Post on 15-Oct-2016
ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS 172, 524-534 (1976)
lmmunochemical Studies on Blood Groups
lmmunochemical Properties of B-Active and Non-B-Active Blood Group Substances
from Horse Gastric Mucosae and the Relative Size Distributions of Oligosaccharides Liberated by Base-Borohydride’, *
WALTER NEWMAN AND ELVIN A. KABAT
Division of Chemical Biology and the Departments of Microbiology, Human Genetics and Development, and Neurology, College of Physicians and Surgeons, Columbia University, and the Neurological Institute,
Prebyterian Hospital, New York, New York 10032
Received July 3, 1975
Horse B-active and non-B-active glycoproteins from gastric mucosae are indistinguish- able in their precipitating abilities with concanavalin A, anti-BPl, type XIV horse antipneumococcal serum, the lectin from Lotus tetragonolobus and a group 1 anti-1 serum, Ma; no Le” or Let’ activity was found. Each was subjected to catalyzed release of ita oligosaccharide chains by 0.05 N NaOH in 1 M NaBH,. Destruction of serine, threonine and 2-acetamido-2-deoxygalactopyranose (DGalNAc) was associated with production of alanine, a-aminobutyric acid and N-acetyl-n-galactosaminitol, as expected for a carbohy- drate to peptide linkage via nGalNAc to serine or threonine. No evidence of base- catalyzed peeling was seen. Bio-Gel P-2 elution patterns of the salt-free oligosaccharides from the two preparations were compared. Unlike results obtained with human ovarian cyst substances, very little material was excluded. The largest-size chains are in the range of deca- or dodecasaccharides, and a reduced octasaccharide was isolated. The four most abundant amino acids in both B-active and non-B-active materials are threonine, serine, proline and glutamic acid, which together account for 60% of the weight of amino acids.
The previous paper (1) described the iso- lation and chemical and immunochemical characterization of 50 fractions of blood group-active glycoproteins from 10 horse gastric mucosae. The 13 fractions with the highest B activity were combined as were 24 fractions showing neither A, B nor H activity or only H activity. Included as well were two B-active and five non-B- active fractions previously described (2).
The present report summarizes serologi- cal and chemical properties of these two
’ Supported by grants from the National Science Foundation, No. BMS-72-02119A02 and 32543 X-l, and a Program Project Grant from the National Institutes of Health, No. 5P0 GM 18153-05.
2 From Part II of a dissertation submitted by W. Newman in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Faculty of Pure Science, Columbia University, New York, This article is No. LX in a series; the previous article appeared in Arch Biochem. Biophys. 172, 510-523 (1976).
glycoprotein materials and elution pat- terns of oligosaccharides released from the peptide backbone by treatment with 0.05 N
NaOH in 1 M NaBH, for 16 h at 50°C. These conditions have been shown (3-5) to result in the p-elimination of carbohydrate from protein with only traces of galactitol, hexenetetrols or reduced chromogens, all indications of degradative peeling (4, 6). Destruction of serine, threonine and nGalNAc3 and formation of alanine, cu-ami- nobutyrate and N-acetyl-n-galactosamini-
” Abbreviations used: GalNAc, 2-acetamido-2- deoxygalactopyranose; GlcNAc, 2-acetamido-2-deox- yglucopyranose; N-acetyl-n-galactosaminitol, 2-ace- tamido-2-deoxygalactitol; Gal, galactopyranose; Fuc, 6-deoxygalactopyranose; mRF, molar response factor; AbN, antibody nitrogen; IM2, isomaltose: IM4, isomaltotetraose; IM6, isomaltohexaose; IM8, isomaltooctaose; R,;, R,, and R,,,, migration dis- tances in paper chromatography relative to galac- tose, lactose and isomaltopentaose, respectively; Con A, concanavalin A.
524 Copyright 0 1976 by Academic Press, Inc. All rights of reproduction in any form reserved.
SIZE DISTRIBUTION OF OLIGOSACCHARIDES 525
to1 by this procedure have been taken as good evidence (7) for glycosidic linkage of the carbohydrate to serine or threonine via nGalNAc(7--10). Amino acid analyses of the two horse glycoprotein materials be- fore and after the NaOH-NaBH, treat- ment confirm this type of linkage.
Elution patterns on Bio-Gel P-2 have provided an estimate of heterogeneity and size distribution of dialyzable reduced oli- gosaccharides formed from ovarian cyst blood group substances by base-borohy- dride (11, 12). Comparable studies with horse B-active and non-B-active materials in the present study provide some insight into species differences with respect to these two parameters as related to blood group activity.
MATERIALS AND METHODS The blood group preparations combined are com-
posed of the most active B fractions from mucosae 7, 8, 9, 13 and 14 (1) and mucosa 2 (2) and non-B-active fractions from mucosae 10, 11, 12 and 16 (1) and 3, 5 and 6 (2).
