Conformational differences between linear α(2➝8)-linked homosialooligosaccharides and the epitope of the group B meningococcal polysaccharide

Publisher’s version / Version de l'éditeur:

Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.

Questions? Contact the NRC Publications Archive team at

PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.

https://publications-cnrc.canada.ca/fra/droits

L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.

Biochemistry, 26, 25, pp. 8399-8405, 1987-12-15

READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.

https://nrc-publications.canada.ca/eng/copyright

NRC Publications Archive Record / Notice des Archives des publications du CNRC :

https://nrc-publications.canada.ca/eng/view/object/?id=6141d6e8-9443-4020-a147-1cdaf926b5fe

https://publications-cnrc.canada.ca/fra/voir/objet/?id=6141d6e8-9443-4020-a147-1cdaf926b5fe

NRC Publications Archive

Archives des publications du CNRC

This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.

For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.

https://doi.org/10.1021/bi00399a055

Access and use of this website and the material on it are subject to the Terms and Conditions set forth at

Conformational differences between linear α(2

➝8)-linked

homosialooligosaccharides and the epitope of the group B

meningococcal polysaccharide

Conformational

Differences

between

Linear

_{a(2—*8)-Linked}

Homosialooligosaccharides

and

the Epitope

of

the Group

B

Meningococcal

Polysaccharide*

Francis Michon, Jean-Robert Brisson, and Harold J. _Jennings*

Division

_of

_BiologicalSciences, NationalResearch Council

_of

Canada, Ottawa, CanadaK1A 0R6 Received

_April

29, 1987; RevisedManuscriptReceivedJune22, 1987

abstract: The _{a-(2-*-8)-linked sialicacid oligosaccharides}

_(NeuAc)„

exhibitan unusual degree

of

het-erogeneity in the conformation

of

theirlinkages. Thiswas _{diagnosedby observation in}their 13C

NMR

_spectra

of

an _{equivalent andunique heterogeneity}inthechemicalshifts

of

their_{anomeric carbons and subsequently} confirmed _bymore comprehensive

'H

and 13C

NMR

studies.

In

these studiesboth one-dimensional and two-dimensional experimentswere carried outon the_{trisaccharide (NeuAc)3}andcolominicacid. Inaddition to_{the unambiguousassignment}

of

the signalsin thespectra, theseexperiments demonstratedthatbothlinkages

of

(NeuAc)3 differed in conformation fromeachother andfromtheinner linkages

of

colominicacid. The

NMR

dataindicatethattheseconformationaldifferences extend tobothterminaldisaccharides

of

oligo-saccharides_{larger than (NeuAc)5,} a resultthathas_{considerable physical and biologicalsignificance.}

In

the context

of

_{the group}B_{meningococcalpolysaccharide,}

it

_providesan _explanationfortheconformational

epitope

of

the groupBmeningococcal polysaccharide, whichwas _proposedon theevidencethat(NeuAc)10,

larger than the optimumsize

of

an _{antibody site,} was _{the smallest oligosaccharide able to} bind _{to group}

B_{polysaccharidespecific}antibodies. Because the two terminal disaccharides

of

_(NeuAc)10

differ

in

con-formation to its inner residues, the immunologically functional part

of

(NeuAc)10residesin its inner six residues. This number

of

residues is now consistent

with

the maximum size

of

an _antibody site.

e_{group B meningococcal polysaccharide}isa_homopolymer

of

a-(2—’ 8)-linked sialic acid residues and is _structurally

identical withbothcolominicacid_{(Bhattacharjeeet al., 1975)} and _{the capsular polysaccharide}

of

Escherichia coli K1

(Jennings, 1983). It is _{poorly immunogenic (Wyle} et al., 1972), and although groupB _{meningococcal organisms} are able to producelowlevelsofgroup B_{polysaccharidespecific}

antibodies inanimals and humans,theseantibodiesare almost

exclusively

IgM

andoflow

_affinity

_{(Mandrell & Zollinger,}

1982). The poor immunogenicity of the group B

poly-saccharide is probably attributable to tolerance due to

cross-reactive tissue components, and this hypothesis is

strengthenedby theidentificationofa-(2— 8)-linked oligomers of sialic acid common to both the group B _{meningococcal}

polysaccharide(Jennings et al., 1985) and the glycopeptides ofhumanandanimaltissue(Finneet al., 1983a). GroupB

polysaccharide specific antibodies are _{induced by}a confor-mationally controlledepitope. Thiswas _{unambiguously} con-firmed _by the fact that an _{unusually largea-(2—8)-linked}

sialic acid oligomer (decasaccharide) was _{required either}to

functionas an inhibitor for_{(Jennings et al., 1985)}or tobind

to _(Finne & _Makela, 1985) these antibodies. The glyco-peptidesofhuman andanimalfetal brainalsocontain thesame

epitopeand inconsequence bindto group B_{polysaccharide}

specific antibodies (Finne et al., 1983b; Finne & Makela,

1985). In thispaper theconformational natureofthe epitope associatedwith theformationof_groupB_{meningococcal}

po-lysaccharide specific antibodieshasbeenconfirmed_by high-resolution NMR studies on colominic acid and some ofits

constituent oligosaccharides. These studiesindicate thatat leastfive residuesare _requiredbeforea_linkagewithasimilar

orientationto thatoftheinternal linkages

of

colominicacid

is_generated.

