Metabolism of Vertebrate Amino Sugars with N-Glycolyl Groups

25 Pages·2012·4.54 MB·English

THEJOURNALOFBIOLOGICALCHEMISTRY VOL.287,NO.34,pp.28898–28916,August17,2012
©2012byTheAmericanSocietyforBiochemistryandMolecularBiology,Inc. PublishedintheU.S.A.
Metabolism of Vertebrate Amino Sugars with N-Glycolyl
Groups
INCORPORATIONOFN-GLYCOLYLHEXOSAMINESINTOMAMMALIANGLYCANSBY
FEEDINGN-GLYCOLYLGALACTOSAMINE*□S
Receivedforpublication,March22,2012,andinrevisedform,May25,2012 Published,JBCPapersinPress,June12,2012,DOI10.1074/jbc.M112.363499
AnneK.Bergfeld‡,OliverM.T.Pearce‡,SandraL.Diaz‡,RogerLawrence‡,DavidJ.Vocadlo¶,BiswaChoudhury‡,
JeffreyD.Esko‡,andAjitVarki‡§1
FromtheDepartmentsof‡CellularandMolecularMedicineand§Medicine,GlycobiologyResearchandTrainingCenter,University
ofCaliforniaSanDiego,LaJolla,California92093-0687andthe¶DepartmentofMolecularBiologyandBiochemistry,SimonFraser
University,Burnaby,BritishColumbiaV5A1S6,Canada
Background:N-Acetylhexosaminesareprecursorsofallmajorvertebrateglycantypes.
Results:MammaliancellscanincorporateN-glycolylgalactosamine(GalNGc)andN-glycolylglucosamine(GlcNGc)intomost
majorcellularglycansandutilizeGalNGcfordenovobiosynthesisofNeu5Gc.
Conclusion: N-Acetylhexosamine biosynthetic pathways and glycan assembly are mostly tolerant toward the N-glycolyl
substituent.
Significance:CatabolismofNeu5GcmightcreatenovelN-glycolylatedglycanepitopesinvivo.
Theoutermostpositionsofmammaliancell-surfaceglycansare poratedintoChinesehamsterovarycellO-glycans.Remarkably,
predominantly occupied by the sialic acids N-acetylneuraminic GalNAcderivativesalteredtheoverallO-glycosylationpatternas
acid(Neu5Ac)andN-glycolylneuraminicacid(Neu5Gc).Todate, indicated by the occurrence of novel O-glycan structures. This
hydroxylation of CMP-Neu5Ac resulting in the conversion into study demonstrates that mammalian N-acetylhexosamine path-
CMP-Neu5Gcistheonlyknownenzymaticreactioninmammals ways and glycan assembly are surprisingly tolerant toward the
tosynthesizeamonosaccharidecarryinganN-glycolylgroup.In N-glycolylsubstituent.
our accompanying paper (Bergfeld, A. K., Pearce, O. M., Diaz,
S. L., Pham, T., and Varki, A. (2012) J. Biol. Chem. 287,
jbc.M112.363549),wereportametabolicpathwayfordegradation The cell surface of mammalian cells contains a variety of
ofNeu5Gc,demonstratingthatN-acetylhexosaminepathwaysare glycoconjugates,whichareinvolvedinmodulatingandmedi-
toleranttowardtheN-glycolylsubstituentofNeu5Gcbreakdown ating events such as cell-cell, cell-matrix, or cell-molecule
products.Inthisstudy,weshowthatexogenouslyaddedN-glyco- interactionsthatarecriticalforthefunctionanddevelopment
lylgalactosamine (GalNGc) serves as a precursor for Neu5Gc de ofacomplexorganism(2).N-Glycansarecovalentlyattached
novobiosynthesis,potentiallyinvolvingsevendistinctmammalian toasparagineresidueswithinaspecificconsensusmotifofpro-
enzymes.FollowingtheGalNAcsalvagepathway,UDP-GalNGcis teinsandshareacommonGlcNAc Man corestructure,which
2 3
epimerized to UDP-GlcNGc, which might compete with the can be processed and/or elongated with additional monosac-
endogenousUDP-GlcNAcforthesialicacidbiosyntheticpathway. charides such as galactose, GlcNAc, fucose, sialic acids, and
Using UDP-N-acetylglucosamine 2-epimerase/N-acetylmannos- lesscommonlybyGalNAc(3).ThemostabundantO-glycans
amine kinase-deficient cells, we confirm that conversion of arecharacterizedbyan(cid:1)-GalNAcresiduecovalentlyattached
GalNGcintoNeu5Gcdependsonthiskeyenzymeofsialicacid to serine or threonine residues of mammalian glycoproteins.
biosynthesis.Furthermore,wedemonstratebymassspectrometry GalNAcmaybefurtherextendedwithmonosaccharidessuch
thatthemetabolicintermediatesUDP-GalNGcandUDP-GlcNGc asgalactose,GlcNAc,fucose,andsialicacids,butnotmannose
serveassubstratesforassemblyofmostmajorclassesofcellular or glucose (4). Many nucleocytoplasmic proteins contain
glycans.WeshowforthefirsttimeincorporationofGalNGcand singleGlcNAcresiduesattachedtoserineorthreonineresidues
GlcNGcintochondroitin/dermatansulfatesandheparansulfates, (seeouraccompanyingpaper(5)).Glycosphingolipidsarethe
respectively. As demonstrated by structural analysis, N-glycoly- mostabundantclassofglycolipidsinmammalsandcommonly
latedhexosamineswerefoundincellulargangliosidesandincor- begin with a galactose (GalCer) or glucose (GlcCer) residue
beinglinkedtoaceramidebackbone.Inhigheranimals,most
*Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth glycosphingolipidsarebuiltonGlcCerandmaycontainaddi-
GrantsR01GM32373andR01CA38701(toA.V.).Thisworkwasalsosup- tional galactose, GalNAc, and sialic acid residues (6). Sialic
ported by Deutsche Forschungsgemeinschaft Research Fellowship BE
acid-containingglycosphingolipidsarereferredtoasganglio-
4643/1-1(toA.K.B.)andaCancerResearchInstitute/SamuelandRuth
sides.Anotherclassofglycansareglycosaminoglycansthatare
EngelbergFellowship(toO.M.T.P.).
□S ThisarticlecontainssupplementalFigs.S1–S5. longunbranchedpolysaccharidesconsistingofrepeatingdisac-
1Towhomcorrespondenceshouldbeaddressed:GlycobiologyResearchand charide building blocks. Examples are chondroitin sulfates
TrainingCenter,Depts.ofMedicineandCellularandMolecularMedicine,
(GalNAc/glucuronic acid), dermatan sulfates (GalNAc/idu-
UniversityofCaliforniaSanDiego,9500GilmanDr.,LaJolla,CA92093-
0687.Tel.:[phone];Fax:[phone];E-mail:[email protected]. ronicacidandglucuronicacid),andheparansulfates(GlcNAc
This is an Open Access article under the CC BY license.
28898 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER34•AUGUST17,2012
MammalianPathwaysforN-Glycolylhexosamines
and iduronic acid or glucuronic acid), which are covalently (35–40).Inthisstudy,weuseexogenousGalNGcformetabolic
attached to serine residues in proteoglycans and represent a labelinganddemonstrateitsincorporationinnearlyallofthe
majorcomponentoftheextracellularmatrix(7). majorglycanspresentinvertebrates.
