THEJOURNALOF BIOLOGICAL

CHEMISTRY
0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 263, No. 2, Issue of January 15, pp. 728-734, 1988

Printed in U.S.A.

Binding Propertiesof a Mannose-specific Lectin from the Snowdrop

(Galanthus nivalis) Bulb*
(Received for publication, April 7, 1987)

Naoto Shibuya$,Irwin J. Goldstein$, Els J. M. Van DammeTII , and WillyJ. Peumansy

From the $Departmentof Biological Chemistry, University of Michigan, Ann Arbor, Michigan 48109 and the TLaboratorium voor
Plantenteelt, Katholieke Uniuersiteit Leuven, Fakulteit der Landbouwwetenschappen, Kardinaal Mercierlaan 92, B-3030 Leuven
(Heuerlee), Belgium

Carbohydrate binding properties of a new plant lectin (GNA) isolated from snowdrop bulbs were studied
using the technique of quantitative precipitation, hapten inhibition, and affinity chromatographyonimmobilized lectin. Purified GNA precipitated highly
branched yeast mannans but did not react with most
glucans. Hapten inhibition experiments showed that Dmannose is an inhibitor of GNA-mannan interaction
but neither N-acetyl-D-mannosamine nor D-glucose is
an inhibitor. Hapten inhibition with various sugars
showed that GNA requires the presence of equatorial
hydroxyl groups at the C-3 and C-4 positions and an
axial group at the C-2 position of the D-pyranose ring.
A nonreducing terminal D-mannose residue is necessary for the interaction of oligosaccharides, and oligosaccharides with terminal Man(cr-1-3)Man units
showed the highest inhibitory potency (10-30 times
greater than D-mannose) among the manno-oligosaccharides tested. The presence of the hydrophobic p nitrophenyl aglycone increased the affinity of D-mannose only slightly. Immobilized GNA boundyeast mannan but did not bind glycogen. The behavior of glycoproteins with high mannose type glycan chains depended onthe density and the structure of their glycan
chains. Glycopeptides which carry Man(cr1-3)Man
units were retarded on the immobilized GNA column
whereas those lacking this unit or with hybrid type
glycan chains were not retarded on the column.
Plant lectinsof established carbohydrate binding
specificity
have proved to be valuable reagents for the detection, isolation, and characterization of glycoconjugates (1, 2). Although
most well characterized plant lectinshave been isolated from
seeds, especially those of legumes, the numberof plant lectins
isolated from the other partsof the plant hasbeen increasing
(3-12). These nonseed lectins have been studied principally
for the interest regarding their physiological role in plants,
but they may also be a good source of new plant lectins of
unique carbohydrate-binding properties. For example, we recently showed that a lectin obtained fromelderberry bark
* This work was supported in partby National Institutes of Health
Grant GM 29470 and grants from the National Fund for Scientific
Research (Belgium), of which W. J. P.is a Senior ResearchAssociate.
The costs of publication of this article were defrayed in part by the
payment of pagecharges. Thisarticlemusttherefore
be hereby
marked advertisement in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
Present address: National Food Research Institute, Ministry of
Agriculture, Forestry, andFisheries, 2-1-2 Kannondai, Yatabe-machi,
Ibaraki 305, Japan.
11 Recipient of a fellowship fromthe Belgian Instituut totAanmoediging van het Wetenschappelijk Onderzoek in Nijverheid en Landbouw.

(10) has a unique carbohydrate binding specificity, recognizing Neu5Ac(a2-6)Gal/GalNAc sequences (13, 14). Also, a
lectin isolated from stinging nettle rhizomes (12) is an extremely small molecule ( M , = 8,500),and itstwo carbohydrate
binding sites exhibit negative cooperativity or are nonequiN,N,N,Nvalent with regard to their affinity toward
tetraacetylchitotetritol, contrary to most plant lectins (15).
Thus,these nonseed plantlectinsshould
be regarded as
unique and useful new carbohydrate binding proteins.
Snowdropbulbs (Galanthus niualis) accumulate alectin
(GNA) which agglutinates rabbit, but not human erythrocytes (16). GNA consists of four identical subunits witha
molecular weight of 50,000; it is devoid of carbohydrate. We
report herein thedetailed carbohydrate binding propertiesof
this new lectin, as examined by quantitative precipitation,
hapten inhibition, and affinity chromatography of its immobilized form.
MATERIALS ANDMETHODS

