ContentslistsavailableatScienceDirect
Journal
of
Molecular
Catalysis
B:
Enzymatic
jo u r n al h om ep age :w w w . e l s e v i e r . c o m / l o c a t e / m o l c a t b
RHA-P:
Isolation,
expression
and
characterization
of
a
bacterial
␣-l-rhamnosidase
from
Novosphingobium
sp.
PP1Y
Federica
De
Lise
a,1,
Francesca
Mensitieri
a,1,
Vincenzo
Tarallo
b,
Nicola
Ventimiglia
a,
Roberto
Vinciguerra
b,
Annabella
Tramice
c,
Roberta
Marchetti
b,
Elio
Pizzo
a,
Eugenio
Notomista
a,
Valeria
Cafaro
a,
Antonio
Molinaro
b,
Leila
Birolo
b,
Alberto
Di
Donato
a,
Viviana
Izzo
d,∗aDipartimentodiBiologia,UniversitàFedericoIIdiNapoli,ViaCinthiaI,80126Napoli,Italy bDipartimentodiScienzeChimiche,UniversitàFedericoIIdiNapoli,viaCinthiaI,80126Napoli,Italy
cIstitutodiChimicaBiomolecolare,ConsiglioNazionaledelleRicerche,ViaCampiFlegrei34,80072Pozzuoli(NA),Italy
dDipartimentodiMedicina,ChirurgiaeOdontoiatria“ScuolaMedicaSalernitana”,UniversitàdegliStudidiSalerno,viaS.Allende,84081Baronissi(SA), Italy
a
r
t
i
c
l
e
i
n
f
o
Articlehistory: Received29July2016 Receivedinrevisedform 23September2016 Accepted3October2016 Availableonline3October2016 Keywords: ␣-l-Rhamnosidases Flavonoids Novosphingobiumsp.PP1Y Glycosylhydrolases Biotransformation
a
b
s
t
r
a
c
t
␣-l-Rhamnosidases(␣-RHAs)areagroupofglycosylhydrolasesofbiotechnologicalpotentialin indus-trialprocesses,whichcatalyzethehydrolysisof␣-l-rhamnoseterminalresiduesfromseveralnatural compounds.Anovel␣–RHAactivitywasidentifiedinthecrudeextractofNovosphingobiumsp.PP1Y, amarinebacteriumabletogrowonawiderangeofaromaticpolycycliccompounds.Inthiswork,this ␣-RHAactivitywasisolatedfromthenativemicroorganismandthecorrespondingorfwasidentified inthecompletelysequencedandannotatedgenomeofstrainPP1Y.Thecodinggenewasexpressedin Escherichiacoli,strainBL21(DE3),andtherecombinantprotein,rRHA-P,waspurifiedandcharacterized asaninvertingmonomericglycosidaseofca.120kDabelongingtotheGH106family.Abiochemical char-acterizationofthisenzymeusingpNPRassubstratewasperformed,whichshowedthatrRHA-Phada moderatetolerancetoorganicsolvents,asignificantthermalstabilityupto45◦Candacatalyticefficiency, atpH6.9,significantlyhigherthanotherbacterial␣-RHAsdescribedinliterature.Moreover,rRHA-Pwas abletohydrolyzenaturalglycosylatedflavonoids(naringin,rutin,neohesperidindihydrochalcone) con-taining␣-l-rhamnoseboundto-d-glucosewitheither␣-1,2or␣-1,6glycosidiclinkages.Datapresented inthismanuscriptstronglysupportthepotentialuseofRHA-Pasabiocatalystfordiversebiotechnological applications.
©2016ElsevierB.V.Allrightsreserved.
1. Introduction
␣-l-Rhamnosidases(␣-RHAs) area groupof glycosyl hydro-lases(GHs)that catalyzethehydrolysis of terminalresidues of
Abbreviations: ␣-RHAs,␣-l-rhamnosidases;GHs,glycosyl hydrolases;GTs, glycosyl transferases; PPMM, potassium phosphate minimal medium; pNPR, p-nitrophenyl-␣-l-rhamnopyranoside;MOPS,3-(N-morpholino)propanesulfonic acid;pNP,p-nitrophenolate;BSA,bovineserumalbumin;LB,LuriaBertanimedium; LB-N,LuriaBertanimediumcontainingafinalconcentrationof0.5MNaCl;LB-BS, LuriaBertanimediumsupplementedwith1mMofbothbetaineandsorbitol; LB-NBS,LuriaBertanimediumcontainingafinalconcentrationof0.5MNaCland1mM ofbothbetaineandsorbitol;IPTG,Isopropyl-d-1-thiogalactopyranoside.
∗ Correspondingauthor.
E-mailaddress:vizzo@unisa.it(V.Izzo). 1 Theseauthorsequallycontributedtothework.
␣-l-rhamnosefromalargenumberofnaturalcompounds[1]. L-Rhamnoseiswidelydistributedinplantsascomponentofflavonoid glycosides,terpenylglycosides,pigments,signalingmolecules,and incellwallsasa componentofcomplex heteropolysaccharides, suchasrhamnogalacturonanandarabinogalactan-proteins[2–7]. In bacteria, l-rhamnose appears to be included in membrane rhamnolipids[8,9]andpolysaccharides[10].Accordingtothe sim-ilaritiesamongtheiraminoacidicsequences,␣-RHAsaregrouped intheCAZy(carbohydrate-activeenzymes)database(www.cazy. org)intofourdifferentfamilies:GH28,GH78,GH106,andNC (non-classified).
Inthelastdecade,␣-RHAshaveattractedagreatdealof atten-tionduetotheirpotentialapplicationasbiocatalystsinavarietyof industrialprocessesandinparticularinthefoodindustry[1]. Sev-eraldietaryproductsarerichinglycosylatedflavonoidsthatshow thepresenceofeitherarutinoside(6-␣-l-rhamnosyl--d-glucose)
http://dx.doi.org/10.1016/j.molcatb.2016.10.002 1381-1177/©2016ElsevierB.V.Allrightsreserved.
ora neohesperidoside (2-␣-l-rhamnosyl--d-glucose) disaccha-ridicunit.Inparticular,naringin,hesperidinandrutin,flavanone glycosidesfoundingrapefruitjuices,lemons,sweetorangesand vegetables,havegainedincreasingrecognitionfortheirpotential antioxidant,antitumorandanti-inflammatoryproperties[11–14]. Theabilitytohydrolyzeglycosylatedflavonoidshasbeenused tomitigatethebitternessofcitrusjuices,whichisprimarilycaused bynaringin.Rhamnoseremovalfromnaringinallowssofteningthe bittertasteof citrus juice[15,16].Moreover,thecorresponding de-rhamnosylatedcompound,prunin,isendowedwith antimicro-bialproperties[14],andshowsanimprovedintestinalassimilation whencomparedtonaringin.Otherapplicationsof␣-RHAsare gain-ingpopularityintheoenologicalindustry,wheretheseenzymesare usedtohydrolyzeterpenylglycosidesandenhancearomainwine, grapejuicesandderivedbeverages[17–19].
Applicationof␣-RHAstoimproveflavonoidsbioavailabilityhas alsobeenrecentlydescribed[20].Inhumans,flavonoids absorp-tionoccursprimarilyin thesmallintestine where theattached glucose(orpossiblyarabinoseorxylose)isremovedby endoge-nous-glucosidases[21–23].Terminalrhamnoseisnotasuitable substrateforhuman-glucosidases.Therefore,unabsorbed rham-nosylatedflavonoidsarriveunmodifiedinthecolon,wherethey arehydrolyzedby␣-rhamnosidaseactivitiesexpressedbythelocal microflora[24].Toimproveintestinalabsorptionofrhamnosylated flavonoids,andthustheirbioavailabilityinhumans,aremovalof theterminalrhamnosegroupcatalyzedby␣-RHAswouldbeindeed beneficial[25–27].
Theabsenceofhuman␣-RHAshasbeenthekeytothe devel-opmentofa noveltargeteddrugdeliverystrategy, indicatedas LEAPT(Lectin-directedenzymeactivatedprodrugtherapy)[28,29], abipartitesystembasedontheinternalizationofanengineered ␣-RHAbearingaglycosidicmoietythatisrecognizedbyspecific lectinspresentonthesurfaceofdifferenteukaryoticcelllines.Inthe LEAPTsystem,theintakeofarhamnosylatedprodrug,which can-notbeprocessedbymammalianenzymes,allowsasite-selective actionofthedrugincellswheretheengineered␣-RHAhasbeen prelocalized.
Todate,microbial␣-RHAshavebeenmainlypurifiedfrom fun-galstrainssuchasPenicilliumand Aspergillus[30–32]; onlyone exampleof␣-RHAisolatedfromaviralsourcehasbeendescribed [33],and itisnoteworthythatonly alimited numberof bacte-rialrhamnosidaseshasbeenfullycharacterized[34–39].Oneof themaindifferencesbetweenfungalandbacterial␣-RHAsistheir differentoptimal pHvalues, with thefungal enzymesshowing moreacidicpHoptimawhencomparedtothebacterial counter-parts,forwhichneutralandalkalinevalueshavegenerallybeen described.Thischaracteristicsuggestsdiverseapplicationsfor fun-galand bacterialenzymes,making bacterial␣-RHAssuitable in biotechnologicalprocessesrequiringgoodactivityinmorebasic solutionssuchas,forexample,theproductionofL-rhamnosefrom thehydrolysisofnaringinorhesperidin,flavonoidswhose solubil-itystronglyincreasesathigherpHvalues[40].Inaddition,bacterial rhamnosidasescouldbeidealcandidatesfordietarysupplements havingactivity acrosstheentire gastrointestinal(GI)tract, and morespecificallyinthesmallintestinewhereflavonoids absorp-tionshouldmostlyoccurtoenhancethebeneficialeffectofthese moleculesonhumanhealth.
Inordertoelucidatetherealbiotechnologicalpotentialof bac-terial␣-RHAs,furtherinvestigationisundoubtedlyneeded.Few detailsonthecatalyticmechanismofbacterial␣-RHAsare avail-able, and most importantly, to the best of our knowledge, no attempttoimprove thecatalytic efficiencyor modifysubstrate specificityoftheseenzymesbymutagenesishasbeenperformed yet.Thisisinpartconsequentialtothefactthatonlyaverylimited numberofcrystalstructuresof␣-RHAsarecurrentlyavailable,such asthe␣-l-rhamnosidaseB(BsRhaB)fromBacillussp.GL1[41],and
the␣-l-rhamnosidasefromStreptomycesavermitilis[42]. There-fore,itisevidentthatbacterial␣-RHAsrepresentayetunexplored reservoirofpotentialbiocatalystsforwhichmorefunctionaland structuraldataarerequired.
