sign in sign up
<<
Home , Isomaltose, Mannans, Trisaccharide

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS 121, 88-95 (1967)

Protein- Interaction

V. Further Inhibition Studies Directed toward Defining the Stereochemical Requirements of the Reactive Sites of Concanavalin A

E. E. SMITH’ AND I. J. GOLDSTEIN2

Department of Biological Chemistry, The University of Michigan, Ann Arbor, Michigan, and the Department of Biochemistry, School of Medicine, State University of New York at Bufalo, Bufalo, New York Received January 16, 1967; accepted February 2, 1967

Inhibition studies on the concanavalin A- interaction have been extended using as inhibitors the members of the -, -, and methyl cy-isomal- toside series, amino and N-acetyl amino , and a number of containing more than one linkage type. The interaction of the protein at the terminal chain ends of the a-linked and has been essentially con- firmed, and the different characteristic inhibition activities of the linkage types is discussed in terms of steric effects. A possible application of the inhibition technique towards structural analysis of carbohydrate molecules is suggested by the authors.

A previous paper in this series (1) elabo- bines with the terminal nonreducing units of rated in some detail the configurational the chains. In this communi- features of low molecular weight oligo- cation further confirmation of this hypothesis saccharides which are essential for binding is provided by a more detailed examination to the reactive sites of the concanavalin A of the inhibition of the dextran-concanavalin protein molecule. On the basis of inhibition A interaction, using the members of several studies it was suggested that the specificity homologous series, a number of oligosaccha- of the protein is directed primarily toward rides containing more than one type of link- the C-3, C-4, and C-6 hydroxyl groups of the age, and some amino sugars as inhibitors. n-mannopyranosyl and n-glucopyranosyl ring forms which possess the a-configuration EXPERIMENTAL PROCEDURE at C-l. The preparation of concanavalin A and the In attempting to account, on a molecular assay procedure used for inhibition studies have level, for the capacity of concanavalin A to already been described (1). Dextran NRRL interact to form a precipitate with a re- B-1355-S was employed as the precipitating poly- stricted group of branched saccharide. [glycogens, , , yeast Many of the saccharides used in this study were mannans, and more recently, bacterial levans gifts (see acknowledgments). All were tested for (a)], it was suggested that the protein com- purity by paper chromatography. Impure prepara- tions were purified by preparative chromatog- 1 Present adress: Department of Biochemistry, raphy on Whatman 3MM or S and S Green Label Royal Free Hospital School of Medicine, London filter paper. W.C. 1. solutions of known concentration were 2 This work was done during the tenure of an prepared by weighing in the case of crystalline Established Investigatorship of the American compounds, the crystals having been previously Heart Association. Present adress: The University dried to constant weight. Noncrystalline sub- of Michigan. stances were determined by the phenol-sulfuric 88 atLid procedure (3). The hetero-oligosaccharides was shown by paper chromatography to contail were also determined by this method lising syil- glucosyl glycerol (III) and glycerol. Glycerol was thetic standards containing a suitable ratio of the characterized as the tri-p-nitrobenzoate derivx- componeilt sugars. tive (m.p. and mixed m.p. 188-191”); the glucosyl Preparalion of nigermy?, erythritol. To an aqrw glycerol (2.O-~-1,.glrlcol)yranosyl-gl?