333: .12 4.. -r ..4 g 1: $33.61... 1 . . f. . . . .; , . 1- .. -31: :4. 1... .. , Livia} . .8215... ..zL.§..4n....4.vu....m...:..¢m14 . . . u o; ” um. 1.4.4543,“ \ .7: n a. . o. . Ala... av.“.1.h.4!...i.uv tvbb-m .. u. r. 4.». «lifts. . Eva}- . 4.!4...n...43. .34 4.1.4.? h , h... u... .1... ._ . 1 . urn . . m, $ «0. .0 v... - ~ .- Io“...ttv (\t 0 Jo: um: I.» I I D I I ‘ n a h . . .. . v.54 #25,. .. u x . {5. , . . c .u . . . . . . .r. . . O I I O O 'u I I. u . a..v- n‘. . u v .139: I v ‘ o-flvohwuucfllluu . . o . 0 0.! Q l D i 19 | A .. . Lantwumwmwuu nut»... N Han" . 0v.- Ivuhbad‘; 35.300. 0.00.6... r u I n . . . “wank“. . tlkufi..(pl.|..o.uln ltglu “‘19-.Vv-v-0HHtHoV‘M-f 051v 44.9‘3‘ .. .Luvfu’n 'r. o. Mut’fl‘.lhl¢..a..~14 .0...\Ivvn4.d u. x . .. :0 u... .n u QM? .firnbuQ... 31$ .- . 34.5.4. 134.13.]...cr... .c. 2.... .3 . $523. “.34....” .u..4.r..........vv.u..wx fin?“ .Tnflfi... ”flu... i. .. “Nun: bfimVuMuq h. .utzumhmflz. 4.45133. ’0‘. “. III. :‘I ‘OV-rln U I I- . v}.o?)umtfl a .vkuu 3 3. Ltd... .1. .4). . . ANN“. . .. .. . a ~u.iv.--u0'z 44 11,“! 1.... : 5...... . z: .1 3.44. 1.1.4.: . . ~1....K.....: .v .. . 1 4... .. . 4.1“” o\ I . . .533... 4.4.1.... . . v I . o . V. O . o . I. n . I V 'l ‘u’ '1'- O. 'I ‘9 I I 243.54%. .v. 3.} .4 a”; “live. .ufiuu.n$m.vu.ll¢....t.. .v ... .IUfil ‘alfwuwnw "VIII-:34” d. ~40. l. I; o I . I4 1: . . I goo .. s v 1.! v. I, . IIQI‘Q: ~vlpbo-UA. . . . d I . .. . I. r . . . TD 1. VQIv A .o . :Vlv. : I u a . .. . {4:8 .5144: . 9%.? 4.... .3... .91....444... 4...4..vv.....o~u {#394 . . . $;~3fl..x¢l;hfluflhd$u "flaky.“ Minna»... nun}... . muahlonlfl 4 v 64 h. .. . .4. . 3! . t. .2 . . 5. f1 . w“ 4 - . It. In. .Illxuuo’ v ~ - lfiulll‘ t . 2 . 0O? 51‘“! I 0": V .6! I. 10 .Rc‘...l4..s.t .C. an??? 43!} . 4-.. . ‘44....3...)..P:4v.h.... Lu-.....c.l.v.z4~ v 1””... \Al Iv “hunt! -i 034Jll‘flol lolcl noifiu’lzIo v. ........ g V.- r I l . ‘II. I 'D ‘ l! U. A ‘ .II I (I .. l I. L.” n 1.64}. .4 unnoanLr. -vLafllfiol'vo o...h.ll. v. - 1:9?“th . .Ioolelhnhi‘Jflt I .4 ‘1..an .r yl..4.lluu..r..ilrx 2.1. 1...! Jill. . tn :2 I I I {‘L‘ v 1.: 4 .w . .4! o '9 o onn\Ithll’ult-qn oo-Jti‘ofiiulllo‘fl. uflth . >6». . -161, :4"! ....r.44:..v444.4 nfxl.2v§t.lx.448} $~111hl1fl§ , 3.1.34 . . Al 11!! . V: II I. Atx"~c‘ll“lfi.’Il ‘I‘kli"'v0l‘ig‘-zl 2. zifi'vltlf 1‘ t .‘I r. I. 11...»... a: 1.44.93!n..rn.4h.4.h4.4.hvv4o4.nP.!u.a-c..zd.flfv I...£..l44.3..\..1u.v.wvl.24...fights... . 5..., .55.... 4 2 . _ 3- £3... .. .. t r .L -4: .. . .Ix - 4o. .gfilo.‘l4fltfiflayf!.f.v;lll.«gingivoh 4o..nva.u:1.¢buwvv.;nu.v f..ulv..qbo¢dflhl.3vhfihflhfl .l ‘4’-.- l 444-: .4311! .181. :1 1.....- .. mafia».-. .4... .1! . .2 . £431.41? .1..h..:(...81.wd.rrv.zlbfu ..n4!fi!...4...z.ll. 4!.i4xtw... . b..rtzz..ou1.mo.¢...n...fipl.: .44.. . 1U . 6 . - .. . 44. .4 .. . . F J . 6|. \' u.fzcl‘ilv‘A'L U‘.O‘I ...... LII-O. In“! IQII. vtsllDOfl‘Hw . ll...l"l- 0! '5 “fir-a: N‘§.t‘nvl > It. 4| nllel¢|I4yllny 001! . Ritz-1.0.... 40".: .4 3.014 MIN. 91 vblK taluro‘vvzl‘v...ovov ". 4. I... NEWS-4.-.. 14.»... .1:lvo.v.4.w..u.ul-n...h....fl..4v¢.§|.u th.x!4..‘..\ v... 4.hX~n2..._.....i... .2... .8. vans”... in... «3. .371 x .I .... 6 94.444.291.14. ....2~.;l4l.|84(.4 4.. lflslcxr.‘ 343%..41... _ -41.} inuflfll.‘.ql}. . 142144!!! Xv.-!..-!l..-.... 4... Li! {II-4...?! 1:41... Iii. 5ulnx4yt\l...l4u.4u\ .-. x .rfvv.!.u.{v.l...¥l!.4. 14434.. E44Xtabi..iu$411!. .b .3»... :4 - . lid. v.44... J 4 .IIJII. 51!. 4.... .v “nituul .l’ I. .. Saw. .n. .thu . :I. t; 424 -l. y . . . LIL ‘ ' v 7 ill" ' Qt-s': |4 "QII I I ‘ v .. '~§-'QII..IQ¢.v.~. .IOI..Ih ‘AIIV‘I‘I s L. 10.” “vol00tvlslotv . VI ’1‘ it .u 3. v.0‘..vo..o.l.;v} . q . a.t| 14‘ Exact. III! p u .4 4 4. . 9..|._otlufoOut.h.| lac... - 44" l I 39%” v . 4... c. to: s I. I z o . .l s ‘0 ooooo till: 6 on .. . .Iluv'lu|luvv(. 0'91. .I . “‘I‘Qc’nun- 3X .v’l.ll‘v|’ 'vu.lv00sn1.lo.." 1 . .. . 4 I)!" 4...! Autl'vvulv‘o\1v.‘v.vof|vvvlkv000lol .. ’.o (110.1th I int i 0 ;J?I‘lplt.D}..le:l<~[‘tltln\no IA -DlrQIQIDIVV4 .IO- \r‘l‘ I’t'-.' IIVOID .ll :1...“ .1-.’ 3.; 0|: I 7‘13- VIII $9.4 39%|. £2.44: 2.! 44.....V.IRD. .flunrfivdvufooéu .v. , II ...o!|¢\f‘|lnl¢h~0“‘t gillinr (‘0‘... £64“ 0!. .1 4 15"}... .6 14 \IHM... 43“; 3-4.5.!!! 0 .34 o . 2 l. a . . 44 v .1... r 3"... 4..V n‘vl‘. ‘Ilnp‘t .(oo 3.. (Itsesfiflvflout‘ #f‘o'clvlvlc.’§ ..hr1".0,01. . f TH.» I .. .n v .. < .‘ t n c .- ~ . .1 r I ?v OgoIOFNII‘CA- .41.» . 41’]. t.-. . Icltidq‘iouk‘. “W’ODLHIIO. I .. Du 4“ I‘I. " lJflV‘INDA J}... Ill..vt.\hl h‘lh4yv. . 4 A, . I I. u'lIIIQ . . 44 u. otfila. girl... I ‘Iol‘oaih fui£§a ; 4. l.- ..tlnlvnou. rt} . 4‘ h... rvquuuvffi..ku4uf.l :31}... I . ' :‘I. v: I. ’i 4 Ulvfln.) ..- “Ej‘lllulw...l..4s.|4 $.11} 44:142..L.£ou.b.»vl.flwnh. 4 I4-4.vv.....!.v. . il.ll.rb. - '0 i A; .Iio'] b I“... oil‘ll uni: ,Ikr“ Vrov‘s ‘ 1 . 00.. 3.! ‘l‘ d, E40001¢i475ll Avo3r.\;O-II|. :IA‘ Q .o“ 3.- OIiII. .J-vv'. .. . u. :91... . ‘F‘hfll 1.410-unlll90llll v13... l’II I.“ n l 10‘ 0’11 I’lvvl’lfi' 4i‘}.¥‘tvi .liNOtJ. v - {0%. .4.» . .Il'lll full I’nl: if .IIUI l‘, i.tc£‘vovln‘t~z.t§|vvl‘t 4 o 1 Id”. Vlvvl. if .lt I’ll-I4 W). 9')... 910‘. {Chill-II u.xbnl..n\ln\lv‘t.4l4. luv). >1 0‘ . 19. \. {7’10}... t..0..l\.‘ u .‘l! alugibnvo «ful;vl:.1.b.lvon4.u.ot. . Evin. . ‘leoI (I... ”.8!!! 4 ,. v!!§'vt§!‘i.lu Ig1gv3;u:ll|o§c ’IV .94 . . Ifu..¢ 4 4 I. .: o .3 [Alli ‘lluc‘.lut“u .NVI 11.1%. ‘31.|\\ 4». I4 1.. H . , 4.13.»! I..- 14.1 ...v¥l|.u4\. .4 i. , .4. r. 126-..: I. 1 4.1.5.4. - dds 1.4.1.... .2. ..l.b.r3..l X. 3.1“”... - -441“: ..ul4112.3r.4|1il.l3!.. ‘Qo¢ .6 \ o on - tQ‘clIO‘AOoO'uI‘r.»‘h|-vlf‘\ t4hhhislu¢.‘lvot\v. vvow\¢!.§¢‘0~vvh.cu‘l‘:.lltd 1.2, the second E map from the sign expansion (Ra = 0.073) revealed the positions of all 13 nonhydrogen atoms. The H atoms were found in a subsequent difference Fourier syn- thesis. The nonhydrogen atoms were refined assuming anisotropic motion, and the 11 atmns on isotropic motion, using full-matrix least-squares methods. Final refinement of the model in the L configuration gave a conventional §_va1ue of 0.0421, and a weighted R value of 0.0401 using a weighting factor, w = 1/ 2(32:0). The final Hamilton 3g factor16 was 0.0425. Neutral atom scattering factors were usedl7. A final difference map showed no significant features, with a maximum residual o _ density of 0.22 e A73. Final refinement of the model in the D 28 configuration from the same starting parameters gave identical values of R, R , and R , making an assignment of the Optical isomer impossible. "W “8 Determination of absolute configuration. - Compound 1 (2 mole) was dried under reduced pressure over P The sample was treated with 205. dry 1N methanolic HCl (1 mL) for 16 h at 85°, neutralized with solid AgCO3, and centrifuged. The insoluble material was washed with 3 ali- quots (1 mL) of methanol, and the supernatants were pooled and evaporated18. The sample was de-esterified with NaOH (pH 12, 2°, 2 h), neutralized with acetic acid, passed through a column (2 mL) of Dowex 50 (H+ form), and then lyophilized, yielding an anomeric mixture of methyl glycosides (7, Figure 10). The methyl glycosides were dissolved in 2.5 mL of 0.05M sodium acetate buffer (pH 5.2). To this was added 0.5 mL of 0.1M NaIO4 in H 0. The sample was incubated in the dark at 220 for 68 h. Excess 2 periodate was eliminated by adding 40 uL of 1M glycerol. The sample was reduced with sodium borohydride, as described above, hydrolyzed in 0.5 mL of 1N HCl for 12 h at 22°, and neutralized with NaOH. The content of L-lactic acid was determined enzymatically in 0.2M glycine buffer (pH 9.2) containing 0.16M hydrazine, 0.8 mg/mL nicotinamide adenine dinucleotide, and 15 units/ml. beef 'heart ‘L-(+)— lactic acid dehydrogenase (Sigma). Total assay volume was 1.5 mL. Substrate-dependent formation of NaDH was determined spectrophotometri- cally (£340). The content of D-lactic acid was determined in the same manner except that 20 units of Iactobacillus leichmannii D-(-)-lactic acid dehydrogenase (Sigma) were used in place of the L-(+)-lactic acid dehydrogenase. Readings of_5340 were corrected by subtracting the_5340 value of a solution containing all assay components except enzyme. 29 RESULTS & DISCUSSION Isolation of compound 1. - Compound 1 (Figure 1) was bound by anion-exchange resins, and it could be eluted only under conditions stronger than those under which uronic acids are eluted. These results demonstrated that l was an acidic compound, and suggested that it could be purified by anion-exchange chromatography of the mixture of monosac- charides produced by acid hydrolysis of RG-II. Compound 1 (25-50 pg) was isolated from acid-hydrolyzed sycamore cell wall RG-II (3 mg) by chromatography in acetic acid on a column of AGl-X8 (acetate) anion- exchange resin. Compound 1 represents about 7% of purified RG-II, and RG-II about 3% of isolated sycamore cell wallsl. Therefore, it was impractical to use sycamore cell wall RG-II as the starting material for the isolation of sufficient l for structural characterization. Fortunately, a more abundant and readily accessible source of 1 was found in our laboratory. Pectinol AC (Corning Inc.), a commercial preparation of the enzymes secreted by the fungus Aspergillus niger when the fungus is grown using plant cell walls as the carbon source, was found to contain 1. Experiments (J. Thomas, id. 8. York, A. G. Darvill, and P. Alber- sheim, unpublished results) have shown that Pectinol AC contains approximately 3Z carbohydrate by weight, about 50% of which is a polysaccharide with a glycosyl-residue composition, glycosyl-linkage composition, and a size very similar to those of sycamore cell wall RG-II. The Pectinol AC RG-II contains apiose, Z-anethyl fucose, 2-9- methyl xylose, and compound 1, the rarely seen glycosyl residues that characterize RG-II. Therefore, Pectinol AC was selected as an inex- pensive, plentiful starting material from which sufficient compound 1 could be isolated to permit its structural characterization. 