Amongthemajormammalianclassesofglycans,N-glycans,
EXPERIMENTALPROCEDURES
O-glycans, and gangliosides typically carry sialic acids at the
outermostpositionsofaglycanchain(8).Sialicacidsareafam- SynthesisandChemicalCharacterization—“Brine”refersto
ily of acidic nine-carbon backbone monosaccharides, with a saturated aqueous solution of sodium chloride. Proton
N-acetylneuraminicacid(Neu5Ac)2andN-glycolylneuraminic nuclear magnetic resonance spectra (d ) were recorded on a
H
acid(Neu5Gc)beingthemostprevalentinmammals.Because JeolECA500(500MHz).500MHzspectrawereassignedusing
oftheirexposedpositionandquantityatthecellsurface,they COSY.Allchemicalshiftswerequotedonthe(cid:2)-scaleinppm,
areinvolvedinthestabilizationofmoleculesandmembranesas usingresidualsolventastheinternalstandard.Lowresolution
wellasrecognitionandmodulationofamultitudeofcell-cell mass spectra, obtained at the University of California San
and cell-environment interactions. The only known cellular Diego, Chemistry and Biochemistry Molecular MS Facility,
pathway for synthesis of Neu5Gc is the conversion of CMP- wererecordedonaMicromassPlatform1spectrometerusing
Neu5Ac into CMP-Neu5Gc catalyzed by the CMP-Neu5Ac electronspray(ES)ionizationwithmethanolascarriersolvent.
hydroxylase (CMAH) enzyme (9–16). In contrast to other Flash column chromatography was performed using Sorbsil
mammals studied to date, humans contain an inactivating C6040/60silicagel.Thinlayerchromatography(tlc)wasper-
frameshiftmutationintheCMAHgene(17–20),andhumans formed using Merck Kieselgel 60F254 pre-coated aluminum
thuslosttheabilitytosynthesizeendogenousNeu5Gc.How- backedplates.Plateswerevisualizedusing5%sulfuricacidin
ever,Neu5Gchasbeenfoundinvarioushumancancers,infetal methanol.
meconium,andeveninnormalhumantissues(21–27),appar- 2-Deoxy-2-[(hydroxyacetyl)amino]-D-galactopyranose(41)—
ently originating from Neu5Gc-rich animal-derived dietary Thehydrochloridesaltof2-amino-2-deoxy-D-galactopyranose
sources such as red meats or dairy products (28, 29). Our (1.0g,4.6mmol)wasdissolvedinwater(60ml)withsodium
accompanying paper (30) demonstrated that mice with a bicarbonate(10.8g,130mmol)andcooledinanicebath.Ace-
human-likedefectintheCmahgeneincorporateNeu5Gcfrom toxyacetylchloride(4.2ml,39mmol)wasaddedslowlytothe
aNeu5Gc-richdiet.Allhumanstestedthusfardisplayapoly- reaction mixture. After 30 min, tlc (EtOAc/MeOH, 7:3, v/v)
clonal circulating anti-Neu5Gc antibody repertoire (31, 32), showedcompleteconsumptionofthestartingmaterial (R (cid:1)
f
whichsuggeststhatNeu5Gcmightbethefirstknownexample 0.0)andtheformationoftheproduct(R (cid:1)0.3).Thereaction
f
ofa“xeno-autoantigen”(33). mixture was neutralized with mixed bed resin and concen-
As the CMAH enzyme is the only known mammalian bio- tratedundervacuum.Theproductwaspurifiedthroughsilica
syntheticpathwaytosynthesizeasugarcarryinganN-glycolyl columnandconcentratedundervacuumtoyieldawhitegum
group,Neu5Gcistheonlylogicalsourceforanyotheramino (1.0g,4.2mM,91%).For1HNMR(D O,500MHz)(assignedfor
2
sugarcarryinganN-glycolylgroup.Thusfar,uptakeandrecy- themajoranomer),(cid:2)(cid:1)3.53(ddd,J(cid:1)1.1Hz,J(cid:1)4.5Hz,J(cid:1)
clingofNeu5Gcthroughthelysosomaltransportersialinhave 8.0 Hz, 1H, H-6), 3.59–3.62 (m, 2H, H-5, H-6(cid:2)), 3.64
beenwellstudied(29,34),butturnoverandputativedegrada- (dd,J (cid:1)3.7Hz,J (cid:1)14.0Hz,1H,H-4),3.82(dd,J (cid:1)2.8
3,4 4,5 3,4
tivepathwaysarejustbeginningtobeexplored.Inouraccom- Hz, J (cid:1) 11.2 Hz, 1H, H-3), 3.97 (s, 2H, CH OH), 4.04 (dd,
2,3 2
panyingpaper(1),wereportapathwayofNeu5Gcdegradation. J (cid:1)3.7Hz,J (cid:1)10.9Hz,1H,H-2),5.09(d,J (cid:1)3.7Hz,1H,
1,2 2,3 1,2
Therein,weshowthatmammaliancellscanconvertNeu5Gc H-1).13C-NMR(D O,500MHz)(cid:2)(cid:1)53.5,61.2,67.4.68.2,70.9,
2
intoManNGc,followedbyepimerizationtoGlcNGcandphos- 74.9,95.1(C-1),175.8(C(cid:1)O).m/z(ESI(cid:3))was260(M(cid:3)Na(cid:3),
phorylationintoGlcNGc6-phosphate,withfinalde-acylation 100%),238(M(cid:3)H(cid:3),20%).TheHRMSm/z(ES(cid:3))calculated
of the latter, resulting in formation of the ubiquitous cellular for C H NO Na (M (cid:3) Na(cid:3)) was 260.0741 and found was
8 15 7
metabolites glycolate and glucosamine 6-phosphate (1). This 260.0740.
finding already suggests a certain level of tolerance of basic 1,3,4,6-Tetra-O-acetyl-2-deoxy-2-[(acetoxyacetyl)amino]-D-galact-
metabolic enzymes toward the N-glycolyl substituent and opyranose—2-Deoxy-2-[(hydroxyacetyl)amino]-D-galactopyranose
raises the question of whether the N-glycolylated breakdown (200mg,0.80mmol)wasmixedwithaceticanhydride(5ml)
productsofNeu5Gcareabletobesalvaged.TheGalNAcsal- andpyridine(5ml)togenerateaslurry.After24h,tlc(petrol/
vagepathwayhasalreadybeendescribedtobeveryefficientin EtOAc, 9:1) showed complete consumption of the starting
introducingartificialGalNAcderivatives,resultinginmodified material (R (cid:1) 0.0) and formation of the product (R (cid:1)
f f
O-glycans on glycoproteins and nucleocytoplasmic proteins 0.3).Theclear/colorlesssolutionwasdriedunderreducedpres-
suretoyieldastickystraw-coloredgum.Theresiduewasdis-
solved in chloroform and washed with sodium bicarbonate,
2Theabbreviationsusedare:Neu5Ac,N-acetylneuraminicacid;CID,collision- water, and finally brine. The solution was then dried with
induceddissociation;CMAH,CMP-Neu5Achydroxylase;CS,chondroitin
sodium sulfite, filtered, and dried under vacuum. The crude
sulfate;DMB,1,2-diamino-4,5-methylenedioxybenzene;DS,dermatansul-
product was purified through a silica column (petrol/EtOAc,
fate; GAG, glycosaminoglycan; GalNAz, N-azidoacetylgalactosamine;
GalNGc, N-glycolylgalactosamine; GlcNGc, N-glycolylglucosamine; GNE, 9:1)andconcentratedundervacuumtoyieldawhitegum(320
UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase; mg,0.72mM,90%).For1HNMR(CD Cl,500MHz)(assigned
Hex, hexose; HS, heparan sulfate; ManNGc, N-glycolylmannosamine; forthemajoranomer),(cid:2)(cid:1)2.01(s,3H,3CH ),2.02(s,3H,CH ),
Neu5Gc,N-glycolylneuraminicacid;UDP,uridinediphosphate;perGalNGc, 3 3
per-O-acetylatedGalNGc;GRIL,glycanreductiveisotopelabeling. 2.03(s,3H,CH ),2.04(s,3H,CH ),2.12(s,3H,CH ),4.08–
3 3 3
AUGUST17,2012•VOLUME287•NUMBER34 JOURNALOFBIOLOGICALCHEMISTRY 28899
MammalianPathwaysforN-Glycolylhexosamines
4.19(m,2H),4.22(t,J(cid:1)6.9Hz,1H),4.52(s,2H),4.65–4.70(m, 1,2-Diamino-4,5-methylenedioxybenzene(DMB)Derivatiza-
1H),5.26(dd,J(cid:1)3.1Hz,J(cid:1)11.4Hz,1H),and6.24(d,J(cid:1)3.8 tion and HPLC—DMB-HPLC was performed as described in
Hz,1H).For13CNMR(CD Cl,500MHz),(cid:2)(cid:1)20.6,20.7,49.7, ouraccompanyingpaper(30).Asstartingmaterial,(cid:6)5%ofcells
3
61.3,62.8,66.7,67.5,70.2,74.0,91.1,169.0,167.5,169.7,170.6, fromaconfluentP-100dishor100(cid:3)lofadensesuspensioncell
171.0,and171.8.m/z(ESI(cid:3))was470(M(cid:3)Na(cid:3),100%)and465 culturewereusedintheprotocol.Celllysatesthereofwerepre-
(M (cid:3) NH (cid:3), 15%). The HRMS m/z (ES(cid:3)) calculated for paredbyrepetitivefreeze-thawfollowedbysonication.Anali-
4
C H NO Na (M (cid:3) Na(cid:3)) was 470.1269 and found was quot (10%) of each lysate was used to determine the protein
18 25 12
470.1266. concentration using the BCA protein assay kit (Pierce). The
Preparation of [glycolyl-3H]GalNGc—To synthesize [glyco- remaining lysates were hydrolyzed with 2 M acetic acid to
lyl-3H]GalNGc, 18.5 MBq of [3H]glycolic acid (prepared as release glycosidically bound Neu5Gc prior to derivatization
describedinthecompanionpaper(1)wereincubatedwith50 withDMBandanalysisbyHPLC.ThepercentageofNeu5Gc
mMgalactosamineinthepresenceof50mMethyl-3-(3-dimeth- wascalculatedfromtotalsialicacidcontentinallsamples.