ThisisNationalResearchCouncilofCanadaPublicationNo.28336.

Experimental Procedures

Materials. _{Colominic acid (E. coli)} was obtained from

Sigma Chemical Co., St. Louis, MO, anda _{high molecular}

weight fractionwas usedin theNMR _experiments. Thiswas

obtained by passing colominic acid (Na+ salt) through a Sephadex G-100 columnwithphosphate-buffered saline (PBS) at _pH7.0as eluant and_isolatingthefraction thatelutedat = _{0.1 (20-25 sialic}acid residues). _{Oligosaccharides}of

DP= 2-6_{were obtained by thepartial}

hydrolysisofcolominic acid andfractionationofthe depolymerized fragments in order

of

_ascending molecular sizeon an _{ion-exchange column} as describedby Jennings et al. (1985).

NMR Methods.

'H

and 13CNMR _spectrawere recorded on BrukerAM500andAM200_{spectrometers at 300}K with acetoneas theinternal chemicalshiftreference (2.225 ppm

for

'H

NMR and31.07_ppmfor13C

_NMR).

Internalacetone

isreferenced to_{external tetramethylsilane. Polysaccharides} and oligosaccharideswere _exchangedtwicewith 99.7%_D20 and then run in 0.4 mL of 99.99% _{D20 (5-mm} tubes) at

concentrations

of

50-100_mg/mL.

Homonuclearshift-correlated 2-D_{NMR (COSY)}

experi-ments and homonuclear /-resolved 2-D

NMR

(JRES) ex-perimentswere carried_{out according to}Aueetal. (1976) and

Nagayamaet al. (1980). COSY experimentswith two- and three-steprelayed coherencetransferwere done_accordingto

Wagner (1983) and Bax andDrobny (1985). The

heteronu-clearshift-correlated 2-D NMR experimentwas carriedout

accordingto Bax et al. (1981),andthechortle _experiment

was donewiththe pulsesequenceofPearsonetal.(1985). All phasecyclingswere _accordingtothestandard software pro-vided byBruker_(DISB86),andallexperimentswere donein

the magnitude mode. Nuclear Overhauser enhancements (NOE)were _{obtained by difference experiments}with multiple irradiationofeachlineofa_multiplet(Neuhaus, 1983; Kinns

&_{Saunders, 1984). Transient}NOEvalueswere obtainedwith

8400 BIOCHEMISTRY MICHON ET AL.

Table I: ChemicalShifts"ofSignals inthe 13CNMR _Spectrumof

a-(2—*8)-Linked1 _[(NeuAc)3] residuesofI* carbon C b a 1 174.19 173.10 175.59 2 101.36 102.87 97.35 3 41.22 41.70 40.06 4 69.27 74.48 68.28 5 52.55 53.19 53.19 6 73.46 68.81 71.50 7 68.99 69.99 68.49 8 72.55 79.05 76.50 9 63.44 62.05 61.89 C=0 175.73 175.73 177.76 ch3 23.07" 23.16" 22.88 “In ppm from internal acetone. _{Assignments confirmed by} 2-D

NMR (C,H)COSY experiments. 6

Contiguousresiduesfromthe

re-ducingresiduea. "Tentative_assignments.

Table II: ChemicalShifts"of_Signalsinthe13C NMR Spectrumof a-(2—*8)-Linked (NeuAc)4 residuesof(NeuAc)/ carbon d c b a 1 174.30 173.89 173.15 175.60 2 101.19 101.90 102.81 97.35 3 41.20 40.89 41.61 40.08 4 69.33 68.97 74.56 68.26 5 52.56 53.20 53.20 53.20 6 73.43 74.01 68.97 71.57 7 68.97 70.24" 69.69" 68.61 8 72.55 78.51 79.16 76.73 9 63.44 62.29" 61.88" 61.88 C=0 175.41 175.71 175.71 177.72 ch3 23.08" 23.08" 23.15" 22.88 "In ppm from internalacetone. _6Contiguousresiduesfromthe

re-ducing residuea. "Tentativeassignments.

figure 1: Structureoftrisaccharide1. _{Arrows depict}NOEspertinent totheconformationof1.

a 180° selective_{pulse(Morris & Freeman, 1978). Proton spin}

simulation witha linewidthof0.5 Hz was _performedwith the _{Bruker program}panic.

Results

and Discussion

Both13Cand

'H

NMR spectroscopic studieson colominic

acid andsome ofits component oligosaccharidesconfirmthat the_{interresidue linkage}conformationofthe_terminallylocated

sialic acidresiduesofthis a-(2— 8)Tinked sialicacid

homo-polymer differsfrom thoseofits inner residues. Colominic acid was used in thisNMR _study becauseit was _easily ob-tainableinamolecularsize_{small enough to yield well-resolved}

spectra. However,because_theyare _{structurally identical with}

colominic acid, the capsular polysaccharides of group B

Neisseria meningitidisandE.coliK1 must alsoexhibit this

phenomenon.

Conformationaldifferenceswere first _diagnosedby

obser-vationin the 13CNMR _spectraofaseriesofa-(2— 8)-linked

sialooligosaccharides,(NeuAc)„,in whichn = _{2-6,extensive} heterogeneityparticularlyin the chemical shiftsof their_linkage

carbons _(C2andC8). The chemical shiftsofthe carbonsof thesialicresiduesof_{the oligosaccharides and}colominicacid are listedinTables

I-IV,

thelatter_beinglistedin Table

III.

