Table Of ContentZOOLOGICAL SCIENCE 8: 39-50 (1991) © 1991 Zoological SocietyofJapan
A 110-kDa WGA-binding Glycoprotein Involved in Cell
Adhesion Acts as a Receptor for Aggregation
Factor in Embryos of the Sea Urchin,
Hemicentrotus pulcherrimus
Jun Kamimura1 Takashi Suyemitsu2 YoichiTsumuraya3
, ,
and YasutoTonegawa4
Department ofRegulation Biology and Department ofBiochemistry,
Saitama University, Urawa, Saitama 338, Japan
—
ABSTRACT Glycoproteins that bind wheat germ agglutinin (WGA-binding glycoproteins) were
isolatedfromaTritonextractofanacetonepowderofblastulaembryosoftheseaurchin,Hemicentrotus
pulcherrimus. The Triton extract was brought to 30% saturation with ammonium sulfate. The
supernatant showed inhibitory activity directed against hemagglutination caused by WGA. The
supernatant was applied to a column of WGA-agarose and the bound fraction was eluted with
N-acetyl-D-glucosamine. The fraction that bound to WGA-agarose was applied to a column of
Sephacryl S-400 after reduction and alkylation. After fractionation on Sephacryl S-400, four
WGA-binding glycoproteins were identifiedwith molecularweights of 110-, 70-, 66- and 60-kDa after
WGA
electrophoresis on sodium dodecyl sulfate polyacrylamide gels and staining with labeled with
horseradishperoxidase(HRPO). StainingwithHRPO-labeledaggregationfactor(AF)showedthat,of
these proteins, only the 110-kDa glycoprotein bound to AF. Moreover, the trypsin fragments ofthe
110-kDaglycoproteininhibitedtheaggregationofdissociatedcellsofseaurchinembryoscausedbyAF.
Thesugarcompositionofthe 110-kDaglycoproteinindicatedthatthisproteincontainedhighlevelsofa
mannose-type oligosaccharide. These results suggest that the 110-kDa WGA-bindingglycoprotein on
the cellsurface maybe involved incell adhesion as a receptorforAF, towhich it bindsbysugar-lectin
type interactions.
mouse, while Ca2+-indipendent cell-adhesion
INTRODUCTION
mechanisms have been reported in chick embryos
Since the discovery of a Ca2+-dependent cell- [5]. Furthermore, the reactions between sugars
aggregation factor in the neural retina of chick and lectins play essential roles in intercellular
embryos [1] and sponge [2], a number of cell interactions in several organisms [6].
adhesion molecules (CAMs) have been isolated Sea urchin embryos can be dissociated into
from various animal cells and characterized [3]. constituent cells in Ca2+- and Mg2+-free seawater
Recently, Ca2+-dependent CAMs, called cadher- (CMF-SW) [7] and such cells aggregate and recon-
ins, have been reported as factors associated with stitute normal embryos upon the addition ofCa2+
morphogenesis in mouse embryos [4]. Ca2+- [8]. An aggregation factorcomplex (AFX), atype
dependent cell-adhesion mechanisms have been of CAM, has been found in the supernatant of
proposed from studies on the sponge and the dissociated cells, and it has been shown to be a
gigantic sugar-protein complex which induces the
Accepted August 10, 1990 aggregation of cells in a Ca2+-dependent manner
Received May 15, 1990 [9].
1 Present address: Department of Rheumatology, In- Recently) a Ca2+-binding protein of 1600-kDa,
Tstsiutuktuebaof30C5l.iniJcaaplanM.edicine, University of Tsukuba, namely aggregati.on factor (°ArF), was i.solated as a
2 To whom all correspondence should be addressed. subunit of AFX [10]. This AF had the ability to
4 Deceased, October 8, 1989. induce the aggregation ofcells in the same manner
40 J. Kamimura, T. Suyemitsu et al.
asdidAFXandwasboundquantitativelytoCa2+ bottom ofthe vessel by chilling in an ice bath and
,
indicative ofelectrostatic bindingbetween AFand collected.