ylaminopropyl)-carbodiimide and 2 mM N-hydroxysulfosuc- Preparation and Analysis of UDP-sugars from Cells—Two
cinimidein50mMMOPSbuffer(pH7.5)inafinalvolumeof2 T-75flasksofBJA-BsubclonesK20andK88weresetupeach
with5(cid:5)106cellsin5mlofculturingmedia.Themediawere
ml. The reaction mixture was incubated overnight at room
supplementedwith1.036MBq[3H]GalNAc(AmericanRadio-
temperaturewithgentlemixing,subsequentlydilutedto10ml
labeled Chemicals, Inc.) or [3H]GalNGc, and cells were har-
withwater,andappliedtoa2mlDowex-50column(AG-50W-
vestedafter3days.Cellpelletswerewashedthreetimeswith
X2, Bio-Rad) pre-equilibrated with water. The column was
washed with 5 column volumes of water; the washes were PBS, resuspended in 30 (cid:3)l of 10 mM Tris-HCl (pH 7.5), and
lysedbyrepetitivefreeze-thaw.Afteradditionof70(cid:3)lofetha-
pooledwiththerun-through,andthepoolwasappliedontoa2
mlAG-3X-4A(OH(cid:4)form,Bio-Rad)columnpre-equilibrated nol (70% final) and incubation at (cid:4)20°C overnight, proteins
and membrane debris were precipitated (20 min/20,000 (cid:5)
withwater.Thecolumnwaswashedwith5columnvolumesof
g/4°C).Thesupernatantsweretransferredintofreshtubes,and
water,andthewasheswerepooledwiththerun-throughand
samples were dried down. For analysis, samples were resus-
dried down. The recovery was monitored by scintillation
pendedinafinalvolumeof110(cid:3)lofwater,whichwassupple-
counting, and the final yield of [glycolyl-3H]GalNGc was 7.2
mented with internal standard [14C]GlcNGc (1). The sample
MBq(39.0%).Theidentityofsynthesized[glycolyl-3H]GalNGc
wassterile-filteredpriortoHPLCanalysisusinghighperform-
wasconfirmedbyHPLCconfirmingco-elutionwiththenonla-
ance anion exchange chromatography-pulsed amperometric
beled chemically well characterized counterpart (data not
detection and scintillation counting. Monosaccharides were
shown).
separated by HPLC (Dionex DX-600 BioLC system equipped
Cell Culture—Human THP-I cells (42) as well as murine
withanED50ElectrochemicalDetector,Dionex)usingaCar-
Emeg32(cid:4)/(cid:4) and Emeg32(cid:3)/(cid:4) embryonic fibroblasts (43) were
boPacPA-1column(Dionex)underalkalineconditions.Sam-
cultivated in Dulbecco’s modified Eagle’s medium (DMEM ples(100(cid:3)l)wereinjected,andanoptimizedgradient(supple-
highglucose,Invitrogen).HumanM-21cells(44)weregrown
mental Fig. S3) was applied at a flow rate of 1 ml/min, and
in(cid:1)-minimalessentialmedium(Invitrogen),andhumanBlym-
0.5-mlfractionswerecollecteddirectlyintoscintillationvials.
phomacelllineBJA-B(Burkitt’slymphoma-like(45),subclones
To each vial, 5 ml of scintillation mixture (Scintiverse BD,
K88andK20(46)waskeptinRPMI1640medium(Invitrogen).
Fisher) was added prior to scintillation counting. Based on
WildtypeCHO-K1andmutantldl-Dcelllines(47)werecul-
the elution of the internal standard, elution profiles were
tured in DMEM/F-12 (Invitrogen). All media were supple-
superimposed.
mentedwith5%Neu5Gc-freehumanserum(heat-inactivated
IsolationofGlycosaminoglycans(GAGs)fromCells—Purifi-
and sterile-filtered, Valley Biomedical Inc.) instead of fetal
cationofGAGswasperformedbasedonestablishedprotocols
bovineserumand2mMglutamine.Cellswerecultivatedina (48). For radiolabeling studies, CHO-K1 and ldl-D cells were
humidified5%CO atmosphereat37°C.
2 split into P-100 dishes and cultured in regular media until 3
FlowCytometry—Experimentswereperformedasdescribed dayspriortoreachingconfluence.Atthistimepoint,themedia
previously (1). For cell feedings, 1 (cid:5) 106 suspension cells weresupplementedwith925kBq[3H]GalNAcor[3H]GalNGc.
(THP-I,BJA-BsubclonesK20orK88)wereseededinto6-well Three days later, cells were detached using 20 mM EDTA in
dishes(2mloffeedingmedia/well),andadherentcells(CHO- PBS.Theculturemediaweresaved,andcellswerewashedthree
K1, ldl-D, Emeg32(cid:4)/(cid:4), and Emeg32(cid:3)/(cid:4) cells) were used in timeswithPBS.Nonradioactiveexperimentswerecarriedout
P-100 dishes 3 days prior to reaching confluence again. The similarly,butP-150disheswereused,andcellswereincubated
feeding media were supplemented with compounds as indi- atafinalconcentrationof10mMGalNAcor10mMGalNGc.