-cerol) had 011s sollrtion of isolichenin (0.513 gm 25 ml) [a] i5 + 125” in water (c, 1.3); Charlsoll P/ trl. (Cjj was added a sollition of sodium metaperiodnte report [a]:’ + 121” in water. (0.23 M, 10 ml) and the volume adjusted to 50 ml. The osidat,ion was allowed to proceed at 25” in the RESULTS dark. After 3 days 0.35 mole of oxidant, per gl\icosyl As in the former study (l), n-c h:n~e :1&i- residllr had been consltmed. Solid BaCOa was trarilv classed 21s noninhibitors thosjc sub- added t.o neutrality and the solids were centri- st,ances of which 20 pmoles gave less than fllged. The sllpernatxrlt so111tion was poured into 10% inhibition of the concanav:~lin A reac- 10 ml of wat.er containitlg sodium borohydride (200 mg). rZfter 24 horlrs at room temperature aI1 addi- t,ion. Tnhle I lists all the substances tested tiollal portion of NaBHJ (100 mg) was added. On in this study. Three homologous series have the following day the solution was made jlrst acid been investlgsted: the (mal- to litmus with acetic acid and poured with stirring tose to maltodecaose inclusively), the iso- illto 957& ethanol (400 ml). The precipitate was maltodextrins (isomnltose, isomnltotriosc, centrifuged, washed several times with 95% -tetraose, and -hept’aose), and the methyl ethanol and successively with absolute ethanol, a-isomaltosides (methyl isomaltosidc to diet,hyl ether, and light petroleum et,her, and dried methyl a-isomaltooctaosicle inclusively). in MZCUOat 25”. Yield: 360 mg of white powder. On a molar basis, the members of the The periodate-oxidized and reduced isolichenin series exhibited the same inhi- (300 mg) was dissolved in 0.2 s sulfuric acid (50 ml) and placed in a wat,er bath at 60” for 30 min- bition activity and fell on a single inhibition utes. The solntion was nent,ralized (BaCOs), curve (Fig. 12.). Similarly, the inhibition filtered, and concent,rated. Paper chromatogra- values for the isomaltodextrins, and the car- phy, with ethyl acet,ate:pyridine:water (10:4:3, responding methyl a-glycosides which were v/v) as the development solvent and silver nitrate- tested, also can be fitt’ed to a single curve sodirlm hydroxide spray reagent, revealed that the (Fig. lb). The rcsult,s of inhibiticjn studies main components of the hydrolyzate were niger- on a number of oligosaccharides containing osyl erythritol (I) and glucosyl erythritol, with more than one type of glycosidic linkage are t.races of glllcose and erythritol. Nigerosyl erythri- given in Fig. 2. H (P:mose), to1 was isolated by preparative paper chromatog- C (7) (see Fig. 3) and make- raphy as a syrup (99 mgj and had [cy]:’ + 184” in water (c, 0.66)) identical with an authentic speci- san (~-0-Lu-n-glucopSranosyl-1, G-anhydro-p- men isolated by Goldsteiii and Whelan (1). A n-glucopyranose) all exhibked an inhibition recent, communication (5) also reports the isola- xtivity equal t.o that of the ,ow series tion of Iligerosyl erythritol from isolichenin. (1.1 pnoles gave 50% inhibit.ion). The inhi- A portion of I (52.8 mg) was oxidized in 0.023 bition w3ivit.y observed for hot,lj met.hvl M sodium periodate (10 ml) at 25” in the dark. After p-malt,oside and trisaccharide E (Tsc,r,nnose\ 48 holux, 3.3 M proportions of periodate had been (2.0 pmoles for 50% inhibition) dlJSdy ap- consumed with the prodltction of 0.9 mole of formic proached that of the maltose series. Tri- acid and 1.1 moles of formaldehyde (theoretical saccharide I’, which differs frnm E in that, ratio: 3:l:l). The reaction mixture was neutral- t)he (1 + A)-glucosidic linkage has the 0 ized (BaC03) and centrifuged, and the super- natant solution was treated with XaBH1 (20 mg). rather than oc-configurat ion, is an inhibitor The redrlced material was deionized by passing somewhat better than malt,ose, but, trisnc- the solution successively through Amberlite IR charide G, in which the (1 -+ 3) bond is also 120 H and IR 45 A exchange resins. Borate was in the P-configuration, is a noninhibitor. removed by several evaporations in the presence The t.\vo branched trisaccharidw? .T (S) and of methanol and the product, was recovered as a I< also differ only in that I< has a 6 (1 + t,)- syrup (II). linkage; hwvevcr, in this case, the lxexnce A portion of syrup II was dissolved in 0.2 N sulfuric acid and allowed to stand at 20” for 16 3 Saccharides in which at least two stlgrrr resi- hours. The solution was neutralized (BaCOz), dries are glyroeitlirally linked to a third sugar filtered, and concentrated to yield a syrnp which residue. 90 SMITH ANI> GOLDSTEIN