30 .mo>fium>wuop HHm cows pom: mos _ vcsoaeoo co nmumoavcfi mEOum conpmo mo wcwuwnesc 0:9 .— vcsoaeoo eoum vmumamua mo>wum>~uwv ocu mo mumesam .~ mpswwm my :0 nxo on: Imo: :o :08 f5 . o All... .8708: o~x- . :00: O . xooo one ”no cum “av Inc ”on rum 6? ”I n an em Iwqu zoo: eqowlamooeu InIIIIIII xownamoox 5qu 034 loo: N efisuoemst Lo AapmfiawamoV pummemmw HmCHOuxm cm cu o>fiumfiou pmmmwmmm whoa muuficm HmO¢Emso .n.a ca ~ vcsoaeoo mo mumeocm mo wusuxfie oxu mo eauuooaw .p.=721z# .a muswwm .Edd n w m o _ _ _ _ _ _ _ _ I oo 3 oomoo 88v an: v+ nu ..A I GO vhnnann:Uanmuonr_ AHAVTA any Au r. 38 TABLE I 13 C-N.M.R. CHEMICAL SHIFTS OF THE MIXTURE OF ANOMERS 0F COMPOUND l Chemical shift8 (p.p.m.) cisb anomer transb anomer C-1 97.3 102.2 C-2 79.8C 83.5 C-3 83.6 81.0 C-3’ 172.2 174.2 C—4 79.2C 79.7 C-5 11.4 14.6 a Chemical shifts were assessed relative to internal dioxane at O 67.4. bThe designation "cis" refers In) the anomer containing a cis-1,2-diol, while "trans" refers to the anomer with a trans-1,2-diol. CThese assignments may be "cis" reversed in the anomer. 39 intensities characteristic of carbonyl carbons (in. this case, carboxyl). These signals were approximately 30Z as intense as those arising from the methyl carbons; low signal intensity is usually observed with carbon atoms that do not have any hydrogen atoms bonded to them. The resonances assigned as C-3 carbons had chemical shifts typical of alcohol carbons, and had intensities comparable to those of the carboxyl carbons. These results indicated that C-3 of each anomer was a tertiary-alcohol carbon, and provided additional evidence for the branched-chain structure of compound 1. Preparation and characterization of reduction products of l (Impound 2. - Reduction of l with borohydride yielded a 3-C- carboxy-S-deoxy-pentitol, which readily formed a Y-lactone upon acidi- fication (2, Figure 1). Evidence for the formation of 2 was obtained by f.a.b.-m.s. The molecular weight of borohydride-reduced l was found to be 162, which is the molecular weight of 2. No ion was observed with mlz_181 that would correspond to the (M + H+) ion of the free-acid form of the hydride-reduced derivative. The 1H-n.m.r. spectrum of 2 in D20 (Figure 5) contained a 3-proton doublet (J = 6.5 Hz) at 6 1.36, and a 1-proton quartet (J = 6.5 Hz) at 5 4.11. These signals arose from the methyl protons and H—4, respectively. The other three nonexchangeable protons (2 protons on C-1, and 1 proton on C-2) gave rise to a complex 13 3-proton multiplet between 6 4.33 and 6 4.70. The C-n.m.r. spectrum of 2 in D 0 contained as expected, resonances at 6 16.5 (methyl); 68.5, 2 68.6, 74.3 (alcohol); 79.2 (30 alcohol); and 180.1 (carbonyl). 4O 0 H 0 00 00 Id DC) H CH3 I 1 I 1 I 1 I 1 I 5 4 3 2 l p.p.m Figure 5. 1H—N.m.r. spectrum of coupound 2 in D20. Chemical shifts were assessed relative to an external standard (capillary) of hexamethyldisilazane (5 0.6995). 41 0(3de 3. - Acetylation of compound 2 yielded a volatile tri-Q-acetyl lactone (3, Figure 1). The molecular weight of 3 was 288, as determined by c.i.-m.s. (isobutane) and f.a.b.-m.s. The molecular weight of the corresponding derivative that was per-Q-acetylated using hexadeuterioacetic anhydride was 297. The difference of nine mass units indicated that three Q-acetyl groups were incorporated. The e.i. mass spectrum of 3 is shown in Figure 3. The ion with mlg_244 (M-44) corresponded to loss of carbon dioxide from the molecular ion. This is characteristic of a lactone28. The M-CO2 ion sequentially eliminated two molecules of ketene to yield the ions with mlg_202 and M160. The base peak, with m_/z_ 142, arose from the elimination of acetic anhydride from the M-CO2 ion. The ion with 3111 231 arose from the sequential losses of a methyl radical and ketene from the molecular ion. The other features of the mass spectrum were typical of those of alditol acetatesz’a. The ion with mlg_229 resulted from the loss of an acetoxy radical from the molecular ion. This ion lost acetic acid to generate the ion with mlg_169, which in turn eliminated ketene to yield the ion with w 127. Loss of acetic acid from the molecular ion yielded the ion with mlg_228, which in turn lost ketene to yield the ion with pig 186. The 1H-n.m.r. spectrum of 3 in CDCl3 (Figure 6) contained a three-proton doublet (J = 6.3 Hz) at 6 1.38 and a one-proton quartet (J = 6.3 Hz) at 5 5.24, arising from the methyl group and H-4, respect- ively. The methyl protons of the IQ-acetyl groups gave rise to three-proton singlets at 5 2.04, 2.11, and 2.18. The other three non- exchangeable protons comprised an AMX spin system, and gave rise to one-proton doublets-of-doublets at 6 4.27 (J =' 9.7 HZ. J ‘ 5-5 HZ). 5 4.74 (J = 9.7 Hz, J = 8.2 Hz), and 6 5.54 (J = 5.5 Hz, J = 8.2 Hz). 42 Carboxyl-reduced alditols. - Borohydride reduction of lactone 2 yielded compound 4a (Figure 1), a branched-chain deoxy alditol. Reduction of 2 with borodeuteride in borate buffer yielded compound 4b (Figure 1), the dideuteriolabeled derivative of 4a. Com- pounds 4a and 4b were per-Q-acetylated, yielding compounds 5a and 5b, respectively (Figure 1), which were analyzed by g.l.c.-m.s. The e.i. mass spectrum of compound 5a is shown in Figure 7A. The highest ions in the spectrum (mlz_317 and mlg_316) were formed by eliminating from the molecular ion an acetoxy radical and acetic acid, respectively. The other primary fragment—ions arose from fission of the alditol chain, as indicated in Figure 7A. Secondary fragment-ions were formed by elimina- tions of ketene, acetic acid, and acetic anhydride from the primary fragment ions. The base peak at m/_z 128 and the ion at 373170 are present in the e.i. mass spectra of all hexa-Q-acetyl hexitols and 29, and are, therefore, of no diagnostic penta-Q-acety1-6-deoxy-hexitols value. Comparison of the e.i. mass spectrum of compound 5a to that of compound 5b (Figure 7B) confirmed the location and nature of the branch in compound 1. The primary ions with M 317, 316, 289, and 231 in the mass spectrum of 5a were shifted to 1113 319, 318, 291, and 233 in the mass Spectrum of 5b. The secondary ions derived from these primary fragments were also shifted by two mass units. The primary ion at m_[_§_ 145 and the secondary ion derived from it were not changed by carbo- xyl reduction with borodeuteride. Compound 5a gave rise to two differ- ent primary fragment-ions with M303. In the mass spectrum of compound 5b, one of these ions had _m_[_z_ 303, while the other had shifted to m/z 305. These results confirmed that compOund 1 had a carboxyl branch at C-3. 43 .E.a.a ~.m Cu 8 scum cowwmu Hmuuooam OLD Lo uoHa moccmaxw cm ma oomuu moan: oFe .Ac~.n 0v Epo ouofisu fimemucH cu w>fium~mp mommwmmm mums wumwzm ~60w2®£0 . Hugo CH n vésoaeoo mo esuuuoam .u.=721=~ .Eda v m a — .o ousmfim Figure 7. 44 E.i. mass spectra of A) compound Sa and B) 5b. Secondary fragment-ions, formed by eliminations of ketene (42 ‘mass units), acetic acid (60 mass units), and acetic anhydride (102 mass units) have been indicated. 45 A. -50 -IOZ r ........... c 13°“ -42 Hie—2019303 AcOCH 145—NOS 129.7—63139'4-4i 231 r ........ J— 7777 | ......... J "2 '60 AcOCH; ' --C|:.-_0_‘:¢ _____ jZBS-a24T—P I87 IAcOCH I I CH, 1 k -42 -80 303—9 261—920I I00-1 128 II2 )- 103 145 .- .70 201 (7) 507 154 z ”7139 E E I u- all ALI.“ llll u 1 l . 1.111 4‘ A AL I: L A o I T 1 I 8 I00 I25 I50 I75 200 - IOO'1 g r. ‘ 5 r. N IO 2 316 d 50-1 303 K 231 261 239 3|? u I 247 l I I O W A Y I I 7 225 250 275 300 nu: B -so -102 /' ........... 0.13.0“ -42 I434— ZO3¢—305 AcOCH 145—.103 -60 -42 /- -------- I """" J -42 ‘50 131 <—19I t—233 AcOCDz t?:-P.Af__ JZSI —->249—-b189 :ACOCH ' I I ”s 1 -4 O )303—3261—‘3201 100 I30 45 )- '03 | 172 t: 50 ”3 2 .55 '89191 203 ='-‘ I I II I I I Z 1711. glhlL- A -Llll in ..Al ‘11.] 1 |lllll Anal Ill - l .111... I l-A— - O l T I I 5 I00 125 ISO 175 200 - 100-7 u ‘5 x10 2 F’ F‘ : 3|8 ..1 50-1 305 m 233 25' I: 291 3'9 249 2T9 I H O ‘4“ 1 7 ‘ ‘1 * *r I ’ 225 250 275 300 av: Figure 7 46 Determination of the relative configurations of the three asymmetric carbons of compound 1. - The results presented above clearly established that compound 1 was a 3-C-carboxy-5-deoxy-pentose. The open-chain form of such a sugar has three asymmetric carbons; therefore, compound 1 could exist as one of eight stereoisomers (four L isomers and four D isomers). The relative configurations of the three asymmetric carbons were established by X-ray crystallography of compound 6 (Figure 1), an oxidation product of compound 1. Compound 6 was prepared by bromine oxidation of compound 1. The lH-n.m. r. and 13C- general structure of compound 6 was confirmed by n.m.r. The 1H-n.m.r. spectrum of 6 in D20 (Figure 8) contained a three— proton doublet (J = 6.6 Hz) at 6 1.40, and a l-proton quartet at 5 4.70, arising from the methyl group and H-4, respectively. H-2 gave rise to a one-proton singlet at 6 4.62. The 13 C-n.m.r. spectrum of 6 contained, as expected, the following signals: 6 15.7 (methyl); 73.4 and 80.5 (alcohol); 81.0 (3° alcohol); 173.9 and 175.9 (carbonyl). Compound 6 was crystallized from a mixture of 2-pentanone and chloroform. X-ray crystallography of compound 6 confirmed all previously determined structural features, and revealed that the sugar has the xylg_configuration (Figure 9)*. It was not possible to deter- mine absolute configuration of 6 from the crystallography data. Determination of the absolute configuration of 1. - The absolute configuration of 1 was determined by periodate oxidation of the anomeric *The other results of the X-ray crystallographic analysis of compound 6 are presented in the Appendix. 