Ca2+ (Ca2+ bridges) in cell-to-cell aggregation
[10]. However, the primary bindingofAFto cells Preparation of dissociated cells and aggregation
was suggested to involve another mecahnism, dis- factor
tinct from the Ca2+ bridges, since I125-labeled AF The concentrated suspension of embryos was
was bound to cell surfaces in the absence of Ca2+ washed twice with 5 volumes of cold Ca2+- and
[10]. The activity of AF was inhibited by specific Mg2+-free seawater (CMF-SW) by centrifugation
sugars, namely N-acetyl-D-glucosamine (GlcNAc) at 500Xg for 2min. Then the embryos were
and mannose [10]. This result demonstrates that suspended in about 200volumes ofCMF-SW sup-
AF has lectin activity and recognizes GlcNAc and plemented with penicillin (lOOIU/ml), purchased
mannose. In addition, the binding activityofcells from Meiji Seika (Tokyo, Japan), and incubated
to AFwas lost as a result ofthe treatment ofcells with gentle stirring at room temperature for 60
with trypsin [11]. Therefore, it can be concluded min, until the cells of the embryos were dissoci-
that a glycoprotein functioning as the receptor for ated. The dissociated cells were collected by
AF is localized on the cell surface and that AF is centrifugation at l,700Xg for 2min and used for
bound to the glycoprotein by asugar-lectin type of bioassays ofthe ability ofcell-surface components
interaction. to aggregate the cells. The cell-free supernatant
Recently, we found that wheat germ agglutinin was used as starting material for purification of
(WGA), a lectin which recognizes GlcNAc, in- AFX. The purification of AF from AFX was
duced dissociation of the cells of embryos of H. carried out as previously reported [10].
pulcherrimus andthatthiseffectwaseliminatedby
the addition of GlcNAc [12]. These results sug- Isolation and purification of WGA-binding gly-
gested that a WGA-binding glycoprotein on the coproteinsfrom dissociated cells
cell surface is involved in cellularadhesionand isa Dissociated cells were washed three times with
possible candidate for the receptor for AF in the cold CMF-SW and were defatted by extraction
cellsofembryosofH. pulcherrimus. Therefore, in seven times with acetone (5 ml acetone/g wet
theexperimentsreportedinthispaper, we isolated weight of cells). The defatted precipitate was
and purified four WGA-binding glycoproteins and collected by centrifugation at 1700xg for 15 min
showed that, among them, the 110-kDa glycopro- and the final pellet was dried in vacuo. The
tein is the receptor for AF on H. pulcherrimus acetone powder was used as the starting material
embryos. In addition, the amino acid and sugar forfurtherpurification ofWGA-bindingglycopro-
composition of this 110-kDa glycoprotein was teins.
analyzed and the involvement of binding of the One gram of the acetone powder was homoge-
mM
sugar-lectin type in intercellular interactions was nizedwith ateflon homogenizerin50mlof10
demonstrated. Tris-HCl buffer, pH8.0, that contained 1 M KC1,
mM
0.1 phenylmethylsulfonyl fluoride (PMSF)
and 1% Triton X-100 (extraction buffer) and the
MATERIALS AND METHODS
mixture was stirred for 24hr at 4°C. After the
extraction, the sample was centrifuged at
Preparation ofsea urchin embryos 15,000Xg for 20min. The supernatant was col-
mM
The eggs of Hemicentrotus pulcherrimus, which lected and dialyzed against 10 Tris-HCl buf-
M mM
were obtained by introduction of 0.5 KC1 into fer, pH8.0, that contained 10 NaCl and
the coelom, were cultured after fertilization at 0.01% Triton X-100.
20°C in normal seawater (NSW), at a concentra- Thedialyzedsamplewasbroughtto30% satura-
tion of6x106 eggs per liter, with gentle agitation. tionwithammoniumsulfateandstirredfor12hrat
At the hatched-blastula stage, the swimming 4°C. Then it was centrifuged at 15,000xg for 20
embryos were immobilized, made to settle at the min. The supernatant (30sup) was dialyzed
Receptor for Aggregation Factor 41
against 10mMphosphate buffer, pH7.2, thatcon-
tained 0.15 M NaCl and 0.01% Triton X-100 and Assay of the inhibition of hemagglutination by
WGA
applied to a column ofWGA-agarose (2.5 cm i.d.