TABLE I SUBSTANCES TESTED FOR INHIBITION ACTIVITY

Inhibitors Xoninhibitors

Methyl B-D-manr~opyranoside~~ 2-Amino-2-deoxy-D- Methyl 2-acetamido-2-deoxy-ol-D-glucopyranoside 2-Amino-2-deoxy-D- Methyl 2-acetamido-2-deoxy-@-D-glucopyrano- Methyl 3-amino-3-deoxy-a-D-glucopyranoside sideb 6-Amino-6-deoxy-D-glucose 2-0-oi-D-Glucopyranosyl-glycerol 6-0-Acetyl-D-glucose Mannosyl-glycerol 6-L>eoxy-6-fluoro-D-glucose 6-0-or-D-~Iannopyrallosyl-D-glucose 4-0-p-D-Mannopyranosyl-D-mannose 0-(2-Acetamido-2-deoxy-a-D-glucopyranosyl)- 4-0-P-D-Glucopyranosyl-D-mannose (1 + 3 or 4).D-galact,itol 0-p-D-Glucopyranosyl-(1 + 4)-@-D-glucopyran- 0-(2.Acetamido-2-deoxy-a-D-glucopyranosyl)- osyl-(1 ---f 6)-D-glucose (1 + 6).2-acetamido-2-deoxy-D-glucose 0-(2.Acetamido-2-deoxy-b-D-glucopyranosyl)- 2-0-or-Nigerosyl-D-erythritol (1 + 4)-2-acetamido-2.deoxy-D-glucose Methyl p-maltoside a-Schardinger 4-O-ol-D-Glucopyranos~-l-l,6-arti~ydro-~-D-gluco- p-Schardinger devtrin Trehalosamine Maltose to maltodecaose, inclusively Galactinol Isomaltose A’-Acetyl neuraminic arid Isomaltotriose Isomaltotetraose Isomaltoheptaose Methyl wisomaltoside to methyl wkomaltooctaoside, inclusively Isopanose Panose Additional oligosaccharides in Fig. 3

0 Relatively poor inhibitor. b Verv poor inhibitor.

100 -

80 -

.z 60- 2 ‘2 C 40- E z f a 20-

Ok-Tk----. . 2.0 3.0 4.0 5.0 0.5 2.0 3.0 4.0 50 Micromoles added FIG. 1. (a) Inhibition of precipitation by the maltose series of oligosaccharides. 0, Maltose; 0, ; A, maltotetraose; A, maltopentaose; 0, maltohexaose; 0, malto- heptaose; 0, malto-octaose; A, maltononaose; n , maltodecaose. (b) Inhibition of pre- cipitation reaction by the isomaltose series of oligosaccharides. 0, Isomaltose; W, iso- maltotriose; q , isomaltotetraose; q , isomaltoheptaose; a, methyl-a-isomaltoside; 8, met,hyl a-isomaltotrioside; 0, methyl a-isomaltotetraoside; 0, methyl a-isomaltopentao- side; 6, methyl a-isomaltohexaoside; Q, methyl-a-isomaltoheptaoside; 0, methyl or-iso- malto-octaose. STEREOCHEMICAL REQUIREMEYTS OF CONCANAVALIS h 91

100 glucosidic units in the form of six- or scvew r membered rings, respectively. The inhibit,ion activity of nigerosyl eryth- ritol (compound I, Ipig. 4) is ahon-n’in Fig. 5 (5.0 pmoles for 50 70 inhibition). The product of pcriodate oxidatjion followed by reduction, compound II, failed to inhibit the tlcxtran- E aJ 40- concanavalin A interaction, \vhereas glucosyl : glycerol (2 - 0 - a - n - glucopy~anosyl - glyc- c 24 erol (III), formed from II by mild acid hy- 20 - drolysis (Fig. 4), cxhibit,ed an inhibit ion (1.2 @moles for 50 % inhibition) compamble to that of t.he isomaltose series. Two u-glucose OM 10.0 derivatives (6.o-ace@-n-glucose :tnd (i-de- Micromoles Oligosocchoride odded oxy-6-fluoro-D-glucose) were found, as cx- FIG. 2. Illhibitioll of precipitation by sac- pected, to be noninhibitors, further confirm- ellarides. Cl, Isomaltose (A); 0, panose (B); 8: ing the requirement for an unmodified tetrasaccharide (C) ; 0, maltose (I)) ; +, isopanose hydroxyl group at, C-6. (E); 0, trisaccharide (F); 0, tetrasaccharide Five amino sugars were tested (after (H) ; +, trisaccharide (J) ; 4, trisaccharide (K) ; neutralizing t’heir solutions to pH 7.0) for a, tetrasaccharide (L) ; 0, pentasaccharide (M); their inhibition of t,he dt~xtran-concanavalin X, methyl p-maltoside; 0, maltosan. A interaction. As cxpcctcd, 6-amino-6-deoxy- n-glucose and methyl 3-nlnino-3-tleox~-cr-n- of the p-linkage resulted in I< being a less effective inhibit,or (9.0 moles for 50 % inhibi- tion) than trisaccharide .J (2.3 pmoles for 50 70 inhibition). The branched pent’asac- charide M (9) has a similar inhibiting activ- ity to that of malt’osc. The CY-and p-Schardinger showed a complete lack of inhibition of the dextran- concanavalin A interaction. These molecules do not possess terminal nonreducing ends I. Nolo4 2. NaBH, because they consist of a-~-(1 + 4)-linked I