Assignmentsmadeon the_{trisaccharide (1)} shown _{in Figure} 1 andon colominic acidwere obtainedfrom2-D

_(H,H)

and

Table III: ChemicalShifts"ofSignalsinthe13CNMR _Spectrumof a-(2—8)-Linked_(NeuAc)sandColominic Acid6

carbon residues_{of (NeuNAc)'s‘} colominicacid e d c b a 1 174.32 173.88 173.88 173.07 175.60 174.00 2 101.18 101.90 101.72 102.92 97.35 101.78 3 41.12 40.78 40.78 41.60 40.02 40.76 4 69.23 68.94 68.94 74.46 68.19 69.13 5 52.53 53.15 53.15 53.15 53.15 53.23 6 73.39 74.04 74.04 68.94 71.40 74.01 7 68.94 70.23* 69.80* 69.71* 68.55 69.86 8 72.55 78.56 78.56 79.06 76.43 78.72 9 63.41 62.28 62.05 61.80 61.80 62.06 C=0 175.70 175.70 175.70 175.70 177.75 175.76 ch3 23.10* 23.24* 23.24* 23.10* 22.89 23.30

"In _ppmfrom internalacetone. 6Assignmentsobtainedfromchortle _{experiment. "Contiguous}residuesfromthereducingresiduea. ^Tentative

assignments.

TableIV: Chemical_{Shifts" of Signals in}the l3CNMR Spectrumofa-(2—-8)-Linked (NeuAc)6

carbon residuesof(NeuAc)66 f e d c b a 1 174.31 173.94 173.94 173.94 173.15 175.61 2 101.17 101.85 101.74 101.74 102.80 97.35 3 41.21 40.74 40.74 40.74 41.60 40.08 4 69.20 69.07 69.07 69.07 74.55 68.29 5 52.56 53.21 53.21 53.21 53.21 53.21 6 73.41 74.02 74.02 74.02 69.07 71.58 7 69.07 70.27 69.84 69.84 69.33 68.66 8 72.54 78.57 78.65 78.65 79.22 76.75 9 63.44 62.29 62.05 62.05 61.83 61.83 C=0 175.71 175.71 175.71 175.71 175.71 177.71 ch3 23.10" 23.29" 23.29" 23.29" 23.10" 22.87 “In ppm from internalacetone. 6Contiguousresiduesfromthereducingresiduea. "Tentativeassignments.

0

n*6

figure 2: 13Cresonances oftheanomeric carbons_{of a-(2-*8)-linked}

homosialooligosaccharides (NeuNAc)„wheren = _2-6.

(13C,H)COSYexperiments and the assignments madeon the

other oligosaccharideswere madewith_{oligosaccharide} 1as a model. Whilethe assignments madeon colominicacidare

in substantial agreementwiththose reported byBhattacharjee

etal._{(1975), they}confirmthatthelatter_{assignments made} on _{the closely spaced C4 and C7 signals should have been} reversed.

The_signalsoftheanomeric carbonsofthe oligosaccharides are shown in Figure2 and reveal_{the heterogeneity in} their chemical shifts. Exceptfor_{the signals}oftheanomeric carbons ofthe_{reducing sialic}acid residues, which remainconstant at 97.35 ppmin_{the spectra}ofall_{the oligosaccharides, coincident} signals donotoccur untilthe hexasaccharideisreached, and these signals at 101.74 _ppmalso coincide with that ofthe

anomeric carbon signal

of

colominicacid. A similar_pattern ofchemicalshiftheterogeneityisalso_{exhibited by}theother linkagecarbons at C8on the_{aglyconic sialic}acid residuesof the oligosaccharides (Tables

I-IV).

Fromtheseresults

it

can beinferredthatheterogeneity inlinkage conformation inthe hexasaccharide is _presentin both its terminaldisaccharides

andthatby analogy thisisalsotrueforcolominicacid. The conformational dependenceof13C

NMR

chemicalshifts of

linkagecarbonswas first_{proposed by Colson et al. (1974) and} was laterused_{empirically to}locate_{conformationally}controlled

determinants in complex polysaccharides (Jennings et al., 1984). However, only recentlyhasthis correlationbetween

glycosidic torsion angle(\p)andchemicalshiftbeenvalidated

withHSEAcalculationson aseriesofoligosaccharideshaving

either a-D-glucopyranosylor _{a-D-galactopyranosyl}residues at _{the nonreducing end(Bocket al., 1986).} Our 1-Dand2-D

’H

NMR studieson trisaccharide1 andcolominicacid also

providefurther evidence forthevalidity ofthis correlation.