X2.5cm). purchased from Seikagaku Kogyo WGA-binding activities of glycoproteins were
(Tokyo, Japan), which had previously been equil- assayed by monitoring the inhibition ofhemagglu-
brated with the same buffer. The bound material tination by WGA, according to the modified
M
was eluted with 0.5 GlcNAc in the same buffer methods of Lis et al. [14, 15]. Hemagglutination
WGA
and the eluate was monitored by measurements of activity of was first assayed by serial, two-
absorbance at 595 nm of samples subjected to fold dilutions in microtiter U-plates, in order to
staining with Coomassie brilliant blue [13]. The determine the minimum concentration of WGA
unbound fraction (WGA-0) and the bound frac- forhemagglutination. For assayofhemagglutinat-
M
tion (WGA-1) were collected, concentrated by ing activity, 25fA of 0.01 phosphate-buffered
ultrafiltration with a YM-10 membrane (Amicon saline, pH7.2, containing 0.8% NaCl and 0.02%
Corp.. Lexington, MA), and the biological activi- KC1 (PBS) and 25/A of a 2% suspension (v/v) of
ties were assayed by the method described below. trypsin-treated and formalin-fixed human type
The bound fraction (WGA-1) was dialyzed erythrocytes in PBS, were added to 25fA of ali-
against distilled water to remove Triton X-100 and quots of serial, two-fold dilutions of a solution of
WGA
then lyophilized. The lyophilized sample, equiva- in PBS. The plates were kept at room
lentto5 mg protein, wassolubilized in 1 ml of 1 M temperature for 1 hr and then the hemagglutinat-
Tris-HCl buffer, pH 8.5, thatcontained6Mguani- ing activities were examined. Subsequently, the
dine-HCl and 2mM ethylenediaminetetraacetic inhibition assay was carried out as follows. Twen-
acid (EDTA). Seven mg of dithiothreitol (DTT) ty-fivefAofPBS, containingthe minimum concen-
WGA
were addedtothe solutiontobreakdisulfidebonds tration of necessary for hemagglutination,
and the solution was then kept for 4hr at room were added to aliquots of25/A ofserial, two-fold
temperature. An aliquot of 17fA of 4-vinyl- dilutions of a solution of WGA-binding glycopro-
pyridine was added into the reaction mixture tein in PBS, and 25fA of human type erythro-
whichwasthen incubated for4hrat room temper- cytes in suspension in PBS were added after a
ature. further 30min. The mixture was kept at room
The reduced and alkylated sample was dialyzed temperature for 1 hr and then the hemagglutinat-
against distilled water to remove the various rea- ing activity was examined.
gents and lyophilized. The lyophilized sample was One unit of inhibitory activity of WGA-binding
solubilized in 1 ml of 0.2M Tris-HCl buffer, pH glycoproteins in the hemagglutination assay was
8.5, that contained 6 M guanidine-HCl and 2 mM defined as the amount ofprotein perone ml in the
EDTA. and then it was applied to a column of sample thatgave complete inhibitionofhemagglu-
Sephacryl S-400 (1.0cm i.d.X100cm), purchased tination.
from Pharmacia Fine Chemicals (Uppsala,
Gel electrophoresis
Sweden), which had been equilibrated with the
same buffer. Thecolumnwaselutedwith thesame Electrophoresis on 8% polyacrylamide gels that
bufferat aflowrateof0.5 ml/min atroomtemper- contained sodium dodecyl sulfate (SDS) was car-
ature. Theeluatewasmonitoredbymeasurements ried out by the method ofLaemmli [16]. The gels
of absorbance at 280 nm. Each fraction was con- were stained with silver [17]. A molecular marker
centrated by ultrafiltration on a YM-10 membrane kit containing fragments of cytochrome c, purch-
and dialyzed first against distilled water and then ased from Oriental Yeast (Tokyo, Japan), was
against 5 mM phosphate buffer, pH6.8, that con- used in orderforestimationsofmolecularweights.
tained 10mM NaCl and 0.01% Triton X-100.