AmemberoftheorderoftheSphingomonadales,recently iso-lated and microbiologically characterized, Novosphingobium sp. PP1Y [43,44], appears tobe a valuable sourcefor the isolation of ␣-RHA activities. Sphingomonadales are a group of Gram-negative␣-proteobacteriawhosegenomesshowthepresenceof a great abundanceof both glycosyl hydrolases (GHs) and gly-cosyltransferases(GTs).Theseactivitiesareprobablyinvolvedin the biosynthesis of complex extracellular polysaccharides and microbialbiofilms[45].TheinterestforNovosphingobiumsp.PP1Y carbohydrate-active enzymeshasgrownas thesequencingand annotationofthewholegenome,recentlycompleted,allowedthe identificationofagreatnumberofgenesencodingforbothGHs(53 orfs)andGTs(57orfs)[46].
Recently,a␣-RHAactivityinNovosphingobiumsp.PP1Ycrude extractwasdescribed,whichshowedanalkalinepHoptimumand amoderatetolerancetoorganicsolvents[47].Novosphingobiumsp. PP1Ycrudeextractexpressingthisenzymaticactivitywasusedfor thebioconversionofnaringin,rutinandhesperidin.Basedonthese preliminaryresults,amoredetailedbiochemicalcharacterization ofthe␣-RHAactivitywasessential.
In thiswork,theisolation, recombinantexpression and par-tialcharacterizationofa␣-RHAfromNovosphingobiumsp.PP1Yis reported.Thisenzyme,namedRHA-P,belongstotheGH106family [48],asubgroupforwhich,accordingtoourknowledge,nocrystal structureisavailableyet.Asevidentfromouranalyses,RHA-Pisa promisingcandidateforseveralbiotechnologicalapplications.
2. Materialsandmethods
2.1. Generals
Generalmolecularbiologytechniqueswereperformed accord-ing to Sambrook et al. [49]. Bacterial growthwas followed by measuring the optical density at 600nm (OD600). pET22b(+)
expressionvectorandE.colistrainBL21(DE3)werefromAmersham Biosciences;E.colistrainTop10waspurchasedfromLife Technolo-gies.N.sp.PP1Ywasisolatedfrompollutedseawaterintheharbor ofPozzuoli(Naples,Italy)aspreviouslydescribed[43].
ThethermostablerecombinantDNApolymeraseusedforPCR amplificationwasTAQPolymerasefromMicrotechResearch Prod-ucts. dNTPs, T4 DNA ligase, and the Wizard PCR Preps DNA purificationsystemforelutionofDNAfragmentsfromagarosegels werepurchasedfromPromega.TheQIAprepSpinMiniprepKitfor plasmidDNApurificationwasfromQIAGEN.Enzymesandother reagentsforDNAmanipulationwerefromNewEnglandBiolabs. OligonucleotidessynthesisandDNAsequencingwereperformed byMWG-Biotech.N-terminusofrRHA-Pwassequencedby Pro-teomeFactoryAG.Thepresenceandlocationofapotentialsignal peptidecleavagesiteontheaminoacidicsequenceofRHA-Pwas analyzedusingSignalP4.1server(http://www.cbs.dtu.dk/services/ SignalP).
Q-Sepharose Fast Flow and p-nitrophenyl-
␣-l-rhamnopyranoside (pNPR)were from Sigma Aldrich;Sephacryl S200HighResolutionwaspurchasedfromAmershamBiosciences. IPTG(isopropyl-d-1-thiogalactopyranoside)wasobtainedfrom Applichem.
SolventsusedinenzymaticassayswereeitherfromApplichem (DMSO)orfromRomil(acetone).TLCsilicagelplateswerefromE. Merck(Darmstadt,Germany).
2.2. GrowthofNovosphingobiumsp.PP1Ycells
N.sp.PP1YcellsweregrowninPotassiumPhosphateMinimal Medium(PPMM)at30◦Cfor28hunderorbitalshakingat220rpm. Apre-inoculuminLBwaspreparedbytransferring50Lfroma glycerolstockstoredat−80◦Ctoa50mLFalcontubecontaining
12.5mLofsterileLB.Thepre-inoculumwasallowedtogrowat 30◦CO/Nunderorbitalshakingandthenusedtoinoculate1Lof PPMMataninitialcellconcentrationof0.01-0.02OD600.0.3mM
naringinewasaddedtothemediumandusedasaninducerofthe ␣-RHAactivity,aspreviouslydescribed[47].
2.3. CloningoforfPP1YRS05470andconstructionofthe pET22b(+)/rha-pexpressionvector
GenomicDNAwasextracted froma50mLsaturatedculture of N. sp. PP1Y as described elsewhere [44]. OrfPP1YRS05470 coding for the ␣-RHA activity was amplified in two contigu-ous fragments, owing to the considerable length of the orf (3441bp). The first fragment, named rha-up (1816bp), was amplifiedusinganinternaldownstream primer,RHA-Intdw(5 -AGGCGGCCATGGGAATGT-3),whichincludedaninternalNcoIsite alreadypresentinorfPP1YRS05470,andanupstreamprimer, RHA-up(5-GGGAATTCCATATGCCGCGCCTTTCGCT-3), designedtoadd aNdeIrestrictionsiteat5 oforfPP1YRS05470.Thesecondhalf of thegene, named rha-dw(1625bp), was amplifiedusing the upstreamprimerRHA-Intup(5-ACATTCCCATGGCCGCCT-3), com-plementarytoRHA-Intdw,andthedownstreamprimerRHA-dw (5-AAAACCGAGCTCTCAATGCCCGCCCGTG-3)thatwasintendedto incorporateaSacIrestrictionsitedownstreamoftheamplifiedorf. Theamplifiedfragments,rha-upandrha-dw,werepurifiedfrom agarosegel,digested,respectively, withNdeI/NcoIandNcoI/SacI, andindividuallyclonedinpET22b(+)vectorpreviouslycutwith thesameenzymes.
Ligatedvectorswereusedtotransform E.coli, strainTop10, competent cells. The resulting recombinant plasmids, named pET22b(+)/rha-up and pET22b(+)/rha-dw,were verified by DNA sequencing. Next, the construction of complete rha-p gene in pET22b(+) was performed. First, both pET22b(+)/rha-dw and pET22b(+)/rha-upweredigestedwithNcoI/SacIrestriction endonu-cleases to obtain, respectively, fragment rha-dw and linearized pET22b(+)/rha-up.
Digestionproductswerepurifiedfromagarosegel electrophore-sis,elutedandligated.LigationproductswereusedtotransformE. coliTop10competentcellsandtheresultingrecombinantplasmid, namedpET22b(+)/rha-pwasverifiedbyDNAsequencing.
2.4. ˛-l-Rhamnosidaserecombinantexpression
ProteinexpressionwascarriedoutinE.coliBL21(DE3)strain transformedwithpET22b(+)/rha-pplasmid.
Allthemediadescribedinthisparagraphcontained100g/mL ofampicillin.
2.4.1. Analyticalexpression
E.coli BL21(DE3) competentcells transformed withplasmid pET22b(+)/rha-pwereinoculatedina sterile50mLFalcontube containing12.5mLofeitherLB[50]orLBcontainingafinal con-centrationof0.5MNaCl(LB-N).Cellsweregrownunderconstant shakingat37◦Cupto0.6–0.7OD600.Thispreinoculumwasdiluted
1:100in12.5mLofeitheroneofthefourfollowingmedia:LB,LB-N, LBsupplementedwith1mMofbothbetaineandsorbitol(LB-BS),or LBcontainingafinalconcentrationof0.5MNaCland1mMofboth betaineandsorbitol(LB-NBS).Cellsweregrownunderconstant shakingat37◦Cupto0.7–0.8OD600.RHA-Precombinant
expres-sionwasinducedwith0.1mMIPTGateither23◦Cor37◦C;growth
wascontinuedin constant shakingfor 3h. Cellswere collected bycentrifugation(5,524×gfor15minat4◦C)andsuspendedin 25mMMOPSpH6.9atafinalconcentrationof14OD600.Cellswere
disruptedbysonication(12timesfora1-mincycle,onice)andan aliquotofeachlysatewascentrifugedat22,100×gfor10minat 4◦C.Bothsolubleandinsolublefractionswereanalyzedby SDS-PAGE.Thesolublefractionwasassayedforthepresenceof␣-RHA enzymaticactivity.
2.4.2. Largescaleexpression
Freshtransformedcellswereinoculatedinto10mLofLB-Nand incubatedinconstantshakingat37◦CO/N.Thepreinoculumwas diluted1:100infour2LErlenmeyerflaskscontainingeach500mL ofLB-NBSandincubatedinconstantshakingat37◦Cupto0.7–0.8 OD600.
Expression of the recombinant protein, named rRHA-P, was inducedwith0.1mMIPTGandgrowthwascontinuedfor 3hat 23◦C.Cellswerecollectedbycentrifugation(5,524×gfor15min at4◦C)andstoredat−80◦Cuntilneeded.
2.5. Nativeandrecombinant˛-l-rhamnosidasepurification Bothnativeandrecombinant RHA-Pwerepurified following threechromatographicsteps.Cellpastewassuspendedin25mM MOPSpH6.9,5%glycerol(bufferA),atafinalconcentrationof100 OD600and cellsweredisruptedbysonication(10timesfora
1-mincycle,onice).Celldebriswereremovedbycentrifugationat 22,100×gfor60minat4◦Candthesupernatantwascollectedand filteredthrougha0.45mPVDFMilliporemembrane.
Afterwards,cellextractwasloadedontoaQSepharoseFF col-umn(30mL)equilibratedinbufferA.Thecolumnwaswashedwith 50mLofbufferA,afterwhichboundproteinswereelutedbyusinga 300mLlineargradientofbufferAfrom0to0.4MNaClataflowrate of15mL/h.Thechromatogramwasobtainedbyanalyzingfractions absorbanceat=280nmandthepresenceoftherecombinant ␣-RHAactivitywasdetectedusingthepNPRassay.Relevantfractions wereanalyzedbySDS-PAGE,pooledandconcentratedatafinal vol-umeof∼0.5mLusinga30kDaAmiconultramembrane,Millipore. ThesamplewasthenloadedontoaSephacrylHRS200equilibrated withbufferAcontaining0.2MNaCl(bufferB).