Calorimetric carbohydrate analyses have been de- scribed elsewhere (13-15). The NaOH-NaBH,-cata- lyzed elimination reaction (14) was used as modified (3). Blood group substance was dissolved in 0.05 N
NaOH in 1 M NaBH, to give a 1% solution and was left at 50°C for 16 h. The reaction was stopped by cooling and neutralizing with dilute HCl. A lo-mg sample, neutralized but not dialyzed (Table I), was removed to be analyzed for amino acids and presence of degradation products and for analytical paper chromatography. The remainder was dialyzed against 10 volumes of water at 4°C and the dialy- sates were collected in two consecutive fractions: Dial I, a pool of live successive dialysates each ob- tained after 30 min and Dial II, a pool of four succes- sive dialysates, three obtained each after a 24-h and the fourth after a 48-h dialysis. Dial I, Dial II and the nondialyzable portion (non dial) were concen- trated to dryness, and borate was removed by re- peated additions of 100 ml of methanol followed by evaporation. Further desalting was carried out on Dowex 5OW-X8 (H+ form), 20-50 mesh, and Dowex l-X8,20-50 mesh (Bio-Rad Laboratories), which was converted from the Cl- to the OH- form immediately before use. Yields of carbohydrate eluted from these resins were low compared with earlier studies (ll), necessitating elution of the resins with 0.1 N NaOH; 5-ml fractions were collected and monitored by pH, periodate uptake and optical rotation at 365 nm. Fractions A, B, C, and D of the B-active and a, c and d of the non-B-active material were all eluted with water or before the NaOH eluate became
strongly basic. Fractions E, F, b and e were eluted under alkaline conditions. After neutralization, materials were desalted on a column (2.8 x 88 cm) of Dow Retardion llA8, eluting with water. Table I is a flow sheet for the NaOH-NaBH, procedure, desalting and Bio-Gel P-2 chromatography.
A 2.8 x 88- or 2.0 x 75-cm column of Bio-Gel P-2 C-400 mesh) was prepared as described by the manu- facturers (Bio-Rad Laboratories) and calibrated with a mixture of dextran NRC 3 (molecular weight, 51,000) (131, isomaltooctaose (IMS), isomaltohexaose (IM6), isomaltotetraose (IM4) and isomaltose (IM2). Fractions of 1.2-1.3 ml were collected and alternate tubes analyzed for periodate consumption, methyl- pentose and optical rotation at 365 nm. Optical rota- tions were performed on a Perkin-Elmer polarime- ter, Model 141.
Hydrolysis for amino acid analyses was done in sealed evacuated tubes for 16 h at 112°C on 1 mg of glycoprotein in 6 N HCl at a concentration of 0.1% or less to minimize destruction of amino acids (16). Analyses were kindly performed by Mr. Walter Schrepel according to Ref. (17). The color factors for each amino acid were used to convert the area under each peak to micromoles.
The N-acetyl-n-galactosaminitol formed in each fraction was quantitated by gas-liquid chromatogra- phy with an F and M Model 810 gas chromatograph and stainless-steel or glass columns packed with 3% ECNSS-M on Gas Chrom Q (Applied Science Labora- tories, State College, Pa.) programmed from 140 to 200°C rising at Z”C/min. The temperature was held until all material had eluted. Peak areas and reten- tion times were measured with a Hewlett-Packard integrator, Model 3370B. Quantitation of a mixture of variously substituted N-acetyl-n-galactosamini- tols presented a problem since the molar response factor (mRF) of N-acetyl-n-galactosaminitol depends upon its degree of substitution (18). Quantitation in Dial I was approximated using the mRF of monosub- stituted N-acetyl-n-galactosaminitol and for Dial II an average mRF of the mono and disubstituted N- aCetyl-D-galaCtOSamiDitOlS. The monosubstituted standard from N-l (18) is RL 1.36: riGalP + 3 N- acetyl-n-galactosaminitol; the disubstituted stand- ard (18) is R, 0.44: nGalP1 + 4~GlcNAcPl --* 6lnGalfll + 3W-acetyl-n-galactosaminitol.
For quantitative precipitin studies, Con A and Lotus lectins were used as described in Ref. (1). Anti-BP1 (191 was produced by immunization of an A,B individual, subject 262, with the nondialyzable portion (Pl fraction) of human saliva blood group B substance that had been subjected to mild acid hy- drolysis. The antiserum was absorbed with the na- tive B substance leaving antibody in the superna- tant precipitable by BPl; 1.0 ml of antiserum was used for each tube. Antiserum Ma is a group 1 anti-1 cold agglutinin (20); each tube contained 30 ~1 of antiserum. Type XIV horse antipneumococcal se- rum was a 1938 bleeding of horse 618 containing 100 pg of AbN/ml. Each tube contained 250 ~1 of antise- rum. Quantitative precipitin curves were performed on a microscale (13). and nitrogen in the washed precipitates was determined by the ninhydrin
TAB
LE
I
FLO
W S
HE
ET
FOR
PU
RIF
ICA
TIO
N
OF
OLI
GO
SA
CC
HA
RID
ES
FR
OM
B-A
CTI
VE
A
ND
NO
N-B
-AC
TIV
E
HO
FSE
GA
STR
IC M
UC
OS
AL
GLY
CO
PR
OTE
INS
k%
B-
ACTI
VE
(127
0 m
g)
or
NON
B-AC
TIVE
(1
924
mg)
m
ater
ials
1.
o.o5
NN
aOH
onM
Na*
H,
1% s
olut
ion
of g
lyca
prot
em,
50°C
, 16
hr
2 ne
ut&z
atio
n,
10m
g sa
t&k
rem
oved
fro
m
each
fo
r an
alys
es
3 di
alys
is
at
4’C
03
DA
1 D
A I
I N
on-d
ialyz
able
I.
J J
I B-
actw
e 22
0 m
g no
n B-
activ
e 42
7 m
g 1
repe
ated
ev
apor
atio
ns
wxh
met
hano
l to
re
mo”
e bo
rate
2.