Western blotting of WGA-binding glycoproteins
WGA
and treatment wih HRPO-labelled
WGA
labelled with horseradish peroxidase
42 J. Kamimura, T. Suyemitsu et al.
(HRPO) was prepared by the modified method of Triton X-100 and then lyophilized. Two mg of
Nakane et al. [18]. Two mg of HRPO (type VI, each lyophilized sample were suspendedin 2ml of
5,000units), purchased from Sigma Chemical Co. PBS, pH7.2, that contained 1 mM CaCl2 and 1
(St. Louis, MO, USA), were reactedwith 2 mgof mM MgCl2 and incubated with 1 mg trypsin
WGA. The proteins separated by SDS-PAGE (10,000IU), purchased from Mochida Seiyaku
were electrophoretically transferred to nitrocellu- (Tokyo, Japan), for7hrat37°C. Toterminatethe
lose sheets for Western blotting [19]. Nonspecific reaction, 0.075 mg of PMSF was added into the
binding was blocked by incubating the sheets with mixture which was then boiled for lOmin. The
0.01 M phosphate buffer, pH7.2, that contained resultant fragments of glycoproteins were precipi-
3% bovine serum albumin (BSA), 0.05% Tween- tated by the addition of ten volumes of acetone
M
20 and 0.15 NaCl (blocking buffer) for 1 hr at and collected by centrifugation at l,700Xg for 10
room temperature. The sheets were then incu- min. Acetone was removed in vacuo and the
bated with the same buffer supplemented with 0.1 trypsin fragments of glycoproteins were dissolved
mg/ml HRPO-WGA as probe for 1 hr at 4°C. A in 400/A of CMF-SW for an examination of their
controlexperimentwasperfomredbytreatmentof biological effects on sea urchin cells.
HRPO-WGA
the sheets with a solution of that
contained0.5 MGlcNAc. Afterthreewasheswith Effects of WGA-binding glycoproteins on the
10mM Tris-HCl buffer, pH7.2, that contained aggregation ofcells caused by AF
M
0.05% Tween-20 and 0.15 NaCl, the bound Effects of trypsin-treated WGA-binding gly-
HRPO-WGA on the sheets was visualized by coproteinsontheaggregationofcellswere assayed
development of color in the enzymatic reaction in 24-well culture plates, purchased from Falcon
with 0.02% diaminobenzidine (DAB) and 0.01% (Lincoln Park, New Jersey, USA), as described
H
2 2 as substrates [20]. below. Dissociated cells, prepared as described
above, were resuspended in CMF-SW adjusted to
Western blotting of WGA-binding glycoproteins pH8.0 with NaHC03 and filtered through two
and treatment with HRPO-labelled aggregation sheets ofnylon mesh (380-mesh). The population
factor ofcellsinsuspension wascountedin ahaemocyto-
Preparation of HRPO-labeled aggregation fac- meter and adjusted to about 5xl06 cells/ml.
tor, SDS-PAGE and Western blotting were car- One hundred /A of the minimum concentration
ried out as described above. After the transfer of of AF in CMF-SW necessary for aggregation of
proteins to nitrocellulose sheets, the sheets were cells, which was determined previously, were
mM
blocked by incubation with 10 Tris-HCl buf- added to 100-//1 aliquots of serial, two-fold dilu-
fer, pH8.0, that contained 3% BSA, 0.05% tions of the trypsin-treated WGA-binding gly-
M
Tween-20and0.15 NaCl (blockingbuffer) for 1 coproteins in CMF-SW, to which 300fA of Milli-
hr at room temperature. The sheets was then pore-filteredseawater (MFSW; 0.45/umporesize)
incubatedwith the same buffersupplementedwith and 100/A of the suspension of cells in CMF-SW
0.1 mg/ml HRPO-AF as probe, for 1 hr at 4°C. were finally added after shaking for 30min. The
mM
After three washes with 10 Tris-HCl buffer, platewassetonagyratoryshakerandrotatedat80
pH7.2, that contained 0.05% Tween-20 and 0.15 rpm. Afterincubation at 4°C for60min, aggrega-
M NaCl, the bound HRPO-AF on the sheets was tion of cells was examined under the light micro-
visualized as described above. scope and recorded photographically.
Treatment of WGA-binding glycoproteins with Chemical analyses
trypsin Protein was estimated by the method ofLowry
The purified WGA-binding glycoproteins were etal. [21] with BSA as a standard. Neutral sugars
treated with trypsin to solubilize them in NSW in were measuredbythe phenol-sulfuricacidmethod
the absence of Triton X-100. The samples were [22] with glucose as a standard. Sialic acid was
dialyzedagainstdistilledwaterfor48hrto remove determinedbytheresorcin-Cu2+-HClmethod [23]
Receptor for Aggregation Factor 43
with N-acetylneuraminic acid (NANA) as a stan- with the extraction buffer that contained Triton
dard. X-100exhibitedinhibitoryactivity directedagainst
Amino acid analysis was carried out with a hemagglutination by WGA, as shown in Table 1.
reaction-liquid chromatography system, (Hitachi However, the extract prepared from the acetone
655. Tokyo, Japan), after hydrolysis of samples powder with the extraction buffer without Triton
with 6N HC1 at 110°C for 24hr [24]. X-100 exhibited no such inhbitory activity.