Purification and Immobilization of GNA-GNA was purified from

crude extracts of snowdrop bulbs by affinitychromatographyon
mannose-agarose as previously reported (16). An aliquot of purified
GNA was coupled to Sepharose 4B using cyanogen bromide by the
method of March et al. (17). The immobilized GNA-Sepharose contained approximately 1.3 mg of protein/ml of the gel.
Oligosaccharides and Glycopeptides-All the monosaccharides and
their methyl or p-nitrophenyl
glycosides were commercially available.
Man(a1-3)Man and its methyl andallyl glycosides were made available by Dr. K. L. Matta of Roswell Park Memorial Institute, Buffalo,
NY; al,3-linked mannooligosaccharides (degree of polymerization =
10 and 11), synthesizedby a chemical procedure, were obtained from
Dr.ConradSchuerch,
Syracuse University.Man(al-G)Man(al6)Man was the gift of Dr. C. E. Ballou of the Universityof California,
Berkeley. Man(al-6)[Man(al-3)]Man-a-O-Me
and Man,GlcNAc,Asn were provided by Dr. J . P .Carver of the University of Toronto.
Man(al-4)[Man(al-2)]-a-O-Me
was given by Dr. T. Ogawa of the
Institute of Physical and Chemical Research, Japan. Glycosyl asparagines, GP-IIIA (Man5GlcNAc4-Asn), GP-IIIB(Man,GlcNAc,-Asn),
GP-IIIC(Man4GlcNAc5-Am),GP-IV(Man6GlcNAcp-Asn),and
GP-V (Man5GlcNAc,-Asn), obtained from ovalbumin, were the gifts
of Dr. H. Nomoto and Dr.
Y. Inoue of Tokyo University, Japan; these
oligosaccharides were used for affinity chromatographic analysis after
labeling with 3Has described below. GP-IV and-V were also provided
by Dr. C. F. Brewer of Albert Einstein College of Medicine, New
York and used for the hapten inhibition studies.Mannooligosaccharide derivatives(Man,-a-0-RorManS-a-O-R;
R=(CH2)8-COOMe)
which carried mannose 6-phosphate as a nonreducing end were supplied by Drs. J. Distler and G. W. Jourdian
of the Universityof
Michigan and treated with alkaline phosphatase before use. Other
sugars and derivatives were available from previous studies.
Polysaccharides and Glycoprotein.-Yeast (ga1acto)mannans from
various strains and a linear a-1,3-linked mannan were supplied by
Dr. P. A. J. Gorin of Priarie Regional Laboratory, National Research
The abbreviations used are: GNA, G. nivalis plant lectin; BSA,
bovine serum albumin.