47 H00 H00 "Sc GOOD DO .— p—— p—- h- I... :— _ p- Figure 8. 1I-I-N.m.r. spectrum of compound 6 in D 0. Chemical shifts were assessed relative to an external standard (capillary) of hexamethyldisilazane (6 0.6995). The upper trace is an expanded plot of the spectral region between 4.6 and 5.0 p.p.m. 48 Figure 9. Structure of compound 6, as determined by X-ray crystallography. 49 mixture of the methyl glycosides (Figure 1). The methyl glycosides were prepared by methanolysis of 1, followed by base de-esterification. The methyl glycosides were subjected to Smith degradation (periodate oxida- tion, followed by borohydride reduction and mild acid hydrolysis)3°. The expected product of this series of reactions was D-lactic acid if 1 had the 1) configuration, or L-lactic acid (Figure 10) if 1 had the L configuration. Enzymatic analysis was used to determine the concentrations of D- and L-lactic acid obtained by Smith degradation of the methyl glycosides of l. L-lactic acid (0.6 umole of L-lactic acid per umole of starting material) was the major product recovered. Therefore, it was concluded that 1 was an L sugar. A small amount of D-lactic acid (0.1 umole/umole starting material) was also detected. The origin of the D-lactic acid is not understood because compound 1 was shown to be optically pure by the method of Gerwig E£.§l: (see above). A possible explanation for the formation of [Plactic acid is that some isomerization occurred during the series of reactions used to form the lactic acid. 50 O HCH HC COOH O Oil-”5"“H 7 3 HO MeOH H9 0 H H.0M0 0M0 COOMO OH H,c “:0 COOH cuzon HO OH 120 NOBH4 0 H H.0Mo T '. COOH OH No 4 Cr OM. INC “3 COOH HC h~ HO 0 Figure 10. Reaction sequence used to determine the absolute configuration of compound 1. 51 GENERAL DISCUSSION Compound 1 it; 3-C-carboxy-5-deoxy-L-xylose. ‘We propose the trivial name "aceric acid" for this sugar because it was first observed1 as a constituent of cell walls isolated from suspension- cultured cells of Acer pseudoplatanus (sycamore). More than a dozen branched-chain sugars have been found in nature, frequently as 31’32. Streptose (3-C-formyl-5-deoxy-L- components of antibiotics 1yxose)1°, a neutral-sugar component of the antibiotic streptomycin, is the most similar Of these to aceric acid, but it has a formyl rather than a carboxyl group at C-3’, and the ly§g_rather than the xylp_configuration. Only two other branched-chain sugars have previ- ously been found in plant tissue33; these are apiose (3-C- (hydroxymethyl)-D-g1ycero-tetrose) and hamamelose (2-C-(hydroxy- methyl)-D-ribose). Aceric acid :hs the first branched-chain acidic sugar to be found as a natural product. 52 REFERENCES 1 10 11 12 13 14 15 16 17 18 19 20 21 A. G. DARVILL, M. MCNEIL, AND P. ALBERSHEIM, Plant Physiol., 62 (1978) 418-422. O. S. CHISOV, L. S. COLOVKINA, AND N. S. WULFSON, IZV. Akad. Nauk SSSR, Ser. Khim. (1966) 1915. J. LONNCREN AND s. SVENSSON, Adv. Carbohydr. Biochem., 29 (1974) 41-1060 M. BARBER, R. S. BORDOLI, R. D. SEDCWICK, AND A. N. TYLER, Nature (London), 293 (1981) 270-275. B. S. VALENT, A. G. DARVILL, M. MCNEIL, B. K. ROBERTSON, AND P. ALBERSHEIM, Carbohydr. Res., 79 (1980) 165-192. M. MCNEIL AND P. ALBERSHEIM, Carbohydr. Res., 56 (1977) 239-248. K. w. TALMADGE, K. KEEGSTRA, w. D. BAUER, AND P. ALBERSHEIM, Plant Physiol., 51 (1973) 158-173. P. D. ENGLISH, A. MAGLOTHLIN, K. KEEGSTRA, AND P. ALBERSHEIM, Plant Physiol., 49 (1972) 293-297. B. LARSEN AND A. HAUG, Acta Chem. Scand., 15 (1961) 1397-1398. Z. DISCHE, Methods Carbohydr. Chem., 1 (1962) 477-512. P. ALBERSHEIM, D. J. NEVINS, P. D. ENGLISH, AND A. KARR, Carbohydr. Res., 5 (1967) 340-3450 T. M. JONES AND P. ALBERSHEIM, Plant Physiol., 49 (1972) 926-936. F. J. KUEHL, Jr., Methods Carbohydr. Chem., 1 (1962) 268-272. G. H. STOUT AND L. H. JENSEN, X-ray Structure Determination, Macmil- lan Publishing, New York, 1968, p. 65. G. M. SHELDRICK, SHELX76. Program for crystal structure determina- tion, University of Cambridge, England, 1976. w. c. HAMILTON, Acta Cryst., 18 (1965) 502-510. D. T. CROMER AND J. B. MANN, Acta Cryst., A24 (1968) 321-324. R. E. CHAMBERS AND J. R. CLAMP, Biochem. J}, 125 (1971) 1009-1018. C. J. GERWIG, J. P. KAMERLING, AND J. F. G. VLIEGENTHART, Carbohydr. Res., 77 (1979) 1-7. N. J. NELSON, if 3101. Chem., 153 (1944) 375-380. Mo LEVER, Anal. BlOChem., 47 (1972) 273—2790 22 23 24 25 26 27 28 29 30 31 32 33 53 N. BLUMENKRANTZ AND G. ASBOE-HANSEN, Anal. Biochem., 54 (1973) 484- 489. R. M. SILVERSTEIN, G. C. BASSLER, AND T. C. MORRILL, Spectrometric Identification p£_Ckganic Compounds, John Wiley & Sons, New York, 1974, p. 13. B. N. AMES, in Methods 13 Enzymology, Vol. 8, Academic Press, New York, 1966, pp. 115-118. J. D. STEVENS AND H. G. FLETCHER, JR., J. Org. Chem., 33 (1968) 1799. D. R. BUNDLE AND R. U. LEMIEUX, Methods Carbohydr. Chem., VII (1976) 79-860 A. S. PERLIN, N. SYR, H. J. KOCH, AND B. KORSCH, Ann. 3. g. Acad. Sci., 222 (1973) 935-942. F. W. MCLAFFERTY, Interpretation pf_Mass Spectra, University Science Books, Mill Valley, California, 1980, p. 283. P.-E. JANSSON, L. KENNE, H. LIEDGREN, B. LINDBERG, AND J. LUNNGREN, Chem. Commun., Univ. Stockholm, 8 (1976) 1-75. I. J. GOLDSTEIN, G. W. HAY, B. A. LEWIS, AND E. SMITH, Methods Carbohydr. Chem., 5 (1965) 361-370. H. GRISEBACH AND R. SCHMID, Angew. Chem. Int. Ed, Engl., 11 (1972) 159-173. N. R. WILLIAMS AND J. D. WANDER, in W. PIGMAN AND D. HORTON (Eds.), The Carbohydrates, Vol. 1B, 2nd ed., Academic Press, New York, 1980, H. GRISEBACH, in J. PREISS (Ed.), The Biochemistry_gf Plants, Vol. 3, Academic Press, New York, 1980, pp. 171-197. CHAPTER II CHARACTERIZATION OF A STRUCTURALLY COMPLEX HEPTASACCHARIDE ISOLATED FROM RHAMNOGALACTURONAN II 54 55 ABSTRACT A heptasaccharide was released from the plant cell wall pectic polysaccharide rhamnogalacturonan II by selective acid hydrolysis of the glycosidic linkages of apiosyl residues. The heptasaccharide was puri- fied to homogeneity by gel-filtration and anion-exchange chroma- tography. Some of the heptasaccharide molecules were found to be mono- and di-Q-acetylated; the location of the acetyl esters was not determined. The heptasaccharide was found to have the following structure (where Rha = rhamnosyl, Ara = arabinosyl, Gal = galactosyl, Fuc = fucosyl, and AcA = aceryl [3-C-carboxy-5-deoxy-L-xylosyl] residues, and Api = apiose): 2. 0 a 6 9 up Z-L-Arap-94-D-Gplpé 2-L-AcAf—>3-L-Rhap—>3'-Api ’Ia L‘FUCp 2 I Methyl 56 INTRODUCTION Rhamnogalacturonan II (RC-II) is :1 complex pectic polysaccharide that is released from the cell walls of suspension-cultured sycamore cells (Acer pseudoplatanuS) by the action of an endopolygalacturonase isolated from Colletotrichum lindemutheanum. RG-II is size-homogeneous, containing about 60 glycosyl residuesl. RG-II yields at least 10 different. monosaccharides upon acid hydrolysis, including the unusual sugar apiose (Api; 3-C-[hydroxy- methyl]-D-g1ycero-tetrose), and the unusual sugar derivatives 2-9- methyl-fucose, and 2-_Q-methyl-xylose1. A new branched-chain sugar, aceric acid (AcA; 3-C-carboxy-S-deoxy-L-xylose) has also been identified as a component of RG-II (Chapter 1). The monosaccharide constituents of RG-II are interconnected by at least 15 different glycosidic linkagesl. The structure of intact RG-II is too complex to be characterized by conventional methods. Therefore, selective acid hydrolysis of the apiosyl glycosidic linkages of RG-II was used to generate a mixture of smaller Oligosaccharides. In this paper, we report the structural characterization of a heptasaccharide that was isolated after selective acid hydrolysis of RG-II. 57 EXPERIMENTAL Isolation of RG-II. - RG-II was isolated from primary cell walls of suspension-cultured sycamore cells, as described in Chapter I. Glycosyl-comppsition analysis. - The glycosyl-residue composition of the heptasaccharide was determined by g.l.c. and g.l.c.-m.s. analyses after hydrolysis in 2N trifluoroacetic acid (TFA) (2 h, 120°), reduction with sodium borodeuteride (NaBDA), and acetylationzo Preparation of 2-O-methyl-L-fucose. - A sample of Z-Q-methyl-L- fucose was prepared from methyl-a-L-fucopyranoside (Sigma Chemical Co.). Methyl-a—L-fucopyranoside (25 mg) was converted to the 3,4-di-9- isopropylidene derivative by the procedure of Liptak §£_§l:3, and then .Q-methylatedé’s. Acid hydrolysis of this product (2-9mnethyl-3,4-di-Q- isopropylidene-a-methyl-L-fucopyranoside) yielded Z-Q-methyl-L-fucose. Glycosyl-linkage composition of the heptasaccharide. - Isolated RG-II heptasaccharide (0.5 mg) was dissolved in 0.4 mL dimethyl sulfoxide. Sodium dimethylsulfinyl anion (75 DL of 4M) was added, and the solution stirred for 2 h. Methyl iodide (75 uL) was added, and the solution stirred for another 2 h. Per-_O_-methylated carbohydrate was recovered by chromatography on a Sep-Pak C1 cartridge (Waters Asso- 8 ciates), as described by waeghe‘££_al,° The glycosyl linkages of the neutral residues were determined by g.l.c.