D
xvex
l-X
8 O
H-,
then
D
owex
SO
W-X
8 H+
to
de
salt
3.
resm
el
utio
n wi
th
waf
er
A
4.
resi
n el
utio
n wi
th
0.1
_N N
aOH
5.
neut
rallz
afio
n
B D
ial
1 -t
non
FJ D
ial
1 +
B D
ial
II no
n B
Dia
l II
B D
ial
II fra
ctio
ns
non
8 D
ial
I fra
ctio
n no
n B
Dia
l II
fract
ions
I A
a B
e 60
mg
209
mg
27 m
g 10
3 m
g 73
mg
70m
g 64
mg
79 m
g 61
mg
151
mg
209
mg
1.
each
fra
cuon
ap
plie
d se
para
tely
to
Bio
-Gel
P
-2
3
2.
peak
s fro
m
BIO
-eel
fra
ctio
natio
n ap
plie
d to
ch
arco
al-C
elite
co
lum
n f:
3 re
appl
icar
lon
to
BLo
-Gel
P-
2 4
prep
arat
we
pape
r ch
rom
atog
raph
y 5.
hi
gh
pres
sure
liq
uid
chro
mat
ogra
phy
f
y;t”d
”,“-
Olig
osac
char
ides
is
olat
ed
from
th
e in
divi
dual
fra
ctio
ns
5 n
A
a B
C
D
E
b
C
d e
x G
I R
J.38
R
,,1.2
6 R
J.38
R
J.3
8 R
J.38
R
(J.2
6 R
,;1.2
6 ID
c3
II
RJ.
41
R<;
0.87
R
J.41
R
J.41
R
G0.
92
R<,
0.87
R
,0.8
7 II
RJ.
37
RJ.
37
RJ.
37
III
R,0
.88
(lost
) R
G0.
85
RJ.
38
RJ.
38
IV
R ,
,O. 65
R
,.0
.65
IV
R&
68
R J
l.68
R ,
,O.S
S
IV
R ,
.0.6
6 V
R
J.04
R
J.0
4 V
I R
,,0
.88
R ,
.0.8
8 V
II R
&l.6
9 R
,0.6
9 V
III
R ,,
,J.
15
R,,,
1.15
R
&.1
5 IX
R
IMS
I. R
,,,l.l
8 R
IMS
l.18
X
R ,
M50
.65
R ,
M50
.65
R ,
,,0.6
5
a E
ach
num
ber
refe
rs
to t
he
sam
e ol
igos
acch
arid
e is
olat
ed
from
a
num
ber
of f
ract
ions
. N
o ol
igos
acch
arid
es
wer
e is
olat
ed
from
fra
ctio
n F.
Pea
ks
from
w
hich
ol
igos
acch
arid
es
I, II,
III
an
d IV
w
ere
isol
ated
ar
e 1,
2,3
an
d 4,
res
pect
ivel
y,
of F
ig.
2. R
,;, R
,, a
nd R
,,,
are
mig
ratio
n di
stan
ces
in p
aper
ch
rom
atog
raph
y re
lativ
e to
ga
lact
ose,
la
ctos
e an
d is
omal
tope
ntao
se,
resp
ectiv
ely.
Tw
o so
lven
ts
wer
e us
ed:
solv
ent
1, 1
-but
anol
:pyr
idin
e:w
ater
, 35
:39:
26,
and
solv
ent
2, I
-but
anol
:pyr
idin
e:w
ater
, 6:
4:3.
O
ligos
acch
arid
es
I, II
and
III
wer
e ch
arac
teriz
ed
in
solv
ent
2, t
he
rest
in
so
lven
t 1.
SIZE DISTRIBUTION OF OLIGOSACCHARIDES 527
method (14) after digestion with sulfuric acid. Type XIV and anti-1 determinations were set up, incu- bated and washed entirely at 4°C. Standards used in precipitin assays were those used in Ref. (1). OG 20% from 2nd 10% from precursor human ovarian cyst fluid (21) and PM phenol insoluble Pl fraction, a mild acid-hydrolyzed B substance from human sa- liva (19) have been described previously. Anti-I, Ma and anti-BP1 reagents are in very short supply and were used therefore only on the combined B-active and non-B-active materials.
RESULTS
Figure 1 shows that the B-active and non-B-active materials do not differ signifi- cantly in their abilities to precipitate hu- man anti-BPl, anti-1 Ma, Con A and the Lotus lectin. Anti-BP1 has been shown by inhibition of precipitation to be most spe- cific for the determinant DGalal + 3DGal linked pl + 3 or 4 to DG~cNAc (22, 23). Precipitation by precursor blood group sub- stance OG of type XIV specific antiserum is inhibitable by oligosaccharides contain- ing terminal nonreducing DGalpl + 4~GlcNAc or DGalPl + 3~GlcNAc (24, 251, the former being somewhat more active, especially when the DG~cNAc is linked @l + 3 to DGal (26). Anti-I Ma has been shown, using milk oligosaccharides conju- gated to poly-L-lysine (27) to contain two populations of antibodies, one I-specific,
complementary to DGalpl -+ 4~GlcNAcPl + 6nGal and the other non-I-specific and reactive with DGalPI + 4~GlcNAcfll -+ 3DGal. Inhibition of precipitation of Con A with hog blood group substances (28, 29) established that terminal nonreducing (Y- linked DG~cNAc was the immunodominant sugar. As shown in Fig. lb, the nondialyza- ble portion from both materials does not precipitate Con A. The Lotus lectin has been found to be specific for ~Fucal + 2DGalBl + 4~GlcNAc with or without a second LFuc-linked al + 3 to the DG~cNAc (30). Goat anti-Lea or Le” (311, kindly pro- vided by Dr. Donald Marcus, did not react with any of the horse materials.