728

729

Carbohydrate Binding Properties

of Snowdrop Lectin
Council of Canada. Dextran B-1355-S, glycogen, and pullulan were
available from previous studies. Elsinan was the gift of Dr. A. Misaki
of Osaka City University, Japan. Fetuin and ovalbumin were purchased from Gibco Laboratories. Porcine thyroglobulin was purchased from Sigma. Orosomucoid was obtained from Dr. G. W.
Jourdian of this university. Lima bean lectin was prepared by Dr. D.
D. Roberts (National Institutes of Health). Soybean lectin was purchased from EY Laboratories, San Mateo, Ca. Miscellaneous synthetic glycoproteins were available from previous studies.
Periodate Oxidation of a-1,3-Linked Mannooligosacchuride-An
al,3-linked mannooligosaccharide (degree of polymerization = 11, 9
mg) was dissolved with 1 ml of 50 mM NaI04 solution and oxidized
for 72 h at 4 "C. After destruction of excess periodate by the addition
of etbyleneglycol, the reaction product was reduced with sodium
borohydride at 4 "C overnight. Oxidized mannooligosaccharide was
purified by gelfiltration on Bio-Gel P-2. Oxidation of the nonreducing
end mannose residue was confirmed by the fact that the oxidized
mannooligosaccharide was not digested by jack bean a-mannosidase
but digested by the same enzyme after the removal of the oxidizedreduced residue by mild acid hydrolysis (pH 1.1 overnight at room
temperature).
Labeling of Glycosylasparagine Glycopeptides-Radioactive glycosylasparagines were prepared by reductive methylation (18) as described by Ohyama et al. (19). Labeled glycosylasparagines were
purified by gel filtration on Bio-Gel P-2.
Precipitation and Hapten Inhibition Reaction-Precipitation reactions were conducted by the method of So and Goldstein (20). Varying
amounts of polysaccharides or glycoproteins were incubated with 20
pgof GNA in a total volume of 150 pl. After 1 h of incubation at
37 "C, the reaction mixture was kept at 4 "C for 48 h, centrifuged,
and analyzed for protein in the precipitates. For hapten inhibition
experiments, varying amounts of sugars were added to the reaction
mixture consisting of 20 pg of GNAand 20 pg of Hansenula capsulnta
mannan in a total volume of 150 pl. Protein was determined by the
method of Lowry et al. (21) using bovine serum albumin as standard
protein. Ouchterlony double diffusion analysis was conducted by the
method of Goldstein and So (22).
Affinity Chromatography-Glycosylasparagines (0.2-0.5 nmol),
glycoproteins (1 mg), or polysaccharides (1 mg) were applied to the
GNA-Sepharose column (0.7 X 15 cm) followedby elution with
phosphate-buffered saline or phosphate-buffered saline containing
0.5 M methyl a-D-mannoside. Elution was monitored either by measuring radioactivity, absorption at 280 nm, or color intensity by
phenolsulfuric acid assay (23). All experiments were performed at
4 "C. To detect the polysaccharides eluted with methyl a-D-mannoside, fractions were dialyzed prior to the phenol-sulfuric acid assay.

2o

15

IO

20

40

60

80

100

20
40
60
80
100
POLYSACCHARIDE
ADDED
(pg)

FIG. 1. Quantitative precipitation of yeast mannans by

snowdrop bulb lectin (GNA). Varying amounts of mannan were
allowed to react with 20 pg of GNA. The amount of protein precipitated in each tube was determined by the method of Lowry et al. (21).
A, S. cereuisiue; 0,
H. capsulata; 0, P. pastoris; A, C. lipolytica; a-

1,3-mannan from 2'. cutaneum.

.,

the highly branchedS. cereuisiae mannan (data not included).

Interestingly, GNA did not react with glycogenor dextran
(Fig. 2), in contrast to concanavalin
A, which reacts with both
branched a-mannans and a-glucans(2).Elsinan, a fungal
glucan having a repeating unit of -3)Glc(al-4)Glc(al-4)GlcRESULTS
(a1-(27),was rather exceptional, reacting weakly with GNA.
Precipitation of Polysaccharides and Glycoproteinswith
Inhibition of Precipitin Reaction by Haptenic Sugars-To
GNA-Preliminary studies of the ability of various polysac- determine the detailed carbohydrate binding specificity
of
GNA, we conducted hapten inhibition studies using H. capcharides and glycoproteins to interact with GNA were made
by the Ouchterlony double diffusion technique. Only various suluta mannan as the precipitating polysaccharide. The H.
yeast (phospho)mannans and some synthetic mannans gave
capsulata mannan was selected for this purpose because
of its
distinctprecipitationlines,whereasnaturalandsynthetic
as alreadydescribed.
high sensitivity to hapten inhibition
glycoproteins such as fetuin, orosomucoid, ovalbumin, p-azo- Concentrations of sugars required for 50% inhibition were
phenyl a-D-mannoside-BSA, p-azophenyl maltoside-BSA, p- obtained from complete inhibition curves
and are listed in
azophenyl lactoside-BSA, p-azophenyl a-D-galactoside-BSA,
Tables I and 11. Among the common monosaccharides tested,
and melibionate-BSA did not react with GNA. Fig. 1 shows only D-mannose was inhibitory. Epimers of D-mannose, i.e.
the precipitation curves
of various polysaccharides with GNA. glucose (C-2), altrose (C-3), and talose (C-4), were all noninHighly branched mannans obtained from Saccharomyces cerehibitory. The 2-deoxy a n d 3-deoxy derivatives of mannose
visiae and Pichia pastoris reacted strongly with GNA; on the werealsononinhibitory,whereasboth6-deoxy-~-mannose
contrary, the reactivity of both a linear a-1,3-mannan from
and -D-lyXOSe were weak inhibitors. These results suggest that
Trichosporon cutaneum (24) and the galactomannanof Can- the configurations of the free hydroxyl groupsat the C-2, -3,
dida lipolytica in which most
of the terminal mannose residues and -4 positions of D-mannose are necessary for interaction
are substituted witha-D-galaCtOSyl residue (25) was verylow,
with GNA. The importanceof the hydrogen atom of the C-2
suggesting the importanceof nonreducing terminal mannosyl hydroxyl group is indicated inasmuch as 2-deoxy-2-fluoro-Dresidues for the interaction. A mannan preparation obtained
of the amino sugars tested
mannose was not inhibitory. None
from H. capsulata also reacted with GNA, although it was
inhibited the precipitation reaction. Reductionof D-mannose
reported to be a linear polysaccharide (26); the much lower
to D-mannitol by sodium borohydride completely destroyed
reactivity of this mannan with GNA compared to other highly
its ability to inhibit the precipitation reaction.
branched yeast mannans was evident from the result of t h e
Methyl a-D-mannopyranose was a somewhat better and
inhibition experiment in which the precipitation of H. capmethyl P-D-mannopyranoside a somewhat poorer inhibitor
a muchlower
sulata mannan with GNA was inhibited by
than D-mannose. Introduction of the nonpolar p-nitrophenyl
concentration (1/40) of methyl a-D-mannoside compared to
aglycone into D-mannose only slightly enhanced its inhibitory