catedusingthefollowingpHneutralstocksolutionseitherpre- Cellpelletswerelysedin2.5mlof0.1MNaOHfor15minat
paredinPBS(pH7.4)(100mMNeu5Gcand1MGalNGc)orin 37°C.Afteradditionof20mlofbufferA(50mMsodiumace-
DMSO(100mMGalNGc,100mMperGalNGc).Regardingcell tate,200mMNaCl(pH6.0)),thepHwasadjustedto8.0with
feedingswithGalNGc,theDMSOstockwasusedwhenaiming aceticacid.ThepHoftheculturemediawasadjustedto8.0as
at a final concentration of 100 (cid:3)M in the media. Cells were well.PronaseE(Sigma)wasaddedatafinalconcentrationof
culturedfor3daysinthefeedingmedia.Thereafter,adherent 100(cid:3)g/ml,andsampleswereincubatedfor24hinaend-over-
cellsweredetachedwith20mMEDTAinPBS,andallcellswere endshaker.Additional50(cid:3)g/mlofPronaseEwasadded,and
washedthreetimeswithPBS. sampleswereincubatedovernight.Sampleswereboiledfor10
28900 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER34•AUGUST17,2012
MammalianPathwaysforN-Glycolylhexosamines
min to inactivate the protease, followed by centrifugation at timeswith2mlofmethanol,fivetimeswith2mlof9:1meth-
3000(cid:5)gfor1h,andpassedthrougha5(cid:3)mlowbindingsyringe anol/aceticacid,andfivetimeswith2mlofmethanol.Finally,
filter.Apolyprepcolumn(1ml)waspreparedwithDEAE-Sep- thesamplewasdissolvedin1mlofwaterandpurifiedusinga
hacel(Sigma)andprewashedwith20mlofbufferB(bufferA(cid:3) C-18column(100mg,Waters),whichwaspre-washedasfol-
0.1%w/vTritonX-100)priortoloadingthesamples.Thereaf- lows:fourtimeswith1mlof10%aceticacid;fourtimeswith1
ter, columns were washed with 30 ml of buffer A, and GAGs ml of 50% methanol; four times with 1 ml of methanol; four
weresubsequentlyelutedwith5mlofbufferC(50mMsodium timeswith1mlof72%isopropylalcohol(cid:3)28%methanol(cid:3)
acetate, 1 M NaCl (pH 6.0)). The eluates were collected and 0.1% formic acid; four times with 1 ml of ethyl acetate; four
desalted on PD-10 columns (GE Healthcare) following the timeswith1mlofchloroform;twotimeswith1mlofmethanol;
manufacturer’s guidelines. Thereafter, samples were frozen fourtimeswith1mlof50%methanol;eighttimeswith1mlof
andlyophilized,andmaterialwasstoredat(cid:4)20°C. water.Afterloading,C-18columnwaswashedthreetimeswith
CL-6BGelFiltrationChromatography—IsolatedGAGsfrom 1mlof0.1%TFAtoelutereleasedO-glycans,andthesample
radiolabeled cells (see above) were resuspended in 500 (cid:3)l of waslyophilizedandstoredat(cid:4)80°C.
water and split equally into two tubes. Samples were diluted IsolationofN-GlycansfromCells—CHO-K1andldl-Dcells
10-foldwithbuffer(0.5MTris-HCl,0.5MNaCl(pH7.9))and20 were split into 2 P-150 dishes each and cultured in regular
milliunitsofchondroitinaseABC(Proteusvulgaris,Sigma)was media until 3 days prior to reaching confluency. At this time
added to only one of the samples followed by incubation at point,themediaweresupplementedwith10mMGalNAcor10
37°C overnight. Thereafter, saturated phenol red solution mMGalNGc.Threedayslater,cellsweredetachedusing20mM
(sterilefiltered)wasaddedtoafinalvolumeof0.5mltoeach EDTAinPBSandwashedthreetimeswithPBS.Forthispro-
sample. A gel filtration column (1.7 cm outer diameter by 42 tocol, 30% of cells from a confluent P-150 dish were resus-
cm, CL-6B, Sigma) was poured and equilibrated in buffer A. pendedin200(cid:3)lofbufferA(20mMHEPES,1%SDS(pH8.2))
Sampleswereloadedandcolumnrunwith(cid:6)40(cid:3)l/minusing andlysedbysonication.Thesamplewasheatedfor20minat
bufferA.Fractionswerecollected(15mineach,(cid:6)120fractions 100°C. Once the sample cooled down to room temperature,
total),andradioactivitywasdeterminedbyscintillationcount- 800(cid:3)lofbufferB(20mMHEPES,1.25%NonidetP-40(pH8.2)),
ing. The elution time of phenol red was set to 1, and relative 2(cid:3)lof(cid:4)-mercaptoethanol,and20(cid:3)lpeptide:N-glycosidaseF
elutiontimesofallfractionswerecalculatedaccordingly. (5 milliunits; prepared as described previously (52, 53)) were
LC/MS Analysis of GAGs—Enzymatically depolymerized added to the sample. The sample was incubated overnight at
GAGpreparationsweredifferentiallymasslabeledbyreductive 37°C in a head-over-head shaker followed by addition of
aminationwith[12C ]anilineasdescribedpreviously(49).Each another5(cid:3)lofpeptide:N-glycosidaseFandfurtherincubation
6
samplewasmixedwithcommerciallyavailablestandardunsat- for4–5hat37°C.AHyperSepHypercarbcharcoalcartridge
urateddisaccharidesthatweretaggedwith[13C ]aniline.Sam- (25 mg, Thermo Scientific) was pre-washed successively four
6
pleswereanalyzedbyliquidchromatography-massspectrom- timeswith1mlofacetonitrile,fourtimeswith1mlof50%
etry (LC/MS) using an LTQ Orbitrap Discovery electrospray acetonitrile, and eight times with 1 ml of 0.1% TFA, and a
ionization mass spectrometer (Thermo Scientific) equipped C-18columnwaspre-washedasdescribedaboveunder“Iso-
with an Ultimate 3000 quaternary high performance liquid lationofO-Glycans.”ThesamplewasloadedontoC-18col-
chromatography micro pump (Dionex) and a C-18 reverse- umn, allowed to elute directly onto the charcoal cartridge,
phase micro-bore column (Higgins Analytical) as described andletpassthroughbothcolumnscompletely.Columnswere
previously(49).DisaccharidescontaininganN-glycolylgroup washedtwotimeswith1mlofwater,andtheupperC-18was
wereidentifiedbasedontheiruniquemass.TheDisaccharide discardedthereafter.Thecharcoalcartridgewaswashedwith
Structure Code previously described (50) was amended to another1mlofwater,andtheN-glycanswereelutedwithtwo
includeN-glycolylmoietiesdenotedwithanuppercaseletterQ timeswith1mlof30%acetonitrile,twotimeswith1mlof60%
in heparan sulfate-derived disaccharides and a lowercase q acetonitrile, and 1 ml of 100% acetonitrile. The eluates were
in chondroitin sulfate-derived disaccharides. Thus, GlcA- pooled, and acetonitrile was removed under nitrogen gas fol-
GlcNGcisdenotedasG0Q0andGlcA-GalNGcasG0q0. lowedbylyophilizingthesample.
IsolationofO-GlycansfromCells—CHO-K1andldl-Dcells PermethylationofGlycans—Permethylationwasperformed
were split into 2(cid:5) P-150 dishes each and cultured in regular basedonpreviouslydescribedreports(54).IsolatedO-glycans
media until 3 days prior to reaching confluency. At this time orN-glycans(seeabove)weredriedinadesiccatorandresus-
point,themediaweresupplementedwith10mMGalNAcor10 pendedin300(cid:3)lofDMSO;thesamplewassealedandstirred
mMGalNGc(from1MneutralstocksinPBS(pH7.4)).Three for45minatRT.Aslurryof2–3NaOHpelletsin(cid:6)500(cid:3)lof
dayslater,cellsweredetachedusing20mMEDTAinPBSand DMSOwasfreshlypreparedusingmortarandpestleandadded
washedthreetimeswithPBS.Forthefollowingprotocol,60%of toeachsample.Afterstirringfor20–30minatRT,thesample
cellsfromaconfluentP-150dishwereresuspendedin600(cid:3)lof wasputonicepriortoadditionof250(cid:3)lofCH Itothefrozen
3
water,rupturedbysonication,andincubatedbystirringover- sample.ThereactionmixturewasputonastirreratRTfor20
nightat45°Cinthepresenceof50mMNaOHand1MNaBH minandthenanother50(cid:3)lofCH Iwasaddedfollowing10min
4 3
(51).Sampleswereneutralizedbyadditionof30%aceticacidon ofstirring.Thereactionwasterminatedbycarefuladditionof1
iceanddesaltedbypassingovera2mlDowexAG50–0W-8X mlofwatertothesampleoniceandstirringfor5min.There-
column. The flow-through was lyophilized, and borates were after,permethylatedN-/O-glycanswereextractedbyadding1
removedbyrepeatedco-evaporationundernitrogenflushfour mlofchloroformfollowedbystirringfor5minandcentrifuging
AUGUST17,2012•VOLUME287•NUMBER34 JOURNALOFBIOLOGICALCHEMISTRY 28901
MammalianPathwaysforN-Glycolylhexosamines
at2000rpmfor5mintoseparatetheorganicandtheaqueous positive and negative ion mode. The MS/MS of selected ions
phases.Thechloroformlayerwaswashedfivetimeswith1mlof was also performed in the negative ion mode using argon as
water to remove salts and DMSO and finally evaporated dry collisiongas.ThecompletesetofdatawereanalyzedusingData
with nitrogen flush. Permethylated samples were stored at Explorer software from Applied Biosystem and GlycoWork-
(cid:4)20°Cuntilused. benchsoftware(56).