Inorderto_{compare the differences in the linkage}

confor-mations between _{the oligosaccharides}andcolominic acid, a 1-D and 2-D

‘H

NMR _studywas carriedout on 1 and

colo-minicacid. Trisaccharide1,shownin Figure 1, was chosen as_{the oligosaccharide to study}becauseit_approximatesto the

Table V: ChemicalShifts0ofSignalsinthe*HNMR Spectrumof

a-(2—*-8)-Linked1 _[(NeuAc)3]andColominicAcid residuesof1

proton c b a colominicacid

H-3a 1.729 1.639 1.748 1.737 H-3e 2.757 2.691 2.201 2.673 H-4 3.668 3.547 3.983 3.600 H-5 3.821 3.782 3.863 3.819 H-6 3.615 3.555 3.877 3.626 H-7 3.576 3.839 3.760 3.896 H-8 3.897 4.113 3.996 4.102 H-9 3.884 4.132 3.940 4.188 H-9'‘ 3.633 3.666 3.733 3.665 ch3 2.028 2.064° 2.060° 2.083 “Spin simulated parameters. In ppm from internal acetone. ‘Primarygeminal protonsare distinguishedbymeans ofa prime for

the proton withthe largest vicinalcoupling constant. 0

Assignments

maybereversed.

TableVI: VicinalandGeminal Coupling Constants(Hz) of a-(2-*8)-Linked1 _[(NeuAc)3]°andColominicAcid‘

residuesof1 /h,h c b a colominicacid 3a,3e -12.2 -12.2 -12.9 -12 3a,4 12.0 12.1 11.5 12 3e,4 4.4 4.4 5.0 5 4,5 10.0 10.2 10.3 10 5,6 10.3 10.3 10.3 10 6,7 1.9 1.0 0.5 <3 7,8 9.1 2.3 6.9 <3 8,9 2.6 4.7 2.8 5 8,9' 6.4 6.9 3.9 6 9,9' -12.1 -12.3 -12.3 -12

"Spin simulated parameters. ‘Taken from resolution-enhanced

spectrum.

smallest structure _{incorporating both terminal}disaccharides ofcolominic acid. _{Strictlyspeaking, the tetrasaccharide would} have been a more _precisemodel but _{the analysis}of its *H

NMR _{spectrum would} havebeen _prohibitive.

The

’H

NMR

spectrum

of

thetrisaccharide (1) isshown

in Figure 3 and was

difficult

_{to assign directly} due to its

complexity; therefore,a_{proton homonuclear correlated 2-D} NMR

_[(H,H)

_{COSY] experiment}withtwo- and_three-step

relaycoherencetransferwas _performed,and the_spectrumis

alsoshown_{in Figure}3. Asaresultofthis_experiment,it was possible to assignallthe protons ofeach sialic acid residue

in 1. _{Improved resolution}ofthe

'H

NMR _spectrumof1 was alsoobtained_{when signals}were further _{separated along the} / axis by two-dimensional /-resolved spectroscopy. Mea-surement ofchemical shiftsandcouplingconstants fromthe contour _plot of the two-dimensional _{/-resolved spectrum} permitted rapidand unambiguousassignmentof individual

protons within the unresolved multiplets observed in the one-dimensional spectrum. The chemicalshifts and_coupling constants ofthe_{protons in} 1 are listedin Tables Vand

VI,

respectively. Additionalconfirmation ofthe assignmentswas also obtained _by a _{proton simulation experiment.} In this

experiment the coupling constants and chemical shifts obtained from the _{/-resolved spectrum} were fed into a _program, a nine-spin systemforeachsialicacid residueof1. _{The signals} associatedwitheachresidueare shown_{in Figure}4,andthe

spectrum obtained by the addition of all these signals is

identicalwiththatoftheresolution-enhanced one-dimensional

'H

NMR spectrumof 1.

Having unambiguouslyassignedalltheprotonresonances, NOE_experimentscouldthen becarriedout. SelectedNOE difference spectrafor 1are shownin_Figure5,andthe relative

8402 BIOCHEMISTRY MICHON ET AL.

TableVII: Nuclear Overhauser Enhancementsfor a-(2—8)-Linked1 _[(NeuAc)3]

observedNOE_(negative)

saturatedsignal0 H3a H3a h4 h5 h6 H, Hg H, H<y

H3e(a) (a) 12 (a)4

H3e(c) (c) 11 _(c) 6 (c)2

H3e(b) (b) 13 (b)9

H3a(c) (c) 14 (c)7 _(b) 1

H3a(a) (a) 10 _(a)2

H3a(b) (b) 17 _(b)4 _(b) 10

Hg(b) (c) 1 (b) 4 (b) 5 (b) 17 (b) +

IHs(b) (c) 2 (b) 5 (b) 5

»H,(b) (a)4 (b) 6

“aisthereducingresidue, bisthemiddleresidue, andcisthe nonreducingresidueof1.

Acetone

,A>L_

withthree-steprelay,of1 in D20₍₃₁₀K)withthe 1-D spectrum

above. Theinitial(r3,t2)matrix of256 X2048pointswas zero-filled to1024X 2048 points andprocessedwithunshiftedsinebell window functions, a _{magnitude calculation,}and_{symmetrization} about the diagonal. Thefinal resolution in bothdimensionswas 1 Hz/point.

NOEsobtainedfromthesespectraare listedin TableVII. An

examinationofthe_couplingconstantsandNOEsof_the pro-tons associatedwiththetwo differentlinkagesin 1indicates

a_{striking difference in conformationbetween these linkages.} Thatconformational differencesare confinedto the_linkage regionsof1is_{confirmed by the}fact that_{the coupling constants}

andNOEs ofthe_{ring protons}of1 indicatethatthe

confor-mationsoftheindividual sialic acid rings remain essentially inthesame conformation_{(5C2) as}describedfortheir_respective monomericunits (Joachims et al., 1967). Thisisalsotruefor the exocyclic chain oftheterminal residue(c)of1,the

pre-ferredconformationofwhichissimilarto that_{proposed by} others_(Brownet al., 1975; Sabesanet al., 1983;Lindonet al., 1984).