H LL0 A B C (isamoltose) (panose) B

D E F G H holtose) (isoponose)

02 0-P 0-L x-0 J K L M

.& Schardinger Dextrin (011 linkages ~4 -Cl+41 0

FIG. 3. Structure of some oligosaccharides tested for inhibition activity. 0, Nonreducing glucose units; 0, reducing glucose nnit,s; (-), 1 --f 4 linkages; 1, 1 --f 6 linkages; all a-linkages except where indicated. FIG. 4 92 SMITH AND GOLDSTEIN

100 -

80 -

s y 40 - b a

20 -

I I I111111 I I I111111 I Ill111d 0’0.1 0.2 0.3 0.5 1.0 2.0 3.0 5.0 IO 20 30 50 100 Micromoles Socchoride added FIG. 5. Inhibition of precipitation by saccharides. V, Methyl 2-acetamido-2-deoxy-or- U- glucopgranoside; <:, methyl 2-acetamido-2-deoxy-fl-n-glucopyranoside; V, 0-(2-aceta- mido 2-deoxy-a-o-glucopyranosyl)-(I - 6)-2-acetamido-2-deoxy-n-glucose; D, nigerosyl erythritol; Q, 2-0-a-n-glucopyranosyl-glycerol; A, 6.0-a-o-mannopyranosyl-r-glucose; \% methyl /3-n-mannopyranoside; T, methyl a-o-glucopyranoside; 17, methyl 6-wgluco- pyranoside; L, n-glucose; 0, 0-(2.acetamido-2.deoxy-a-D-glucopyranosyl)-(I + 3 or 4)- n-galactitol. glucopyranoside failed to inhibit’, but sur- tive than the corresponding glucoside, is a prisingly, 2 - amino - 2 - deoxy - D - glucose, relatively poor inhibitor and corresponds 2-amino-2-deoxy-n-mannose, and trehalosa- closely to the inhibition activity of glucose mine (TV- D - glucopyranosyl - 3 - amino - 3 - (11 pmoles for 50 % inhibition), whereas two deoxy-a-n-glucopyranoside) also failed to P-linked , 4-0-fl-n-mannopy- inhibit the interaction. ranosyl-n-mannose (mannobiose) and 4-O-p- On a molar basis, the inhibition activities n-glucopyranosyl-n-mannose, can be classed of the w and P-methyl glycosides of 2- as noninhibitors. On the other hand the acetamido-2-deoxy-n-glucose were approxi- a-linked , 6-O-a-n-mannopy- mately 50% of t,he corresponding parent ranosyl-n-glucose, was one of t,he most methyl LY- and ,8-n-glucopyranosides. The potent inhibitors tested (0.22 pmole for 50 70 disaccharide O-(2.acetamido-2-deoxy-a-D- inhibition). glucopyranosyl) - (1 --f 6) - 2 - acetamido - 2- Galactinol and N-acetyl deoxy-n-glucose is somewhat less effective as were both found to be noninhibitors. an inhibitor (1.9 pmoles for 50 % inhibition) than isomaltose (1.1 pmoles for 50 % inhibi- DISCUSSION t.ion, whereas N , N’-diacetylchitobiose is a From earlier inhibition (1) studies it was noninhibitor. A sample of 0-(2-acetamido-2- anticipated that the D-glucose derivatives deoxy-a-n-glucopyranosyl) (1 -+ 3 or h)-D- modified at C-3 and C-6 (6-O-acetyl-n- gala&it,01 showed an inhibition of 35 % for glucose, 6-deoxy-6-fluoro-n-glucose, methyl 3.2 pmoles, demonstrat,ing the importance of 3-amino-3-deoxy a-D-glucopyranoside, and the wglycosidic linkage. 6-amino-6-deoxy-n-glucose) would not be in- Of the mannose-containing sugars tested, hibitory, but the finding that 2-amino-2- it was again observed that the presence of the deoxy-n-glucose and trehalosamine also were &configuration at C-l of the nonreducing inactive was unexpected inasmuch as sugar residue greatly reduced or complet’ely 2 - acetamido - 2 - deoxy - D - glucose previ- abolished inhibition activity. Thus, methyl ously had been shown to be an inhibitor fl-n-mannopyranoside, although more effec- equivalent in activity to n-glucose (1). The occurrence of a positive charge on the free tion activity of methyl ~-rJ-niaf~~lol)~“arlr,- amino group (pk’ 7.S) (10) at pH 7.0 ma!. side. In contrast are the very high act&ties accounts for the qparent failure of these of the uc-1inlz.d compounds (Fig. ;5), methyl sugars tjo interact with the