IsolationofGangliosides—M-21cellsweresplitinto2P-150
RESULTSANDDISCUSSION
dishes and cultivated in standard media until 3 days prior to
reaching confluency. At this point, the media were supple- Mammalian Cells Are Capable of Converting Incorporated
mentedwithafinalconcentrationof10mMGalNAcor10mM GalNGc into Neu5Gc—In our accompanying paper (1), we
GalNGc(from1MneutralstocksinPBS(pH7.4)).After3days describeametabolicpathwayfordegradationofNeu5Gc.We
of incubation, cells reached confluency and were detached alsodemonstratedthattheNeu5Gc-degradingpathwayispar-
using 20 mM EDTA in PBS, and the pellet was washed three tiallyreversibleinthatexogenouslyaddedGlcNGcorManNGc
times with PBS. After addition of 20 volumes of chloroform/ resultedindenovobiosynthesisofNeu5Gc(1).Asapresumed
methanol,2:1,cellsweremixedonicefor30susingthePoly- “negative control,” we had synthesized GalNGc, i.e. galactos-
tronathighspeed.Thesamplewasincubatedonicefor10min aminecarryinganN-glycolylgroup.Surprisingly,weobserved
priortomixingcellsagainonicefor30susingthePolytronat that supplemented GalNGc was not only incorporated by
highspeed.Thesamplewasspunfor10minat2000(cid:5)g,andthe mammaliancellsbutalsoresultedindenovoNeu5Gcbiosyn-
organic supernatant was transferred into a clean tube. The thesisinvariouscelllines.Thisunanticipatedfindingprompted
remainingpelletwasextractedagainasdescribedabovewith20 anadditionalstudytofurtherinvestigatethetoleranceofmam-
volumesofchloroform/methanol,1:1,followedby20volumes malian N-acetylhexosamine pathways toward the N-glycolyl
of chloroform/methanol, 1:2, and finally with 20 volumes of substituent,aspresentedinthispaper.
chloroform/methanol/water,10:10:1.Allextractswerepooled Human THP-I cells were kept under Neu5Gc-free condi-
(pool 1). The remaining pellet was dissolved in 3 ml of 95% tions,andthemediaweresupplementedwitheitherGalNGcor
ethanol,stirredat4°Covernight,andspun10minat2000rpm. perGalNGc for 3 days. In parallel, non-fed cells were kept in
Thesupernatantwasaddedtopool1,andtheoriginaltubewas culturetoserveasanegativecontrol,andNeu5Gc-loadedcells
washedwith2mlofchloroform/methanol,2:1,priortoadding wereusedasapositivecontrol.Loadedcellswereanalyzedby
ittothepelletandmixing.Thetubewasspun10minat2000 flow cytometry using anti-Neu5Gc IgY (1, 27). These human
rpmagain,andthesupernatantwasaddedtopool1anddried cellswereabletouseGalNGcasaprecursorfordenovobio-
undernitrogengas.Totallipidextractswerethenresuspended synthesisofNeu5Gc.Although100(cid:3)MGalNGcinthefeeding
in500(cid:3)lofpotassiumphosphatebuffer(pH7.5),sonicatedfor media was not sufficient to achieve detectable levels of cell-
10minpriortoadditionof6unitsofphospholipaseC(Sigma), surface-exposedNeu5Gc,100(cid:3)MperGalNGcresultedinasig-
andincubatedfor3hat37°C.Thereafter,thesamplewasdia- nificantshiftofthesignal(Fig.1A).Increasingtheunmodified
lyzed using 500 molecular weight cutoff tubing (Spectrapore, GalNGcconcentrationinthemediumto10mMalsoresultedin
cellulose acetate) against 4 liters of water at 4°C overnight. significant synthesis of cell-surface Neu5Gc, confirming that
Water was changed once, and dialysis was continued for peracetylationsignificantlyenhancesuptakeandutilizationof
another10h.Samplewasrecoveredfromtubing,lyophilized, sugars (Fig. 1A) (57). This finding was further confirmed by
anddissolvedin1mlofmethanol.Samplewasloadedontoa analyzing acid-treated total cell lysates by DMB-HPLC, in
1 ml DEAE-A25 column, which had been pre-equilibrated whichNeu5Gcwaschemicallydetected(supplementalFig.S1).
three times with 1 ml of methanol. Neutral glycolipids were PossibleMammalianCellularPathwayfortheConversionof
elutedfivetimeswith1mlofmethanol;monosialylgangliosides GalNGcintoNeu5Gc—Basedonthewellstudiedmammalian
wereelutedfivetimeswith1mlof10mMammoniumacetatein pathwaysofN-acetylhexosamines,weproposeaputativemet-
methanol,anddisialylgangliosideswereelutedwithelutedfive abolic pathway for enzymatic conversion of GalNGc into
timeswith1mlof200mMammoniumacetateinmethanol.An Neu5Gc(Fig.2)includingthefollowingsevensteps.1)Phos-
equalvolumeofwaterwasaddedtoeachoftheeluates;samples phorylation of supplemented GalNGc by GalNAc kinase
werepurifiedusingaC-18columnasdescribed(48),andC-18 (GALK2, EC 2.7.1.157) resulted in the formation of GalNGc-
eluatesweredrieddownwithnitrogengas. 1-P(58).2)SubsequentconversionofGalNGc-1-PintoUDP-
MALDI Analysis—Permethylated or native N-/O-glycan GalNGc potentially catalyzed by UDP-N-acetylglucosamine
samplesweredriedtocompletionbeforedissolvinginmetha- pyrophosphorylase (AGX2, EC 2.7.7.23) (59); followed by
nol(55).Theglycansamplewasthenmixed1:1(v/v)withsuper 3) epimerization into UDP-GlcNGc by action of the
DHBmatrix(Sigma;preparedbymixinga1:5molarmixtureof UDP-GlcNAc4-epimerase(EC5.1.3.7)(60).4/5)Next,bifunc-
5-methylsalicylicacidand2,5-dihydroxybenzoicacidinabso- tional UDP-N-acetylglucosamine 2-epimerase/N-acylman-
lute methanol) and spotted on a stainless steel MALDI plate. nosaminekinase(GNE,EC5.1.3.14andEC2.7.1.60)mightcat-
The MALDI was run on an Applied Biosystem Q-StarXL alyze conversion of UDP-GlcNGc to ManNGc-6-P via
instrument in positive mode. The collision-induced dissocia- ManNGc(61).6)AldolcondensationofManNGc-6-Pwaswith
tionwasdoneusingargonascollisiongas.Thegangliosidesam- phosphoenolpyruvate, which may be catalyzed by Neu5Ac
plesweredissolvedin2:1chloroform/methanolandmixedwith 9-phosphatesynthase(EC2.5.1.57)(62).7)Thefinalstepwould
0.5 M trihydroxyacetophenone in a 1:1 (v/v) ratio. The mass involvedephosphorylationofNeu5Gc9-phosphatetoresultin
spectraldataofthegangliosidesampleswereacquiredinboth Neu5Gc, potentially catalyzed by Neu5Ac 9-phosphate phos-
28902 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER34•AUGUST17,2012
MammalianPathwaysforN-Glycolylhexosamines
FIGURE1.MammaliancellscanincorporateGalNGcfromthemedium,convertitintoNeu5Gc,andexpressitonthecellsurface.A,humanTHP-Icells
werecultivatedunderNeu5Gc-freeconditionsusing5%humanserum.Thereafter,thegrowthmediumwassupplementedwitheither5mMNeu5Gc(positive
control;shadeddarkgray),100(cid:3)MGalNGc(lightblueline),10mMGalNGc(magentaline),or100(cid:3)Mper-O-acetylatedGalNGc(perGalNGc,darkblueline).In
parallel,cellswerekeptinculturewithoutsupplementation(negativecontrol;blackline).After3days,cellswereharvestedandanalyzedbyflowcytometry
usingpolyclonalchickenanti-Neu5GcIgY(27)forsensitivedetectionofcell-surfaceglycosidicallyboundNeu5Gc.Asanadditionalnegativecontrol,cellsfed
5mMNeu5GcwerealsostainedwithcontrolchickenIgYantibody(controlIgY;shadedlightgray)toverifytheabsenceofNeu5GconhumanTHP-Icells.Also,
EMeg32(cid:3)/(cid:4)(B)andEMeg32(cid:4)/(cid:4)cells(C)wereanalyzedfortheirabilitytoconvertsupplementedGalNGcintoNeu5GcandsubsequentlyexpressNeu5Gconthe
cell-surfacebyflowcytometry.