Proteinswereelutedfromthegelfiltrationcolumnwith250mL ofbufferBataflowrateof12mL/h.Fractionswerecollected, ana-lyzedandscreenedforthepresenceof␣-RHAactivityaspreviously described.Atthisstage,NaClwasremovedfrompooledfractions byrepeatedcyclesofultrafiltrationanddilutionwithbufferA.The samplewasthenloadedonaQSepharoseFFcolumn(30mL) equi-libratedinbufferA.Thecolumnwaswashedwith50mLofbuffer A,afterwhichboundproteinswereelutedwith300mLofalinear gradientofbufferAfrom0to0.25MNaClataflowrateof13mL/h. Fractionswerecollected,analyzedbySDS-PAGEandscreenedfor thepresenceof␣-RHAactivity.Relevantfractionswerepooled, concentrated,purgedwithnitrogen,andstoredat−80◦Cuntiluse.
2.6. Analyticalgelfiltration
Analyticalgel-filtrationexperiments werecarriedout as fol-lows:100Lofa40pMproteinsamplewasloadedonaSuperdex 200HR10/300columnpreviouslyequilibratedinbufferB,installed onanAKTATMFPLCTM(GEHealthcareLifeScience).Sampleswere
elutedisocraticallyatRTataflowrateof0.5mL/min.Protein elu-tionwasmonitoredat=280nm.Amolecularweightcalibration wasperformedinthesamebufferwiththefollowingproteinsof knownmolecularweight:-amylase(200kDa), glyceraldehyde-3-phosphatedehydrogenase(143kDa),carbonicanhydrase(29kDa).
2.7. Enzymeactivityassays
␣-RHAactivitywasdeterminedusingpNPRassubstrate(pNPR assay). Otherwise stated, all activity assays were performed at RT. The reaction mixture contained, in a final volume of 0.5mLof50mMMOPSpH6.9,pNPRatafinalconcentrationof 600M and variable amounts of the sample tested. The reac-tionwasblockedafter10 and20minbyadding0.5MNa2CO3;
the product, p-nitrophenolate (pNP), was detected spectropho-tometrically at =405nm. The extinction coefficient used was 405=18.2mM−1cm−1.Oneunitofenzymeactivitywasdefined
astheamountoftheenzymethatreleasesonemicromoleofpNP permin.
2.7.1. Kineticparametersdetermination
KineticparameterswereobtainedatpH6.9usingapNPR con-centrationintherange0.025–1mM.Allkineticparameterswere determinedbyanon-linearregressioncurveusingGraphPadPrism (GraphPadSoftware;www.graphpad.com).
2.7.2. pHoptimum
pHoptimumfor␣-RHAactivitywasdeterminedintherange 4.7–8.8.Enzymeassayswereperformedasdescribedabove,using the following buffers: 50mM potassium acetate (pH 4.7–5.7), 50mMMOPS/NaOH(pH5.7–7.7)and50mMTris/HCl(pH7.7–8.8). 2.7.3. Temperatureoptimumandstability
Optimumtemperaturewasevaluatedbyperformingthe stan-dardpNPRassayandincubatingthereactionmixtureatdifferent temperatures,intherange25–55◦C.
Thethermalstabilityoftheenzymewasdeterminedby incubat-ingtheenzymeforoneand3hat30,40,50,60◦Candmeasuring, aftereachincubation,theresidualspecificactivity.
2.7.4. Organicsolventstolerance
The tolerance of the enzymatic activity to the presence of organicsolventsinthereactionmixture,suchasDMSO,acetone orethanol,wasevaluatedbyperformingthestandardpNPRassay in50mMMOPSpH6.9towhicheither10%or50%ofsolventwas added.
2.8. Substratespecificity
Reactionswerecarriedoutin0.6mLof50mMNa-phosphate bufferpH7.0undermagneticstirringat40◦Cinthepresenceof 20mM ofeither arylglycoside and 0.25U ofrRHA-P. Reactions weremonitoredover time(0–24h)byTLCanalysis(system sol-vent:EtOAc:MeOH:H2O70:20:10).CompoundsonTLCplateswere
visualizedunderUVlightorcharringwith␣-naphtholreagent. Hydrolysisreactionsofmaltose,pullulan,starch,amylopectin, sucrose,raffinose,lactose,xylanfrombirchwood,xylanfromoat spelt,hyaluronicacid,␣−cellulose,cellobiose,chitosan,-Glucan from barley, laminarin, curdlan,fucoidan from Fucus versiculo-sus,rhamnogalacturonan,rutinosewereperformedusing2.5mg ofeachsubstrate,whichwassuspendedin0.5mLof50mM Na-phosphatebufferpH7.0.Thereactionwascarriedoutat40◦Cunder magnetic stirring,in thepresence of 0.25U of rRHA-P. Hydrol-ysis productswere monitored by TLCanalysis (system solvent HCOOH:HAc:H2O:2-propanol:EtOAc:1:10:15:5:25).
Flavonoidiccompoundssuchasnaringin,rutin,neohesperidin dihydrochalcone,werescreenedaspossiblesubstrates.Inthiscase, a6mMsolutionofeachcompoundinafinal volumeof1mLof 50mMNa-phosphatebufferpH7wasincubatedat40◦Cinthe presenceof0.25UofrRHA-P.Reactionswerecheckedovertime byTLCanalysis(t=0,15,30,60,90,120,150,180,24h)with
thefollowingsolventsystem:EtOAc:MeOH:H2O70:20:10.An
addi-tionalhydrolysisreactionofnaringinwasperformedinconditions similartothosereportedaboveinthepresenceof10%DMSO,and thereactionwasmonitoredovertime.
Inallexperiments,TLCstandardsolutionsofpurereagentsand productswereusedforcomparison.
2.9. Massspectrometryanalysis
IdentificationofnativeandrecombinantRHA-Pbymass spec-trometrywasperformedbyenzymaticdigestion,eitherinsolution orinsituafterseparationbySDS-PAGE.
Whendigestinginsolution,theproteinsamplewaslyophilized and then dissolved in denaturing buffer (300mM Tris/HCl pH 8.8,6Murea,10mMEDTA).Cysteineswerereducedwith10mM dithiothreitolindenaturingbufferat37◦Cfor2h,andthen car-bamidomethylated with55mMiodoacetamide dissolved in the same bufferat RT for 30min, in the dark.Protein sample was desaltedbysizeexclusionchromatographyonaSephadexG–25M column (GE Healthcare, Uppsala,Sweden). Fractions containing RHA-Pwerelyophilizedandthendissolvedin10mMNH4HCO3
buffer(pH8.0).Enzymedigestionwasperformedusingproteases withdifferentspecificitysuchastrypsin,chymotrypsinand endo-proteinaseGluC(V8-DE)(Sigma),usingaprotease/RHA-Pratioof 1:50(w/w)at37◦Cfor16h.
For theenzymatic digestion insitu, Coomassie blue stained bands excised from polyacrylamide gels were destained by severalwasheswith0.1MNH4HCO3pH7.5andacetonitrile.
Cys-teines were reduced for 45min in 100L of 0.1M NH4HCO3,
10mMdithiothreitol,pH7.5.Carbamidomethylationofthiolswas achieved after 30min in the dark by adding 100L of 55mM iodoacetamidedissolvedinthesamebuffer.Enzymaticdigestion wascarriedoutondifferentbandsofRHA-P byusing100ngof proteases with different specificity (same as above) in 10mM NH4HCO3 buffer,pH7.5.Sampleswereincubatedfor2hat4◦C.
Afterwards,theenzyme solutionwasremovedanda fresh pro-teasesolutionwasadded;sampleswereincubatedfor18hat37◦C. Peptideswereextractedbywashingthegelspots/bandswith ace-tonitrileand0.1%formicacidatRT,andwerefilteredusing0.22um PVDFfilters(Millipore)followingtherecommendedprocedures.
Peptidemixtureswereanalyzedeitherbymatrix-assisted laser-desorption/ionizationmassspectrometry(MALDI-MS)orcapillary liquidchromatographywithtandemmassspectrometrydetection (LC–MS/MS).
MALDI-MSexperimentswereperformedona4800PlusMALDI TOF/TOF mass spectrometer (Applied Biosystems, Framingham, MA)equippedwithanitrogenlaser(337nm).Thepeptidemixture (1L)wasdiluted(1:1,v/v) inacetonitrile/50mMcitratebuffer (70:30v/v) containing10mg/mL of␣-cyano-4-hydroxycinnamic acid.Masscalibrationwasperformedusingexternalpeptide stan-dardspurchasedfromAppliedBiosystems.
Spectra were acquired using a mass (m/z) range of 300–4000amuandrawdatawereanalyzedusingData Explorer Software provided by the manufacturer. Experimental mass valueswerecomparedwithcalculatedmassesderivedfromanin silicodigestionofRHA-Pobtainedwithdifferentproteases,using MS-Digest, a proteomics tool from ProteinProspector software (prospector.ucsf.edu/).
PeptidemixtureswereanalyzedbynanoLC–MS/MSonaCHIP MS6520QTOFequippedwithacapillary1200HPLCsystemanda chipcube(AgilentTechnologies,PaloAlto,Ca).Aftersample load-ing,thepeptidemixturewasfirstconcentratedandthenwashedat 4nL/minina40nLenrichmentcolumn(AgilentTechnologieschip) with2%acetonitrile+0.1%formicacid.
ThesamplewasthenfractionatedonaC18reverse-phase cap-illarycolumnataflowrateof400nL/minbyusingatwo-solvents
systemconsistingof2%acetonitrile+0.1%formicacid(solventA) and95%acetonitrile+0.1%formicacid(solventB).Separationwas carriedoutwitha50minlineargradientfrom5to60%ofsolvent B.
Peptideanalysiswasperformedusingdata-dependent acquisi-tionofoneMSscan(massrangefrom300to2,000m/z)followed byMS/MSscansofthefivemostabundantionsineachMSscan.
MS/MSspectrawereacquiredautomaticallywhentheMS sig-naldetectedupraisedthethresholdof50,000counts.Doubleand triplechargedionswereisolatedandfragmented.Theacquired MS/MSspectraweretransformedinMascotGenericformat(.mgf) andusedtoqueryeitheraNCBInr02-2015(21,322,359,704amino acidsequences;10,835,265,410residues)oraN.sp.PP1Y(4635 aminoacidsequences;1,493,994residues)sequencedatabase.