-m.s. after hydrolysis, reduc- tion with NaBD and acetylation. The glycosyl linkage of aceric acid 49 was determined by a modified procedure. An aliquot (0.2 mg) of the per- _(_)_-methylated carbohydrate was hydrolyzed and reduced with NaBD“, as described7, deionized by passage through a column (2 mL) of Dowex 50 (H+ 58 form), and then lyophilized. This procedure yielded a mono-Q-methyl lactone derivative of aceric acid (compound 2, Figure 5). The mono-O- methyl lactone was carboxyl reduced, using the procedure of JOnes and Albersheim8. This process yielded compound 4 (Figure 5), which was analyzed by g.l.c.-m.s. after per-Q-acetylation (compound 5, Figure 5). Per-O-deuteriomethylation. - The 'heptasaccharide-alditol (3 mg) was dissolved in dimethyl sulfoxide (2 mL), and the solution was stirred for 4 h. Sodium dimethylsulfinyl anion (75 ML of 4M) was added, and the mixture was stirred overnight. Trideuteriomethyl iodide (20 uL; Stohler Isotopic Chemicals) was added, and the solution was stirred for 2 h. Sodium dimethylsulfinyl anion and trideuteriomethyl iodide were added twice more; 100 DL of trideuteriomethyl iodide was used for the final addition. L.c.-m.s. separation of the mixture of per-O-deuteriomethylated Oligosaccharides and Oligosaccharide-alditols generated by degradation of the aceryl residue during per-O-deuteriomethylation of the heptasac- charide-alditol. - The mixture of per-Q-deuteriomethylated oligomers, generated by partial degradation of the aceryl residue of the heptasac- charide, was separated by l.c.-m.s. on a Dupont Zorbax ODS column using a linear, 45-min gradient of 50 to 75% acetonitrile. All other l.c.- m.s. conditions were as describedg. Preparation of per-O-alkylated Oligosaccharide-alditols from per- O-deuteriomethylated fragments generated by degradation of the acetyl residue of the heptasaccharide-alditol. - The fractions containing material that eluted from the Dupont Zorbax ODS column with retention times between 4 and 10 min (Figure 7) were pooled and evaporated to 59 dryness under a stream of filtered air. The material was partially hydrolyzed by treatment with 88% formic acid for 2 h at 50°, and the formic acid was evaporated under a stream of filtered air. The sample was reduced with NaBDA, deionized by passage through a column (2 mL) of Dowex 50 (H+ form), lyophilized, and then dried overnight in a vacuum oven at 40°. The sample was dissolved in 0.5 mL of dimethyl sulf- oxide. Sodium dimethylsulfinyl anion (100 DL of 4M) was added, and the solution was stirred for 4 h. Pentadeuterioethyl iodide (100 uL; Stohler Isotopic Chemicals) was added, and the mixture was stirred for 3 h. The per-Q-alkylated Oligosaccharide-alditols were recovered from the reaction mixture by chromatography'cn1.a Sep-Pak C18 cartridge, as described6. G.1.c.-m.s. of per-O-alkylated Oligosaccharide-alditols derived from the heptasaccharide. - Partially-Q-deuteriomethylated, partially-9- deuterioethylated, Oligosaccharide-alditols were analyzed by g.l.c.-m.s. on a 151m DB1 (J&W) capillary column (0.32 mm i.d.), as described10 Isolation (fl? per-O-alkylated Oligosaccharide-alditols [a], [d’], and [f] (Tables II and III) by l.c. - An aliquot (2 mg) of the per-9- alkylated Oligosaccharide-alditols derived from the heptasaccharide was separated by l.c. on a Dupont Zorbax ODS column using a linear, 30-min gradient of 50 to 657. acetonitrile. Per-O-alkylated Oligosaccharide- alditols were detected in the l.c. effluent by g.l.c.-m.s. (e.i.) analy- sis 7 of aliquots of the column fractions. 1H—N.m.r. — 1H-N.m.r. spectra were recorded on a Bruker WM—250 Fourier-transform n.m.r. spectrometer operated at 250 MHz. Water- soluble samples were lyophilized twice from deuterium oxide (99.7 atom Z 60 D), then dissolved in deuterium oxide (99.997 atom Z D). Chemical shifts of aqueous samples were assessed relative to internal dioxane (6 3.70). Methylated samples were run in hexadeuterioacetone (99.997 Z D) and chemical shifts were assessed relative to internal pentadeuterio acetone (5 2.04). 61 RESULTS AND DISCUSSION Isolation of an oligosaccharide released by partial acid hydrolysis of RG-II. - Purified RG-II (125 mg) was treated with 0.1N TFA for 24 h at 40°, and the extent of hydrolysis of glycosyl residues was determined as describedlo. Under these conditions, approximately 35Z of the apiosyl glycosidic linkages and approximately 5Z of the methyl fucosyl and tflua arabinofuranosyl glycosidic linkages were hydrolyzed. There was no detectable hydrolysis of any other glycosidic linkage. The partially hydrolyzed RG-II was applied to a heated (65°) column (1.5 x 85 cm) of Biogel P-lO (200-400 mesh) that had been equili- brated in 50mM sodium acetate buffer (pH 5.2). Two major peaks of carbohydrate-containing material were resolved (Figure 1). The first peak had the same approximate elution volume as untreated RG-II. The second peak (shaded) was pooled, deionized by passage through a column (1 x 6 cm) of Dowex 50 (H+ form), and then lyophilized. Lyophilized peak II from the Biogel P-lO column was chromatographed on a column (1 x 2 cm) of QAE Sephadex Q-50-120 that had been equilibrated in 30mM NH HCO (Figure 2). The column was washed 4 3 with 15 bed volumes of 30mM NHAHCO3, 250mM NH4HC03. 1A peak of orcinol-positive11 material (shaded region, Figure 2) was eluted from the column in the 30mM NHAHCO3 wash. This material, which did not give a positive response in the m- hydroxydiphenyl assay12 for uronic acids, was the Oligosaccharide then with 15 bed volumes of characterized in the remainder of this paper. The 250 mM NHAHCO3 wash eluted material that gave positive responses in both the orcinol and the 62 1? I 12— /\ 7 i; . C) o: 4'“ _ _ V1 " I .8” -1 3; 1‘3 co F 5 <1 8 Z .41- - <1 00 m --1 C) _ (I) CD q . 21-11:}? 25 I00 ELUTION VOLUME (ml) Figure l. Chromatography of partially hydrolyzed RG-II on a heated (650 column (1.5 x 85 cm) of Biogel P-10 (200-400 mesh). The column was equilibrated in 50 mM sodium acetate buffer (pH 5.2). Collected fraction volume was 2.5 mL. Fractions were assayed for neutral-sugar content by the orcinol method 1 (A6 5)12nd for uronic acid content by the mrhydroxydiphenyl method (A O). The shaded fractions were pooled for further puriIIcation. 63 .0 N I ’I. A520 (°'"’°I .0 1 ABSORBANCE A565 (- Figure 2. I6— 30 mM NH4HC03 a: a 250 mM NH4HC03—-9I / )1 b----..9 ELUTION VOLUME (ml) Chromatography, on QAE-Sephadex, of pooled material from the Biogel P-lO column. A column (1 x 2 cm) of QAE-Sephadex Q-50-120 was equilibrated with 30 mM NH HCO . After sample application, the column was washed with 5 bed volumes of 30 mM NH HCO followed by 15 bed volumes of 250 mM NHAHCOB. Collected fraction volume was 1.5 mL. Fractions were assayed as described in the Figure 1 legend. 64 mfhydroxydiphenyl assays. This material has not been further character- ized. Determination of glycosyl composition and size of the isolated Oligosaccharide. - The material that eluted from the QAE Sephadex column in 30mM NHQHCO3 (shaded region, Figure 2) was pooled and lyophilized. This material was chromatographed in water on a heated (65°) colimin (1 x 75 cm) of Biogel P-6, and eluted as a symmetrical, partially included peak (Figure 3). Aliquots were taken from the fractions com- prising the peak, and the glycosyl-residue composition was determined by g.l.c. and g.l.c.-m.s. analysis of the products of reduction. with borohydride, hydrolysis (2N TFA, 120°, 1 h), reduction with borodeuteride, and acetylation. The glycosyl-residue composition was constant across tflua peak (Figure 3, inset). The glycosyl composition data were consistent with those of a heptasaccharide. Only apiose was reducible before acid hydrolysis; this indicated that apiose was at the reducing end of the heptasaccharide, as would be anticipated from the partial hydrolysis results. Fast atom bombardment-mass spectrometry (f.a.b.-m.s.) provided strong evidence that the isolated Oligosaccharide was a heptasaccharide. The negative-ion f.a.b. mass spectrum contained ions corresponding to [PU-H (mlz_1056), [M + acetyl]-H (mlg_1097), and [M + 2 acetyl]-H (mlg_ll39). The results were consistent with the glycosyl composition data, and also indicated the presence of one and of two acetyl esters on some of the heptasaccharide molecules. The lH-n.m.r. spectrum of the underivatized heptasaccharide in D20 (Figure 4) confirmed the presence of acetyl esters on some of the hepta- saccharide molecules. The spectrum contained singlets, at 5 2.16, 2.12, 65 .uxmu ozu c“ wopfiuommv mm vOCHELOuOm whoa Auomcfiv m—Icfi mco«uomuw mo mcoHufiwonEoo Cavemen Ifimmouxfiw och .N pogum=_ flocfiouo ozu mp ucoucou umwsmlfimuuam: pom vwxmmmm muo3 erauumpm .AE m.~ mm: OESHO> cofiuumuu kuomHHou .o : cw kumpanHSUO mm3 Aswwn_ ccwloofiv elm Hoaofim mo AEU mm x 3 5.38 A69: 689:. < .moozezz :2 cm 5 55:8 8.6665863 mLu Eoum kusflm umzu Hmflpoume may mo .clm Homewm co .>;amuwoumeoucu .m muswwm 20:05.... 0.. m . quoudufiuououludj o\ v on n: 0 Mu on v t m. m. m. m. 2 38.8.5 m N. o. o. .. N. .. .raea . . . 3 n. v. e. e. N. N. Bee... ..0 v m. m. m. m. m. t .8583 m... AN 8 N 8 MN 8 .8852”. (.7 t m. m. m. m. m. assist... o\o mmoz 701 n. v. m. N. .. o. 2.23:. o c 20.83.... 3.60830 /\ q . O) ..Nd Figure 4. 6 .4 66 HOD l I 1 p.p.m. lH-N.m.r. spectrum of the underivatized heptasaccharide in D 0. Chemical shifts were assessed relative to internal donane (<8 3.70). The resonances arising from the methoxy protons of the Z-Q-methyl-fucosyl residue (8), the methyl protons of the Q—acetyl esters (b), and the methyl protons of the deoxy-glycosyl residues (c) are indicated. 