The composition of sugars was analyzed by The extract with Triton X-100 was brought to
gas-liquid chromatography (GLC) of alditol ace- 30% saturationwithammoniumsulfate andcentri-
tates derived from sugars by hydrolysis ofsamples fuged. The supernatant (30sup) showed the in-
with 4 N trifluoroacetic acid (TFA) at 121°C for 1 hibitory activity but the pellet (30ppt) showed no
hr [25]. Agaschromatograph, model GC-6Afrom activity. The supernatant contained 91% of the
Shimadzu (Tokyo, Japan), was used and was starting activity and represented a 1.3-fold
equipped with a 3% ECNSS-M column (0.3cm purification of the activity (Table 1).
i.d.X200cm) at 190°C, for analysis of neutral The supernatant (30sup) was applied to a col-
sugars, and a 3% OV-17column (0.3cm i.d.X100 umn of WGA-agarose. The elution profile is
cm) programmed from 150°C to 205°C (2°C/min), shown in Figure 1. The activities of the unbound
for analysis of amino-sugars, as described pre- and bound fractions, which designated WGA-0
viously [26]. Individual derivatives ofhexoses and and WGA-1, respectively, were assayed. Only
hexosamines were quantitated with a Hitachi data WGA-1 contained the inhibitory activity. Affinity
processor 833. chromatography allowed recovery of 89% of the
activity of the starting material with a 14-fold
Chemicals
purification of the activity (Table 1).
Triton X-100, DAB and GlcNAc were purch- We tried to purify WGA-1. However, we were
ased from Wako Pure Chemicals (Osaka, Japan). unsuccessfulinourattemptstopurifyintactWGA-
o-metheyl-D-mannoside was purchased from Flu- binding glycoproteins by chromatography on va-
ka AG. (Buchs, Switzerland). WGA was purch- riousgel-filtrationcolumns. Therefore,we triedto
ased from E. Y. Laboratories, Inc. (San Mateo, fractionate the subunits of WGA-binding gly-
CA, USA). Tween-20 was purchased from Bio- coproteins after reduction ofdisulfide bonds. The
Rad Laboratories(Richmonds, CA, USA). PMSF reduced and alkylated WGA-1 was applied to a
and BSA were purchased from Signma Chemical column of Sephacryl S-400. Figure2 shows the
Co. (St. Louis, MO, USA). All other reagents elution profile. Only eight tubes, numbered from
were of either HPLC or analytical grade. 59 to 66 (SC-59 through SC-66), contained inhibi-
tory activity (Fig. 2). The gel filtration allowed
recovery of 62% of the activity of the starting
RESULTS
material. Thetotal amountoftheproteinin SC-59
through SC-66 corresponded to 2.9% of the pro-
Purification ofWGA-binding glycoprotein tein in the starting material.
The extract prepared from the acetone powder
Table 1. Purificationof WGA-binding glycoprotein
Total protein Total activity Specific activity Purification Yield of activity
Fraction (mg) (IU) (IU/mg) (fold) (%)
Crude extract 208.9 169.8 0.8 1 100
30% ammonium sulfate 153.1 154.6 1.0 1.3 91
30 sup
WGA-agarose 13.6 150.9 11.1 14 89
WGA-1
Sephacryl S-400 0.4 7.4 18.5 23 4.4
SC-60
J. Kamimura, T. Suyemitsu et al.
Gel electrophoresis
Figure3 shows the results ofSDS-PAGE ofthe
fractions obtained at each step of the purification.
The profile after SDS-PAGE of the 30sup was
nearly as the same as that of crude extract. The
profile of WGA-1 gave four major bands that
corresponded to molecular weights of 110-, 70-,
66- and 60-kDa and minor bands that corres-
pondedto lowermolecularweights. Theprofileof
the pool of fractions 56 through 70 from the
chromatography on the column of Sephacryl S-
400, after reduction of disulfide bonds, showed
that four WGA-binding glycoproteins were sepa-
rated from other proteins with lower molecular
Fig. 1. Profile ofthe elution from a column ofWGA-
agarose of the supernatant (30sup) after fractiona-
tionwith30% ammoniumsulfate. Tenmlof30sup,
which contained 30mg protein, were applied to a
column ofWGA-agarose (2.5cm i.d.X2.5cm) and
eluted with 10mM phosphate buffer, pH7.2, that
M
contained0.15 NaCl,0.01%TritonX-100and0.5
M
GlcNAc. Each 10-ml fraction was collected and
10 20 30 monitored by measurements of absorbance at 595
FRACTION NO. nm (•) by the Coomassie-brilliant blue method.