A licensed version (2.4.0.) of MASCOT software
(www.matrixscience.com) was used with the following search parameters: trypsin, chymotrypsin or GluC endoproteinase as enzyme; 3 or 2 as allowed number of missed cleavages; car-bamidomethylation of cysteines as fixed modification; 10ppm MStoleranceand0.6Da MS/MStolerance;peptidechargefrom +2to+3.Oxidationofmethionineandformationofpyroglutamic acidfromglutamineresiduesattheN-terminalpositionof pep-tides were considered as variable modifications. Ions score is −10*Log(P),wherePistheprobabilitythattheobservedmatchisa randomevent.Individualionsscore>14indicateidentityor exten-sivehomology (p<0.05). Protein scoresderive from ion scores as a non-probabilisticbasis for ranking protein hits. Individual ionscorethresholdprovided byMASCOTsoftware,necessaryto evaluatethequalityofmatchesinMS/MSdataandtocalculate proteinscore,was52inNCBInrdatabaseand14inPP1YDatabase. InTable1onlypeptideswithionscore≥10arereported.
2.10. NMRAnalysis
All the experiments were performed at 298K on a Bruker 600MHzDRXspectrometerequippedwithacryoprobe.pNPRwas dissolvedinD2Otoafinalconcentrationof3mMina5-mmNMR
tube.Theligandprotonresonanceswereassignedbya combina-tionof1Dand2DNMRexperimentsincludingCOSY,TOCSYand HSQC.1HNMRspectrawereacquiredwith32kdatapoints.
Dou-blequantum-filteredphasesensitiveCOSYandTOCSYexperiments wereperformedbyusingdatasetsof2096×256(t1xt2)points. TotalCorrelation Spectroscopy (TOCSY)spectrum wasrecorded witha spin locktimeof 100ms.Heteronuclear singlequantum coherence(HSQC)experimentwasmeasuredinthe1H-detected
modevia singlequantum coherencewithprotondecouplingin the13Cdomain,byusingdatasetsof2048×256points.
Experi-mentswerecarriedoutinthephase-sensitivemodeaccordingto themethodofStatesandcoworkers[51].Inallhomonuclear spec-trathedatamatrixwaszero-filledintheF1dimensiontogivea matrixof4096×2048pointsandresolutionwasenhancedinboth dimensionsbyacosine-bellfunctionbeforeFouriertransformation. Aftertheacquisitionoftheinitialspectraoftheligandalone, rRHA-Pwasaddedtoafinalconcentrationof∼1M(∼60g)in theNMRtubefroma∼5.9Msolutionin25mMMOPS contain-ing5%glycerol.TheNMRsamplewasimmediatelyplacedinthe spectrometerprobetorecordexperimentsonthemixture.
DataacquisitionandprocessingwereperformedwithTOPSPIN 2.1software.
2.10. Analyticalmethods
Proteinconcentrationwas measuredusing theBio-Rad Pro-teinSystem[52]usingbovineserumalbumin(BSA)asstandard. Polyacrylamidegelelectrophoresiswascarriedoutusingstandard techniques[53].SDS–PAGE15%Tris–glycinegelswererununder
denaturingconditionsandproteinswerestainedwithCoomassie brilliantblueG-250.“Widerange”(200–6.5kDa)molecularweight standardwasfromSigma(ColorBurstTMElectrophoresisMarker).
3. Results
3.1. Isolationandidentificationofa˛-RHAfrom Novosphingobiumsp.PP1Y
Inapreviouswork[47],a␣-RHAactivitywasdetectedinthe crudeextractofN.sp.PP1Ygrowninminimalmediumandinthe presenceof0.3mMnaringin.Inordertoidentifyandcharacterize thisenzymaticactivity,N.sp.PP1Ycellsweregrownunderthesame experimentalconditionsdescribedbytheauthors[47].Cellswere collectedbycentrifugationafter24hofgrowthinPPMM(optical densityof∼1OD600)at5,524×gfor15minat4◦C.␣-RHA
purifi-cationwascarriedoutfollowingtheproceduresdescribedinthis workinMaterialsandMethods.Inallthreechromatographicsteps usedforthepurificationofN.sp.PP1Ycrudeextract,auniquepeak of␣-RHAactivitywasalwaysdetected,butnomajorproteinband waseverevidentfromtheSDS-PAGEanalysisofthecorresponding activefractions(datanotshown).
Peak fractions obtainedfrom the first two chromatographic steps were digested with trypsin and analyzed by LC–MS/MS analysis. Raw data were used tosearch NCBI database (Bacte-riaastaxonomyrestriction)withtheMS/MSionsearchprogram of Mascot software. Among the identified proteins, a mem-ber of the glycoside hydrolase family (Novosphingobium sp. PP1Y:WP013837086.1)wasidentifiedwith10(20%ofsequence coverage)and32peptides(46%ofsequencecoverage)inthefirst and inthesecond chromatographicstep,respectively (datanot shown).TheputativeMW,deducedfromtheaminoacidsequence, wasof124,225Da.
Peakfractionsobtainedfromthelastchromatographicstep (Q-sepharose)showedthepresenceofarelativelylimitednumberof proteinbands(Fig.S1,Supplementarymaterial).Fivebands migrat-ing asexpected for proteins of MWhigher than 100kDa were selectedandexcisedfromthepolyacrylamidegelstainedwith Col-loidalCoomassieBlue.Bands weredigestedinsituwithtrypsin, andtheresultingpeptidesmixturewasanalyzedbyLC–MS/MS. RawdatawereusedtosearchNCBIdatabase(Bacteriaastaxonomy restriction)withMASCOTsoftware.
Oneoftheprotein bandsanalyzed,indicated in Fig.S1by a redarrow, wasidentified(52 peptides,71%sequence coverage, Table1)astheproteinencodedbytheorfPP1YRS05470;thislatter islocatedinoneoftheextrachromosomalelementsofN.sp.PP1Y, theMegaplasmidreferredtoasMpl[44].Thisprotein,fromnow onindicatedasRHA-P,hadbeenpreviouslyannotatedasa mem-beroftheGHsfamilyandiscomposedby1,146aaforacalculated molecularweightofabout124,225Da.
3.2. Cloning,recombinantexpressionandpurificationof recombinantRHA-P
The recombinant plasmid pET22b(+)/rha-p (Materials and Methods),in which orfPP1YRS05470wascloned, wasusedto transformE.coliBL21(DE3)strain.Firstattemptsofexpressionof therecombinantproteinat37◦Cshowedthepresenceofaprotein bandwiththeexpectedmolecularweightofrecombinantRHA-P (rRHA-P)almostexclusivelyintheinsolublefractionoftheinduced culture(Fig.S2).AsevidentfromFig.S2oftheSupplementary mate-rial,whenE.coliwasinsteadinducedat23◦C[54],thepresenceof aproteinbandwiththeexpectedMWforrRHA-Pinthesoluble fraction(lane5)appearedtobemarkedlyincreasedwhen com-paredtothesolublefractionobtainedafteraninductionperformed
Table1
IdentificationofnativeRHA-PbyinsitudigestionfollowedbyLC–MS/MSanalysis.
Protein[Species]NCBInr Accessionnumber
Mass Proteinscore Sequencecoverage(%) n◦ofpeptides Ionscoreaandpeptidesequence
Glycosidehydrolasefamily protein[Novosphingobiumsp. PP1Y] gi|334145605|ref| WP013837086.1| 124,225 2318 71 52 34R.TGDAIK.A 31R.VGGFYGK.G 38R.GVVQNEK.R 28R.LDNELVR.G 21R.GPEIEPVK.A 11R.VLPITGRK.A 19KVGSKNEPR.G 37R.AVAEAAFDK.V 42R.GVVQNEKR.Q 48KLSSPNAMALR.A 13KLSSPNAMALR.A+Oxidation(M) 37R.DITDQVPVLR.L 13R.ITVVGDITMDR.L 41R.ITVVGDITMDR.L+Oxidation(M) 32R.LGRPEDYYVK.L 17KAPIVSVTLYYK.V 24KVEKWFGQIPR.G 55KVGLHILGGLASSR.L 52R.VKAQAEAELAQK.L 61R.LDATAIAEAQER.L 35R.TFVVSAPVQADR.T 28R.MGWLLPAVTQDK.L 15R.MGWLLPAVTQDK.L+Oxidation(M) 27R.KAPIVSVTLYYK.V 52R.DLAAMAALTPAQVK.A 70R.DLAAMAALTPAQVK.A+Oxidation(M) 74R.AVVSTLSGEIGGLER.V 32R.LKPMLEQAFGAWK.A 31R.LKPMLEQAFGAWK.A+Oxidation(M) 74KGATLAEGQVYSDDPAK.Y 28R.RAVVSTLSGEIGGLER.V 50KFVLANGLTTIVHTDR.K 18KSAAPVKPLGAAVPAAQGR.I 23R.NWTAPGINDPDTPALK.V 37R.MGWLLPAVTQDKLDK.Q 33R.MGWLLPAVTQDKLDK.Q+Oxidation(M) 47R.AVAEAAFDKVLADYLR.D 65R.ALPNQFETNAAVQAAIR.K 25KGWSYGVYSSVTQPTGPR.T 29KWMSRPVYALNVVPGPR.T 26KWMSRPVYALNVVPGPR.T+Oxidation(M) 20R.TFVVSAPVQADRTGDAIK.A 18KGATLAEGQVYSDDPAKYK.R 36R.IVLIDRPNSPQSVIMAGR.V 43R.IVLIDRPNSPQSVIMAGR.V+Oxidation(M) 58KADTEALGLANDVLGGGFLSR.L 47R.GPEIEPVKAGPVTLPAPLSR.D 55R.HATIGSMADLDAATLTDVRK.W 56R.HATIGSMADLDAATLTDVRK.W+Oxidation(M) 21R.EEKGWSYGVYSSVTQPTGPR.T 46KAIIADVGAFPGQKPITQEELQR.V 70R.TNYVETVPTGALDLALFMESDR.M 50R.AIEASANEFLRPGGMTYVVVGDR.K 59R.AIEASANEFLRPGGMTYVVVGDR.K+Oxidation(M) 65R.AIEASANEFLRPGGMTYVVVGDRK.I 54R.QGDNQPYGLFDYAQADGLLPVGHPYR.H+ Gln->pyro-Glu(N-termQ) 71R.QGDNQPYGLFDYAQADGLLPVGHPYR.H 23R.LNKDLREEKGWSYGVYSSVTQPTGPR.T 38KWFTEHYGPNNVVLVLSGDIDAATARPK.V 35KIVEPQLEGLGLPIDVQQAPQAAADDQSVE.-55 R.KIVEPQLEGLGLPIDVQQAPQAAADDQSVE.-40R.ALGSTLFGQHPYAQPTDGLGNAASLAALTPAALR.A aInthetableonlypeptideswithionscore≥10arereported.
at37◦C(lane3).NeithertheuseofdifferentIPTGconcentrations northevariationoftheinductiontimeseemedtoimproverRHA-P solubility.