67 and 2.08, having typical chemical shifts for the methyl protons of O- acetyl esters”. Integration of these resonances showed that the 9: acetyl esters were present in less than one mole per mole of heptasaccharide: the largest resonance (6 2.16) corresponds to 0.6 mole acetyl ester per mole heptasaccharide; each of the other two resonances corresponds to less than 0.3 mole acetyl ester per 'mole heptasac- charide. It is likely that the .Q-acetyl esters were partially hydrolyzed by the acidic conditions used to release the heptasaccharide from RG-II. Acidic conditions are also known to cause migration of 9- acetyl esterslé No attempt was made to determine the points of attach- ment of these substituents. Determination of the absolute configurations of the glycopyl residues of the heptasaccharide. - The absolute configurations of all of the constituent sugars of the heptasaccharide except apiose were deter- mined by the methods of Gerwig 3£“31315’l° and Leontein pp £1.17. Both methods utilize solvolysis reactions in which the solvent is an optically active alcohol. When a pair of enantiomers is treated with a chiral reagent, the products are diastereomers, which can usually be resolved by g.l.c. The procedure of Gerwig £5 Elf uses optically active butanol; the products of butanolysis are then. converted into their trimethylsilyl derivatives, and analyzed by g.l.c.-m.s. The procedure of Leontein 35 El: is analogous, but the solvolysis reaction is run in optically active octanol, and the acetyl derivatives of the products are analyzed by g.l.c. The galactosyl residue was shown to have the D absolute configura- tion, and the rhamnosyl and Z-anethyl fucosyl residues L absolute con- figurations by the method of Gerwig EL 31. Authentic Z-O-methyl-L- 68 fucose was synthesized for use in these experiments, as described in the "Experimental" section. The arabinosyl residue was shown to have the L absolute configuration by the method of Leontein SEAS}! Aceric acid was previously shown to have the L absolute configuration (Chapter 1). It was not possible to assign the absolute configuration of apiose by the method of Gerwig g£_§l: or by the method of Leontein g£_§l: because the diastereomeric products were not sufficiently resolved by g.l.c. The L 18 Therefore, the optical isomer of apiose has not been found in nature apiose in the heptasaccharide has been assumed to have the D absolute configuration. Determination of the glycosyl-linkage composition of the heptasac- charide. - The heptasaccharide was reduced with NaBD4 to yield the heptasaccharide-alditol. The glycosyl-linkage composition of the heptasaccharide-alditol was determined by per-anethylation4’? followed by hydrolysis, reduction with NaBD and acetylation (Table I). The 4. glycosyl linkage of aceric acid was determined by a modified procedure, which will be discussed separately. Prereduced apiose was recovered in low yield, presumably because some of the volatile tetra-Qfinethyl derivative was lost in the workupzo. Aceric acid was partially degraded under the conditions used to methylate the heptasaccharide and consequently was also recovered in low yield. Degradation of aceric acid during the alkaline conditions of methylation became apparent from the results of experiments performed to determine the glycosyl-residue sequence of the heptasaccharide, as will be discussed later. Alkaline degradation of similar compounds has been reportedzl. 69 TABLE I GLYCOSYL-LINKAGE COMPOSITION OF THE HEPTASACCHARIDE Glycosyl Points of Deduced Mole Za residue attachment glycosidic of Q-methyl linkage groups Z-Q-methyl Fucosyl 2,3,4 terminal 17 Rhamnosyl 2,3,4 terminal 19 Rhamnosyl 2,4 3-linked 18 Arabinosyl 3,4 2-1inked 19 Galactosyl 3,6 2,4-linked 15 Aceryl 3b 2-linked 6 Apiose 1,2,3,4 31-1inked (alditol) 5 3 Calculated using "Effective Carbon Ratio"2°. bDetermined in a separate experiment, as described in text. 70 The glycosyl linkage of aceric acid was determined as follows. Per-Q-methylation of the heptasaccharide, followed by hydrolysis, reduc- tion, and acetylation, yielded a di-Q-acetyl-mono-Qjmethyl lactone (compound 3, Figure 5), which was identified by g.l.c.-m.s.(e.i.). Because compound 3 is cyclic, it was not possible to determine the position of the .9-methyl group using the established fragmentation patterns of partially-Q-methylated alditol acetateszo. Therefore, compound 2 was reduced with sodium borohydride in borate buffer° to yield the acyclic, carboxyl-reduced alditol (compound 4, Figure 5). This derivative was per-_O_-acetylated, yielding compound 5, which was analyzed by g.l.c.-m.s. The e.i.-m.s. fragmentation pattern of compound 5 (Figure 6) clearly established that aceric acid was 2-linked in the heptasaccharide. Determination of the sequence of glycosyl residues in the heptasaccharide. - The formation, resolution, and characterization of overlapping per-Q-alkylated Oligosaccharide-alditol fragments constitute a general procedure for determining the glycosyl-residue sequence of a 9’10 such as the heptasaccharide. The sequencing complex carbohydrate method was modified in the present study by using deuterated alkylating reagents in place of the normal (i.e., nondeuterated) alkylating reagents. The reason for this modification will become evident in the following discussion. The present study was complicated by degradation of the aceryl residue during per-Qfdeuteriomethylation of the heptasac- charide-alditol. The partial degradation of the aceryl residue made necessary an extra resolution step to recover the per-Q-deuteriomethyl derivative of the intact heptasaccharide-alditol from the mixture of per-Q-deuteriomethylated Oligosaccharides and. Oligosaccharide-alditols that resulted from this degradation. 71 .A—v oswwmoh Hhumom vmxcwalm wcu scum vocwmuno mw>wum>wuom wzu mwumuumafifia oewzom wash .AOqufimIocHumsuommmuaoc ecu av oaewmou Haydon 0.3 no wwmxcz T3093w wcu mcaeuwuwv 3 com: mocoavom cowuommm .m muswwm an :wox .zolwloooz e o 5.3: [a z :oozo 0.2 n . :o .20 :o :08 o: _N o.. o xo;. 0 o 2.1a Ne t5.09 QIDOZ O 4 (mPZN nxo nxo 064 0.: . £6: 064 .20 o o .208 an: v 5on198: o... o m \ :wo: xooxo o o 084 38 10min nzo o: . 10084 n .... o o .68 o: m .20.“? :82 \ — :oouq . _ eaooxu o 72 CHOOAC AcOCH -42 -60 ------------- _ o ’60 -42 101<——143<——203f ACOCHZI (Ii-0M2 Jzezigoz—emz—aloo I —————————— . ~42 {ACOIH I60 ; CH3 I K -60 -42 -60 276—92I6——>I74——>II4 I00 101 .1 >- E: 1:12 50‘ 143' 203 1‘3 "4 142 202 Z! l I '50 I74 "' 0 I ........ 11.11 ‘ 1 1 - - l.c--3-- .I---- “"11 .- .. -.-.--.... ‘ 11-... Z 100 125 150 175 200 o — 100'} ‘5’ .06 X ; r" <1 d 50— 262 276 0 1.--: Il...1.-i--. 11....“ -. ...... ‘ - 4111.. I1.. 4 ‘ 225 250 275 300 m/z Figure 6. E.i. mass spectrum of compound 5. Compound 5 is the per-Q- acetylated branched-chain-deoxy-alditol produced from the aceryl residue of the ‘heptasaccharide: by the sequence of reactions outlined in Figure 5. Secondary fragment-ions formed by eliminations of ketene (42 mass units) or acetic acid (60 mass units) are indicated. 73 The first step of the sequencing procedure was to reduce the heptasaccharide with NaBDA to form the corresponding heptasaccharide- alditol, which was then per-Q-deuteriomethylated . As mentioned above, the acetyl reside was partially degraded during per-gfdeuteriomethyla- tion of the heptasaccharide-alditol, generating a mixture of per-Q- deuteriomethylated Oligosaccharides and Oligosaccharide-alditols. This mixture was partially resolved by l.c. on a twpont Zorbax ODS column. The elution was monitored by chemical ionization (c.i.)-m.s. of 3% of the l.c. effluent (Figure 7). The remaining 97% of the l.c. effluent was collected as fractions (0.5 mL). The material that eluted with a retention time of 11.3 min (shaded region, Figure 7) was found to be the per-Q-deuteriomethyl derivative of the intact heptasaccharide-alditol. This was established by the results of hydrolysis, reduction, and acetylation of aliquots of the l.c. frac- tions comprising the peak (data not shown), and by f.a.b.-m.s., as will be discussed later. The fractions containing the per-Q-deuteriomethyl- ated ‘heptasaccharide-alditol. were conbined_ and evaporated In) dryness under a stream of filtered air, and saved for subsequent f.a.b.-m.s. and n.m.r. analyses. Most features of the sequence of glycosyl residues in the heptasaccharide were elucidated by forming and characterizing overlapping per-Q-alkylated Oligosaccharide-alditol fragments of the heptasaccharide. The per-Q-alkylated Oligosaccharide-alditol fragments ‘were produced by partial acid hydrolysis, reduction, and per-Q-deuterio- ethylation of the mixture of per-Q-deuteriomethylated Oligosaccharides 23nd Oligosaccharide-alditols that resulted from degradation of the ,aceryl residue of the heptasaccharide (material with retention times TOTAL ION INTENSITY Figure 7. 74 I 1 J J 5 6 7 8 9 10 II I2 l3 RETENTION TIME (min) Reverse-phase l.c. elution profile of the per-Q-deu- teriomethylated Oligosaccharides and Oligosaccharide- alditols generated by degradation of the aceryl residue during per-Q-deuteriomethylation of the heptasaccharide- alditol. The profile is the c.i.