibacterialaction.
aromaticcompoundshavebeenencapsulatedinananomedicine systemintensivelydescribedintheliterature(Donsìetal.,2011;
KeawchaoonandYoksan,2011;Lioliosetal.,2008).
Forthefirsttime,wesynthesizedblank-LNCsandCarCin-LNCs to compare their physicochemical properties. We observed a difference in the size distribution between blank-LNCs and CarCin-LNCs.Thisphenomenonwasexplainedbythestericand amphiphilicpropertiesofcarvacrol,which increasedthesizeof Car-LNCsandCarCin-LNCscomparedtoCin-LNCs.Concurrently,an increaseofthezetapotential(ZP)wasobservedforCarCin-LNCs compared to Car-LNCs, Cin-LNCs and blank-LNCs, indicating a possiblemodificationoftheLNCsurfaceduetothepresenceofthe activesintheinterfacialzone.
TheantibacterialactivityofCar,Cin,CarCinandtheencapsu- lated actives was evaluated by MIC against A. baumannii to investigate a possible synergistic effect between carvacrol and cinnamaldehyde.WenotedthattheMICsoftheactivesandactives mixturewerecomparable,withvaluesbetween0.16mg/mland 0.31mg/ml. It is difficult to anticipate the in vitro and invivo interactionsbetweenthesetwocomponents,buttheaimsofthis studyweretoextendtheactionmechanismagainstbacteriaandto limitthepotentialriskofresistanceemergence.
TheMICsofactivesandtheactivesmixturewerecomparedto the MICs of active(s) encapsulated by LNCs. The MICs of the actives and actives-loaded LNCs were comparable for the cinnamaldehyde and CarCin actives, except for carvacrol. The Car-LNCs had a MIC larger MIC than Car alone. This can be explainedbythefactthatcarvacrolalonewasdirectlydeposited onthemembrane,causingbetterbacterialinteractionscompared toCar-LNCs,whichpresentedadelayedreleasecomparedtoCar alone.Nevertheless,thissituationcannot bereproducedinvivo duetothelipophilicnatureofcarvacrol.Weconcludedthatthe encapsulationprocesshadnosignificanteffectontheantibacterial activityoftheactives.Moreover,aftercontactwithaneffluxpump inhibitorCCCP,there wasa synergisticeffectbetweencarvacrol andEPIandbetweenCar-LNCsandEPIbasedontheFBC index, which could be explained by carvacrol accumulation in the bacteria(Fig.4(A/B)).Moreover,wenotedasynergisticantibacte- rialeffectthatwasgreaterforCar-LNCs,whichcouldbeexplained by thebetterpenetration of these LNCs intothe bacteria. This synergisticantibacterialeffectunderlinedtheactionmechanism ofbacteriaresultingaCarorCar-LNCsreleasebytheeffluxpump system.Thiseffectcouldbeassociatedwiththeactionmechanism of carvacrol indicating a potential bacterialtarget blocking the release of Car-LNCs. These targets could be proton-dependent efflux pump systems, such as RND (resistance nodulation cell division),MATE(multidrugandtoxiccompoundextrusion),MFS (majorfacilitatorsuperfamily) andSMR (smallmultidrugresis- tance) (Poole, 2002). An previous study (Fadli et al., 2014) demonstrated the stimulation of membrane-associatedmecha- nismsofresistancebynaturalextracts(carvacrolandthymol)in Gram-negative bacteria. Furthermore, there was no synergistic effect withcinnamaldehyde in thepresence of CCCP (datanot shown),whichisnotsurprisingbecausethis activeactedatthe bacterialmembrane.
Inthisexperiment,weusedCarCin-LNCsattheMIC,twicethe MICandtheproportionofeachencapsulatedactiveintheCarCin mixture,Car-LNCsandCin-LNCsforkillkineticstudiestoestablish anantibacterialeffectagainstA.baumanniiovertime.Theresults didnotdemonstrateabactericideeffectforCar-LNCsandCin-LNCs (withtheCar proportionand Cinproportionin activemixture) comparedtothecontrolexceptforCarCin-LNCs.Thesedifferences canbeexplained bythefact thatthe activeconcentrationsare belowtheMICsofCin-LNCsandCar-LNCs.