ToincreasetheyieldofactiverRHA-Pinthesolublefractionof cellculturesinducedat23◦C,anovelsetofanalyticalexpression experimentsusingeitherLBorLB-N(MaterialsandMethods) sup-plementedwith1mMofbothbetaineandsorbitol(twoosmolytes),
wereperformed. Datain literaturesuggest that theadditionof thesemoleculestothegrowthmediumofinducedrecombinant cellsofE.colimightincreaseproteinsolubility.Thiseffectshould beimprovedbytheco-presenceinthemediumofahighsalt con-centration(LB-Nmedium),whichmighti)increasetheuptakerate in thecytoplasm of betaineand sorbitolfrom theextracellular
Table2
PurificationTableofrRHA-P.
TotalUnitsb SpecificActivity
(Units/mg)b P.F.a Yield(%) Celllysate 13,728.8 49.8 IQ-Sepharose 5,939.6 54.7 1.1 43.3 Gelfiltration 1,370.6 66.1 1.3 9.9 IIQ-Sepharose 920.9 346.9 6.9 6.7 aPurificationFactor.
b Standarddeviations,expressedaspercentageerrors,werealwayswithin20%.
Fig.1.SDS-PAGEanalysisofrRHA-Ppurificationsteps.Lane1:MWstandard.Lane 2:celllysateofE.colistrainBL21(DE3).Lane3:peakfractionafterthefirst Q-sepharose.Lane4:peakfractionafterS200-gelfiltrationchromatography.Lane5: purifiedrRHA-P;theredarrowinthislineindicatesthecontaminantproteinfound attheendofpurification.Foreachsampleanamountofca.4goftotalproteins wasloaded.(Forinterpretationofthereferencestocolourinthisfigurelegend,the readerisreferredtothewebversionofthisarticle.)
environment[55],and/orii)inducetheexpressionofadditional heat-shockproteinsassistinginproteinfolding[56].
Fortheseexperiments,IPTGconcentrationandinductiontime wererespectively0.1mMand3h.Activityassaysshowedthatthe bestexpressionofactiverRHA-Pwasobtainedinducingthe cul-tures at 23◦C and using LB-NBS(LB-N towhich 1mM of both betaineandsorbitolwereadded)asgrowthmedium.Inthiscase, specificactivity of cell extract was46 times higher than what obtainedwiththesolublefractionofrecombinantcellsgrownin LBandinducedat37◦C.
Large-scalerecombinantexpressionwasperformedbyusingthe optimizedconditionssuggestedbytheanalyticalexperiments.
rRHA-Ppurificationwascarriedoutfollowingthreedifferent chromatographicsteps.Thepurificationtable(Table2)showeda finalpurificationfactor(PF)of6.9.SDS-PAGEanalysisofpeak frac-tionsrepresentativeofeachpurificationstep(Fig.1)showedinthe finalpurifiedsampleofrRHA-Ptheadditionalpresenceofamajor contaminantproteinwithanapproximateMWof∼35kDa.
The 35kDa protein bandwas excised, digestedin situ with trypsin and analyzed by LC–MS/MS. The protein, identified by searchingtheE.colisequencedatabase(MaterialsandMethods), resultedtobeglyceraldehyde3-phosphatedehydrogenase,as iden-tifiedby17peptidesmatchesforasequencecoverageof71%.
PurifiedrRHA-P amino acidicsequence was verifiedupto a 94%coverage,byMSMappinganalysiscarriedoutbyMALDI-TOF andLC–MS/MS.However,itisworthnotingthattheprotein N-terminus,detectedbyMSanalysis,lacksthefirst23aminoacids fromtheexpectedclonedsequence(Fig.2andFig.S3).
ToconfirmthelackoftheputativeN-terminalpeptide,purified rRHA-PwassubjectedtobothaMALDI-TOFandLC–MS/MSanalysis afteraninsitudigestionwitheithertrypsinorchymotrypsin,and totheN-terminussequencing.Massspectrometryanalysisallowed the identification of peptides 23ESRDDAAEVAPSTRPEPSLEQAF45
(MH+ 2502.17 ppm) and 23ESRDDAAEVAPSTR36 (MH+ 1503.70)
whenrRHA-Pwasdigestedwithchymotrypsinortrypsin,
respec-Table3
TLCresultsofRHA-P-catalyzedreactionusingdifferentpNP-␣-andpNP-- deriva-tivesassubstrates.
Substratesa Hydrolysis Transglycosylation
3h 20h 3h 20h 1 pNP-␣-d-Glcp – – – – 2 pNP−-d-Glcp – – – – 3 pNP−␣-d-Galp – – – – 4 pNP−-d-Galp – – – – 5 pNP−␣-d-Manp – – – – 6 pNP−-d-Manp – – – – 7 pNP−-d-Xylp – – – – 8 pNP−␣-l-Fucp – – – – 9 pNP−-d-Fucp – – – – 10 pNP−␣-l-Araf – – – – 11 pNP−-l-Araf – – – – 12 pNP−␣-l-Rhap + + – – 13 pNP-GlcAp – – – – 14 pNP-d-NAGlcp – – – – 15 pNP-d-NAGalp – – – –
aAra,arabinose;Fuc,fucose;Gal,galactose;Glc,glucose;Man,mannose;NAGlc, N-acetylglucosamine;NAGal:N-acetylgalactosamine,GlcA:glucuronicacid,Rha, Rhamnose;Xyl,xylose.
tively.ThisevidencewasalsoconfirmedbyN-terminussequencing results,whichshowedthatthefirst5aminoacidsofthe recom-binantproteinareESRDD.Interestingly,alsoinnativeRHA-Pthe sameN-terminalpeptide,obtainedbyfragmentationspectrumof thetriplechargedionat834.73m/zfromtheinsitudigestionofthe proteinwithchymotrypsin(Fig.S4),hadbeendetectedinthefirst place.Thescarceamountofnativeprotein,however,didnotallow ustoconfirmtheeffectivelackoftheN-terminuspeptidebydirect aminoacidicsequencing.
3.3. BiochemicalcharacterizationofrRHA-P 3.3.1. Analyticalgelfiltration
TheoligomericstateofrRHA-Pwasverifiedbyanalyticalgel filtration using a Superdex 200 analytical column; the protein eluted as a single peak with an apparentmolecular weight of 101,500±5,000Da,whichindicatesthatrRHA-Pisamonomer. 3.3.2. Substratespecificity
ToinvestigaterRHA-P substratespecificity inhydrolysis and self-condensationprocesses,severalpNP-␣-andpNP-- deriva-tiveswereusedassubstratesinenzymaticassays performedas describedinMaterialsandMethods.Reactionswerefollowedover timebyTLCanalysisanddataarereportedinTable3.Asevident fromTable3,thepurifiedenzymeshowedactivityonlyonpNPR, confirmingthatrRHA-Pisindeeda␣-l-rhamnosidase.Moreover, noactivityonpNP--d-Glcwasdetected,excludingthatrRHA-P mightactasanaringinase,aclassofglycosylhydrolasehavingboth ␣-l-rhamnosidase and -d-glucosidase activities [40]. Further-more,oligo-andpolysaccharidessuchasrutinose,maltose,sucrose, lactose,pullulan,starch,amylopectin,raffinose,xylanfrom birch-wood,xylanfromoatspelt,hyaluronicacid,␣-cellulose,cellobiose, chitosan,-Glucanfrombarley,laminarin,curdlan,fucoidanfrom Fucusversiculosus,werealsotestedassubstratesforrRHA-P.TLC analysisofallthesereactions,performedasdescribedinMaterials andMethods,showedthelackofactivityofthisenzymeonanyof thesecompounds.
3.3.3. Mechanismofaction
TounderstandwhetherrRHA-P behavesasaninverting ora retainingglycosidase[1],1HNMRspectroscopywasusedto
mon-itor the stereochemical course of pNPR cleavage. As shown in Fig.3, after theadditionof the enzyme,a mixtureof ␣and 
Fig.2.rRHA-PsequencecoveragebyMSanalysis.Highlightedingreyarethepeptides,identifiedbyMSanalysis(MALDI-TOFand/orLC–MSMS),oftheenzymaticdigestions ofrRHA-Pcarriedouteitherinsituorinsolutionwithproteasesofdifferentspecificity:trypsin,chymotrypsinandendoproteinaseGluC(V8-DE).Aglobalsequencecoverage of94%wasobtained.CrossedistheN-terminussequenceexpectedforrRHA-P(referencesequenceWP013837086.1)andnotretrieved,eitherbyN-terminusorbyMS analysis.
Fig.3. 1HNMRanalysisofpNPRafterrRHA-Pcatalyzedhydrolysis.A.Spectrumoftheligandaloneinsolution(3mM).B.Spectrumfollowingtheadditionoftheenzymeto theNMRtube(∼60g),whichshowedtheappearanceofCH3signalsrelativeto␣-and-anomers.C.Spectrumobtainedaftera30minreaction;amixtureof␣-and -anomerswithadifferentrelativeratiowasobservedduetoemiacetalequilibrium.
Fig.4. rRHA-Pkineticbehavior.ReactionrateexpressedinM/minisplottedasa functionofpNPRconcentration.
rhamnoseanomerswasobservedinsolution,asevidentfromthe presenceoftwodifferentmethylgroupresonances.TOCSY spec-trumallowedtheassignmentofboth␣andspinsystemsupto themethylgroupresonanceslocatedatposition6ofeachrhamnose residue.Therefore,thelessshieldedmethylsignalwasassignedto -rhamnose.Theanalysisoftheintensityofthedifferentmethyl protonsignalsshowedthatthethermodynamicallyunfavorable -rhamnoseprevailedsoonaftertheadditionofrRHA-P(Fig.3B). Afterwards,asaconsequenceoftheequilibriumestablishedbya reducingmonosaccharideinwater,therelativeratioofthetwo peakschangeduptothepointinwhichthe␣anomericproductwas predominant(Fig.3C).Hence,thedatasuggestedthatrRHA-Pacts asaninvertingenzymeasitcleavesthe␣-glycosideligandyielding a-anomer,whichinwatersolutionequilibratesaccordingtothe anomericeffect.