Despite the factthat thelinkageregions between the

res-iduesof1are _{complex, both}are _composedoffourbondswith

potential for unrestricted rotational freedom, the coupling

constants_(Table

_VI)

andNOEdata(Table

VII)

indicatethat

thelinkagecarbons adopt arather_preciseconformation. For

.jaIUwxJi—ll_,11

JiLii

a)

a. I 4.0 5.9 58 3.’ 3.C

figure 4: _Comparisonofthe_{simulated and experimental spectrum}

of1. Simulationforeachindividual_{residue (a,}b,and c)isshown

in (a), (b),and(c)andtheirsum in (d). The resolution-enhanced

experimentalspectrumisshownin (e).

the exocyclic chainofresidue bof1,a_strongNOEbetween

Hjjand_{Hjj (Figure5) and the small}vicinal_{coupling constants} (Table

VI)

associatedwith _{these protons} indicatethat both

sets ofrotamers about the_{C8-C7 bond and the C7-C6 bond}

are confined toa _preferredrotamer with<t>(H8,C8, C7, H7)

= +60° forthe C8-C7 bond and

<j>(H7, C7, C6, H6) = -60°

forthe C7-C6 bond. In addition the orientationofthe

non-reducing terminal sialicacid residue (c)in relation toresidue b can be determined from the NOE between _{H|a and H8,}

which_{puts these two protons in}close _proximity.

Theconformationofthelinkage between residuesbanda of1 _{(Figure 1) iscompletely different}to the_linkagebetween residuescandb. Residuea,whichisin_{its preferred}/3-dform

as determined_by

'H

and 13C chemicalshift data (Tables I and_{V) (Brownet al.,} 1975;Jennings&Bhattacharjee, 1977),

is bent back toward the middle residue (b). This can be

deduced _{by the} enhancement of _{the H7 signal when H9} is

irradiated, indicatingthe_proximityofthese_{two protons.} The couplingconstantsare alsoin_agreementwitha _{large change}

in the_{linkage conformation}between residues b anda. Inthis case thecouplingconstant betweenHjjand H7isalsosmall, indicating that they are in _{gauche orientation, while} the

couplingconstant betweenH7CandH|cis_{large, showing}that

these two _protons are trans to eachother.

The chemical shiftsand_couplingconstants associatedwith

the protonsofthesialic_{acid repeating}unitofcolominicacid were derivedfromits resolution-enhanced

'H

NMR _spectrum andare alsolistedinTablesVandVI. Similarvaluesforthese

parameterswere also_reportedforsome ofthese_protonsby

TableVIII: Nuclear Overhauser EnhancementsforColominicAcid

saturated signal

observedNOE_(negative)0

H3a H3e h4 h6 Hy h5 h7 h8 h9 H3a 32 9 7 1 4 H3e 32 15 3 1 2 h5 5 2 8 1 h7 <1 8 8 8 3 h8 3 1 9 6 9 14 h9, 4 17 23 0_{InterresidueNOE}

intensityvaluesare underlined.

2LA

(iv)

figure 5: NOE_{difference spectra}of1. Resonancesofresidueaare notunderlined,thoseofresidue bare underlinedonce,and thoseof

residuec are underlined twice. The spectra represent theirradiation

of(i)H3aofb, (ii) H3aofaand_H3aofc,(iii) H3eofb,(iv)H3eof c, (v)H8and H9ofb,and(vi)H8 (mostly) ofb. _{The spectrum}of 1 isshown_{in (vii).}

directlyon colominicacid. SelectedNOEdifference_spectra

are shown_{in Figure}6, andtherelativeNOEs obtained from

these_spectraare listedin Table

VIII.

Fromthisinformation

it

can be determined that the conformation of the _linkage betweentwoconsecutivesialic acid_{residues (a and b}of2 as shown _{in Figure 7)} located _internally in colominic acid is

different from thatofboth linkagesin1. _{Althoughthe linkage}

conformation

of

2ismore similartothatbetween residuesc and b

of

1than thatbetween residues banda of1,the

dif-ferencesin the formerare still_{strikingand show, despite the}

possibilityofaveraging,aconsiderabledifference in the ori-entationofthesialic acid residues. That theorientationof residuesa and bof2 isdifferenttothatofresiduescand b of1can bedeterminedfromtheNOEdataof2_(Table

_VIII)

wherean enhancementon _Hgwas not _{onlygeneratedby}

ir-radiatingH3a,as was thecase withthe equivalent signals _(Hb,

H3a) in 1_(Table

_VII),

butalso_{byirradiatingH3e.} Therefore

for _{2, H8 must} be _{situated midway} between, and in close proximityto, both geminalH3 protonsofresiduea. Onthe

basisofcoupling constants(Table

VI)

and_{NOEs (Table}

_VIII)

assigned to Hb, Hb, and Hg, the remainder of the linkage regionof2 appears to besimilarto thatbetween residuesc

7 CH3

' 1 ' 1 1 1 1 1 1

1 1 1 ! 1 1 1 1 1 ' 1 1 1 1 1 '

4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2,2 2.0 1.8 l.S PPM

figure 6: TransientNOE_{difference spectra}ofcolominic acidwith

50-ms delayafterselectiveinversion_{of (i)}theH3aresonance and_(ii)

theH3e resonance. The reference spectrum isshownin (iii),and

interresidueenhancementsare underlined.