concanavalin ,\ cu-n-maIlnopSrarlosi(le (I) and the disaccha- protein. In this regard, 2-:tmiIlo-2-deox~,-l)- ride (i - 0 - a! - 1) - I1i:lririoI)Sr:iIios~l - 1) - glu- mannose was also a noninhibitor. N-Acetyla- cose. These data prompt the suggestion that tion precludes the possibility of such a glycoproteiw containing terminal a-n- charge effect in the case of 2-:LmiIlo-2-deox)-- nlannop~rarlos~l units should interact Ivith u-glucose, :md it was noted t,hat, t.hc anomerlc concanavalin A. methyl glycosidcs of 2-acetnmitlo-2-deoxr- The hypothesis that it is predominantl\, I)-glucose inhibited tlextran-cor!carlavalili A the terminal nonreducing a-D-g~ycO~qTaIlO- interackion, the wanomcr showing 50 % in- syl residues of simple and complex a-glucans hibition for 2.3 pmolcs and the /% and wmannans with which concanavalin A gvmg 50 5 inhibition for GO pm&s; these interacts (1, I 1) received addit ion al support values are approximately OIIC half those from the fact that higher members of the found for the parent methyl a- and &gluco- three homologous series of saccharides tested sides. Previously it, had been shown that have the same inhibition activities for the 2-acetrtmido-2-dcox~-u-mannose was a mm- concanavalin A interaction as the corrc- inhibitor (1). These result’s indicate that sponding disaccharide members. The identi- although modification at the C-2 position of cal inhibition values of panose and ktra- n-glucose is t,oleratcd, a 50 % reduction in wccharide C (E‘ig. 2) \vith that of the iso- inhibiting potency is noted when the hy- maltose ad methyl cr-isomaltose series rlros~l group at C-2 of methyl cr- ad p-glu- (1.1 qnoles for 50% inhibition) can be re- copyranoside is transformed into a ‘1. lated to the observation that, both of t’hese acetarliido-L’-tleoxy group. oligosaccharides possess an isomalt,ose unit Two disaccharides and a disaccharide alco- ((i-0-Lu-r,-glucop~ra~los~l-u-gluc~)~e) at, t’heir hol, all cont.aining 2-acet,amido-2-deoxy-w nonreducing ends. Similarly, the possession glucose rcsiducs in t)hc nonrcducing positions, of a maltose residue (~-O-cu-n-glucopyrano- \vere examined. As expect’ed, by virtue of its syl-r-glucose) at the terminal end of methvl @glycosidic linkage, N, N’-diacet,yl chito- @-maltwide :md isopanose (trisaccharide k, biose [O-(2-:lcetamitlc,-“-deoxy-P-u-glucop~- Fig. 3) confers on these oligosaccharides a11 ranosyl) - (1 ---f 1) - 2 - acetamido - 2 - deoxy- inhibition activity (2.0 qwles for 50 54 inhi- n-glucose] was a noninhibitor. On the other bition) very close t-o that of the maltose hand, O-(l>-acctamido-2-deox~-~-n-glucopy- series (1.9 ~moles for 30 !G inhibition). Al- ranosyl) - (1 + 6) - 2 - acetamido - 2 - dcox:-- though kisaccharide F (P’ig. 3) and maltosnn D-glucose, which contains an wglycosldlc also contain a maltose-like struct’ure, t’hcy bond, is a good inhibitor, being GO% that of exhibited inhibition activities (1.3 and 1.1 isomaltose. Similarlv 0-(2.acetamido-2- ,.moles, respectively, for 50 5% inhibition) tleox~-a-,,-gluco~~~ranosJ-1).(1 -+ 3 or 4)-r)- more ch:wact,cristic of the isomalt.ose series. galact,it,ol proved t.o inhibit. dextran-con- No explanation can be offered for the high canavalin A interaction. These observations inhibition of t.rie:rccharide F, but, it may be suggest, t,hat glycoproteins which contain significant that in maltosan the C-6 h!-droxvl nonreducing terminal 2-acetamido-2-deoxJ.