We demonstrated that a bactericidal effect on kill kinetic studieswaspresent from3hin thepresenceofCarCin-LNCsat
theirCMI.Toovercomethepresenceofdeadbacteria,whichcan altertheresultsinterpretation,flowcytometrywasconductedat theMICofCarcin-LNCsandtheproportionsofeachencapsulated activeinthemixture.Weassumedthatbacteriafluorescencewas duetothecontactofDiO-LNCswithbacteriabecausenosignificant DiOreleasecanoccuroutsideoftheparticles(Bastiatetal.,2013).
Moreover,arecentstudy(Gravieretal.,2014)demonstratedthat LNCs were internalized rapidly in cultured cells and remained intactfor3h.
Thebehaviorofblank-LNCsandCarCin-LNCslabelledwithDiO wasevaluatedattheMICoftheseLNCs.Theresultsofthisstudy supported the potential of the encapsulated actives (CarCin mixture) compared tounloadedLNCs. Thelipophilic properties of aromatic molecules allowed a better interaction with the bacterial membrane composed of a lipid bilayer. Moreover, unloadedLNCs tended toadsorbin a random and non-specific manneratthebacterialmembrane.Thedecreaseoftheblank-LNCs fluorescenceisexplainedbythereleaseoftheseDiO-LNCsorDiO alone(afterdisintegrationofLNC)intheextra-bacterialenviron- mentovertime,andtheseLNCswereeliminatedinthewashes.The blank-LNCs release in the medium is related to efflux pumps (Fig.4(C)).WhenDiOwasaloneinthebacteria,itsaffinitywiththe bacterialmembranewouldbestrongerduetoitslipophilicnature, andbacteriafluorescencewouldremainconstant.Thisisnotwhat wefoundinthisstudy.
Itisimportanttoknowwhichactive(s)is(are)responsiblefor these loaded LNC-bacteria interactions. The results indicated a major attractionandpenetrationofcarvacrolintobacteria.This phenomenonwasduetothehydroxylgroupofcarvacrol,which destabilizedthemembraneanddelocalizedtheelectronsinthe aromaticringsystem(Ulteeetal.,2002).Thus,theOHfunctionsof carvacrolcouldbeaccessibleattheLNCsurfaceandcouldallow directcontact withthebacterialmembrane(Fig.4(A)).Further- more,forCin-LNCs,thepenetrationislimited.Evenifittendedto adsorbonthebacterialmembrane,this moleculeis notableto insertintothemembranestructure(Helanderetal.,1998).These results are consistent with the literature, which shows an antibacterial modeof action related to bonds with membrane proteins.These resultsunderlinedtheimportanceofcombining thesetwoactivestoobtainenhancedeffectsagainstA.baumannii forsuitableinvivoadministration.
TodemonstratethecrucialroleoftheOHfunctionofcarvacrol for antibacterial activity, modifications of this compound were performedbysubstitutionhydroxylfunctionsbyfattyacids(acetic acid,palmiticacid),andthemodifiedcompoundswereencapsu- lated.
TheresultsindicatedasmallersizeforCarPA-LNCsandCarAA- LNCscomparedtoCar-LNCs.Thisphenomenoncanbeexplainedby theabsenceofOHfunctionalgroupswhichcouldincreasethesize ofLNCsbyswelling.Thedifferenceofmodifiedcarvacrol-loaded LNCswascausedbybulkstericoffattyacidchainsandthelengthof theradicalsaddedattheestergroup.
Theencapsulation ofCarPAandCarAA didnotinducea size decrease or ZP modifications (close to 8mV). Thesemodified activesdidnotmodifythePEGconformationattheLNCsurfaceas was the case for carvacrol. Both LNCs were evaluated by flow cytometryaftercontactwithA.baumannii.Theresultshighlight theactivityoftheOHfunctionalgroupsofcarvacrolcomparedto modifiedcarvacrol.Indeed,thedecreaseof theCarPA-LNCs and CarAA-LNCs fluorescence couldbe explained by the release of theseDiO-LNCsintheextra-bacterialenvironmentovertimeand LNCseliminationafterwashing.TheCarPA-LNCsandCarAA-LNCs didnotpresentantibacterialeffectsandtheyarereleasedintothe mediumbytheeffluxpump(similartothemodeofactionofblank- LNCs)(Fig.4(C)).Apreviousstudyshowedthatthehydroxylgroup of this compound and thepresence of a system of delocalized
electronsareimportantfortheantimicrobialactivityofcarvacrol comparedtothymol,cymene,andmenthol(Ulteeetal.,2002).