3.3.4. Enzymaticcharacterization
Allenzymaticassaysdescribedinthissubsectionaredetailed inMaterialsandMethodssectionandwereperformedusingpNPR assubstrate.KineticconstantsonpNPRweredeterminedusinga substrateconcentrationrangingfrom0.025to1mM.Thereaction rate(M/min),plottedasafunctionofthesubstrateconcentration (M),showedatypicalMichaelis-Mententrendandisreportedin Fig.4.KMconstantwasof160.2(±17.3)M,avaluesignificantly
lower than most other ␣-RHAs described in literature [32,57], indicating ahigher affinity ofrRHA-P for pNPR.A Vmax of21.1
(±8.0)M/min,akcatof734.4(±212.9)sec−1andaKsof4.6(±1.7)
sec−1M−1wereobtained.
Fig.6.rRHA-Ptolerancetoorganicsolvents.Dataarereportedaspercentageof residualspecificactivity(S.A.)comparedtothecontrol(25mMMOPS,pH6.9).MOPS bufferisusedatafinalconcentrationof25mMandatpH6.9.
rRHA-Prevealedanoptimalactivityintherange6.0-7.5withan optimumatpH6.9(Fig.5A),retaining47%and36%ofactivityatpH 7.9and8.8,respectively.Inaddition,rRHA-Pactivitywasassayed inthetemperaturerange25◦C–55◦C.DataarepresentedinFig.5B andshowanoptimalactivitytemperatureatca.40.9◦C,witha44% retainingofitsinitialvalueupto50◦C.Toevaluatethethermal stabilityofrRHA-Pinthesamerangeusedtodeterminetheoptimal temperature,anincubationofthepurifiedproteinforeitherone or3hwasperformedatdifferenttemperatures;afterwards,the residualrhamnosidaseactivityoftheenzymewascalculated(Fig. S5).Theenzymaticactivityappearedtobestablebetween25and 40◦C,independentlyfromthedurationoftheincubation.At40◦C, correspondingtothetemperatureoptimum,theactivityisstable upto24h(datanotshown).
In orderto establishthetolerance ofthe enzymaticactivity of rRHA-P to the presence of organic solvents, pNPR assay in 50mMMOPSpH6.9containing10%or50%ofeitherDMSO, ace-toneorethanolwasperformed(Fig.6).Whileactivitydramatically decreasedinallconcentrationofacetoneandin50%ofethanoland DMSO,a66%ofactivitywasstillrecoveredinthepresenceof10% ofeitherDMSOorethanol.
3.3.5. Activityonflavonoids
TheabilityofrRHA-Ptohydrolyzenaturalflavonoidswas eval-uated.Enzymaticassaysusingnaringin,rutinandneohesperidin
Fig.5.rRHA-Pactivityassays.pH(A)andtemperature(B)optimumcurves.Inbothgraphsspecificactivity(S.A.)isplottedasafunctionofpHandtemperature,respectively. ThemaximumvalueofS.A.obtainedisreportedinbothgraphsandindicatedbyablackarrow.
dihydrochalconewereperformedand followedby TLCanalysis, asdescribedinMaterialsandMethods.Tothis purpose,a6mM solutionofeachcompoundwasincubatedwithpurifiedrRHA-P in50mMNa-phosphatebufferpH7.0at40◦C,andthereaction wasmonitoredbyTLCanalysesovertime.After3hofincubation, TLCanalysisofthereactionmixturesshoweda40–60%hydrolysis ofneohesperidindihydrochalconeandrutininthecorresponding derhamnosylatedneohesperidindihydrochalconeand quercetin-3--glucopyranoside(isoquercitrin).Besides,atotalconversionof naringinwasobservedasthecorrespondingproducts,rhamnose andprunin,weretheonlycompoundsobservedontheTLCplate. Reactiononnaringinwasalsocarriedoutinthepresenceof10% DMSO.Inthiscase,inourexperimentalconditions,thecomplete conversionofnaringininpruninandrhamnoseoccurredafterjust 1h.
4. Discussion
␣-RHAsare a group of glycosyl hydrolases (GHs) that have attracteda great deal of attentiondue totheirpotential appli-cationas biocatalystsina varietyofindustrial processes.These enzymesare of particular interest for the biotransformation of severalnaturalcompoundsusedinpharmaceuticalandfood indus-try. Enzymatic derhamnosylation catalyzed by ␣-RHAs can be used, for example, in functional foods and beverages contain-ing molecules with enhanced health-related properties. Some examplesencompass thebiotransformation of natural steroids, antibiotics,flavonoids,andterpenylglycosidesresponsibleforwine aromas[2,40].
Bacterial ␣-RHAs are still poorly characterized, but their biochemical properties might be of key importance for bio-transformationprocessesinvolvingreactionsconditionsthatare unfavorable for fungal rhamnosidases. In this work, we have successfullyexpressedandpartiallycharacterizedanovel recom-binant␣-RHAfromNovosphingobiumsp.PP1Y,namedRHA-P.The ␣-RHAactivityhasbeenisolatedfromthenativesourceandthe correspondingproteinsequencewasverifiedbyMSanalysis.The sequencewasusedforaBLASTsearchagainstallproteindatabases, and resultedtobehomologoustoGHspresentin thegenomes ofseveralSphingomonads;amongthese,a48%sequenceidentity ofRHA-Pwiththe␣-RHA(rhaM)fromSphingomonaspaucimobilis FP2001,wasretrieved[58].Noteworthy,ananalysisoftheamino acidicsequenceofRHA-Phighlightedthepresenceofa23amino acidslongsignalpeptide[59]locatedattheN-terminusofthe pro-tein.
Thenew␣-RHAidentified isencodedbyorfPP1YRS05470,a 3,441bpgene locatedinMegaplasmid Mplof N.sp.PP1Y [44], which was amplified and cloned into pET22b (+) plasmid to allowfortherecombinantexpressioninE.coliBL21(DE3)strain. Firstattemptsofexpressionoftherecombinantprotein,rRHA-P, resultedintheforemostpresenceoftheproteinintheinsoluble portionoftheinducedcultures. Alowerinductiontemperature, shiftedfrom37◦Cto23◦C,alongwiththeuseofahigh-saltLB for-mulationcontainingbothbetaineandsorbitol,whichpresumably actas“chemicalchaperones”,efficientlyconcurredinimproving theyieldofsoluble,activerRHA-P[54–56].
rRHA-Pwaspurifiedfollowingathree-steppurification proto-col.Therelativelypooryieldobtainedattheendofthepurification wasnotunexpectedatthisstage,aswechoseaftereachstepto avoidcollectingrRHA-Pactivefractionsthatshowedthepresence oftoomanyproteincontaminants.
PurifiedrRHA-P sequence wasverifiedbyMS mapping.It is interestingtounderlinethattheproteinN-terminusexpectedfrom theclonedsequence wasnever detected,either byMSanalysis (bothnativeandrecombinantRHA-P)orbyN-terminus
sequenc-ing (only recombinant RHA-P).The lack of this peptide, which hadbeenexpectedbyapreliminaryanalysisoftheaminoacidic sequence,confirmedthepresenceofasignalpeptidepresumably cleavedthroughapost-translationalproteolyticprocessing.A sim-ilarevidencehasbeendescribedforthe␣-RHAisolatedfromS. paucimobilisFP2001andrecombinantlyexpressedinE.coli[58]. TakeshiM.andcoworkersreportedinthiscasethepresenceofa putativesignalpeptidecleavedbetweenpositionsA24andN25 [58].Similarly,inrRHA-PthecleavagesiteislocatedbetweenA23 and E24. Theamino acidicsequence of theN-terminusof both ␣-RHAfromstrainFP2001andstrainPP1Yrevealedseveral pecu-liaritiesownedbybacterialsignalpeptides,suchasthepresenceof achargedregion(2–5residues)followedbyahydrophobicstretch of∼12aa,alongwiththepresenceofsmallandapolarresidues located1and3aaupstreamofthecleavagesite[60].The func-tionalroleofthiscleavage,aswellasthepossiblesortingofthese processedproteinsintheperiplasmicspaceofthenativebacteria, needstobefurtherinvestigated.
ApreliminarystructuralcharacterizationshowedthatrRHA-P isamonomericproteinwithanapproximatemolecularweightof 101,500±5,000Da.Mechanismofaction,kineticparametersand optimumreactionconditionsweredeterminedusingasynthetic substrate,pNPR.
Asforthereactionmechanism,GHsaregenerallygroupedintwo mainclasses,invertingandretaining.Atypicalinverting glycosi-daserequiresthepresenceofacatalyticacidresidueandacatalytic basicresidue,whereasatypicalretainingglycosidasecontainsa generalacid/baseresidueandanucleophile.Thesedifferent hydrol-ysismechanismslead,respectively,toaninversionortoaretention oftheanomericoxygenconfigurationintheproductcomparedto thesubstrate[61].NMRexperimentsshowedthatrRHA-Pactsas aninvertingGH;furtherinvestigationwillgiveadditionalinsights inidentifyingresiduesintheactivesiteofthisenzyme,whichare directlyinvolvedincatalysis.
rRHA-P catalytic efficiency, defined by a Ks value of
4.6s−1M−1,underlinedthattheactivityofthisenzymeresulted
to be comparable or even higher than that of other bacterial ␣-RHAsdescribed in literature. Whenclearly reported,suchas in lactic acid bacteria Lactobacillus plantarum, and Pediococcus acidilactici,Ks values onpNPRrangein factfrom0.01s−1M−1
to0.019s−1M−1 [38,39].In addition,Ks ofrRHA-P resultedto
be approximately3 times higher than the Ks of 1.74s−1M−1
reported for the ␣-RHA isolated from S. paucimobilis FP2001 [57].ThecomparisonamongtheKMvaluespresentinliterature,
rangingfrom1.18mMforS.paucimobilisFP2001to16.2mMfor the␣-RHAisolatedfromP.acidilactici,showedahigherapparent affinityofrRHA-PforpNPR(KMof160M),whichmightbeinpart
responsibleforthegreatdifferencesfoundinKSvalues.