mannoside

Carbohydrate Binding Properties of Snowdrop Lectin

730

!F 5
O
Iw

1 20!

40
#

s 60 5 80s

1i00

.,

POLYSACCHARIDE ADDED ( r u g )

FIG. 2. Quantitative precipitation of glucans by GNA. Varying amountsof glucans were allowed to react with20 pg of GNA and
analyzed as in Fig. 1. A, elsinan; pullulan; 0, glycogen or dextran.
TABLE
I
Inhibition of GNA-H. capsulata mannan precipitation by methyl and
p-nitrophenyl glycosides
There was no inhibition by 100 mM D-glucose (C-2 epimer), Daltrose (C-3 epimer),D-talose (C-4 epimer), D-galactose, D-fucose, Lrhamnose, L-arabinose, D-XylOSe, 2-deoxy-~-mannose,2-deoxy-2-fluoro-D-mannose,3-deoxy-D-mannose, N-acetyl-D-mannosamine, N acetyl-D-glucosamine, N-acetyl-D-galactosamineand D-mannitol.
Concentration for
50% inhibition

Sugar

mM

D-Mannose
D-Lyxose
6-Deoxy-D-mannose

17
12% inhibition a t 100 mM

Methyl
Methyl 8-D-mannoside
p-Nitrophenyl a-D-mannoside
p-Nitrophenyl p-D-mannoside
p-Nitrophenyl a-D-glucoside
p-Nitrophenyl P-D-glucoside
p-Nitrophenyl a-D-galactoside
p-Nitrophenyl 8-D-galactoside
p-Nitrophenyl2-O-methyl-aD-mannoside

11
21

No
No
No
No

No

6
9
inhibition at 15mM
inhibition at 25 mM
inhibition a t 25 mM
inhibition a t 25 mM
inhibition a t 15 mM