Weenvisageperformingatranscriptomicandproteomicstudy to verify the action of carvacrol on the efflux pumps of A. baumannii. Previous studies used extracts of essential oils (thymol and carvacrol) as substrates for the efflux pumps of Escherichiacoli(Fadlietal.,2014).
Using flowcytometry,thisstudyhighlightedtheinteractions betweenbacteriaandLNCsovertimebasedonthefluorescenceof DiO-LNCs and the properties of trypan blue to determine the physicochemical mechanisms occurring at the level of the biologicalmembrane.Thisisthefirsttimethatastudyhasused thepropertiesoftrypanbluetoquenchfluorescentLNCsatthe bacterialmembrane.Thistechniqueenabledbetterunderstanding of the action mechanisms of aromatic compound-loaded LNCs with bacteria. CarCin-LNCs demonstrated the attractiveness of encapsulated actives compared to unloaded LNCs. This study showed improved interaction and internalization of Car-LNCs comparedtoCin-LNCs.Modificationsofcarvacrolaftersubstitu- tionofhydroxylfunctionalgroupsbyfattyacidsdemonstratedthe crucialroleofhydroxylgroupsforantibacterialactivity.Finally,a synergistic antibacterial effect was demonstrated between Car-LNCsandCCCP.Thiseffectunderlinedtheactionmechanism ofbacteriaresultingaCarorCar-LNCsreleasebytheeffluxpump system.
Acknowledgements
The authors would like to acknowledge the Eydo Pharma CompanysupportbytheAssociationNationaledelaRechercheet de la Technologie, and the Bacteriology Service of Angers University Hospital and INSERM (U1066 “Biomimetic Micro- nanomedicine”).
References
Bajpai,V.K.,Baek,K.H.,Kang,S.C.,2012.ControlofSalmonellainfoodsbyusing essentialoils:areview.FoodRes.Int.45,722–734.
Bakkali,F.,Averbeck,S.,Averbeck,D.,Idaomar,M.,2008.Biologicaleffectsof essentialoils—areview.FoodChem.Toxicol.46,446–475.
Bastiat,G.,Pritz,C.O.,Roider,C.,Fouchet,F.,Lignieres,E.,Jesacher,A.,Glueckert,R., Ritsch-Marte,M.,Schrott-Fischer,A.,Saulnier,P.,Benoit,J.P.,2013.Anewtoolto ensurethefluorescentdyelabelingstabilityofnanocarriers:arealchallengefor fluorescenceimaging.J.Control.Release170,334–342.
Berenbaum,M.C., 1978.Amethodfortestingforsynergywithanynumberofagents.
J.Infect.Dis.137,122–130.
Burt,S.,2004.Essentialoils:theirantibacterialpropertiesandpotential applicationsinfoods—areview.Int.J.FoodMicrobiol.94,223–253.
Donsì,F.,Annunziata,M.,Sessa,Mariarenata,Ferrari,Giovanna,2011.
Nanoencapsulationofessentialoilstoenhancetheirantimicrobialactivityin foods.FoodSci.Technol.44,1908–1914.
Eveillard,M.,Soltner,C.,Kempf,M.,Saint-Andre,J.P.,Lemarie,C.,Randrianarivelo,C., Seifert,H.,Wolff,M.,Joly-Guillou,M.L.,2009.Thevirulencevariabilityof differentAcinetobacterbaumanniistrainsinexperimentalpneumonia.J.Infect.
60,154–161.
Fadli,M.,Chevalier,J.,Hassani,L.,Mezrioui,N.E.,Pages,J.M.,2014.Naturalextracts stimulatemembrane-associatedmechanismsofresistanceinGram-negative bacteria.Lett.Appl.Microbiol.58,472–477.
Gaumet,M.,Gurny,R.,Delie,F.,2009.Localizationandquantificationof biodegradableparticlesinanintestinalcellmodel:theinfluenceofparticlesize.
Eur.J.Pharm.Sci.36,465–473.