Table I is a flow sheet for the isolation procedure for oligosaccharides from the B- active and non-B-active materials after al- kali-borohydride treatment. Oligosaccha- rides listed in Table I are described in de- tail in Ref. (32).
A small-scale NaOH-NaBH, p-elimina- tion was first carried out on each material. After methanolysis and 0-acetylation, gas chromatographic analyses of the reaction products showedN-acetyl-n-galactosamini- tol, indicating that at least some of the carbohydrate was linked via DGalNAc to the peptide backbone by serine and threo-
020406080100I201400 102030405060708090 0 5 IO 15M25303540
MICROGRAMS ANTIGEN ADDED
FIG. 1. Quantitative precipitin curves of B-active (a) and non-B-active (0) horse blood group substances with (a) type XIV antipneumococcal horse serum; (b) Con A; (c) anti-1 Ma; (d) lectin from Lotus tetrugonolobus; and (e) anti-BPl. Standards are CO), OG 20% from 2nd 10% ; (a), dextran B 1355-S-4; (V), JS phenol insoluble; and CO), PM phenol insoluble, Pl fraction. The nondialyzable fraction of the B-active (r) and the non-B-active (m) were tested for Con A activity.
528 NEWMAN AND KABAT
nine (7-10). After large-scale NaOH- NaBH4 treatment, neutralization, dialysis and removal of borate, the dialyzable frac- tions were passed over the H+ and OH- ion-exchange resins to remove salt. The low recovery of carbohydrate by water, fractions A and a of Tables I and II, necessi- tated further elution with 0.1 N NaOH; this was possible since all the carbohy- drate adsorbed to the resins was termi- nated at the reducing end by N-acetyl-D- galactosaminitol and would not be af- fected. Fractions B-F of the B-active and b-e of the non-B-active in Tables I and II were eluted sequentially with 0.1 N NaOH, neutralized with HCl and desalted on Retardion llA8.
Fractions A-F and a-e were applied to Bio-Gel P-2 columns. Figure 2 is a compari- son of the elution patterns from Bio-Gel P- 2 of fraction D from the B-active and frac- tion d from the non-B-active material. Dex- tran in Fig. 2 marks the boundaries of the excluded volume, in which little sugar was found for the B-active sample and none for the non-B-active; on purification the same three oligosaccharides were obtained from both. These came from peaks 1, 2 and 3 in Fig. 2 and correspond, respectively, to I, II and III in Table I. A B-specific tetrasac- charide (IV) was isolated from peak 4 of the B-active sample. Oligosaccharides are described in detail in Ref. (32). Analytical paper chromatography in l-butandl:pyri- dine:water, 35:39:26, for 14 h or the ma- terials obtained immediately after the base-borohydride treatment and before dialysis revealed that from both almost all carbohydrate had migrated from the ori- gin. The slowest moving spot from the B- active sample had RIMsO. and from the non-B-active R,,,0.38.
Peak 1 in Fig. 2 was largely free N- acetyl-n-galactosaminitol (32). Little of this compound was isolated from either fraction. Gas chromatography of directly 0-acetylated material obtained immedi- ately after base-borohydride treatment and as well of fractions A-F and a-e also indicated very small amounts of free N- acetyl-n-galactosaminitol. Methanolysis plus 0-acetylation of these same fractions revealed that almost all the N-acetyl-D- galactosaminitol was in glycosidic linkage
to other sugars. A small amount of fucose- containing material eluted after IM2 but in amounts too small to be characterized. After fractionation on Bio-Gel P-2, oligo- saccharides were further purified by char- coal-Celite, preparative paper and high pressure liquid chromatography as de- tailed in Ref. (32).
Table II gives yields of the individual sugar components after final desalting and before application to Bio-Gel P-2. Percent- age of recovery is expressed in two ways: recovery including the yield of the nondi- alyzable portion and recovery excluding this material showing essentially the yield from the ion-exchange resins by elution with water followed by 0.1 N NaOH. From the latter, yields of fucose and galactose were in the 40-60% range. The lower val- ues for hexosamines, N-acetylhexosa- mines and DGalNAc reflect conversion by NaBH, of nGalNAc to N-acetyl-n-galacto- saminitol. If the yield of N-acetyl-n-galac- tosaminitol is calculated as percentage of original DGalNAc, recovery is 50-70%. Peptide N yields were the lowest. From specific optical rotations at 589 nm and differences in composition of fractions B-F and b-e, some fractionation may have oc- curred in the elution with NaOH.
Table III compares amino acid composi- tions of the B-active and non-B-active ma- terials before and immediately after expo- sure to NaOH-NaBH,. The increase in alanine and appearance of cY-aminobutyric acid result from destruction of serine and threonine, respectively. In both materials, moles of a-aminobutyrate recovered are less than threonine destroyed (about 64%) and moles of alanine produced are less (about 27%) than moles of serine destroyed (33-36). In the native glycoproteins, the most abundant amino acids in B-active and non-B-active materials are threonine, serine, proline and glutamic acid which together total about 60% of the amino acids. Recovery of amino acids after base- borohydride treatment is 74% for B-active and 78% for non-B-active.