ToinvestigatetheeffectivepotentialofusingrRHA-Pin biotech-nologicalprocessesinvolvingthebiotransformationofflavonoids, optimalreactionconditionsintermsofpH,temperatureand pres-enceoforganicsolventswereevaluated.Inparticular,rRHA-P,in linewithotherbacterial␣-RHAs[34–38,55],hasanoptimalactivity atpH6.9,a valuethathighlights amarkeddifferencewith fun-gal␣-RHAsactingat pHvaluesranging from4to5 [17,31,32]. rRHA-Pshowsanoptimalactivityat40.9◦Candasignificantoverall thermalstability,retaining∼66%oftheactivityupto45◦C,a tem-peraturerangemainlyusedinmostindustrialprocesses.Inthis respect,alsoagreatmajorityofotherfungalandbacterialenzymes exhibitsimilarvaluesoftemperaturestabilityandoptimalactivity [17,32,34,37,38,55],whereasonlyalimitednumberof␣-RHAsis morethermophilicandhasanoptimumoftemperatureatca.60◦C [31,36,37,39].rRHA-Phasamoderatetolerancetoorganicsolvents, usuallyemployedforthebiotransformationofflavonoidsthatare poorlysolubleinwater,retainingthe66%ofactivityonpNPRin solutionscontaining10%ofeitherDMSOorethanol.Naringin
con-versionis evenfaster in10%DMSO thanintheabsenceof this solvent,probablyowingtothefactthatthisflavonoidismore sol-ubleinthepresenceofDMSO.Noteworthy,flavonoidsconversion catalyzedbyother␣-RHAhasbeendescribedinthepresenceof 2–5%DMSO[37–39];higherconcentrationofDMSO,astheone usedforourbioconversion,mightfurtherincreasethesolubility ofmanyflavonoidsofcommercialinterest.Thesedataaltogether suggestahighertoleranceofrRHA-Ptothepresenceofsolvents comparedto␣-RHAsdescribedinliteratureforwhicharesidual20 −30%activityin12%ethanolanda24%activityin25%DMSOhas beenreported[17,39].ItisworthmentioningthatrRHA-Presidual activityinethanol-containingbuffersmightbeofadditional impor-tanceasmanybiologicallyrelevantrhamnosylatedflavonoidsare foundinwineandpossiblycitrusderivedalcoholicbeverages.
Aspreviouslyunderlined,thekineticdataobtainedsofarfor thisenzymeundoubtedlyencourageitsuseforthebioconversion ofglycosylatednaturalflavonoids thataremoresolubleathigh temperature (ca.50◦C) [62], basic pH values, and in the pres-enceoforganicsolvents.Inthiscontext,experimentstoverifythe hydrolyticpropertiesofrRHA-Ponfewnaturalflavonoids contain-ingeither␣-1,2or␣-1,6glycosidiclinkagesbetween␣-l-rhamnose and-d-glucosewereperformed.TLCanalysisonnaringin(␣-1,2), rutin(␣-1,6)andneohesperidin(␣-1,6)showedatotalconversion ofnaringinin thecorrespondingpruninand rhamnose,while a partialhydrolysisofrutinandneohesperidinwasobserved,thus confirmingtheabilityofthisenzymetohydrolyzeboth␣-1,2and ␣-1,6linkages.Likewise,themajorityofotherbacterialand fun-gal␣-RHAsexhibitsimilarsubstratespecificity,eventhoughthese enzymesoftenshowaslightpreferenceforthehydrolysisof␣-1,6 linkages[31,32,37–39].
In conclusion, the results described in this work boost the interesttowardafurtherbiochemicalcharacterizationofRHA-P activityusingforexampleawidearrayofglycosylatedflavonoids, toinvestigatetheeffectivepotentialofthisnovelenzymefor vari-ousbiomassandfood-processingapplicationsaimedatrecovering high-valueaddedcompoundsfromcomplexvegetablematrixes.
Acknowledgment
ThisresearchwassupportedbyFARB2014ORSA132082.The funderhadnoroleinstudydesign,datacollectionandanalysis, decisiontopublish,orpreparationofthemanuscript.
AppendixA. Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttp://dx.doi.org/10.1016/j.molcatb.2016.10. 002.
References
[1]B.Henrissat,Aclassificationofglycosylhydrolasesbasedonamino-acid sequencesimilarities,Biochem.J.280(1991)309–316.
[2]P.Manzanares,S.Vallés,D.Ramón,M.Orejas,␣-l-Rhamnosidases:oldand newinsights,industrialenzymes,structure,functionandapplication,in:J. Polaina,A.P.MacCabe(Eds.),IndustrialEnzymes,Springer,2007,pp.117–140. [3]M.Mutter,G.Beldman,H.A.Schols,A.G.J.Voragen,Rhamnogalacturonan
␣-l-rhamnopyranohydrolase.Anovelenzymespecificfortheterminal nonreducingrhamnosylunitinrhamnogalacturonanregionsofpectin,Plant Physiol.106(1)(1994)241–250.
[4]D.C.Crick,S.Mahapatra,P.J.Brennan,Biosynthesisofthe
arabinogalactan-peptidoglycancomplexofMycobacteriumTuberculosis, Glicobiology11(9)(2001)107R–118R.
[5]E.G.Maxwell,I.J.Colquhoun,H.K.Chau,A.T.Hotchkiss,K.W.Waldron,V.J. Morris,N.J.Belshaw,RhamnogalacturonanIcontaininghomogalacturonan inhibitscoloncancercellproliferationbydecreasingICAM1expression, Carbohydr.Polym132(2015)546–553.
[6]E.Nguema-Ona,M.Vicré-Gibouin,M.Gotté,B.Plancot,P.Lerouge,M.Bardor, A.Driouich,CellwallO-glycoproteinsandN-glycoproteins:aspectsof biosynthesisandfunction,Front.PlantSci.5(2014)499.
[7]W.Hashimoto,O.Miyake,H.Nankai,K.Murata,Molecularidentificationofan ␣-l-rhamnosidasefromBacillussp.strainGL1asanenzymeinvolvedin completemetabolismofgellan,Arch.Biochem.Biophys.415(2)(2003) 235–244.
[8]R.Rahim,U.A.Ochsner,C.Olvera,M.Graninger,P.Messner,J.S.Lam,G. Soberón-Chávez,CloningandfunctionalcharacterizationofthePseudomonas aeruginosarhlCgenethatencodesrhamnosyltransferase2,anenzyme responsiblefordi-rhamnolipidbiosynthesis,Mol.Microbiol.40(3)(2001) 708–718.
[9]-W.Nereus,I.V.Gunther,A.Nun˜ez,W.Fett,D.K.Y.Solaiman,Productionof rhamnolipidsbyPseudomonaschlororaphis,anonpathogenicbacterium,Appl. Environ.Microb.71(5)(2005)2288–2293.
[10]E.Champion,I.Andre,L.A.Mulard,P.Monsan,M.Remaud-Simeon,S.Morel, Synthesisofl-rhamnoseandN-acetyl-d-glucosaminederivativesenteringin thecompositionofbacterialpolysaccharidesbyuseofglucansucrases,J. Carbohydr.Chem.28(2009)142–160.
[11]E.Jr.Middleton,T.C.Theoharides,Theeffectsofplantflavonoidson mammaliancells:implicationsforinflammation,heartdisease,andcancer, Pharmacol.Rev.4(52)(2000)673–751.
[12]F.L.Perez-Vizcaino,J.Duarte,Flavonolsandcardiovasculardisease,Mol. AspectsMed.31(6)(2010)478–494.
[13]P.Fresco,F.Borges,M.P.M.Marques,C.Diniz,Theanticancerpropertiesof dietarypolyphenolsanditsrelationwithapoptosis,Curr.Pharm.Des.16 (2010)114–134.
[14]G.Ce´ıliz,M.C.Audisio,M.Daz,Antimicrobialpropertiesofprunin,acitric flavanoneglucoside,anditsprunin6-O-lauroylester,J.Appl.Microbiol.109 (4)(2010)1450–1457.
[15]M.Puri,S.S.Marwaha,R.M.Kothari,J.F.Kennedy,Biochemicalbasisof bitternessincitrusfruitjuicesandbiotechapproachesfordebittering,Crit. Rev.Biotechnol.16(1996)145–155.
[16]R.González-Barrio,L.M.Trindade,P.Manzanares,L.H.deGraaff,F.A. Tomás-Barberán,J.C.Espín,Productionofbioavailableflavonoidglucosidesin fruitjuicesandgreenteabyuseoffungalalpha-l-rhamnosidases,J.Agric. FoodChem.52(20)(2004)6136–6142.
[17]M.V.Gallego,F.Pi ˜naga,D.Ramón,S.Vallés,Purificationandcharacterization ofan␣-l-rhamnosidasefromAspergillusterreusofinterestinwinemaking,J. FoodSci.66(2)(2001)204–209.
[18]Z.Gunata,S.Bitteur,J.M.Brillouet,C.Bayonove,R.Cordonnier,Sequential enzymatichydrolysisofpotentiallyaromaticglycosidesfromgrape, Carbohydr.Res.184(1988)139–149.
[19]P.Manzanares,M.Orejas,J.V.Gil,L.H.deGraaff,J.Visser,D.Ramón, Constructionofageneticallymodifiedwineyeaststrainexpressingthe AspergillusaculeatusrhaAgene,encodingan␣-l-rhamnosidaseofenological interest,Appl.Environ.Microbiol.69(12)(2003)7558–7562.
[20]M.E.MeloBrancodeAraújo,Y.E.MoreiraFranco,T.GrandoAlberto,M.Alves Sobreiro,M.A.Conrado,D.Gonc¸alvesPriolli,A.C.H.FranklandSawaya,A.L.T.G. Ruiz,J.E.deCarvalho,P.deOliveiraCarvalho,Enzymaticde-glycosylationof rutinimprovesitsantioxidantandantiproliferativeactivities,FoodChem.141 (2013)266–273.
[21]V.D.Bokkenheuser,C.H.Shackleton,J.Winter,Hydrolysisofdietaryflavonoid glycosidesbystrainsofintestinalBacteroidesfromhumans,Biochem.J.248 (3)(1987)953–956.
[22]A.J.Day,J.M.Gee,M.S.DuPont,I.T.Johnson,G.Williamson,Absorptionof quercetin-3-glucosideandquercetin-4-glucosideintheratsmallintestine:
theroleoflactasephlorizinhydrolaseandthesodium-dependentglucose transporter,Biochem.Pharmacol.65(7)(2003)1199–1206.
[23]S.Wolffram,M.Block,P.Ader,Quercetin-3-glucosideistransportedbythe glucosecarrierSGLT1acrossthebrushbordermembraneofratsmall intestine,J.Nutr.132(4)(2002)630–635.
[24]A.Scalbert,G.Williamson,Dietaryintakeandbioavailabilityofpolyphenols,J. Nutr130(2000)2073S–2085S.
[25]P.C.Hollman,J.M.vanTrijp,M.N.Buysman,M.S.vanderGaag,M.J.Mengelers, J.H.deVries,M.B.Katan,Relativebioavailabilityoftheantioxidantflavonoid quercetinfromvariousfoodsinman,FEBSLett.418(1997)152–156. [26]K.Németh,G.W.Plumb,J.G.Berrin,N.Juge,R.J.Hassan,H.Y.Naim,G.