Gravier,J.,Sancey,L.,Hirsjarvi,S.,Rustique,E.,Passirani,C.,Benoit,J.P.,Coll,J.L., Texier,I.,2014.FRETimagingapproachesforinvitroandinvivocharacterization ofsyntheticlipidnanoparticles.Mol.Pharm.11,3133–3144.
Helander,K.L.-K.,Mattila-Sandholm,T.,Pol,I.,Smid,E.J.,Gorris,L.G.M.,vonWright, A.,Alakomi,H.-L.,1998.Characterizationoftheactionofselectedessentialoil componentsonGram-negativebacteria.J.Agric.FoodChem.46,3590–3595.
Heurtault,B.,Saulnier,P.,Pech,B.,Proust,J.E.,Benoit,J.P.,2002.Anovelphase inversion-basedprocessforthepreparationoflipidnanocarriers.Pharm.Res.
19,875–880.
Karageorgopoulos,D.E.,Falagas,M.E.,2008.Currentcontrolandtreatmentof multidrug-resistantAcinetobacterbaumanniiinfections.LancetInfect.Dis.8, 751–762.
Keawchaoon,L.,Yoksan,R.,2011.Preparation,characterizationandinvitrorelease studyofcarvacrol-loadedchitosannanoparticles.ColloidsSurf.B84,163–171.
Liolios,C.C.,Gortzi,O.,Lalas,S.,Tsaknis,J.,Chinou,I.,2008.Liposomalincorporation ofcarvacrolandthymolisolatedfromtheessentialoilofOriganumdictamnusL.
andinvitroantimicrobialactivity.FoodChem.112,77–83.
Michiels,J.,Missotten,J.,Fremaut,D.,DeSmet,S.,Dierick,N.,2007.Invitrodose–
responseofcarvacrol,thymol,eugenolandtrans-cinnamaldehydeand interactionofcombinationsfortheantimicrobialactivityagainstthepiggut flora.Livest.Sci.109,157–160.
Montagu,A.,Saulnier,P.,Casissa,V.,Rossines,E.,Eveillard,M.,Joly-Guillou,M.L., 2014.Aromaticandterpeniccompoundsloadedinlipidicnanocapsules:
activityagainstmulti-drugresistantacinetobacterbaumanniiassessedinvitro andinamurinemodelofsepsis.J.Nanomed.Nanotechnol.5.
Pei,R.S.,Zhou,F.,Ji,B.P.,Xu,J.,2009.Evaluationofcombinedantibacterialeffectsof eugenolcinnamaldehyde,thymol,andcarvacrolagainstE.coliwithan improvedmethod.J.FoodSci.74,M379–83.
Peleg,A.Y.,Seifert,H.,Paterson,D.L.,2008.Acinetobacterbaumannii:emergenceof asuccessfulpathogen.Clin.Microbiol.Rev.21,538–582.
Poole,K.,2002.Outermembranesandefflux:thepathtomultidrugresistancein Gram-negativebacteria.Curr.Pharm.Biotechnol.3,77–98.
Srivastava,G.K.,Reinoso,R.,Singh,A.K.,Fernandez-Bueno,I.,Hileeto,D.,Martino, M.,Garcia-Gutierrez,M.T.,Merino,J.M.,Alonso,N.F.,Corell,A.,Pastor,J.C.,2011.
TrypanBluestainingmethodforquenchingtheautofluorescenceofRPEcellsfor improvingproteinexpressionanalysis.Exp.EyeRes.93,956–962.
Ultee,A.,Bennik,M.H.,Moezelaar,R.,2002.Thephenolichydroxylgroupof carvacrolisessentialforactionagainstthefood-bornepathogenBacilluscereus.
Appl.Environ.Microbiol.68,1561–1568.
Vila,J.,Marti,S.,Sanchez-Cespedes,J.,2007.Porins,effluxpumpsandmultidrug resistanceinAcinetobacterbaumannii.J.Antimicrob.Chemother.59,1210–1215.
Ye,H.,Shen,S.,Xu,J.,Lin,S.,Yuan,Y.,Jones,G.S.,2013.Synergisticinteractionsof cinnamaldehydeincombinationwithcarvacrolagainstfood-bornebacteria.
FoodControl34,619–623.doi:http://dx.doi.org/10.1016/j.
foodcont.2013.05.032.