2

figure 7: Structureofdisaccharide2located_{internally in}colominic acid. Arrows_{depictNOEs pertinent}totheconformation of2.

and bof1,although minorchangeswouldstillbecompatible

withthe data. TheNOEbetween Hb_{and Hg,} is_{also present}

in 2,andthe couplingconstants between H^b,

H?

and

_H7\

H|bare ofthesame orderof_magnitudeasthosefoundforthe equivalent protonsin1_(Table

_VI).

Aminor difference inthe

NOE_patternbetween 1and2was that irradiationof_H3aand H3a also causeda smallenhancement on _{H7b; however,} this may not be _important to the _{linkage conformation.} CPK modelsof2showthattheclose_{proximity of}both_geminalH3 protonsof2toH7bis_unlikelyandthatthereforetheseNOEs couldbe attributedto three_{spin effects.}

The aboveNMR analyses demonstratethatthe linkagesin

1 are _{conformationally different}toeachother and to theinner linkagesofcolominicacid. Theextent ofthisdifferencecan

be seen _{by examining} CPK modelsof 1 and an _equivalent

internallylocated trisaccharidesequenceofcolominicacid. The modelsare shown_{in Figure}8,and two dominantfeatures

8404 BIOCHEMISTRY

figure 8: _{Space-filling}models(CPK) oftrisaccharide1(left)and

thesame trisaccharidelocated_internallyincolominicacid(right).

readily visualized. First, due to_{the folding} backofthe re-ducing sialic acid residueof

I,

trisaccharide 1ismuchshorter

thantheinternal trisaccharide. Second,the carboxylate groups of1 _{appear to} be_{randomly arranged,} whereas those

of

the

internal trisaccharideare _{essentially linedup}on onesideof the molecule. Thisis_{exemplifiedbycomparing}the_coupling constantsof_{the H9 protons}of1 and colominic acid (Table

VI).

For I the_{vicinal coupling}constants ofthe H9protons

ofeachresidueare all different,whereas thoseofcolominic

acid are all thesame.

Alltheseobservationshave_{importantbiological implications} in_{that the unusual length requirement}of«-(2—*-8)-linked sialic

acid_{oligosaccharides} (dccamer) to bind to _groupB menin-gococcalpolysaccharide specific antibodies (Jenningset al., 1985; Finne

&

Makela, 1985)can be rationalized. The fact thatadecasaccharide istheminimumsize _requiredis

com-pletely consistent with our

NMR

studies because we have shown that the two terminal disaccharides of the deca-saccharidewoulddifferinconformationfromits innerresidues.

Thus, theimmunologically functional moietyofthe decamer

istheinnersix residues. This numberisnow consistentwith the inhibition studies of Kabat and _{Mayer (1961) using} a

dextran-antidextransystemin whichtheinhibition _power

of

dextran-derived _{oligosaccharides} maximized at the hexa-saccharide. Also consistentwiththelackofa_{conformationally}

controlled determinantinthe dextransisthefactthat linear

a- and_{/9-D-(l-*4)-linked}anda- and _{/9-d-(1—6)-linked}

oli-gosaccharidesrelated to dextranexhibitno conformational heterogeneity intheirlinkagesasdetermined_{by the}absence

ofanomericand_linkage l3Cchemicalshift_{heterogeneity} in

their13C NMR _spectra (Bocketal., 1984). It is_interesting

to notethatthe dependenceoftheconformationofthe group

B_{meningococcal polysaccharide}on _{chain length probably}also

explainsa similar unusually_{largeoligomer (eight sialic}acid

residues) requirement

of

a_{bacteriophage endosialidase,}which

cleaves_{colominic acid (Finne} & Makela, 1985).

Thatthe_{groupB polysaccharide contains}a_preferred

con-formationnot foundin_{the smaller oligosaccharides}could be

due_{to cooperative}stabilization, whichis_dependenton chain length. Theforcesinvolved in thisstabilizationare not known with _{certainty but would}bea _{very important factor inany} theoretical calculationsoftheconformationofthedeterminant. Obviously anypotential energycalculationswould requirean oligosaccharideofatleastsix residuesto_{adequately describe}

the _{preferred conformation}

of

colominic acid. _Charge is

MICHON ET AL.