potency (about 2-fold compared to thecorresponding methyl

glycosides, Table I).
Among the oligosaccharides tested, those carrying terminal
Man(a1-3)Man units were the most potent inhibitors and
exhibited 10-30 times greater potency for GNA compared to
D-mannose (Table 11). Mannooligosaccharides linked by a1,6 linkages were also fair inhibitors, being 4-6 times better
than D-mannose. On the other hand, mannooligosaccharides
with an a-1,2 or a-1,4 linkage showed only 2-3 times higher
inhibitory potency compared to D-mannose. Mannooligosaccharides with fi-linkage generally showed much lower inhibitory potency compared to those with a-linkages.
Various branched oligosaccharides or glycosylasparagines
containing more than two terminal mannoseresidues showed
an affinity similar to the unbranched oligosaccharides with
the same terminal structural units and
did not show any
enhancement of affinity which has been reported for the
interaction of some branched oligosaccharides and plant lectins such as theDatura stramonium seed lectin (28).
Substitution of a terminal mannose residue by glucose or
galactose completely destroyed the inhibitory potencyof the
original sugars, suggesting a strict requirement for terminal
nonreducing mannose units. This was further confirmed by
comparing the inhibitorypotency of an a-1,3-linked mannooligosaccharide before andafterperiodate
oxidation and

NaBH4 reduction.After the oxidativemodification of the

terminal mannose residue, the oligosaccharide was no longer
inhibitory.
Binding Characteristicsof Immobilized GNA-An aliquot of
GNA was immobilized onto cyanogenbromide-activated
Sepharose 4B (GNA-Sepharose), and its ability to bind
polysaccharides,glycoproteins, and glycopeptides was investigated. Yeast mannan bound strongly to the
GNA column and
was eluted with methyla-D-mannoside whereas glycogen
passed throughthesame
column (Fig. 3). Both of these
polysaccharides bind to a concanavalin A-Sepharose column
(29, 30).
Various glycoproteins carrying high mannose type glycan
chains behaved differently on GNA-Sepharose (Fig. 4). Porcine thyroglobulinwhich carries numeroushigh mannose type
glycan chains in its molecule (31) bound strongly to GNASepharose and was eluted by methyl a-D-mannoside. Lima
bean lectin which carries a terminal Man(a1-3)Man unit on
both of its two glycan chains in each subunit (32) also bound
tothis column. Ontheotherhand,
soybeanlectin,each
subunit of which also contains two glycan chains terminating
in Man(a1-2)Man units (33), was only slightly retarded on
the GNA column. Ovalbumin which carries only one glycan
chain of various high mannose or hybrid type chains in its
molecule (34-36) did not shown any retardation on theGNASepharose column.
Elution profiles of a series of glycosyl asparagine glycopeptides which were obtained from ovalbumin and labeled with
3H by reductive methylation (18, 19) areshown in Fig. 5, and
their structures areshown in Table 111. Among these glycosyl
asparagines, GP-V which contains two Man(a1-3)Man units
showed the greatest retardation on GNA-Sepharose, with a
very broad peak. GP-IIIB and GP-IV, both of which carry
only one Man(a1-3)Man unit,also were retarded, but less so
than GP-V. On the other hand, Man9GlcNAcz-Asn which
contains only Man(a1-2)Man units in its peripheral portion
showed no retardation. Interestingly,two hybrid-type glycosyl
asparagine glycopeptides, GP-IIIAandGP-IIIC,
were not
retarded although they each contain one Man(a1-3)Man unit.
It is possible that the bisecting or terminal GlcNAc residues
in these molecules sterically hindertheinteraction
of
Man(a1-3)Man units withGNA.
DISCUSSION

The results obtained in this study demonstrate that GNA

is specific for D-mannosecompared to the
so-called mannose/
glucose-specific lectins such as concanavalin A, or the pea,
lentil, andVicia faba lectins (2) in which D-mannose was only
2.5-5 times more inhibitory than D-glucose (2). On the other
hand, D-glucose (100 mM) did not inhibit GNA-yeast mannan
precipitation.This conclusion was also supported by the
selective reactivity of GNA with mannans or mannose-containing glycoproteins by quantitative precipitation and affinity chromatography. Elsinan, a fungal a-D-glucan (27), was
the only exception, reacting very weakly with GNA (Fig. 2),
perhaps because of its content of Glcal,3Glc linkages; nevertheless, nigerose (Glcal,3Glc) was a noninhibitor.
A strict requirementof GNA for nonreducing end mannose
residues is also evident from the fact that the
oligosaccharides
in which terminal mannose residues were blocked by other
sugar residues were not inhibitory (Table 11). This requirement was also supported by the destruction of inhibitory
potency of an 1,3-linked mannooligosaccharides by modification of its nonreducing terminal mannose residue by periodate oxidation.
The resultsof hapten inhibition experimentsindicated that
the Man(a1-3)Mandisaccharide unit is most complementary