phatase (EC 3.1.3.29) (63). A recent publication revealed that Incontrast,supplementationwith100(cid:3)MperGalNGc(Fig.1,B
metabolic cross-talk allowed exploitation of the GalNAc sal- andC,darkblueline)or10mMGalNGc(magentaline)resulted
vagepathwaytotagandidentifyO-GlcNAcylatedproteins(64). in a significant shift of the signal in both EMeg(cid:3)/(cid:4) and
This study clearly demonstrated that synthetic GalNAz (N- EMeg(cid:4)/(cid:4) cells, into the range of the positive control (5 mM
azidoacetylgalactosamine)isconvertedintoUDP-GalNAz,fol- Neu5Gc).Interestingly,aslittleas100(cid:3)MGalNGc(turquoise
lowedbyepimerizationtoUDP-GlcNAz(64).Giventhesimilar line)wassufficienttoachieveNeu5Gcdenovobiosynthesisin
size of GalNAz and GalNGc, these data suggested that the EMeg(cid:4)/(cid:4) cells (Fig. 1C), whereas EMeg32(cid:3)/(cid:4) cells did not
N-glycolyl group might be tolerated as well by the first three revealdetectableamountsofnewlysynthesizedNeu5Gcunder
stepsofthepredictedpathwayfromGalNGctowardNeu5Gc. theseconditions(Fig.1B).Thedifferencesbetweenthesetwo
To further study this putative pathway, we next analyzed celllinessuggestthatthepathwayofconversionfromGalNGc
Emeg32(cid:4)/(cid:4)mouseembryonicfibroblasts,whichlacktheglu- toNeu5GcisaffectedbyendogenousUDP-GlcNAclevels.This
cosamine-6-phosphate acetyltransferase (emeg32) activity (5, resultfurtherconfirmsthatinadditiontohumanTHP-Icells,
43).BecauseofamarkedlyreducedUDP-GlcNAcpool,thesug- mousecellsarealsocapableofincorporatingGalNGcandcon-
gestedpathwayforconversionofGalNGcintoNeu5Gcmight vertingitintoNeu5Gc.Apotentiallyvaluablecelllinetofurther
beaffectedaswell.EMeg32(cid:3)/(cid:4)cells,whichwerefoundtohave investigate the putative pathway would be CHO-K1-derived
normalUDP-GlcNAclevels,andmutantEMeg32(cid:4)/(cid:4)cellswere mutant cell line ldl-D (47, 65), which is deficient in
cultured under Neu5Gc-free conditions using human serum, UDP-GlcNAc 4-epimerase activity (step 3 of the predicted
and the medium was supplemented with GalNGc or per- pathway, Fig. 2), and therefore should not be able to convert
GalNGcfor3days.Loadedcellswereanalyzedbyflowcytom- GalNGcintoNeu5Gc.Unfortunately,thesehamstercellswere
etryusinganti-Neu5GcIgYasdescribedabove(1,27).Despite foundtoalreadyhavehighendogenousNeu5Gclevelsdueto
the presence of an intact Cmah gene in murine cells, Cmah CMAHexpression,asthenon-fedcellsandthepositivecontrol
expression and base-line endogenous Neu5Gc levels were (5 mM Neu5Gc) have almost comparable levels of Neu5Gc
foundtobenegligible,asdemonstratedbytheoverlyingflow (supplementalFig.S2).
cytometry signals of non-fed cells with the negative control ConversionofGalNGcintoNeu5GcIsGNE-dependent—An-
cells(controlIgY-stainedNeu5Gc-fedcells,seeFig.1,BandC). other key enzyme of the predicted pathway would be the
AUGUST17,2012•VOLUME287•NUMBER34 JOURNALOFBIOLOGICALCHEMISTRY 28903
MammalianPathwaysforN-Glycolylhexosamines
FIGURE2.ConceivablemechanismforconversionofGalNGcintoNeu5Gc.BasedonthewellstudiedmammalianpathwaysofN-acetylhexosamines,the
depictedmetabolicpathwayforsuccessiveenzymaticconversionofGalNGcintoNeu5Gc,includingthefollowing7steps,isproposed:1)phosphorylationof
supplementedGalNGcbyGalNAckinase(GALK2,EC2.7.1.157)resultingintheformationofGalNAc-1-P(58);2)subsequentconversionofGalNAc-1-Pinto
UDP-GalNAcpotentiallycatalyzedbyUDP-N-acetylglucosaminepyrophosphorylase(AGX2,EC2.7.7.23,(59))followedby3)epimerizationintoUDP-GlcNAcby
actionoftheUDP-GlcNAc4-epimerase(EC5.1.3.7,(60)).4/5)Next,bifunctionalUDP-N-acetylglucosamine2-epimerase/N-acylmannosaminekinase(GNE,EC
5.1.3.14andEC2.7.1.60)wouldcatalyzeconversionofUDP-GlcNActoManNGc-6-PviaManNGc(61).6)AldolcondensationofManNGc-6-Pwithphospho-
enolpyruvate,whichmightbecatalyzedbyNeu5Ac9-phosphatesynthase(EC2.5.1.57(62)).7)FinalstepwouldinvolvedephosphorylationofNeu5Gc
9-phosphatetoresultNeu5Gc,potentiallycatalyzedbyNeu5Ac9-phosphatephosphatase(EC3.1.3.29(63)).