-m.s. total-ion response of 3Z of the effluent from the l.c. column introduced directly into the source of the mass spectrometer. The mass spectrometer scanned from.mlg_150 to 1000 once every 3 s. 75 between 4 and 10 min, Figure 7). These experiments are described below. The eluent fractions (M? the l.c. column (Figure 7) collected between 4 and 10 min were combined and evaporated to dryness. Prelimi- nary analyses of this material were carried out as described10 to deter- mine conditions for further hydrolysis to form di-, tri-, and tetrasac- charide fragments. Treatment of the per-Q-deuteriomethylated Oligosac- charides and Oligosaccharide-alditols in the 4 to 10 min eluent with 88% formic acid for 2 h at 500 was found to yield useful mixture of frag- ments. The partially-O-deuteriomethylated Oligosaccharide and Oligosac- charide-alditol fragments generated by this partial hydrolysis were reduced with NaBD4 and then per-Q-deuterioethylated, yielding a mixture of partially-Q-deuteriomethylated, partially-Q-deuterioethylated oligo- saccharide-alditols. The Q-deuterioethyl groups of these derivatives mark the points of attachment of other residues in the intact hepta— saccharide-alditol“). The term "per-_O_-alkylated Oligosaccharide- alditols" will be used to refer to components of this mixture in the remainder of this paper. An aliquot (200 ug) of the mixture of per-Q-alkylated Oligosaccha- ride-alditols was resolved and analyzed by g.l.c.-m.s. (Figure 8). The structures of the per-Q-alkylated Oligosaccharide-alditol fragments were elucidated from diagnostic ions in their e.i. mass spectra. This is il- lustrated in Figure 9, in which the e.i. mass spectrum of per-9- alkylated tetrasaccharide-alditol [f] is shown and the A- and J-series of fragment ions, which are typical of these moleculeszz, are indi- cated. The Ar and J-series of ions, in conjunction with the glycosyl- linkage composition of the intact heptasaccharide, established, with one Figure RELAHVEIONINTENSHY p.- 76 [(1 1°] [61'] 51 “a [d] \ 1:1 l I l J I 1 8 9 IO IN I2 I3 I4 15 I6 RETENTION TIME (min) — b 8. G.1.c.-m.s. elution profile of the per-O-alkylated Oligosaccharide-alditols derived from the heptaSaccharide. The per-Q-alkylated Oligosaccharide-alditols were produced by partial hydrolysis, reductions, and per-9- deuterioethylation of the Oligosaccharide fragments that resulted from partial degradation of the aceryl residue during per-Q-deuteriomethylation of the heptasaccharide- alditol. the mass spectrometer scanned from mlg_100 to 1000 every 1.5 s. Each per-Q-alkylated fragment structurally characterized has been assigned a letter to show where it eluted from the g.l.c. column (Tables II, III, and IV and Figure 9). 77 Figure 9. E.i. mass spectrum of per-Q-alkylated tetrasaccharide-alditol [f]. The g.l.c.-m.s. conditions are described in the Figure 8 legend. 78 -55 -35 294(ch3) .__ 329 (ch2)<—— 364 (mm) [692=och.] _35 62910b¢J2I 163(cA2)<-—1981cA1) Cnooczoa CH3 03cc 0 03CO 0 \ H? 0 0 0003 CH; 03COCH \ 000; 1 O CO 0 O-L———CH 3 000, 0003 \) "°°c2°5 OCH, H20C03 c b a b' I ~ I I -35 I \6261055 .12) \ 460 (ch .12) \ 195(bA.)——>166(5A21 [689=ODUJJ [523:65Jd 100- 198 x 5 x 20 > 1" r* P- 160 364 ‘7,- 504 195 2: 163 u: an 1- 294 l 5 Li .1 .1 ..l L . . - . 0 1 l r ”m_— mocfinmum 6cm Hamooau dzzuoELmrN ozu mo ucwazumuum mo ucfioa 6:99 .womuu .u.H.m no ucoewmuw mo cowumooH pow m ouamum com a 6: a 0mm b fiancee. Anaema. Asmvme. Aoo.cma_ AmvaNn Aaveem A.VmNm Amvoee _e. aoamoe.nuHumawuv N\B ucm mum mpwumzoommowaao mcoHlucwawmum Hmuuuwam new: uomnanlcouuuon l I. monHmmo wAOHHQqA4uwmno 6005609 no mcofiufimom Humoomau .. moaHmmN asmoeoH m4m<9 86 l98 364 754 948 ”28 RhfiArfiGfl-Acfishfi-Api I95 MeFuc (I G 193 33 195 G x” W 3. G r‘ ‘ H'Hhii'. ' “’"' ‘ - F ' * A ~ - s‘ H i" “KN"!M‘IJK H 1: 2m 0| 6 G i ! ‘H H!” m” w" . I“ ‘ ll” “mil; M H ”MIN: HM; h II|H ., m; H Ii.“ WWW ': . NIH 0 IO '. w A - A‘ 4 L._. A _— i A - V Y m m 1371 lHoan' M4... 1!!) Figure 10. Positive-ion f.a.b. mass spectrum of the per-_Q-deu- \ teriomethylated heptasaccharide-alditol. The signals labeled G represent glycerol polymers generated from the glycerol matrix on the f.a.b. target. The per-Q- deuteriomethylated heptasaccharide-alditol was isolated by l.c. (retention time of 11.3 min, Figure 7). 87 or pyranoid (6-membered) rings. The apiosyl and aceryl residues of the heptasaccharide can form only furanoid rings. Methylation analysis of the heptasaccharide (Table I) established the pyranoid ring-forms of the 2-_(_)_-methyl fucosyl, arabinosyl, and the two rhamnosyl residues of the heptasaccharide because the partiallngfinethylated alditol acetates derived from these residues had methyl groups at 0-4. The partiallngfmethylated alditol acetate derived from the galac- tosyl residue of the heptasaccharide was 1,2,4,5-tetraj97acetyl-3,6-di- Efmethyl galactitol (Table I). This derivative could have arisen from a 2,4-linked galactopyranosyl residue or from a 2,5-linked galactofurano- syl residue. This ambiguity was resolved in the experiment that revealed the point of attachment of the arabinosyl and ngfmethyl fucosyl residues 11) the galactosyl residue. In that experiment, the heptasaccharide was partially hydrolyzed, reduced, and perfigf methylated. The resulting fragments were then completely hydrolyzed, reduced, and acetylated. One of the partially _g-methylated alditol acetates resulting from this series of reactions was 1,2,5-trijgfacetyl- 3,4,6-tri-Q-methyl galactitol (Table IV). The methyl group at 0-4 of this derivative established that the galactosyl residue in the heptasac- charide was in the pyranoid ring-form23. Determination of the anomeric configurations of the glycosyl residues of the lmmmasaccharide. - The anomeric configurations of the glycosyl residues of the heptasaccharide were determined by lH-n.m.r. 1 The H-n.m.r. spectrum of the per-_g-deuteriomethylated heptasaccharide alditol (Figure 11) contained resonances from the six anomeric protons. Four of those resonances were assigned to specific glycosyl 1 residues of the heptasaccharide with the assistance of the H-n.m.r. 88 III I! I III II I I l l 1 l l I l l 5.5 5.0 4.5 4.0 p.p.nL Figure 11. Spectral region from 3.75 to 5.75 p.p.m. of the 1H—n.m.r. spectrum of the perjgfdeuteriomethylated heptasaccharide- alditol in CDCl . Chsnical shifts were assessed relative to internal hexadeuterio-—acetone (6 2.04). The resonances arising from the six anomeric protons have been designated I-IV (Table V). The per—O—deuteriomethylated hepta- saccharide-alditol was isolated-by l.c. (retention time of 11.3 min, Figure 7). 89 spectra of perjgfalkylated Oligosaccharide-alditol fragments [a], [d'], and [f] (Table V). These fragments were isolated by l.c. of an aliquot (2 mg) of the mixture of perfgfalkylated Oligosaccharide-alditols, and were detected by g.l.c.-m.s.(e.i.) of aliquots of the column fractions. The 1H-n.m.r. spectrum of perfigfalkylated disaccharide-alditol [a] contained an anomeric resonance, from the anomeric proton of the 3- linked rhamnosyl residue, at 6 4.70 with a 1 Hz coupling constant (Table V). A small (0-2 Hz) coupling constant is typical of a rhamnopyranosyl anomeric resonanceza, and the chemical shift indicated that the anomeric proton of the 3-linked rhamnopyranosyl residue was axial. There- fore, the 3-linked rhamnopyranosyl residue had the B anomeric configura- tion. It should be noted that the chemical shift of the anomeric-proton resonance of a particular glycosyl residue may differ when the residue is situated :hi different Oligosaccharide-alditols. For example, the chemical shift of the B—rhamnosyl residue was <5 4.64 in the per-Q- deuteriomethylated heptasaccharide-alditol and 5 ‘4.70 in fragment [a]. This was as expected, because the anomeric proton was not in an identical chemical environment in both Oligosaccharide-alditols. The 1H-n.m.r. spectrum of per-_Q-alkylated trisaccharide-alditol [d’] contained a resonance at 6 5.34, with a coupling constant of less than 1 Hz, and another resonance at 6 4.61 with a 7 Hz coupling constant. These resonances arose from the anomeric protons of the terminal rhamnOpyranosyl residue and the arabinOpyranosyl residue. The resonance with a 7 Hz coupling constant could not have arisen from a 9C) .HHH mam HH moanmh wmmu .qo.N e um wcouwumofiumuswcmucwo Hmcumqu cu w>wumfiwu vwmmmwm< ._~ musmam mums p Aumxcasunv assocawsmuoum am a ch.s _mm s: a sc.< ‘ H> axmocapmu n: n so.s ..v_ asuouH Aamcsaumev assocawcmuouo a: _ v m~.m .um u: a v so.n HHH n: _ v «Wm 73 imouamAifioFmATfa a: 5 Gd .3 a: s 83 Z ameuumHmunoua a: a mm.m H . A.a.n.nv . A.a.o.ov N an nuwfinm Hanuaucu oucwaampm N - numwzm Hmouausu «wocncomux Houwvamuwvaumnuummowuao HeuacamlwkumcuummMuaos ucoacwumm< kumfimxamnmruom wounfixzuwso«uou:ovnMVuum .l mAOPHQqMAHmmo ho 02¢ AOPHQA mqm<8 91 rhamnopyranosyl residue because H-1 and H-2 of both o- and B-rhamnopyra- nosyl residues are in gauche orientation and exhibit coupling constants of less than 2 H224. Therefore, the resonance at 6 4.61 with a 7 Hz coupling constant was assigned to the 2-linked arabinopyranosyl residue. The chemical shift and coupling constant of this resonance indicated that the 2—linked arabinopyranosyl residue had the a anomeric configuration. The resonance at 6 5.34 with a coupling constant of less than 1 Hz must have arisen from the terminal rhamnopyranosyl residue. The chemical shift indicated that the anomeric proton of this residue was equatorial and that the residue had the a anomeric configu- ration. The 1H—n.m.r. spectrum of perfgfalkylated tetrasaccharide-alditol [f] contained resonances arising from the anomeric protons of the ngf methyl fucopyranosyl residue, the terminal rhamnopyranosyl residue, and the arabinopyranosyl residue. Two of the resonances corresponded to those identified in the spectrum of perfgfalkylated trisaccharide aldi- tol [d']. Therefore, the third resonance (6 5.21, with a 4 Hz coupling constant) must have arisen from the anomeric proton of the 2-_Q-methyl fucopyranosyl residue. The chemical shift and coupling constant of this resonance indicated that the 2-_Q-methyl fuc0pyranosyl residue had the a anomeric configurationzs. The assigmnents of the 2-_Q-methyl fucopyranosyl and the arabino- pyranosyl anomeric configurations were confirmed by a nuclear Overhauser 26,27 effect (n.0.e.) difference n.m.r. experiment that was performed with underivatized RG-II heptasaccharide in D20. The anomeric region of l the H—n.m.r. spectrum of the underivatized heptasaccharide (trace B, Figure 12) was more complex than the same region of the spectrum of the 92 HOD (— ll 4?” :1 l' || H H H 1| 'l t_____;:::: 1 l 1 l 1 l 5.5 5.0 4.5 p.p.m. Figure 12. lH-N.m.r.