- group of the “reducing” glucosyl residue is a-r)-glucop?ratio~yl units may also interact involved in anhydro ring formation. It has with concunavalin A. been previously reported (1) that isomaltose, The specific requirement for the oc-con- in which the C-G hydroxyl of t,he reducing figuration at, C-l of the interacting sugar glucose residue is engaged in the glucosidic residue is further confirmed by the observed linkage, has a higher inhibition activity than lack of inhibitionof thedextran-concanavalin any of the other cr-linked glucose disaccha- A reaction by m:rnnobiose (4-0-/3-n man- rides tested. 011 the basis of t,hese observa- nopprnrlos~l-r)-maurlosc) and trisaccharide tions and of the similarity of the isomaltose CT (Fig. 3), a11t1 by t.hc rather lo\v inhibi- inhihition activity to that of mct.hyl WD- 94 SMITH AND GOLDSTEIN glucoside, it was suggested that the different stituent residue at C-3 of the remaining apparent affinities of the disaccharides for intact cr-n-glucosidic residue, failed com- the concanavalin A receptor sites could be pletely t,o inhibit,. Removal of t,his C-3 explained by steric considerations, rather substituent, by mild acid , again than by any real specificity difference for the presented an unsubstituted terminal OC-D- linkage types. It is possible that the unex- glucosidic residue for interaction nith con- pectedly high inhibition activity observed canavalin A, with the results that’ the for maltosan may also be explained by steric product, 2-0-a-n-glucopyranosyl-glycerol factors. (compound III), exhibited a high inhibit’ion The branched trisaccharide .J (Fig. 3), al- activity. though possessing two terminal nonreducing Although the possibility still exists that a-n-glucopyranosyl residues, has an inhibit- the interaction of the protein may extend in ing activity surprisingly rather less than that certain cases beyond the terminal chain end of maltose. It seems probable that the of a carbohydrate molecule (indeed this is greater part of the inhibition is a function of indicated by the high inhibition activities the constituent isomaltose unit, because in observed for 3-O-a-n-glucopyranosyl-n- trisaccharide K, in which t,he 1 + 6 bond is and for ) (l), the data in the &configuration, the inhibition is less presented in this and the previous paper pro- by nearly four times. It is again possible to vide a fairly consistent conception of the rationalize this relatively low inhibition ac- stereochemical requirements of the interac- tivity of J in terms of a steric effect if one tion between concanavalin A and the termi- regards this trisaccharide either as an iso- nal nonreducing chain-ends of the a-linked maltose derivative with a large substituent glucose- and mannose-polysaccharides. At radical at C-4 of the reducing glucose unit, this point, therefore, it was felt that these or as a maltose molecule carrying a large data could be employed as a 0001in the struc- radical at C-6 of the reducing glucosyl resi- tural analysis of oligosaccharides, in which due. In pentasaccharide M (Iqig. 3) one of certain structural features are left in doubt the nonreducing ends may also be regarded by more orthodox methods. Two such studies as a maltose residue substitut,ed in the re- will be described briefly. ducing glucosyl moiety; the second non- The first involved a mannosyl-glycerol, reducing end may be considered as an un- obtained by alkaline hydrolysis from M. substituted maltose residue. Only the latter Zysodeikticus (12, 13). This compound was possibility would be expected to contribute found to inhibit the dextran-concanavalin A significantly to the inhibition activity of the precipitation reaction to the same extent as molecule as a whole, and this