bifunctional GNE (UDP-GlcNAc 2-epimerase/ManNAc asdistinctlyasthepurestandards(Fig.4),therewerestrikingly
kinase; steps 4/5, Fig. 2) that acts in an irreversible manner. differentelutionpatternsincellsincubatedinthepresenceof
Therefore,wepredictedthathumanB-celllymphomacellline [3H]GalNGc or [3H]GalNAc for the two subclones of BJA-B
BJAB-K20, which is devoid of UDP-GlcNAc 2(cid:2)-epimerase cells.Themajorityofrecoveredcountsfrom[3H]GalNGc-fed
activity(46)andlacksendogenousNeu5Gc,wouldbeavaluable cells co-eluted with UDP-GalNGc/UDP-GlcNGc standards,
tooltostudythepredictedpathway.Inparalleltothemutant whereas the majority of counts aligned with UDP-GalNAc/
BJAB-K20 cells, hypersialylated BJAB-K88 cells were also UDP-GlcNAcinsamplesfromcellsincubatedinthepresence
grownunderNeu5Gc-freeconditionstoserveasapositivecon- of[3H]GalNAc(Fig.4B).Thisfindingshowedthathumancells
trol.ThemediaweresupplementedwithGalNGc,perGalNGc, are capable of incorporating GalNGc and converting it into
orNeu5Gcfor3days,andloadedcellswereanalyzedforcell- UDP-GalNGcaswellasUDP-GlcNGc.Thisobservationiscon-
surfaceNeu5Gcbyflowcytometryusinganti-Neu5GcIgYas sistent with separate observations made in the preceding
describedabove.ControlBJAB-K88cellswerefoundtosynthe- accompanying papers, wherein we showed that GlcNGc is a
size cell-surface Neu5Gc after incubation with either 100 (cid:3)M metabolite of the Neu5Gc-degrading pathway (1) and that
perGalNGcor10mMGalNGc(Fig.3A),whereasnosignificant exogenouslyaddedGlcNGccanbeactivatedtoUDP-GlcNGc
amountofNeu5GcwasdetectableintheGNE-deficientBJAB- andUDP-GalNGctoserveasasubstratetoO-GlcNActrans-
K20cells(Fig.3B).ThisresultwasconfirmedbydirectDMB- ferase(5).
HPLCanalysisofsialicacids(supplementalFig.S1).Thus,con- CellsIncorporateUDP-GalNGcintoGAGs—Wenextasked
version of exogenously supplied GalNGc to cell-surface ifGalNGccouldbeincorporatedintovariouscellularglycocon-
Neu5GcisGNE-dependent,whichrendersthepredictedpath- jugatesbysupplementingthemediaofwildtypeCHO-K1and
wayverylikely. ldl-D cells with [3H]GalNGc and control [3H]GalNAc. ldl-D
UDP-GalNGc and UDP-GlcNGc Are Formed in Cells—To cells are deficient in UDP-GlcNAc 4-epimerase activity, and
furthersubstantiatetheputativepathway,wenextanalyzedthe therefore, they should only synthesize [3H]UDP-GalNGc but
occurrenceofcytosolicUDP-GalNGcandUDP-GlcNGc(steps not [3H]UDP-GlcNGc from [3H]GalNGc, whereas wild type
2and3ofthepredictedpathway,Fig.2).Therefore,amethod CHO-K1cellscanpotentiallysynthesizebothactivatedsugars.
wasdevelopedtoseparatecytosolicUDP-sugarsbyHPLCusing As only chondroitin sulfate (CS) and dermatan sulfate (DS)
a PA-1 column under alkaline conditions (supplemental Fig. polymers contain GalNAc residues among GAGs, only ldl-D
S3).Asstandards,UDP-GalNAc,UDP-GlcNAc,UDP-GalNGc, cellsshouldincorporateradioactivityintoCS/DS,whereaswild
andUDP-GlcNGcseparatedwellundertheseconditions(Fig. typeCHO-K1cellsmightincorporatecountsintobothCS/DS
4A).TotracktheformationofUDP-sugars,theculturemedia andHS.WildtypeCHO-K1andmutantldl-Dcellswerethere-
ofBJABcellsweresupplementedwith[3H]GalNGcor[3H]Gal- forecultivatedunderNeu5Gc-freeconditions,andtheculture
NAcfor3days.Harvestedcellswerelysedfollowedbyprotein mediumwassupplementedwith[3H]GalNGcor[3H]GalNAc
precipitation, and the supernatants were analyzed by HPLC. for3days.Thereafter,cellswereharvestedandlysed,andGAGs
AlthoughUDP-sugarsinthecrudecelllysatedidnotseparate were purified using DEAE-Sepharose chromatography after
28904 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER34•AUGUST17,2012
MammalianPathwaysforN-Glycolylhexosamines
FIGURE4.MammaliancellsincorporateexogenousGalNGcintoUDP-
GalNGcandconvertittoUDP-GlcNGc.A,UDP-GalNAc,UDP-GlcNAc,UDP-
GalNGc,andUDP-GlcNGcwereanalyzedbyHPLCunderalkalineconditions
using a PA-1 column with an optimized gradient (supplemental Fig. S3).
B,humanBJA-BK20(GNE-deficient)andBJA-BK88(hypersialylated)cells
wereculturedunderNeu5Gc-freeconditions.Themediaweresupplemented
with[3H]GalNGcor[3H]GalNAcfor3days.Cellswereharvested,washedwell,
lysed,andsubjectedto70%ethanolfollowedbyprecipitationat(cid:4)20°Cover-
night.ThesupernatantswereanalyzedbyHPLCasdescribedinA.Fractions
(0.5 ml) were collected, and radioactivity was determined by scintillation
counting.
into CS/DS, likely in the form of GalNAc or GalNGc. About
90%oftheradioactivityfromwildtypeCHO-K1cellslabeled
with [3H]GalNGc was sensitive to chondroitinase ABC treat-
ment(relativeelutiontime(cid:1)0.9),whereasaportionelutedas
FIGURE 3. Conversion from GalNGc into Neu5Gc is GNE-dependent. intact chains (relative elution time (cid:1) 0.6–0.7) (gray symbols
HumanBJA-BK20(A,GNE-deficient)andBJA-BK88(B,hypersialylated)cells
werecultivatedunderNeu5Gc-freeconditions.Thereafter,thefeedingmedia andline;Fig.5,CandD).Whencellswerelabeledwith[3H]-
weresupplementedwitheither5mMNeu5Gc(positivecontrol;shadeddark GalNAc,alargeportionofmaterialwasresistanttochondroiti-
gray),100(cid:3)MGalNGc(lightblueline),10mMGalNGc(magentaline),or100(cid:3)M
naseABC,consistentwithitsidentificationasheparansulfate.
peracetylatedGalNGc(perGalNGc,darkblueline).Inparallel,cellswerekeptin
culturewithoutfeeding(negativecontrol;blackline).After3days,cellswere Thesefindingssuggestthatsome[3H]GalGcwasincorporated
harvestedandanalyzedbyflowcytometryusinganti-Neu5GcIgYfordetec- intoheparansulfate,whichweconfirmedbymassspectrome-
tionofcell-surfaceglycosidicallyboundNeu5Gc.Asanadditionalnegative
control,cellsfed5mMNeu5GcwerealsostainedwithcontrolIgYantibody try(seebelow).
(shadedlightgray)toverifytheabsenceofNeu5Gconthesehumancells. Exogenous GalNAc Derivatives Incorporate into Cellular
ChondroitinSulfateandDermatanSulfatePolymers—Tofur-
proteolysis(48).DesaltedGAGswerethentreatedwithchon- thersubstantiateourfindingthatexogenouslyaddedGalNGcis
droitinaseABCtospecificallydepolymerizeCS/DSintodisac- incorporated into cellular CS/DS, samples were analyzed by
charides,whichwereanalyzedbygelfiltrationandscintillation mass spectrometry. Wild type CHO-K1 and ldl-D cells were
counting.BothwildtypeCHO-K1andldl-Dcellsappearedto culturedfor3daysinthepresenceofeither10mMGalNAcor
incorporatetheradiolabeledsubstratesintoGAGsasseenby 10 mM GalNGc. No significant growth difference of the cell
theradioactivityelutingatarelativeelutiontimeof(cid:6)0.5–0.8 lineswasobservedinthepresenceofthesecompoundscom-
(blacksymbolsandline)forsamplesthatwerenottreatedwith pared with non-fed cells (data not shown). GAGs of cultures
chondroitinaseABC(Fig.5).Treatmentofsamplesfromldl-D supplemented with GalNGc and control cultures incubated
cells with chondroitinase ABC caused a significant shift of withGalNAcwereisolatedanddepolymerizedbychondroitin-
countselutedatarelativeelutiontimeof(cid:6)0.9,wheredisaccha- ase ABC, and released disaccharides were analyzed by glycan
rideselute(graysymbolsandline;Fig.5,AandB).Thisfinding reductive isotope labeling-liquid chromatography mass spec-
indicatesthatallradioactivityofthissamplewasincorporated trometry(GRIL-LC/MS)(49).Toidentifytheresultingdisac-
AUGUST17,2012•VOLUME287•NUMBER34 JOURNALOFBIOLOGICALCHEMISTRY 28905
MammalianPathwaysforN-Glycolylhexosamines
FIGURE5.MammaliancellsincorporateexogenousGalNGcintoGAGs.WildtypeCHO-K1andldl-DcellswereculturedinNeu5Gc-freemedium.Themedia
werethensupplementedwith[3H]GalNGcor[3H]GalNAcforthelast3daysbeforecellsreachedconfluence.GAGswereisolatedbyDEAEchromatographyand
treatedwithchondroitinaseABC(graysymbolsandline)orbuffer(blacksymbolsandline).ThedisaccharideswereseparatedfromtheintactchainsbyCL-6Bgel
filtrationchromatography.Relativeelutiontimesofallsampleswerecalculatedbysettingtheelutiontimeofphenolredto1.