—n.0.e. difference spectrum of the underivatized heptasaccharide in D 0. Trace A: spectral region from 4.25 to 5.75 p.p.m. of an n.0.e. difference spectrmn obtained with presaturation of the methoxy protons of the 2-0-methy1 fucosyl residue ("a", Figure 4). Trace B: spectral region from 4.25 to 5.75 p.p.m. of the normal Fourier transform spectrum. 93 perjgfdeuteriomethylated heptasaccharide-alditol. There were two causes for most of the increased complexity of the former spectrum: the hepta- saccharide was not reduced to the heptasaccharide-alditol; therefore, some of the heptasaccharide molecules had a-apiose, while others had 8— apiose, at their reducing termini. Some of the heptasaccharide mole- cules also contained Q-acetyl esters. The inductive effect of an _Q- acetyl ester can cause the resonance of a non-anomeric proton to appear in the anomeric region of the spectrum. The n.0.e. difference n.m.r. experiment was designed to discriminate between the anomeric protons of the ngrmethyl fucosyl and arabinosyl residues by taking advantage of the_gnmethy1 group on C-2 of the fucosyl residue. The experiment entailed presaturating the methoxy protons of the ZjQfinethyl fucosyl residue (Figure 4, 5 3.5) of the un- derivatized heptasaccharide. Under the conditions of the experflment, any protons that were within approximately 4 X of the presaturated protons would experience a slight (I-SZ) enhancement in signal intensity. The n.0.e. enhancement is extremely distance—dependent, diminishing with increasing distance as R-627. Trace A in Figure 12 shows the anomeric region of the n.0.e. difference spectrum. The n.0.e. difference spectrum was generated by obtaining free induction decays with and without presaturation of the methoxy protons of the Zfomethyl fucosyl residue and then performing a Fourier transformation on the dif- ference between the free induction decays. The resonance at 6 5.3 with J of 4 Hz, which was the only anomeric resonance with significant 1,2 n.0.e. enhancement, was attributed to the anomeric proton of the 2-0~ methyl fucosyl residue. 'nua result confirmed the assignment of this resonance made on the basis of 1H-n.m.r. analysis of fragments [d'] and [f]. 94 Two resonances of the lH--n.m.r. spectrum of the per-_g- deuteriomethylated heptasaccharide alditol (5 5.53, 4 Hz coupling constant 6 4.87, 2 Hz coupling constant) were not assigned by 1H-n.m.r. analysis of perfigfalkylated alditols [a], [d’] and [f]. These resonances were attributed to the anomeric protons of the 2,4-linked galactopyranosyl residue and the 2-linked acerofuranosyl residue because these were the only glycosyl residues whose anomeric resonances were not identified in the above experiments. A. 2 Hz coupling constant is smaller than that usually observed with either anomeric configuration ofa galactopyranosyl residue, but is typical of that observed when H91 and H—2 are trans in a furanoid ring25’28. Therefore, the resonance at 6 4.87 was attributed to the 2-linked acerofuranosyl residue in the 8- configuration. The resonance at 6 5.53 with a 4 Hz coupling constant was in a typical range of chemical shift, with a coupling constant typical for a galactosyl residue with the a anomeric configuration. This resonance was assigned to a 2,4-linked-o-galactosy1 residue. Thus, the anomeric configuration of each glycosyl residue of the heptasaccharide was determined. 95 GENERAL DISCUSSION The heptasaccharide released from RG-II has the structure shown in the abstract figure. All aspects (H? the primary structure have been elucidated except the points of attachment of the gracetyl groups. The heptasaccharide contains six different glycosyl residues, including aceric acid, a hitherto unobserved glycosyl residue. The complexity of this heptasaccharide is tummecedented among structures that have been elucidated for primary cell wall constituents; this finding reinforces our belief that RG-II is a well-defined, structurally complex polysac- charide. The glycosyl-linkage composition of RG-II is compared to that of the heptasaccharide in Table VI. The value given in Table VI for the 2- 1inked aceryl residue is an estimate because this residue cannot be quantitated reliably after 1nethylation. The glycosyl residues that comprise the heptasaccharide are present in approximately equal numbers in intact RG-II, and, with the exception of the 4-linked galactosyluronic acid residues, are the most abundant glycosyl residues of RC-II. The data in Table VI also reveal that a molecule of RG-II probably contains more than one heptasaccharide unit. Three heptasaccharide units would account for approximately 30%, and four heptasaccharide units for more than 40% of the glycosyl residues of RG-II. Release of the heptasaccharide(s) does not change significantly the P-10 elution volume of the remainder of the molecule (unpublished results). This suggests that the radius of gyration of RG- 11 is not changed significantly by release of the heptasaccharide, and implies that the heptasaccharide units may be side chains on a 4-linked galactosyluronic acid-rich backbone. RG-II is believed to be linked to TABLE VI COMPARISON OF THE GLYCOSYL-LINKAGE COMPOSITION OF INTACT RG-II TO THAT OF THE HEPTASACCHARIDE Glycosyl residue Linkage Number per Number per RG—II moleculea heptasaccharide molecule Galactosyluronic acid 4-linked 12 Terminal 2 3,4-linked 1 Rhamnosyl Terminal 4 1 3-linked 4 l 2,4-linked 1 3,4-linked 1 2,3,4-linked 1 Arabinosyl 2-1inked(pyranose) 4 1 Terminal (furanose) 2 Galactosyl 2,4-linked 5 1 Terminal 2 3-linked 1 Apiosyl 31-linked 4 1 Aceryl 2-linked (4)b 1 ZfiQfMethyl Fucosyl Terminal 3 l Fucosyl 3-linked 3,4-linked 1 Zingethyl Xylosyl Terminal 2 Glucosyluronic Acid 2-linked 2 Glucosyl 4-linked 2 TOTAL 60 a The values given for intact RG-II are calculated, using the data of Darvill 1 .ggugl. bThe value given for the 2-1inked aceryl residue is an estimate because this residue cannot be quantitated reliably after methylation. 97 other cell wall polysaccharides by such a 4-linked galactosyluronic acid chainl. The structure of the heptasaccharide provides the first informa- tion about the locations of the Zfomethyl fucosyl and apiosyl residues in the primary cell wall of dicots. Methyl fucose has been long recog— 29’30, and was nized as a minor constituent of pectic polysaccharides identified as a component of RG-III. The heptasaccharide is the first plant cell wall Oligosaccharide characterized that contains a 2j9fmethyl fucosyl residue. Apiose-containing polysaccharides from monocots have been partially characterized. The best-characterized of these is the apioga- lacturonan of 321133132, which consists of a chain of 4-linked c-D— galactosyluronic acid residues, with apiobiose units attached to 0-2 or 31,32. 0—3 of some of the galactosyluronic acid residues Zosterin, an apiose-containing polysaccharide isolated from Zosteraceae33, is struc- turally more complex than the Lgmna_apiogalacturonan, but is degraded by pectinase In) a molecule very similar to the lfmgg_apiogalacturonanBa. The apiosyl residue in the RG-II heptasaccharide occurs in a chain with other neutral glycosyl residues, the first such example. It is possible that the apiosyl residues of RG-II are gycosidically linked to 4-linked galactosyluronic acid residues. If this is true, then RG-II could be structurally related to zosterin and to apiogalacturonans. The heptasaccharide molecule has many hydrophobic functional groups: four of the glycosyl residues are deoxy sugars and contain methyl groups instead of hydroxymethyl groups at C-6; the arabinosyl residue is in the pyranoid ring-form and contains no hydroxymethyl group; and the heptasaccharide has one endogenous Efmethyl group and at 98 least two endogenous Qfacetyl esters. The hydrophobicity and structural complexity of the heptasaccharide are important because recent evidence indicates that the interactions between carbohydrates and their protein 35’36. These properties of receptors are governed by hydrophobic bonding the heptasaccharide are made more significant by recent reports that Oligosaccharide fragments of the cell walls function_ as regulatory molecules in plant—pathogen interactions and in plant growth and development37’38. 99 REFERENCES 1 10 11 12 13 14 15 16 17 18 A.G. DARVILL, M. MCNEIL, AND P. ALBERSHEIM, Plant Physiol., 62 (1978) 418-422. P. ALBERSHEIM, D. J. NEVINS, P. D. ENGLISH, AND A. KARR, Carbohydr. Res., 5 (1967) 340-345. A. LIPTAK, J. IMRE, AND P. NANASI, Carbohydr. Res., 92 (1981) 154- 156. s.