FRACTION NO.
Fig. 2. ProfileoftheelutionfromthecolumnofSephacrylS-400ofWGA-1. WGA-1,equivalentto5mgofprotein,
was reduced and alkylated. The resultant material was dissolved in 1 mlof0.1 MTris-HCl buffer, pH8.5, that
contained6Mguanidine-HCland2mMEDTA,andwasappliedtothecolumnofSephacrylS-400(1.0cmi.d.X
100cm). Thecolumnwaselutedwiththesamebuffer. Each1-mlfractionwasmonitoredbymeasurementsofthe
absorbance at 280nm (•). WGA-binding activity of each fraction was shown by the inhibitory activities of
WGA
hemaggulutination by (O).
Receptor for Aggregation Factor 45
M A B C 56 58 606264666870
kDa
74.4-
49-6—
I
37.2-
24.8—
Fig.3. Results ofSDS-PAGE (8% gels) at eachstep ofpurificationofWGA-bindingglycoproteins. Eachsample,
equivalent to 10-100//g of protein, was solubilized in sample buffer that contained /?-mercaptoethanol and
subjected to electrophoresis on an 8% gel. The gelswere stained with silver. M, molecularmarkers; A, crude
extract;B,30% ammoniumsulfatesupernatant (30sup);C,thefraction(WGA-1)boundtoWGA-agarose;56—
70, fractions eluted from the column ofSephacryl S-400 after reduction ofWGA-1.
weights (Fig. 3, lane 56—70) and that a 110-kDa were stained with HRPO-AF after SDS-PAGE
WGA-bindingglycoprotein could be isolated (Fig. and Western blotting.
3. lane 60). As shown in Figure4, the band at 110-kDa was
stainedwith HRPO-AF, but the other three bands
Western blotting of WGA-binding glycoproteins were notstained. In addition, HRPO-AFthat had
and treatment with HRPO-WGA been previously incubate with 0.5 M GlcNAc re-
The nitrocellulose sheets onto which the gly- tained the ability to bind to the 110-kDa glycopro-
coproteins in WGA-1 were transferred from SDS- ABC
polyacrylamide gels, after electrophoresis, were
treated with HRPO-WGA. In consequence, the
major bands which corresponded to molecular
weights of 110-, 70-, 66- and 60-kDa were stained
with HRPO-WGA and the staining was inhibited
M
by 0.5 GlcNAc (data not shown). In order to
IIOkDa
confirm that the 110-, 70-, 66- and 60-kDa gly-
kDa
coproteins were WGA-binding glycoproteins, re-
74.4—
duced WGA-1 was loaded again onto a column of
WGA-agarose. The boundfractioncontainedfour
49.6-
bands that corresponded to molecular weights of
110-, 70-, 66- and 60-kDa, but the unbound frac-
tion did not generate these bands (data not 37.2-
shown). It was clear, therefore, that the four 24.8-
glycoproteins of 110-. 70-, 66- and 60-kDa bound
to WGA-agarose. Fig. 4. Staining of WGA-binding glycoproteins with
HRPO-AF (see text for abbreviations). The frac-
Western blotting of WGA-binding glycoproteins tion (WGA-1) that bound to WGA-agarose was
and treatment with HRPO-AF subjected to SDS-PAGE on 8% gels. Separated
To examine which WGA-binding glycoprotein proteins were transferred to nitrocellulose sheets.
The sheets were then incubated with 0.1 mg/ml
among the four glycoproteins is the receptor for HRPO-AF (B) and with HRPO-AF plus 0.5M
AF, the 110-, 70-, 66- and 60-kDa glycoproteins GlcNAc (C). A, WGA-1 stained with silver.
46 J. Kamimura, T. Suyemitsu et al.
tein. In order to examine the effects ofsugars on AF (data not shown). These results suggest that
the bindingbetweenAFandthereceptor, WGA-1 the 110-kDa WGA-binding glycoprotein is the
was treated with HRPO-AF with or without AF-binding protein and that AF does not recog-
sugars. After spotting ofWGA-1 onto nitrocellu- nize GlcNAc but recognizes a-mannose in the
lose sheets, the sheets were treated with HRPO- 110-kDa WGA-binding glycoprotein.