DISCUSSION
As shown in Fig. 1, both B-active and non-B-active blood group glycoproteins of horse gastric mucosae contain multiple se-
SIZE DISTRIBUTION OF OLIGOSACCHARIDES 529
rological activities. The non-B-active gave a small amount of precipitate with anti-B serum, but much less than did the B-active material (1, Fig. 2e). In a previous study, (37),, substances with weak B activity and materials previously considered as inac- tive from horse mucosae 3, 5 and 6 precipi- tated about 50% of the antibody from an anti-B serum produced by alloimmuniza- tion of pregnancy, though they were very poor in inhibition of B hemagglutination.
Earlier studies with anti-BP1 (22, 23) showed that an individual B mucosa (horse 4, 25%) did not react while two A- active samples (horse 1, 15 and 25%) did, although they precipitated little antibody and large amounts had to be used (19). The present findings show the B-active and non-B-active materials to be esserstially equivalent with anti-BPl; much more ma- terial was also needed for precipitation compared to human saliva BP1 substance. These results indicate that the A, B and BP1 specificities of horse blood group glyco- proteins are independent and probably not a consequence of loss of fucose either in the mucosal cavity or during isolation. Horse B-active blood group substances contain immunogenic determinants unrelated to B specificity. An earlier study (19) on the cross reactions of human, cow and horse B substances concluded that an antiserum to horse B substance, horse 4, 25%, contains some precipiating antibodies directed against determinants shared by cow and horse B substances related to BP1 but not to B specificity.
B-active and non-B-active materials pre- cipitated equally well with anti-1 Ma, and with type XIV antiserum. However, rela- tive to the same standard in both assays, horse materials precipitated essentially all the anti-1 but very little of the anti-type XIV. Anti-I and anti-type XIV specificities are associated (20, 27, 38). The cross reac- tion of type XIV antibodies with blood group substances is inhibited almost as well by DGalPl ---, 3~GlcNAc . . . as by DGalPl ---, 4~GlcNAc . . ., the former not being specific for anti-1 Ma. Weak anti-1 activity in horse mucosal glycoprotein has been demonstrated previously (20) by inhi- bition of hemagglutination of 01 erythro- cytes.
Con A reactivity was not found in the nondialyzable portions (Fig. lb). Hence these determinants were probably at- tached by a base-labile linkage of DGalNAc to serine or threonine. In con- trast, NaOH-NaBH, treatment of MN-ac- tive sialoglycoproteins from erythrocyte membranes (33) resulted in loss of MN activity and increase in Con A activity of the glycoprotein portion, suggesting a rela- tively base-stable glycosyl-asparagine linkage of Con A determinants to the pro- tein backbone.
Evidence for fucosyl-substituted type 2 chain structures: ~Fuccxl + 2DGalpl --* 4~GlcNAc . . . is provided by reactivity of both materials with the Lotus lectin (30), although neither was as active as any of the fractions from H-active mucosa 11 (1, Fig. 3).
The retention of substantial amounts of reduced carbohydrate by the ion-exchange resins resulted in generally low yields com- pared to those obtained previously with cyst materials passed over a mixed-bed resin (11). This has been noted previously (3, 36) in studies with reduced carbohy- drate chains. Adsorption to the resins may have been due to complexing of carbohy- drates via borate which had not been com- pletely removed by evaporations with methanol. Alternatively the base-borohy- dride procedure may have caused some N- de-acetylation of the amino sugars (39, 36) which under the acidic conditions of elu- tion from the resin can cause the produc- tion of RNH,+ groups in the amino sugar moieties of the carbohydrate resulting in adsorption to the resins. There is no way to distinguish between these two alterna- tives for the present case. Gas chromato- graphic analyses of the base-borohydride- treated material before and after elution with 0.1 N NaOH showed no evidence of peeling or other degradative reactions. Re- tardion llA8 proved a much more effective tool in removing salt from solutions of car- bohydrate and should be used in future studies. Yields of individual sugars (Table II) after recovery from the resins indicate that some carbohydrate was not eluted.
The elution patterns of dialyzable B-ac- tive and non-B-active oligosaccharides on Bio-Gel P-2 (Fig. 2) are strikingly different
TAB
LE
II
CO
MP
OS
ITIO
N A
ND
YIE
LDS
OF
DIA
LYZA
BLE
A
ND
N
ON
DIA
LYZA
BLE
FR
AC
TIO
NS
OF
CO
MB
INE
D
B-A
CTI
VE
O
R N
ON
-B-A
CTI
VE
H
OR
SE
BLE
ND
GR
OU
P
SU
BS
TAN
CE
S B
EFO
RE
AN
D A
FTE
R N
aOH
-NaB
H,
TRE
ATM
EN
T A
ND
D
ES
ALT
ING
Frac
tion
Yie
ld
Lalg
M
ethy
l- H
exos
e N
-ace
tyl-
Hex
os-
Tota
l P
eptid
e N
G
alac
tos-
(m
g)
(deg
rees
) pe
ntos
e N
-ace
tyl-
(gal
ac-
hexo
s-
(fuco
se)
amin
es
amin
es”
N (
mg)
+
N-a
cety
l- am
ine”
,’ ga
lact
os-
tose
) (m
g)
(mg)
ga
lact
os-
Cm
g)
amin
itol”
B-ac
tive
glyc
opro
tein
N
ativ
e m
ater
ial
Rec
over
ies
afte
r N
aOH
- N
aBH
I
A B
C
D
E F Non
dial
To
tal
(mg)
R
ecov
ery
(%)
I S
um A
-F
II S
tarti
ng
mat
eria
l m
inus
no
ndia
l (I/
II)
x 10
0
1270
60
+6.3
3.