-1. HAKOMORI, J. Biochem., 55 (1964) 205-208. H. E. CONRAD, Methods Carbohydr. Chem., 6 (1972) 361-364. T. J. WAEGHE, Ph. D. Dissertation, 1982, University of Colorado, Boulder, Colorado. P. XMAN, M. MCNEIL, L.-E. FRANZEN, A. G. DARVILL, AND P. ALBERSHEIM, Carbohydr. Res., 95 (1981) 263-282. T. M. JONES AND P. ALBERSHEIM, Plant PMsiol” 49 (1972) 926-936. M. MCNEIL, A. G. DARVILL, P. XMAN, L.-E. FRANZEN, AND P. ALBERSHEIM, in V. GINSBERG (Ed.), Methods 31 Enzymology, Vol. 83, Academic Press, New York, 1982, pp. 3-45. B. S. VALENT, A. G. DARVILL, M. MCNEIL, B. K. ROBERTSEN, AND P. ALBERSHEIM, Carbohydr. Res., 79 (1980) 165-192. Z. DISCHE, Methods Carbohydr. Chem., 1 (1962) 477-512. N. BLUMENKRANTZ AND G. ASBOE-HANSEN, Anal. Biochem., 54 (1973) 484- 489. G. M. BEBAULT AND G. G. s. DUTTON, Carbohydr. Res., 64 (1978) 199- 213. A. N. DE BELDER AND B NORRMAN, Carbohydr. Res., 8 (1968) 1-6. C. J. GERWIG, J. P. KAMERLING, AND J. F. G. VLIECENTHART, Carbohydr. Res., 62 (1978) 349-3570 G. J. GERWIG, J. P. KAMERLING, AND J. F. G. VLIEGENTHART, Carbohydr. Res., 77 (1979) 1-7. K. LEONTEIN, B. LINDBERG, AND J. LONNGREN, Carbohydr. Res., 62 (1978) 359-362. R. R. WATSON AND N. S. ORENSTEIN, Advan. Carbohydr. Chem. Biochem., 31 (1975) 135-184. 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 100 D. P. SWEET, R. H. SHAPIRO, AND P. ALBERSHEIM, Carbohydr. Res., 40 (1975) 217-225. H. BJORNDAL, C. C. HELLERQVIST, B. LINDBERG, AND S. SVENSSON, Angew. R. U. LEMIEUX AND M. L. WOLFROM, Advan. Carbohydr. Chem. 3(1948) 337-384. N. K. KOCHETKOV AND 0. S. CHISOV, Advan. Carbohydr. Chem., 21 (1966) 39—930 A. G. DARVILL, M. MCNEIL, AND P. ALBERSHEIM, Carbohydr. Res., 86 (1980) 309-315. G. M. BEBAULT, G. G. S. DUTTON, N. A. FUNNELL, AND K. L. MACKIE, Carbohydr. Res., 63 (1978) 183-192. D. R. BUNDLE AND R. U. LEMIEUX, Methods in Carbohydr. Chem., 7 (1976) 79—86. G. WAGNER AND K. WUTHRICH, l. Magnetic Resonance, 33 (1979) 675-680. A. KUMAR, R. R. ERNST, AND K. WUTHRICH, Biochem. Biophjs. Res. Commun., 95 (1980) 1-6. J. D. STEVENS AND H. G. FLETCHER, JR., _J_. o g. Chem., 33 (1968) 1799. G. o. ASPINALL AND A. CANAS-RODRIGUEZ, :1. Chem. Soc. (1958) 4020— 4027. ”“— G. O. ASPINALL AND R. S. FANSHAWE, _J_. Chem. Soc. (1961) 4215-4225. D. A. HART AND P. K. KINDEL, Biochem. J” 116 (1970) 569-579. D. A. HART AND P. K. KINDEL, Biochemistry, 9 (1970) 2190-2196. R. G. OVODOVA, V. E. VASKOWSKY, AND Y. S. OVODOV, Carbohydr. Res., 6 (1968) 328-332. Y. S. OVODOV, R. G. OVODOVA, 0. D. BONDARENKO, AND I. N. KRASIKOVA, Carbohydr. Res., 18 (1971) 311-318. E. A. KABAT, Structural Concepts in Immunology and Immunochemistry, 2nd. ed., Holt, Rinehart, and Winston, New York, 1976, pp. 156-161. O. HINDSGAUL, T. NORBERG, J. LE PENDU, AND R. U. LEMIEUX, Carbohydr. Res., 109 (1982) 109-142. M. G. HAHN, A. G. DARVILL, AND P. ALBERSHEIM, Plant Physiol., 68 (1981) 1161-1169. 101 38 P. ALBERSHEIM, A. c. DARVILL, M. MCNEIL, B. s. VALENT, 11. G. HAHN, G. LYON, J. K. SHARP, 1%; E. DESJARDINS, M. W.’SPELLMAN, L. M. ROSS, B. K. ROBERTSEN, P. AMAN, AND L.-E. FRANZEN, Pure and Applied Chemistry, 53 (1981) 79-88. APPENDIX 102 APPENDIX Crystal Structure of 3-C-Carboxy-5-Deoxy-L-Xylono-1,4-Lactone This appendix contains the results of the X-ray crystallographic analysis of 3-C-carboxy-5-deoxy-L-xylono-1,4-lactone (compound 6, Chapter I). Experimental details were included in the "Experimental" section of Chapter I. TABLE AI 103 FRACTIONAL ATOMIC COORDINATES AND ANISOTROPIC THERMAL PARAMETERS (Ueq) FOR THE NONHYDROGEN ATOMS 0F 3-C-CARBOXY-5-DE0XY-L-XYLON0-l,4-LACT0NE'H20 b Atoma Fractional Atomic Coordinates Ueq _12 _X. .3 C-1 175(10) 5385(13) 702(10) 29(4) C-2 1610(8) 6378(14) 2031(9) 28(4) C-3 3271(8) 5235(14) 1952(9) 25(4) C-4 2868(8) 4590(13) -108(9) 25(4) 0-5 901(6) 4597(12) -630(6) 34(4) 0-1 -1376(6) 5281(13) 689(7) 40(3) 0-2 1203(6) 6483(12) 3826(6) 34(3) 0-3 4865(5) 6229(13) 2340(6) 28(3) C-31 3377(8) 3649(14) 3270(9) 26(4) 0-32 4649(5) 3353(12) 4513(7) 38(3) 0-31 1901(6) 2663 13) 2930(7) 37(3) C-41 3579(10) 5814(15) -1403(10) 44(5) O-lw -1794(5) 4944(13) 4768(6) 35(3) a The numbering system is shown in Figure 9, Chapter I. Estimated standard significant digit. deviations, CValues are A2 x 104. bValues are x104. in parentheses, refer to the least- Estimated standard deviations, in parentheses, refer to the least-significant digit. 104 TABLE AII FRACTIONAL ATOMIC COORDINATES AND ISOTROPIC THERMAL PARAMETERS (U180) FOR THE HYDROGEN ATOMS OF 3-C-CARBOXY-5-DEOXY-L-XYLONO-l,4-LACT0NE'H20 b Atoma Fractional Atomic Coordinates Uisoc _x_ _L z H-2 182(7) 760(10) 135(8) 4(2) H-zo 190(7) 746(10) 427(8) 12(2) H-3o 512(7) 641(10) 357(8) 5(2) H-31o 195(8) 166(10) 339(8) 9(2) H—4 332(7) 323(9) —20(8) 1(2) H-41a 305(7) 543(10) -268(8) 3(2) H-41b 487(8) 587(10) -110(8) 4(2) H-41c 358(7) 720(10) -145(8) 9(2) H-low -106(7) 536(11) 392(8) 7(2) H-Zow —291(7) 528(10) 445(8) 12(2) a The numbering systan is shown in Figure 9, Chapter 15 bValues are x103. Estimated standard deviations, in parentheses, refer to the least- significant digit. cValues are A2 x 103. Estimated standard deviations, in parentheses, refer to the least-significant digit. 105 TABLE AIII BOND LENGTHS OF 3-C-CARBOXY-S-DEOXY-L-XYLONO-l,4-LACTONE’H20 Bonda Lengthb Bonda Lengthb C-1--C-2 1.508(9) C-1--0-5 1.357(9) C-1--0-1 1.195(9) C-2--C-3 1.545(11) C-2--0-2 1.422(9) C-3--C-4 1.560(10) C-3--0-3 1.409(10) C-3--C-31 1.517(13) C-4--0-5 1.483(7) C-4--C-4l 1.499(12) C-31--0-31 1.331(10) C-31--0-32 1.215(7) C-2--H-2 1.06(7) C-4--H-4 1.08(7) 0-2--H-20 .92(6) 0-3--H-30 .90(6) 0-31--H310 .81(7) C-41--H-41a .99(6) C-41--H-41b .98(6) C-41--H-4lc 1.03(7) 0-1w--H-10w .98(6) 0-1w-H-20w .88(5) a The numbering system is shown in Figure 9, Chapter I. bValues are in A. Estimated standard deviations, given in parentheses, refer to the least-significant digit. 106 TABLE AIV BOND ANGLES FOR 3-C-CARBOXY-5-DEOXY-L-XYLON0—l,4-LACTONE'H20 Anglea Degreesb Anglea Degreesb 0-5--C-1--C—2 108.9(6) 0—1--C-1-—C—2 129.3(7) 0-1--C-1--0-5 121.8(6) C—3--C—2--C—1 102.0(7) 0-2--C-2--C—1 111.0(6) 0-2--C-2--C-3 114.1(6) C-4--C-3—-C-2 101.8(5) 0-3--C—3--C-2 113.5(8) 0-3—-C-3--c—4 110.4(6) C-31--C-3--C-2 109.7(6) C—31-—C-3—-C-4 111.0(8) C-3l--C3--0-3 110.2(5) 0-5--C-4--C-3 103.8(5) C—41--C—4-—C-3 114.0(7) C-41--C-4--0-5 108.6(6) C-4--0-5--C-1 111.4(5) 0—31——C-31--C-3 112.0(6) 0-32--C-31--C-3 123.5(8) 0-32--C—31--0-31 124.5(9) H-2--C—2--C-3 104(3) H-2--C-2--C-1 106(3) H-4--C-4--C-3 110(3) H—2--C-2--0-2 118(3) H—4--C—4--C-41 111(3) H-4--C-4--0-5 108(3) H-30--0-3--C-3 106(4) H-2o-—0-2-—C-2 99(4) H—41a--C-41--C-4 107(4) H-310--0-31--C-31 117(4) H—41b--C-41--H-4la 115(5) H-41b--C-41--C-4 111(4) H-41c--C-41--H-41a 105(5) H-41c--C-41--C-4 129(4) H-20w--0-1w--H-10w 114(5) H-41c--C-41--H-41b 88(5) a The numbering system is shown in Figure 9, Chapter I. bEstimated standard deviations, given in parentheses, refer to the least-significant digit. 107 «N« 9N« N N NI «n Nm « N m an mN « v cl «mN mvN « « «I mm 0m a c o N«« cs« N « c an ac « N N Na we « v NI N«n MNM « « NI 36 NN a Q m mo« 00« N « m Ns« ««« « N 9 no ea « n N NN« eN« « « ml on an s 0 ¢ mo« oo« N « v NN« os« « N «I am 0N « n o o«« oa« « « cl «m an a v m NN« 0N« N « n mm Nv « N ml om am « m m nm« on« « « ml «N« NN« a c N to no N « N mm no « N ml Ne 0v « n v mm mm « « cl SON NMN 9 0 « m§« ««« N « « ow «m « o v meN th « n n no «a « « al «N« mN« 8 ¢ 9 mo« NN« N « 9 No on « o N N«« o«« « n N No Nm « « GI NN oN a n N va MMN N « «I No on « o « nnN NMN « n « am so « a a fin no a m a 03 mm N « NI MN No « a a amN mmN « N a Na 00 « 9 N co co 3 n m mm «m N « ml Nm Nn « o «I mNN N«N « n «I on av « 9 m mnN «mN s m e «a «a N « VI no No « o NI naN NaN « n NI vv nv « s v No 93« s m m N«« an« N « ml 99 cm « 0 ml mN« om« « m ml amN mvN « a n vm« ma« 5 m N mN «N N « OI Na Na « 0 Cl Nn Ne « n NI nv Nv « a N an oeN s n « «N« o«« N « NI an «o « 9 nl N«« c«« « N a ma Na « o « «0 SN 8 N o No at N « al on so « 9 cl oN« NN« « N m Nov cmv « a «I an 9e 3 N N 00 at N s o no we « n n mv« ov« « N v ma em « o NI VN ON a N o nN« oN« N s 0 mm« va« « m M No mm « N n mmN NwN « 3 ml NN «N G N m NN vN N s m on «a « n « nan omN « N N N«« NN« « 6 cl «a «a a N c Nm« nn« N s n on Nm « m a Nnn ONM « N « N«« NN« « 5 ml NN« so« a N n 9v« nv« N s N mm mm « m «I mNn mom « N a so «9 « 3 ml 0N NN s N N «No one N s « «N 0N « m NI SNN msN « N «I «n« NN« « & NI oSN N«N a N « o«« v«« N s s vm« nm« « m nl omN NmN « N NI mm mm a a n va emQ a N a «m Nm N s «l ov «m « m vl mm mm « N nl no on s a. « on n@ s « a on on N @ NI an no « m ml «a NN « N vl «o «a s a a mo Nm 9 « N o«N «NN N s ml 1o NN « n cl N«« m«« « N ml we we a N n mm Nm a « 0 Ne an N 5 0| so ca « v a an on « N cl am an a c m «N «N a « m cm «m N a ml as« vs« « v n «v N¢ « N NI mn« 0n« & o m nNN «MN 3 « v NN« «n« N s NI co co « v v «o co « « N cm« on« s o a so« no« 9 « M Na an N 3 ml NN« «n« « v n «o« «o« « « a an we N m N NON «ON 9 « N to NS« « a « sn« NN« « 4 N «m Nm « « a No em a m o N«n can 9 « « Nv an « o a co mo « v « N«« N«« « « e No no a n n om« mm« a a w an vv « m «I «N« mN« « v s NN« mm« « « m co no a n 0 nc« NV« 9 a m me No « a NI nN 0N « v «I nn 0N « « N «M Na 9 m n NN« NN« a 9 ¢ Nm mm « N m «v «v « v «I N«N msN « « « mnN @MN N m N mNn mNn a a n «o «o « N v v«« v«« « v nl Nam «NM « « s N«« o«« a m « Nm Nm 3 a N uua« ous« a z : uma« ous« J z : uus« ous« J x : ums« oua« J x : Uua« oua« J z : oNz.MZOPU<.«I.~.«IOZOEIAINkOmn—Imltnommdnulnvlm mom $5933 EUDMHm Ewan—bog 92¢ Emmmmo >< mqm4 unm