Carbohydrate
Binding
Properties

of Snowdrop
Lectin

731

TABLEI1
Inhibition of GNA-H. capsuluta mannan precipitation by oligosaccharides
Oligosaccharide

Concentration for
50% inhibition
mM

Man(a1-2)Man
Man(al-2)Man(al-2)Man
Man(p1-2)Man
Man((31-2)Man@l-2)Man
Man(a1-3)Man
Man(a1-3)Man-a-0-Me
Man(a1-3)Man-a-0-allyl
Man(al-3)Man(pl-4)GlcNAc
Man(a1-3)Man(al-2)Man(al-2)Man
Man(a1-4)Man(S-benzylglycoside)
Man(p1-4)Man
Man(/31-4)Man(@l-4)Man
Man(a1-6)Man
Man(al-G)Man(al-G)Man
Man(a1-6)Glc

8
5
68
42
1.4
1.2
1.4
0.6
1.6
9
18
27
4
3
9

Glc(a1-2)Man
Glc(B1-4)Man
Gal(al-2)Man(a1-2)Man
Glc(c~l-3)Glc
Glc(al-4)Glc
Gal(a1-6)Glc

No inhibition at 20 rnM
No inhibition at 30 m M
No inhibition at 10 mM
No inhibition at 40 mM
No inhibition at 100 mM
No inhibition at 100 mM

a-1,3-MannooligomeP
Oxidized a-1,3-Mannooligomerb

3.6
No inhibition a t 2 mM

Man(al-2)Man(al-3)Man-R
Man(al-2)Man(al-6)Man-R

2.8
3.5

Man(al-2)Man(a1-3)

Man-R

10% inhibition at 1 mM

Man(al-2)Man(al-6)
Man(a1-4)
\

Man-a-0-Me

24% inhibition at 4 mM

Man(a1-2)
Man(a1-6)

\
Man-a-0-Me

0.6

Man(a1-3)
Man(a1-6)

\
Man(a1-6)

Man(a1-3)

0.6

Man(@l-4)GlcNAc(@l-4)GlcNAc-Asn

Man(a1-3)
Man(a1-6)

\
Man(a1-6)

Man(a1-3)

0.55
Man(/31-4)GlcNAc(@l-4)GlcNAc-Asn

/
Man(al-P)Man(al-3)
D.P. = 10 (Ref. 39).
Periodate-oxidized and NaBH4-reduced.
e R = -O-(CH,),-COOCH,.

732

Carbohydrate Binding Properties of Snowdrop Lectin

S. cerevisiae mannan

TABLE
111

Structures of glycosylasparagine glycopeptides assayed for their

behauwr on an immobilized G. niualis lectin column
R = GlcNAc(B1.4)GlcNAc(~l.~Asn.
Olieosaccharide

Structure

Man9GlcNAcs Man(a1,2)Man(al,G)

Man(a1,G)

Man(@l,4)R
Man(al,P)Man(al,B)

Man(al,2)Man(a1,2)Man(al,3)
FRACTION NUMBER
(30 drops /tube)

GP-IIIA

FIG. 3. Elution profile of yeast mannan and glycogen on

GNA-Sepharose. Polysaccharides (1mg) were applied to the GNASepharose column (0.7 X 15 cm)followed by elution with phosphatebuffered saline. The arrow indicates the addition of 0.5 M methyl aD-mannopyranoside intheeluent.Themethyl
glycoside ineach
fraction was removed by dialysis before the assay for total carbohydrate.

GlcNAc((31,4)
Man(a1,G)

Man(a1,G)

Man(al,3)

Man(/31,4)R

GlcNAc(Bl,Z)Man(al,3)
GP-IIIC

GlcNAc(B1,4)

Man(al,3)Man(al,G)

Ovalbumin

GlcNAc(o1,4)

Me u -E-mannoside

Man(pl,4)R

Man(al,3)

GlcNAc(B1,2)
GP-IIIB

Man(a1,2)Man(al,G)

Man(a1,G)

Man(al,3)

Man(al,B)Man(a1,3)

GP-IV

co

Man(a1,G)

(u

Man(a1,G)