4 10
.6
10.2
12
.5
3.7
2.9
1.7
6.1
27
+2.2
1.
9 5.
4 6.
1 6.
8 1.
0 0.
6 0.
7 5.
6 10
3 -2
6.0
18.3
37
.6
11.8
13
.4
2.9
2.1
3.1
26.0
73
-8
.1
9.4
25.8
8.
8 10
.0
2.2
1.6
2.4
17.4
70
+8
.1
5.4
18.0
8.
5 9.
6 2.
9 2.
3 2.
2 9.
9 64
+4
.8
4.1
10.4
5.
4 6.
9 1.
7 1.
3 1.
7 4.
4 22
1 -1
1.5
5.7
26.2
55
.7
34.4
12
.2
10.0
6.
1 1.
8 61
8 48
.2
134.
0 10
6.5
93.6
26
.6
20.8
17
.9
71.2
” 48
.7
41.3
44
.5
39.6
29
.2
33.8
35
.6
12.9
51
.4
397
42.5
10
8 50
.8
59.2
14
.4
10.8
11
.8
69.4
10
49
111.
1 27
5 21
3 28
7 66
.5
48.4
13
2.4
136.
7
37.8
38
.3
39.3
23
.8
20.6
21
.6
22.3
8.
9 50
.8
12.9
(mg)
(m
g)
i7.7
11
6.8
301
269
321
78.7
58
.4
138.
5 -
amin
itol
N”
(mg)
(m
g)
Pep
tide
Nf
(mg)
z 8
58.4
5 z
2.5
0.2
2 0.
5 x
0.5
$ > 1.
7 6
1.0
8.9
15.3
” 26
.2
6.4
49.5
9---L
-c---
----
__
Non
-B-a
ctiv
e gl
ycop
rote
in
Nat
ive
mat
eria
l 19
24
+10.
0 79
.3
363.
6 40
4.2
Rec
over
ies
afte
r N
aOH
- N
aBH
,
i=t
209 79
-1
3.7
+5.6
11
.9
3.7
49.5
10
.3
66.9
5.
7 i
151 61
-29.
3 +6
.2
17.5
2.
7 41
.4
13.9
24
.0
18.4
e 20
9 +1
1.7
11.9
44
.5
39.7
N
ondi
al
427
+11.
0 7.
7 46
.5
55.9
To
tal
(mg)
11
36
55.4
20
6.1
210.
6 R
ecov
ery
(55 )
59
.0
69.9
56
.7
52.1
I S
um
a-e
709
47.7
15
9.6
154.
7 II
Sta
rting
m
ater
ial
min
us
1497
71
.6
317.
1 34
8.3
nond
ial
WII)
x
100
47.4
66
.6
50.3
44
.4
(I C
alcu
late
d as
N-a
cety
lhex
osam
ines
.
470
141.
2 11
2.2
210.
9 -
112.
2
77.2
10
.7
5.8
5.9
50.2
2.
7 7.
4 1.
3 0.
8 1.
1 11
.0
0.1
21.4
2.
6 1.
2 1.
6 14
.3
0.3
28.7
5.
1 3.
3 2.
5 41
.7
0.7
48.5
10
.2
7.1
4.7
26.3
5.
3 35
.8
31.2
28
.9
19.0
1.
3 28
.8
219.
0 61
.1
47.1
33
.8
144.
8 37
.9
46.6
43
.3
42.0
16
.0
68.7
33
.8
183.
2 29
.9
18.2
15
.8
143.
5 9.
1 43
4.2
110.
0 83
.3
192.
9 20
9.6
83.4
42.2
27
.2
21.8
8.
2 68
.5
10.9
’ C
alcu
late
d as
diff
eren
ce
betw
een
mill
igra
m
of t
otal
N
and
m
illig
ram
of
hex
osam
ine
N.
r C
alcu
late
d ca
lorim
etric
ally
(1
5).
” C
alcu
late
d by
gas
-liqu
id
chro
mat
ogra
phy
(see
Mat
eria
ls
and
Met
hods
). p
Per
cent
age
of o
rigin
al
gala
ctos
amin
e.
’ C
alcu
late
d as
diff
eren
ce
betw
een
mill
igra
ms
of t
otal
N
an
d m
illig
ram
s of
hex
osam
ine
N
+ N
-ace
tylg
alac
tosa
min
itol
N.
” P
erce
ntag
e of
orig
inal
pe
ptid
e N
.
532 NEWMAN AND KABAT
FIG. 2. Elution profiles from Bio-Gel P-2 of the dialyzable materials from B-active (top) and non-B-active (bottom) substances obtained after treatment with 0.05 N NaOH in 1 M NaBH, at 50°C for 16 h, neutralization, ion-exchange desalting, elution of resins with 0.1 N NaOH, neutralization and desalting on Retardion llA8. Arrows indicate elution volumes for dextran, IM8, IM6, IM4 and IMZ. Numbers 1, 2 and 3 refer to peaks which yielded identical oligosaccha- rides from both B-active and non-B-active materials; peak 4 yielded a B-specific tetrasacchar- ide not found in the non-B-active material. See text footnote 3 for abbreviations.