would seem to methyl a-n-mannopyranoside (cf. t’he equal be borne out by the similarity of the ob- inhibition activities of 2-O-a-n-glucopyrano- served inhibition (2.3 pmoles gave 50 % inhi- syl-n-glycerol and methyl a-n-glucopyrano- bition) to that of maltose. side). The a-configurat,ion was therefore The failure of the o(- and P-Schardinger assigned to the mannosyl-glycerol linkage. dextrins to inhibit is consistent with the Subsequent investigations have supported absence of any terminal glucose residues in this assignment (13). their ring-like molecules. Possibly the best Second, two , which on direct evidence for the interaction of con- the basis of their mode of preparation from canavalin A with the saccharide chain-ends panose (14) could be assigned either struc- is provided by the inhibition studies carried ture H or L (Fig. 3), were tested as inhibitors out on nigerosyl erythritol and its oxidation in the dextran-concanavalin A system. One and acid hydrolysis products. Xigerosyl of the tetrasaccharides inhibited t,he reaction erythritol (compound I, Fig. 3), having an (1.6 pmoles for 50% inhibition; Fig. 2) close unsubstituted terminal a-n-glucosidic resi- to the value found for maltose and was there- due, inhibited (50% for 5.0 pmoles) the fore tentatively assigned structure H, dextran-concanavalin A precipitation reac- whereas the second tetrasaccharide was as- tion, whereas compound II, carrying a sub- signed structure 1, because it’ displayed an inhibition (3 pmoles for 50% inhibition, 2. GOLDSTEIN, 1. k, AND SO, LUCY L., Arch. Fig. 3) similar t.o the inhibition of t.risaccha- Hiorham. Hiophys. 111, 407 (1965). ride J. 3. I>UBOJS, &I., GILLEs, K. A., HAMILTON, J. K., It, is suggested that, with the examination REBERS, P. A., ASI) S>lrrrr. F., Anal. Chettc. of :I greater number of model compounds, 28, 350 (1956). GOLDSTEIN, I. J., AND WHELAN, W. d., (in this method may fired a general application preparation). in t,he structjural analysis of saccharides con- FLEMING, M., AND MASSEI~S, II. J., Biochem. taining n-glucose, n-mannose, Zacetamido- J. loo, 241’ (lxxi). ‘-deoXv-D-glUCOSe, and certain of their CHARLSON, A.J., GoRIN,~'. $.J.,AND PERLIN, derivatives. A. S., Can. J. Chew 34, 1811 (1956). Bnowx, I). H., ILLING\VOI~TH, BARBARA, AND ACKNOWLEDGMENTS KORNPELI), I~~~AI,J~J~, Biochemisfq/ 4, 486 This work was supported by grants from the (1965) National Instit.utes of Healt,h. We are grateful to 8. DE&Jl-ZA, 1:ANC:A, AND (;OI,I~SI’EIS, I. J., the following for many of the sugars used in this Telrahedron Let/e,x 1215 (1964). I:. C., SMITH, E. E., AND WHELAN, study: P. %. Allen, 1). H. Brown, R. deSouza, 9. Hoort~s, J. Distler, R. Durr, I. Ehrenthal, H. G. Fletcher, W. J., D&kern. J. 88, 63P (19G3). A. B. Foster, R. Hoover, T. 1,. Hullar, Barbara 10. Ksxr, P. W., AND WIIITEHOUSE, M. W., Illingworth, R. Iyer, R. Jeanloz, J. Klingman, II. “Biochemistry of the Amino Sugars,” p. 202. Lemieus, W. Lennarz, Betty Lewis, K. Lloyd, Academic Press, New York (1955). A. S. Perlin, N. K. Richtmyer, S. Roseman, F. 11. GOLDSTEIN, I. J., HOI,I,EILMAN, C. E., AND Penti, F. Smith, T. F. Timell, and W. J. Whelati. DERRICK, J. &I.: Bi~chinz. Hi~php. Ada 97, We are indebted to Dr. Allene Jeanes for the 08 (1965). sample of dextran NIlRL B-1355-S. 12. LENNAILZ, W. J., J. Sol. Chern. 239, PC3110 (1961). REFER.ENCES 13. LENNAILZ, W. J., AND TALAMO, BAI~BARA, J. Bid. Chenz. 241, 2707 (1966). 1. GOI,~S~EIS, 1. J., HOLLEKYAS, C. E., ANI) 14. FRENVII, I)., SMITH, l3. E., ASD WHELAS, W. J., SMITH, E. E., ~%U&e??liS&/ 4, 8% (1965). (in press).