AF or with HRPO-AF plus either 0.5 M GlcNAc
or 0.5 M a-methyl-D-mannoside. The staining of Treatment of WGA-binding glycoproteins with
WGA-1 with HRPO-AF was not inhibited by 0.5 trypin
M GlcNAcbut itwas inhibitedby0.5 Ma-methyl- Two mg of fyophilized WGA-1 were treated
D-mannoside (data not shown). On the other with 1 mg of trypsin. The resultant fragments of
hands, the spots of 30 ppt and WGA-0 onto glycoproteins were precipitated by the addition of
nitrocellulose sheetswerenotstainedwithHRPO- acetone and dissolved in 2ml of PBS. Trypsin-
t
»*T'-V-,'."•". '">'•
A-?_• ». * . fE*'ky • I
•*>?.?•..-' "£*•,V,>- r ,.,- •:-, -. ..:••. .*.** v::„ t r1
v
B j
*?. «1.*'«
*\
?jf*. %>
c K
m •!*
£ ' -'4, :"%,
*** •'*«• *,
r»s »'
•H*C
#:.
#
H .JE-
Fig.5. Involvement of WGA-binding glycoproteins in the aggregation of cells caused by AF. Fraction 60,
containingthe 110-kDaWGA-bindingplycoprotein (Fig. 3, lane60) andfraction66, containingthe70-, 66-and
60-kDaWGA-bindingglycoproteins(Fig. 3,lane66)weretreatedwithtrypsintoallowthemtodissolveinNSW
withoutTritonX-100. Thetrypsinizedfragmentsoffraction60orthoseoffraction66weresubjectedtotwo-fold
serialdilutionandaddedtodissociatedcellssimultaneouslywiththeminimumconcentrationofAFnecessaryfor
aggregation ofcells. A, dissociated cells; B, cells with AF added at aconcentration of3.125X10~2mg/ml; C,
cells with the trypsinized fragments ofthe 110-kDa glycoprotin added at a concentration of0.5mg/ml; D, cells
withthetrypsinizedfragmentsofthe70-,66-and60-kDaglycoproteinsaddedatatotalconcentrationof0.5mg
/ml;-3E, F, G, H, cellswith the 110-kDa glycoprotein added at concentrations of0.5, 0.125, 3.125X10"2, 7.8X
10 mg/ml, respectively, plus AF; I, J, K, L, cells with the 70-, 66- and 60-kDa glycoproteins added at total
concentrations of0.5, 0.125, 3.125X10"2, 7.8X10"3mg/ml, respectively, plus AF.
Receptor for Aggregation Factor 47
treated WGA-1 had the same inhibitory effect on ability of AF to aggregate cells.
WGA
hemagglutinationcausedby asthatofintact
WGA-1 solubilized in a solution of 0.01% Triton Chemical composition of the 110-kDa WGA-
X-100atthe same concentration. Thus, theinhibi- binding glycoprotein
tory activitiesofWGA-bindingglycoproteinswere The chemical composition of the purified 110-
retained after the treatment with trypsin. kDa WGA-binding glycoprotein, in fraction 60
from the column of Sephacryl S-400 (Fig. 3, lane
Involvement of WGA-binding glycoproteins in the
aggregation ofcells caused by AF Table2. Amino acid composition ofthe 110-kDa
WGA-binding glycoprotein
The effects of WGA-binding glycoproteins on
the aggregation of cells caused by AF were ex- Amino acid number/1000 amino acids"
amined. Eithertrypsinized fragmentsofthe mate- Asx 128
rials in fraction 60, which contained the 110-kDa Thr 79
WGA-binding glycoprotein, or of that in fraction
Ser 82
66, which contained 70-, 66- and 60-kDa WGA- Glx 100
binding glycoproteins, were added to dissociated
Pro 39
cells simultaneously with the minimum concentra- Gly 68
tion of AF that caused the aggregation of cells. Ala 58
The results are shown in Figure 5. AF had the Val 64
ability to induce the aggregation of cells, but Met 36
neitherthetrypsinizedfragmentsoffraction60nor
He 61
those offraction 66 had any such ability. Howev- Leu 81
er,whenthetrypsinizedfragmentsoffraction60at Tyr 32
a concentration of more than 0.125 mg/ml were Phe 55
added to the dissociated cells with AF, the Lys 59
aggregation of the cells was inhibited, while the
His 16
trypsinized fragments offraction 66did not inhibit Arg 42
the aggregation ofcells caused by AF even at the
high concentration of 0.5mg/ml. These results Total 1000
suggest that the trypsinized fragments of the 110- 3 Values obtained after 24-hr hydrolysis with 6N
kDa WGA-binding glycoprotein neutralized the HC1 at 110°C.