from those of oligosaccharides from human ovarian cyst (11, 12) with respect to overall heterogeneity and sizes of the largest chains released. While the B-active sam- ple contains some excluded oligosaccha- rides, all the carbohydrate from the non-B- active is included. This apparent differ- ence in the upper size limits of oligosaccha- rides released was confirmed by paper chro- matography of material obtained immedi- ately after base-borohydride treatment and before dialysis. Almost all carbohy- drate migrated from the origin in both samples, the nondialyzable portion con- taining only about 10% of the total carbohy- drate. The largest B-active and non-B-ac- tive materials had values of RIMsO. and RIM50.38, respectively, confirming that any oligosaccharides not eluted by NaOH from the ion-exchange resins were probably not of higher molecular weight since paper chromatography was carried out before de-
salting. Very little N-acetyl-ngalactosa- minitol was found in the nondialyzable por- tion. Thus all released chains were dialyza- ble, consistent with the upper size limits based on Bio-Gel P-2 and paper chromatog- raphy for included materials in the octa- to decasaccharide range and perhaps some- what larger for the oligosaccharides of the B-active material excluded from Bio-Gel P-2.
The sizes of the largest oligosaccharides released by base-borohydride from human ovarian cyst blood group glycoproteins have not been established but are probably in the range of 16-18 sugars (18). How- ever, the bulk of the released carbohydrate from horse materials is included in Bio-Gel P-2 while 61-65% of the reduced oligosac- charides from human HLe” or Lea ovarian cyst (11) are either nondialyzable or ex- cluded from Bio-Gel P-2. The oligosaccha- ride moieties of blood group A- and H-
SIZE DISTRIBUTION OF OLIGOSACCHARIDES 533
TABLE III
AMINO ACID COMPOSITION OF B-ACTIVE AND NON-B-
ACTIVE HORSE BLOOD GROUP SUBSTANCES BEFORE
AND AFTER TREATMENT WITH 0.05 N NaOH IN 1 M
NaBH,
Amino acid Grams/l00 grams of blood group substance
B-active Non-B-active
Before After Before After
Lysine 0.7 0.5 1.1 0.8 Histidine 0.8 0.7 0.9 0.9 Arginine 1.6 0.8 1.5 0.9 Aspartic acid 1.0 0.8 1.7 1.3 Threonine 5.2 0.8 5.0 1.1 Serine 3.6 0.8 3.3 0.9 Glutamic acid 2.3 1.7 3.3 2.7 Proline 3.6 3.3 3.5 2.8 Glycine 0.9 0.8 1.1 1.0 Alanine 1.5 2.1 1.7 2.3 a-Aminobutyrate - 2.4 - 2.2 Valine 1.5 2.1 1.7 2.3 Isoleucine 0.3 0.3 0.3 0.3 Leucine 0.8 0.5 0.8 0.6 Phenylalanine 0.4 0.3 0.4 0.4
active glycoproteins from hog gastric mu- cosae have been estimated to be about 14 sugar residues or larger (40, 41). Studies of human gastric mucosal glycoproteins by sequential Smith degradations indicated carbohydrate chains at least as long as those found in human ovarian cyst mate- rials (35, 42). The existence af.*substan- tially smaller carbohydrate chains of horse blood group-active glycoproteins raises the possibility that such differences contribute to species specificity in the immunochemi- cal properties of B-active materials from cow and horse noted earlier (2, 19, 37, 43).
The nGalNAc in the dialyzable plus non- dialyzable fractions is 12.9% of that origi- nally present in the B-active and 16.0% of that in the non-B-active materials. These NaOH-NaBH,-resistant residues might represent chains linked to N-terminal-or C-terminal serine or threonine (44).
The less than quantitative conversion of serine to alanine and threonine to cu-amino- butyric acid has generally been found in NaOH-NaBH, treatment of glycoproteins containing oligosaccharides linked through nGalNAc to serine or threonine
(9, 33, 34, 36, 42) and also with model glycodipeptides (44). It was suggested that NaOH-NaBH, causes conversion of unsat- urated derivatives of serine and threonine to the corresponding amino alcohols (44). Moreover, under relatively mild condi- tions, NaBH, may cause reduction of pep- tide bonds and hydrolysis would then yield amino alcohol derivatives (45). This could explain the destruction of amino acids by base-borohydride treatment in this and a previous study (9). Alternatively, the alka- line conditions used here may have caused destruction of unsubstituted hydroxy- amino acids (46).
ACKNOWLEDGMENTS
We acknowledge the helpful advice of Dr. Allen M. Gold with the amino acid analyses.
REFERENCES
1. NEWMAN, W., AND KABAT, E. A. (1976) Arch. Biochem. Biophys. 172, 510-523.
2. BAER, H., KABAT, E. A., AND KNAUB, V. (19501 J. Exp. Med. 91, 105-114.
3. IYER, R. N., AND CARLSON, D. M. (1971) Arch. Biochem. Biophys. 142, 101-105.
4. ANDERSON, B., Rovrs, L., AND KABAT, E. A.
(1972) Arch. Biochem. Biophys. 148, 304-314. 5. CARLSON, D. M. (1974) in Methodologie de la
Structure at du Metabolisme de Glycoconju- gues, pp. 249-254, Centre National de la Re- cherche Scientifique, Paris.
6. LLOYD, K. O., AND KABAT, E. A. (1967) Car- bohyd. Res. 4, 165-177.
7. ANDERSON, B., HOFFMAN, P., AND MEYER, K. (1963) Biochim. Biophys. Actu 74, 309-311.
8. TANAKA, K., BERTOLINI, M., AND PIGMAN, W. (19641 Biochem. Biophys Res. Commun. 16, 404-409.
9. KABAT, E. A., BASSETT, E. W., PRYZANSKY, K., LLOYD, K. O., KAPLAN, M., AND LAYUG, E. (1965) Biochemistry 4, 1632-1638.