Williamson,D.M.Swallow,P.A.Kroon,Deglycosylationbysmallintestinal epithelialcell-glucosidasesisacriticalstepintheabsorptionand metabolismofdietaryflavonoidglycosidesinhumans,Eur.J.Nutr.42(2003) 29–42.
[27]P.C.Hollman,M.N.Bijsman,Y.vanGameren,E.P.Cnossen,J.H.deVries,M.B. Katan,Thesugarmoietyisamajordeterminantoftheabsorptionofdietary flavonoidglycosidesinman,FreeRadic.Res.31(6)(1999)569–573. [28]P.Garnier,X.T.Wang,M.A.Robinson,S.vanKasteren,A.C.Perkins,M.Frier,
A.J.Fairbanks,B.G.Davis,Lectin-directedenzymeactivatedprodrugtherapy (LEAPT):synthesisandevaluationofrhamnose-cappedprodrugs,J.Drug Target.18(10)(2010)794–802.
[29]M.A.Robinson,S.T.Charlton,P.Garnier,X.T.Wang,S.S.Davis,A.C.Perkins,M. Frier,R.Duncan,T.J.Savage,D.A.Wyatt,S.A.Watson,B.G.Davis,LEAPT: lectin-directedenzyme-activatedprodrugtherapy,Proc.Natl.Acad.Sci.101 (40)(2004)14527–14532.
[30]N.M.Young,R.A.Z.Johnston,J.C.Richards,Purificationofthe ␣-l-rhamnosidaseofPenicilliumdecumbensandcharacteristicsoftwo glycopeptidecomponents,Carbohydr.Res.191(1989)53–62.
[31]T.Koseki,Y.Mese,N.Nishibori,K.Masaki,T.Fujii,T.Handa,Y.Yamane,Y. Shiono,T.Murayama,H.Iefuji,Characterizationofanalpha-l-rhamnosidase fromAspergilluskawachiianditsgene,Appl.Microbiol.Biotechnol.80(6) (2008)1007–1013.
[32]P.Manzanares,H.C.vandenBroeck,L.H.deGraaff,J.Visser,Purificationand characterizationoftwodifferentalpha-l-rhamnosidases,RhaAandRhaB, fromAspergillusaculeatus,Appl.Environ.Microbiol.67(5)(2001)2230–2234. [33]J.E.Chua,P.A.Manning,R.Morona,TheShigellaflexneribacteriophageSf6
tailspikeprotein(TSP)/endorhamnosidaseisrelatedtothebacteriophageP22 TSPandhasamotifcommontoexo-andendoglucanasesandC-5epimerases, Microbiology145(7)(1999)1649–1659.
[34]W.Hashimoto,H.Nankai,N.Sato,S.Kawai,K.Murata,Characterizationof ␣-l-rhamnosidaseofBacillussp.GL1responsibleforthecomplete depolymerizationofgellan,Arch.Biochem.Biophys.415(2)(1999)56–60. [35]A.G.Orrillo,P.Ledesma,O.D.Delgado,G.Spagna,J.D.Breccia,Cold-active
␣-l-rhamnosidasefrompsychrotolerantbacteriaisolatedfroma sub-Antarcticecosystem,Enz.Microbiol.Technol.40(2)(2007)236–241. [36]V.V.Zverev,C.Hertel,K.Bronnenmeier,A.Hroch,J.Kellermann,W.H.
Schwarz,Thethermostable␣-l-rhamnosidaseRamAofClostridium stercorarium:biochemicalcharacterizationandprimarystructureofa bacterial␣-L-rhamnosidehydrolaseanewtypeofinvertingglycoside hydrolase,Mol.Microbiol.35(2000)173–179.
[37]M.Avila,M.Jaquet,D.Moine,T.Requena,C.Peláez,F.Arigoni,I.Jankovic, Physiologicalandbiochemicalcharacterizationofthetwo
alpha-l-rhamnosidasesofLactobacillusplantarumNCC245,Microbiology155 (2009)2739–2749.
[38]J.Beekwilder,D.Marcozzi,S.Vecchi,R.deVos,P.Janssen,C.Francke,J.van HylckamaVlieg,R.D.Hall,Characterizationofrhamnosidasesfrom LactobacillusplantarumandLactobacillusacidophilus,Appl.Environ.Microbiol 75(2009)3447–3454.
[39]H.Michlmayr,W.Brandes,R.Eder,C.Schümann,A.M.delHierro,K.D.Kulbe, Characterizationoftwodistinctglycosylhydrolasefamily78
␣-l-rhamnosidasesfromPediococcusacidilactici,Appl.Environ.Microbiol.77 (18)(2011)6524–6530.
[40]M.Puri,Updatesonnaringinase:structuralandbiotechnologicalaspects, Appl.Microbiol.Biotechnol.93(1)(2012)49–60.
[41]Z.Cui,Y.Maruyama,B.Mikami,W.Hashimoto,K.Murata,Crystalstructureof glycosidehydrolasefamily78␣-l-rhamnosidasefromBacillussp.GL1,J.Mol. Biol.374(2)(2007)384–398.
[42]Z.Fujimoto,A.Jackson,M.Michikawa,T.Maehara,M.Momma,B.Henrissat, H.J.Gilbert,S.Kaneko,ThestructureofaStreptomycesavermitilis
␣-l-rhamnosidaserevealsanovelcarbohydrate-bindingmoduleCBM67 withinthesix-domainarrangement,J.Biol.Chem.288(17)(2013) 12376–12385.
[43]E.Notomista,F.Pennacchio,V.Cafaro,G.Smaldone,V.Izzo,L.Troncone,M. Varcamonti,A.DiDonato,ThemarineisolateNovosphingobiumsp.PP1Y showsspecificadaptationtousethearomaticfractionoffuelsasthesole carbonandenergysource,Microb.Ecol.61(3)(2011)582–594. [44]V.D’Argenio,M.Petrillo,P.Cantiello,B.Naso,L.Cozzuto,E.Notomista,G.
Paolella,A.DiDonato,F.Salvatore,Denovosequencingandassemblyofthe wholegenomeofNovosphingobiumsp.strainPP1Y,J.Bacteriol.193(16) (2011)4296.
[45]W.Hashimoto,K.Murata,Alpha-l-rhamnosidaseofSphingomonassp.R1 producinganunusualexopolysaccharideofsphingan,Biosci.Biotechnol. Biochem.62(6)(1998)1068–1074.
[46]V.D’Argenio,E.Notomista,M.Petrillo,P.Cantiello,V.Cafaro,V.Izzo,B.Naso, L.Cozzuto,L.Durante,L.Troncone,G.Paolella,F.Salvatore,A.DiDonato, CompletesequencingofNovosphingobiumsp.PP1Yrevealsa
biotechnologicallymeaningfulmetabolicpattern,BMCGenomics15(2014) 384.
[47]V.Izzo,P.Tedesco,E.Notomista,E.Pagnotta,A.DiDonato,A.Trincone,A. Tramice,RhamnosidaseactivityinthemarineisolateNovosphingobiumsp. PP1Yanditsuseinthebioconversionofflavonoids,J.Mol.Catal.B:Enzym. 105(2014)95–103.
[48]B.L.Cantarel,P.M.Coutinho,C.Rancurel,T.Bernard,V.Lombard,B.Henrissat, Thecarbohydrate-ActiveEnZymesdatabase(CAZy).anexpertresourcefor glycogenomics,Nucl.AcidsRes.37(2009)D233–D238.
[49]J.Sambrook,E.F.Fritsch,T.Maniatis,MolecularCloning.ALaboratoryManual, 2nded.,ColdSpringHarborLab.Press,1982,pp.545.
[50]P.Gerhardt,R.G.E.Murray,W.A.Wood,N.R.Krieg,MethodsforGeneraland MolecularBacteriology,ASMPress,WashingtonD.C,1994.
[51]D.J.States,R.A.Haberkorn,D.J.Ruben,Atwo-dimensionalnuclearoverhauser experimentwithpureabsorptionphaseinfourquadrants,J.Magn.Reson.48 (1982)286–292.
[52]M.M.Bradford,Arapidandsensitivemethodforthequantitationof microgramquantitiesofproteinutilizingtheprincipleofprotein-dyebinding, Anal.Biochem.72(1976)248–254.
[53]F.He,Laemmli-SDS-PAGE,Bio-protocol.2011.Bio101:e80. http://www.bio-protocol.org/e80.
[54]J.R.Blackwell,R.Horgan,Anovelstrategyforproductionofahighlyexpressed recombinantproteininanactiveform,FEBSLett.295(1–3)(1991)10–12. [55]N.Oganesyan,I.Ankudinova,S.H.Kim,R.Kimb,Effectofosmoticstressand
heatshockinrecombinantproteinoverexpressionandcrystallization,Prot. Expr.Purif.52(2)(2007)280–285.
[56]M.Kilstrup,S.Jacobsen,K.Hammer,F.K.Vogensen,Inductionofheatshock proteinsDnaK,GroEL,andGroESbysaltstressinLactococcuslactis,Appl. Environ.Microbiol.63(5)(1997)1826–1837.
[57]F.Miake,T.Satho,H.Takesue,F.Yanagida,N.Kashige,K.Watanabe, Purificationandcharacterizationofintracellularalpha-l-rhamnosidasefrom PseudomonaspaucimobilisFP2001,Arch.Microbiol.173(1)(2000)65–70. [58]M.Takeshi,K.Nobuhiro,S.Tomomitsu,Y.Tadatoshi,A.Yoichi,M.Fumio,
Cloning,Sequenceanalysis,andexpressionofthegeneencoding SphingomonaspaucimobilisFP2001a-l-rhamnosidase,Curr.Microbiol.51 (2005)105–109.
[59]T.N.Petersen,S.Brunak,G.vonHeijne,H.Nielsen,SignalP4.0:discriminating signalpeptidesfromtransmembraneregions,Nat.Methods8(2011)785–786. [60]L.M.Gierasch,Signalsequences,Biochemistry28(1989)923–930.
[61]B.Cobucci-Ponzano,M.Moracci,Glycosidasesandglycosynthases,in: BiocatalysisinOrganicSynthesisofFaberK,FessnerW-DTurnerNJ,2016,pp. 483–506.
[62]T.Sathishkumar,R.BaskarR,S.Shanmugam,P.Rajasekaran,S.Sadasivam,V. Manikandan,Optimizationofflavonoidsextractionfromtheleavesof TabernaemontanaheyneanaWallusingL16Orthogonaldesign,Naturesci6(3)