/
Man(@l,4)R

Man(B1,4)R

Man(al,3)

Man(a1,2)Man(al,3)
GP-V

Man(a1,G)

Man(a1,G)

Man(pl,4)R

Man(al,3)

Man(al,3)

10

20

30

40

50

FRACTION NUMBER
FtG. 4. Elution profile of some glycoproteins on GNA-Seph-

arose. Conditions were similar to those of Fig. 3. The arrow indicates

the addition of 0.5 M methyl a-D-mannopyranoside in the eluent.

to the carbohydrate binding site of GNA, being 10-30 times

more potent than D-mannose. Neither the elongation of the
1,3-linked chain nor thepresence of a second Man(a1-3)Man
unit in a branched oligosaccharide or glycopeptide increased
its affinitysignificantly.

The preference of GNA for 1,3-linkages was also demonstrated by the affinity chromatographic analysis of several
high mannose type glycosyl asparagine compounds in which
glycopeptides containing Man(a1-3)Manlinkages showed retardation on GNA-Sepharose(Fig. 5 ) . It is interesting to note
that the degree of retardation was significantly different depending on the numberof these disaccharide units, although
the inhibition potency
of these glycosyl asparagines estimated
from hapten inhibition experiments showed no significant
differences in their affinity for GNA. There may be some
differences inthe binding mechanism
between the interaction
of the lectin with the glycopeptides in solution (hapten inhibition) and ona solid surface (affinity chromatography).
Thus, GNA is a unique plantlectin recognizing terminal D-

Carbohydrate Binding
Properties

of Snowdrop
Lectin

733

mannose groups, especially those possessing Man(a1-3)Man

units. Recently, Oda and Minami (37) reported that a lectin
isolated from tulip bulbs was also specific forD-mannose,
although this lectin isdifferent fromGNA initsdetailed
binding specificity. Forexample,it
preferred a-1,6-linked
mannooligosaccharides over 1,3-linked ones, and it also reacted specifically with the mannans isolated from the yeasts
belonging to thegenus Saccharomyces. A bacterial cell surface
lectin, Escherichia coli type I lectin, which is present as a
component of fimbriae, appears to have a binding specificity
(38) similar to GNA. From its unique carbohydrate binding
specificity, GNA shouldprove useful for the analysis and
purification of glycoconjugates, especially those which carry
high mannose type glycan chains.

4
2

REFERENCES

2
rr)