charide residues, differentially stable isotope-labeled internal CS/DS disaccharides may result from incorporation of
standardsofN-acetylatedspecieswereaddedtoeachsample, GalNAc-containingserumcomponents.Together,thesefind-
andpeakswereassignedbasedonco-elution(49).Insamples ings show for the first time that exogenously added GalNAc
from wild type CHO-K1 cells cultured in the presence of derivativescanbesuccessfullyusedtoincorporateintocellular
GalNAc,threeprimarydisaccharideswereobserved,whichall GAGsviatheGalNAc-salvagepathway.Significantincorpora-
harboranN-acetylgroupanddifferonlyinthepresenceand/or tion of exogenous GalNGc into cellular CS/DS was even
locationofsulfategroups(Fig.6A).Incontrast,CS/DSdisac- observed in wild-type cells despite competition with UDP-
charides derived from GalNGc-fed cells give rise to two GalNAc derived from the metabolic flux of the cellular
additionaldisaccharideswithm/zvaluesconsistentwith(cid:7)UA- UDP-GlcNAcpool(Fig.6A).
GalNGc (D0q0) and (cid:7)UA-GalNGc4S (D0q4). Further struc- Exogenous GalNGc Incorporates as GlcNGc into Cellular
turalanalysisofthemostpredominantspecieswascarriedout HeparanSulfatePolymers—Asdescribedabove,onlyasubfrac-
bytandemmassspectrometry.Theproductionsderivedfrom tion of radiolabeled GAGs isolated from CHO-K1 cells fed
fragmentationofD0q4(Fig.6D)weremostsimilartotheprod- [3H]GalNGcorcontrol[3H]GalNAcwassusceptibletochon-
uct ions of (cid:7)GalNAc4S (D0a4, Fig. 6C), suggesting that it is droitinaseABCtreatment,suggestingthatthechondroitinase-
structurallysimilar,consistentwithitsassignmentasD0q4and resistantradioactivityhadbeenincorporatedintoGAGsother
notanotherisobaricspecies.Bothspectrashowaproduction thenCS/DS(i.e.as[3H]GlcNGcor[3H]GlcNAcintoheparan
with identical m/z value consistent with the B ion from the sulfate). To verify this conclusion, GAGs were isolated from
1
nonreducingend(Fig.6,CandD,seeinsets).Alloftheremain- wild type CHO and ldl-D cells fed either 10 mM GalNAc or
ing product ions found in D0a4 have corresponding product GalNGc as described above. Purified GAGs were incubated
ionsinD0q4thatdifferby16massunitsasexpectedforspecies withheparinlyasestospecificallydepolymerizeheparansulfate
thatdifferonlyinthesubstitutionofanN-glycolylgroupforan (HS)polymersfollowedbymassspectrometryanalysistodeter-
N-acetyl group. Compared with wild type CHO-K1 samples, minethestructureoftheresultingHSdisaccharides.Compo-
CS/DS disaccharides from control GalNAc-fed ldl-D cells sitionalanalysisbyGRIL-LC/MSwasperformedusingdiffer-
revealedasimilarextractedioncurrentchromatograph(Fig.6, entially stable isotope-labeled standards of the N-acetylated
AandE).Incontrast,ldl-DcellsfedGalNGcshouldnotcom- speciesasinternalstandards.ThreeHSdisaccharidesfromwild
poseanyGalNAc-containingCS/DSdisaccharidesduetothe type cells supplemented with 10 mM GalNAc were detected
lackofUDP-GlcNAc4-epimeraseactivity.Asexpected,thetwo whichdifferedonlyinthepresenceand/orpositioningofsul-
CS/DS disaccharides harboring an N-glycolyl group, which fategroups(Fig.7A).Interestingly,HSdisaccharidesfromwild
weredetectedinsmallamountsinsamplesfromwildtypecells typecellssupplementedwithGalNGccontainedtwoadditional
fedGalNGc(Fig.6B),becamethepredominantspeciesfoundin species, suggesting that UDP-GlcNGc had been synthesized
CS/DSdisaccharidesfromthecorrespondingldl-Dcellsample from exogenous GalNGc and was incorporated into cellular
(Fig.6F).Asthecelllineswereculturedinthepresenceof5% HS. The identity of all peaks was confirmed by tandem mass
human serum at all times, low levels of GalNAc-containing spectrometryusingCIDasdescribedabove.Thefragmentation
28906 JOURNALOFBIOLOGICALCHEMISTRY VOLUME287•NUMBER34•AUGUST17,2012
MammalianPathwaysforN-Glycolylhexosamines
FIGURE6.DisaccharideanalysisofchondroitinsulfateextractedfromwildtypeandmutantCHOcellsafterfeedingGalNGcorGalNAc.Wildtype
CHO-K1andldl-DcellswerekeptunderNeu5Gc-freeconditionsandsupplementedwith10mMGalNGcor10mMGalNAcforthelast3daysbeforecellsreach
confluence.GAGswerepurifiedfromtheculturemediausingDEAEchromatography,andchondroitinsulfatewasenzymaticallydepolymerizedandsubjected
tocompositionaldisaccharideanalysisbyLC/MS.TheextractedioncurrentforinternaldisaccharidecontainingbothGalNGcandGalNAcisshownforeachcell
lineandfeedingcondition.ThedisaccharidestructurecodewasadaptedtoincludelowercaseqtodesignateN-glycolylwiththerestofthecodeleftas
described(50).PeakassignmentsforN-acetylatedspeciesweremadebasedonco-elutionandmassanalysisofdifferentiallystableisotope-labeledstandards
addedtoeachsample.A,extractedioncurrentchromatographforCHO-K1cellsfedGalNAc.B,extractedioncurrentchromatographforwildtypeCHO-K1cells
fedGalNGc.ThepeakmarkedwithanasteriskisisobaricwithD0q4.CandD,tandemmassspectraofD0a4andD0q4inchondroitinsulfatefromGalNGc-fed
CHOcells.ThedaughterionprofilesofbothD0a4inGalNAcfedldl-Dcells(C)andtheputativeD0q4detectedinGalNGcfedldl-Dcells(D)werecompared.The
overallCIDspectrumoftheputativeD0q4ionismostsimilartothatofD0a4(49).Bothspectrashowaproductionwithidenticalm/zvalueconsistentwiththe
B ionfromthenonreducingend(seeinsets).AlloftheremainingproductionsfoundinD0a4havecorrespondingproductionsinD0q4thatdifferby16mass
1
unitsexpectedforspeciesthatdifferonlyinthesubstitutionofanN-glycolylgroupforanN-acetylgroup.E,extractedioncurrentchromatographforldl-Dcells
fedGalNAc.F,extractedioncurrentchromatographforldl-DcellsfedGalNGc.ThepeakmarkedwithanasteriskisisobaricwithD0q4.
AUGUST17,2012•VOLUME287•NUMBER34 JOURNALOFBIOLOGICALCHEMISTRY 28907