probably involved, butdue to the factthat the isomeric

a-(2—»9)-linkedgroupC meningococcalpolysaccharide forms

aconventionaldeterminant(Jennings et

al„

1985),additional environmental factors mustbe_{important. Perhaps the unique}

dispositionofthecarboxylateandacetamidogroupsin

colo-minic acid, each being lined _up on _{opposite sides} of the molecule _{(Figure 8), could} be a factor in the stabilization

energy. This alignmentofthe carboxylate groups leavesthem

all inclose_proximity to_{9-(hydroxymethyl)}groups,which is consistentwiththe_{reported(Lifely}etal., 1981)easeoflactone

formationbetweenthese_{groups in thegroup B polysaccharide.}

The closeproximity ofthese _{groups could also}result in hy-drogen bonding, possibly with the participation of water

molecules, which could also be a factor in _stabilizing the

conformation

of

the_group B_{polysaccharide.}

Inconclusion,itis_{interestingto speculate}astowhether the groupBpolysaccharideconformationaldeterminantisunique. Certainlythereissome _preliminaryevidence that would in-dicatethatitisnot. For instance, thereare _{reports(Mehmet}

etal., 1986)ofasimilarsizedependenceforoligosaccharides

obtained from keratan sulfatein_bindingto _{keratan-specific}

monoclonal antibodies. Also,the largemolecularsize_(12-14 residues)ofdermatan sulfate oligosaccharides required to bind

to _{heparin cofactor} II seems _{unusually large} forthe entire

oligosaccharide tobeincludedin a_bindingsite(Tollefsenet

al., 1986).

Registry No. (NeuAc)3, 76152-09-5; (NeuAc)4, 96425-83-1;

(NeuAc)j, 110935-75-6; (NeuAc)6, 96425-82-0; colominic acid, 9013-15-4.

References

Aue, W. P„ Bartholdi, E. & Ernst, R. R. (1976) J. Chem. Phys. 64,2229-2246.

Bax,A., & Morris,G. (1981)J._Magn. Reson.42,501-505.

Bax,A.,& Drobny,G. _{(1985)J. Magn.}Reson.61, 306-320. Bhattacharjee, A. K.,Jennings, H.J.,Kenny,C.P., Martin,

A.,& Smith,I.C.P._{(1975) J.}Biol.Chem. 250, 1926-1932. Bock, K., & Pedersen, C. (1984) Adv. Carbohydr. Chem.

Biochem. 42, 193-225.

Bock,K., Brignole,A.,& Sigurskjold,B.W. _{(1986) J. Chem.}

Soc.,Perkin Trans. 2, 1711-1713.

Brown, E. B., Brey, W.S.,Jr., & Weltner, W., Jr. (1975) Biochim. Biophys. Acta 399, 124-130.

Colson,P., Jennings,H.J., & Smith, I. C.P. (1974) J. Am.

Chem. Soc. 96, 8081-8087.

Finne, J., & Makela, P. H. (1985) J. Biol. Chem. 260,

1265-1270,

Finne, J., Finne, V., Deagostini-Bazin, H., & Goridis, C.

(1983a)Biochem.Biophys. Res.Commun. 112,482-487. Finne, J.,Leinonen, M., & Makela, P._{H. (1983b)}Lancet2

(No. 8346), 355-357.

Flippen, J. _{L. (1973) Acta Cryslallogr.,} Sect. B. Struct.

Crystallog. Cryst.Chem. 29, 1881-1886.

Jennings, H.J.(1983) Adv. Carbohydr.Chem.Biochem. 41,

155-208.

Jennings, H.J., & Bhattacharjee,A. K. (1977) Carbohydr.

Res. 55, 105-112.

Jennings,H.J.,Katzenellenbogen, E.,Lugowski,

C„

Michon,

F., Roy, R., & Kasper, D. L. (1984) PureAppl. Chem. 56,

893-905.

Jennings, H.J., Roy,R., & Michon, F.(1985)J. Immunol.

134, 2651-2657.

Joachims,J.,Taigel,G., Seeliger,A.,Lutz,P., & Driesen,H. (1967) Tetrahedron Lett., 4363-4369.

Kabat,E. A.,& _{Mayer, M. M. (1961) in Experimental}

Kinns,M.,

&

Saunders,J.K.M. ₍₁₉₈₄₎J._Magn.Reson. 56,

518-520.

Lifely,M.R.,Gilbert, A.S.,

&

Moreno, C.0981) Carbohydr. Res. 94, 193-203.

Lindon, J. _{C., Vinter,} J. _{C., Lifely, M. R.,}

&

_Moreno, C.

(1984) Carbohydr. Res. 133, 59-74.

Mandrell,R. E., &_{Zollinger, W.}D. (1982) J.Immunol. 129,

2172-2178.

Mehmet,H.,Scudder, P.,Tang,P._{W., Hounsell,}E. F., Ca-terson, B., & Feizi, T. (1986) Eur. J. Biochem. 157,

385-391.

Morris, G. A., & Freeman, R. (1978) J.Magn. Reson. 29,

433-436.

Nagayama, K., Kumar, A., Wiithrich, K., & Ernst, R. R.

(1980)J. Magn. Reson. 40, 321-334.

Neuhaus, D. (1983) J. _{Magn. Reson. 53,} 109-114.

Pearson, G. A. (1985) J. _{Magn. Reson. 64,} 487-500.

Sabesan, S., Bock, K., & Lemieux, R. _{U. (1984)} Can. J. Chem. 62, 1034-1045.

Tollefsen,D.M., Peacock, M. E., & Monafo, W. J. ₍₁₉₈₆₎

J. Biol. Chem. 261, 8854-8858.

Wagner, G. (1983) J. Magn.Reson. 55, 151-156.

Wyle, F. A.,Artenstein, M. S.,Brandt, B. _{L., Tramont, D.}

L.,Kasper, D.L.,Altieri,P., Berman,S.L., &Lowenthal,

J. P. (1972)J. _{Infect. Dis.} 126, 514-522.

Biosynthesis

of

a

_{Specifically Deuteriated Diunsaturated}

_{Fatty Acid}

_(18:2A6,9)

for

2H

NMR

Membrane

Studies'^

John E. Baenziger,*’*’******8 Ian C. P.