10. BERTOLINI, M., AND PIGMAN, W. (19701 CUF- bohyd. Res. 14, 53-63.
11. ROVIS, L., ANDERSON, B., KABAT, E. A.,
GRUEZO, F., AND LIAO, J. (1973) Biochemistry 12, 1955-1961.
12. MCAULIFFE, F. M. (1975) Ph.D. Thesis, Colum- bia University.
13. KABAT, E. A. (1961) Kabat and Mayer’s Experi- mental Immunochemistry, 2nd ed., C. C Thomas, Springfleid, III.
14. SCHIFFMAN, G., KABAT, E. A., AND THOMPSON,
W. (1964) Biochemistry 3, 113-120. 15. LUDOWIEC, J., AND BENMAMAN, J. D. (1967)
Anal. Biochem. 19, 80-88.
534 NEWMAN AND KABAT
16. EASTOE, J. E. (1972) in Glycoproteins (Gotts-
chalk, A., ed.), 2nd ed., Part A, pp. 158-207,
Elsevier, Amsterdam.
17. SPACKMAN, D. H., STEIN, W. H., AND MOORE, S.
(1958) Anal. Chem. 30, 1190-1206.
18. Rows, L., ANDERSON, B., KABAT, E. A.,
GRUEZO, F., AND LIAO, J. (1973) Biochemistry
12, 5340-5354.
19. ALLEN, P. Z., AND KABAT, E. A. (1959) J. Immu-
nol. 82, 340-357.
20. FEIZI, T., KABAT, E. A., VICARI, G., ANDERSON,
B., AND MARSH, W. L. (1971) J. Exp. Med. 133,
39-52.
21. VICARI, G., AND KABAT, E. A. (1969) J. Zmmu- nol. 102, 821-825.
22. KABAT, E. A., AND SCHIFFMAN, G. (1962)J. Im-
mu&. 88, 782-787. 23. LUNDBLAD, A., HAMMARSTROM, S., LICERIO, E.,
AND KABAT, E. A. (1972) Arch. Biochem. Bio- phys. 148, 291-303.
24. WATKINS, W. M., AND MORGAN, W. T. J. (1956)
Nature (London) 178, 1289-1290.
25. ROVIS, L., KABAT, E. A., PEREIRA, M. E. A., AND
FEIZI, T. (1973) Biochemistry 12, 5355-5360. 26. KABAT, E. A. (1962) Arch. Biochem. Biophys.
Suppl. 1, 181-186.
27. ZOPF, D. A., AND GINSBURG, V. (1975) Arch. Biochem. Biophys. 167,345-350.
28. LLOYD, K. O., KABAT, E. A., AND BEYCHOK, S.
(1969) J. Immunol 102, 1354-1362.
29. ETZLER, M. E., ANDERSON, B., BEYCHOK, S.,
GRUEZO, F., LLOYD, K. O., RICHARDSON, N.
G., AND KABAT, E. A. (1970) Arch. Biochem. Biophys. 141, 588-601.
30. PEREIRA, M. E. A., AND KABAT, E. A. (1974)
Biochemistry 13, 3184-3192.
31. MARCUS, D. M., AND GROLLMAN, A. P. (1966)J.
Immunol. 97, 867-875. 32. NEWMAN, W., AND KABAT, E. A. (1976) Arch.
Biochem. Biophys. 172. 535-550.
33. FUKUDA, M., AND OSAWA, T. (19731 J. Biol. Chem. 248. 5100-5105.
34. DONALD, A. S. R. (1973) Biochim. Biophys. Acta 317, 420-436.
35. HOUGH, L., JONES, J. V. S. AND Ko, A., (1974) in Methodologie de la Structure et du Metabo-
lisme des Glycoconjugues, pp. 255-262,
Centre National de la Recherche Scientifique,
Paris.
36. ROUSSEL, P., LAMBLIN, G., DEGAND, P.,
WALKER-NASIR, E., AND JEANLOZ, R. W.
(1975) J. Biol. Chem. 250, 2114-2122.
37. LESKOWITZ, S., AND KABAT, E. A. (1955) J. Im- munol. 75, 171-177.
38. FEIZI, T., AND KABAT, E. A. (1974) J. Immunot.
112, 145-150.
39. WALKER, E., ROUSSEL, P., JEANLOZ, R. W., AND
REINHOLD, V. N. (1974) Carbohyd. Res. 35, 270-279.
40. SLOMIANY, B. L., AND MEYER, K. (1972)J. Biol.
Chem. 247, 5062-5070. 41. SLOMIANY, B. L., AND MEYER, K. (1973) J. Biol.
Chem. 248, 2290-2295. 42. OATES, M. D. G., ROSBOTTOM, A. C., AND SCHRA-
GER, J. (1974) Carbohyd. Res. 34, 115-137.
43. BEISER, S. M., AND KABAT, E. A. (1952) J.
Immunol. 68, 19-40. 44. WAKABAYASHI, K., AND PIGMAN, W. (1974) Car-
bohyd. Res. 35, 3-14.
45. PAZ, M. A., HENSON, E., ROMBAUER, R.,
ABRASH, L., BLUMENFELD, 0. O., AND GAL- U~P, P. M. (1970)Biochemistry 9,2123-2127.
46. MCGUIRE, E. J., AND ROSEMAN, S. (1967) J. Biol.
Chem. 242, PC3745-3747.