0
*

E
Q
V

1. Lis, H., and Sharon, N. (1984) in Biology of Carbohydrate (Ginsburg, V., and Robbins, P., eds) Vol.2, pp. 1-85, John Wiley
and Sons, New York
2. Goldstein, I. J., and Poretz, R. D. (1986) in The Lectins (Goldstein, I. J., and Sharon, N., eds) pp. 33-247, Academic Press,
Orlando, FL
3. Talbot, C. F., and Etzler, M. E. (1978) Biochemistry 17, 14741479
4. Lamb. J. E.. Shibata.. S.,. and Goldstein, I. J. (1983) Plant Physiol.
7 1,'879-887
5. Kiluatrick. D. C. (1980) Biochem. J . 185. 269-272
6. Allen, A. K., and'Neuberger, A. (1973) Biochemistry 135, 307314
7. Dazzo,F. B., Yanke, W. E., and Brill, W. J. (1978) Biochim.
Biophys. Acta 539,267-286
8. Peumans, W. J., Nsimba-Lubaki, M., Carlier, A. R., and Van
Driessche, E. (1984) Planta 160, 222-228
9. Horejsi, V., Haskovec, C., and Kokourek, J. (1978) Biochim.
Biophys. Acta 532,98-104
10. Broekaert, W. F., Nsimba-Lubaki, M., Peeters, B., and Peumans,
J. (1984) Biochem. J. 2 2 1 , 163-169
11. Peumans. W. J.. Nsimba-Lubaki, M., Peeters, B., and Broekaert,
W. F. (1984) Planta 164, 75-82
12. Peumans. W. J.. DeLev. M.. and Broekaert. W. F. (1984) FEBS
Lett. -177,99-103
13. Shibuya, N., Goldstein, I. J., Broekaert, W. F., Nsimba-Lubaki,
M.. Peeters,. B... and Peumans, W. J. (1987) J . Biol. Chem. 262,
1596-1601
14. Shibuva. N.. Goldstein. I. J.. Broekaert. W. F., Nsimba-Lubaki.
M.,"Peeters, B., a n d Peumans, W. J. (1987) Arch. Biochem:
Biophys. 254, 1-8
15. Shibuya, N.,Goldstein, I. J., Shafer, J. A,, Peumans, W. J., and
Broekaert, W. F. (1986) Arch. Biochem. Bzophys. 249,215-224
16. Van Damme, E. J. M., Allen, A. K., and Peumans, W. J. (1987)
FEBS Lett. 215,140-144
17. March, S. C., Parikh, I., and Cuatrecasas, P. (1974) Anal.
Biochem. 60, 149-152
18. Means, G.E., and Feeney, R. E. (1968) Biochemistry 7,21922201
19. Ohyama, Y., Kasai, K., Nomoto, H., and Inoue, Y. (1985)J. Biol.
Chem. 260,6882-6887
20. So, L. L., and Goldstein, I. J. (1967) J. Biol. Chem. 2 4 2 , 16171622
21. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.
(1951) J. Biol. Chem. 193, 265-275
22. Goldstein, I. J., and So, L. L. (1965) Arch. Biochem.Biophys.
111,407-414
23. Dubois, M., Gilles, K.A., Hamilton, J. K., Rebers, P. A., and
Smith, F. (1956) Anal. Chem. 28, 350-356
24. Gorin. P. A. J.. and SDencer. J . F. T. (1967) Can. J. Chem. 45,
1543-1549 '
25. Gorin, P. A. J., Spencer, J. F. T., and Magus, R. J. (1969) Can.
J. Chem. 47,3569-3576
26. Gorin, P. A. J., and Spencer, J. F. T. (1972) Can. J. Microbiol.
18,1709-1715
27. Tsumuraya, Y., Misaki, A., Takaya, S., andTorii, M. (1978)
Carbohydr. Res. 66,53-65
28. Crowley, J. F., Goldstein, I. J., Arnarp, J., and Lonngren, J.

~"
~

4
2

0.5
I
"

10

20

30

FRACTION NUMBER

(20dropsltube)
FIG. 5. Elution profile of glycosylasparagines on GNASepharose. Conditions were the same as those of Fig. 3. The arrow
indicates the addition of 0.5 M methyl a-D-mannopyranoside in the
eluent. Structures of the glycosylasparagine glycopeptides are shown
in Table 111. The dashed line indicates the elution position of D mannose.

~~~~

734

Carbohydrate
Binding
Properties

(1984) Arch. Biockm. Biophys. 231,524-533

29. Lloyd, K. 0.(1970) Arch. Biochem. Biophys. 137,460-468
30. Kennedy, J. F., and Rosevear, A. (1973) J. Chem. SOC.Perkin
Trans. I. 2041-2046
31. Tsuji, T., Yamamoto, K., Irimura, T., and Osawa, T. (1981)
Biochem. J . 195,691-699
32. Misaki, A., and Goldstein, I. J. (1977) J. Biol. Chem. 2 6 2 , 69956999
33. Dorland, L., van Halbeek, H., Vliegenthart, J. F. G., Lis, H., and
Sharon, N. (1981) J. Biol. Chem. 256,7708-7711

of Snowdrop
Lectin
34. Tai, T., Yamashita, K., Ogata-Arakawa, M., Koide, N., Muramatsu, T., Iwashita, S., Inoue, Y., and Kobata, A. (1975) J.
Bioi. Chem. 250,8569-8575
35. Tai, T., Yamashita, K., Ito, S., and Kobata, A. (1977) J. Biol.
Chem. 252,6687-6694
36. Yamashka, K., Tachibana, Y., and Kobata, A. (1978) J. Biol.
Chem. 253,3862-3869
37. Oda, Y., and Minami, K. (1986) Eur. J. Biochem. 159, 239-245
38. Lis, H., and Sharon, N. (1986) Annu. Rev. Biochem. 55,35-67
39. Kong, F., and Schuerch, C. (1984) Macromolecules 17,983-989