Smith,*’8 and Robin J. Hill8

Department

of

Biochemistry, University

of

Ottawa, andDivision

_of

_BiologicalSciences, NationalResearch Council

_of

Canada,

Ottawa, Ontario, CanadaK1A0R6

Received_May 13, 1987; RevisedManuscriptReceived_August 14, 1987

abstract:

A

_{unique procedure}

for

the biosynthesis and subsequent isolation

of

a series

of

_specifically

deuteriated c/y,m-octadeca-6,9-dienoicacidshas been_developed.

An

_auxotroph

of

_{Tetrahymena, which}

lacks A9 and A12 desaturase_activity,is_{supplementedwith specifically}deuteriated oleic acid and converts

it

to _{the corresponding deuteriated dy,c/y-octadeca-6,9-dienoic acid,} 18:2A6,9. Thedeuteriated

_fatty

acid

is _{subsequently isolated by argentationchromatography} and

HPLC.

Todemonstrate the

_utility

of

the procedure, we describe here_{the biosynthesis}

of

cw,c/j-octadeca-6,9-dienoicaciddeuteriated _{at positions}

9 and 10. Gas and _{thin-layer chromatography}

of

the isolated

_fatty

acid showed that

it

was _{greater than} 99% _purewhile 13C

NMR

and mass _spectrometry

of

the

_{G-(trimethylsilyl)}

derivativeconfirmedthatthe 18-carbon

_fatty

acid contains two doublebonds_{located at positions}6and 9. _{The yield,}froman 11-L culture, was

_typically

100 _mg

of

which 35% was found to be deuteriated at both the 9- _{and 10-positions. The} deuteriated

_fatty

acidwas esterified to _{l-hexadecanoyl-Jn-glycero-3-phosphocholine, and aqueous,}

multi-lamellar_dispersions

of

the_lipidwere _{studied by}2H

NMR.

_{Each spectrum}consists

of

two overlapping powder patterns andthereforeyields twoquadrupolar splittings. Over a_{temperature range}from0to 40 °C,one splitting decreasesfrom6.6to 1.8

kHz

whiletheotherincreases from4.5 to 5.3 kHz. The magnitudes

of

the two splittings are _equivalent between 10 and 15 °C. The values

of

the_splittings, and their _response

to _temperature,

differ

_{significantly from} those

of

_{the corresponding} deuteriated oleic acid in microbial

membranes [Ranee,

M.,

Jeffrey, K. R.,Tulloch, A.P.,Butler, K. W.,

&

Smith, I. C. P. ₍₁₉₈₀₎ Biochim. Biophys. Acta 600, 245-262] and _{in bilayers}

of

l-hexadecanoyl-2-m-octadec-9-enoyl-j«-glycero-3-phosphocholine (POPC) [Seelig, J.,

&

Waespe-Sarcevic,

N.

(1978) Biochemistry 17,3310-3315], The

results suggest

that

a

_fatty

_{acyl chaincontaining} two double bonds has _{physicochemical properties very}

different

from those

of

_{the corresponding acyl}chain

with

a _{single double bond.}

In

recent years considerable evidencehasaccumulated

sug-gestinga_Uniquerolefor polyunsaturated lipids in eukaryotic membranes. These_{lipids modulate}a_varietyof

membrane-associated processes [for a review see _Spector and Yorek (1985)] and are _thought to _playan essential rolein neural

tissue_(Lamptey &_Walker, 1976;Crawfordet al., 1984) and

in theretina _(Neuringer et al., 1984).

The effectofpolyunsaturatedlipidsis _{generally attributed}

to _{their ability} to “fluidize” membranes. A high degreeof fThisworkwas _{supportedby the}NaturalSciencesand_Engineering ResearchCouncilandtheNationalResearchCouncilofCanada. J.E.B.

isarecipientofan _{Ontario graduatescholarship.}

’Addresscorrespondence to thisauthor atthe NationalResearch

CouncilofCanada.

*

UniversityofOttawa.

8_{NationalResearchCouncilofCanada.}

0006-2960/87/0426-8405$01.50/0

unsaturation inamembraneis_thoughttocorrelatewithalow gel toliquidcrystal transition temperatureanda_highdegree of mobilityand disorder

of

thelipids. Although the double

bonditselfis a_{relativelyordered,}immobile_{structure (Seelig}

& Waespe-Sarcevic, 1978;Ranee et al., 1980;Dufourcet al., 1984), thiscorrelation appears to hold formembranes

con-tainingsaturated andmonounsaturated acyl chains(Davis & Keough, 1983; Stubbset _{al., 1981; Seelig} & _{Seelig, 1977).}

The extension to_{highly unsaturated}systems,however,hasno

rigorous physicochemical basis. For _example, differential scanningcalorimetryhasshownthat_liposomescontaininga

variety ofpolyunsaturated lecithins havesimilar transition temperatures(Coolbearet al., 1983), and fluorescence

depo-larizationof_{diphenylhexatriene, in similarbilayers,suggests}

thatthe order and ratesofmotionofthe_{lipidacyl chains} are

very similar(Stubbset al., 1981). Both _{techniques suggest}