Table3. Sugar composition of the 110-kDa WGA-binding glycoprotein
Sugar mole % Residues/molecule ^g/lOOO^g of protein
Neutral sugars3
Glucose 1.4 0.18 2.6
Galactose 2.0 0.26 3.6
Mannose 79.0 10.4 144.2
Fucose 2.5 0.32 4.2
Amino sugars3
GlcNAc 15.2 2.0 34.0
GalNAc
Sialic acidb 3.3 0.44 10.4
Total 103.4 13.6 199.0
3 Valuesobtained byGLC analysisofhydrolyzatespreparedbytreatmentwith4NTFA at 121°Cfor 1 hr.
b Estimated by the resorcin-Cu2+-HCl method.
c Values calculated from mole % taking the value for GlcNAc as 2.0.
48 J. Kamimura, T. Suyemitsu et al.
60), was examined. The amino acid and sugar glycoproteins in cell membranes are involved in
compositions are shown in Tables2 and 3, respec- theadhesionofcellsinseaurchinembryos. There-
tively. As shown in Table2, the 110-kDa gly- fore, we examined the effects of WGA-binding
coprotein was rich in Asx and Glx, a standard glycoproteins in vivo. Thetrypsin-treatedWGA-1
feature ofglycoprotein. As shown in Table 3, the also inhibited the aggregation of cells caused by
110-kDaglycoproteinwascomposedofabout20% AF (data not shown). These result suggests that
sugars. The major sugar was mannose which was WGA-binding glycoproteins are involved in cell
presentin amolarratioof10.4:2.0withrespectto adhesion as receptors for AF. It is probable that
WGA
GlcNAc, indicating the presence of typical high- dissociation of embryos by is due to the
mannose type sugar chains which are usually com- inhibition of binding between AF and the recep-
posed of 8 mannose residues and 2 GlcNAc re- tors for AF.
sidues [27, 28]. The levels of gluocose, galactose, I125-labeled AF binds to the surface of cells in
fucosewere verylow, andGalNAcwasnot detect- the absence of Ca2+ [10], and so the primary
able. binding of AF to cells was thought to involve a
mechanism distinct from the Ca2+ bridge. Furth-
ermore, the activity of AF is inhibited sugar-
DISCUSSION
specifically by GlcNAc and mannose [10], and
The extract prepared from the acetone powder erythrocytes after preincubation with AF are not
WGA
with the extraction buffer without Triton X-100 aggregated by [10]. These results demons-
exhibited no inhibitory activity against hemagglu- tratethatAFhaslectin-like activityandrecognizes
tination by WGA. This result suggests that the GlcNAc and mannose. In addition, since the
WGA-binding glycoprotein is a hydrophobic pro- bindingofAFtocellsislostaftertreatmentofcells
tein in cell membranes. with trypsin [11], it seems likely that receptors for
When reduced WGA-1 was incubated with AF are glycoproteins exposed to the cell surfacce.
HRPO-WGA afterthe Western blotting and SDS- Fromthese results, itisstronglysuggestedthatAF
PAGE, bands that corresponded to molecular binds to receptors on the cell surface by a sugar-
weights of 110-, 70-, 66- and 60-kDa were stained lectin type of binding. When WGA-1, containing
(data not shown). Furthermore, when reduced the 110-, 70-, 66- and 60-kDa glycoproteins was
WGA-1 was applied to a column of WGA- incubated with HRPO-AF, only band at 110-kDa
agarose, only four glycoproteins of 110-, 70-, 66- was stained with HRPO-AF, and the other three
and 60-kDa were bound to the affinity column. bands were not stained. The binding between the
These results demonstrate that the 110-, 70-, 66- 110-kDaglycoproteinandHRPO-AFwasnotinhi-
and 60-kDa glycoproteins are WGA-binding gly- bitedby0.5 M GlcNAcbutwasinhibitedby0.5 M
coproteins. However, unless WGA-1 was re- a-methyl-D-mannoside. From these results, it
duced, WGA-1 did not enter in 8% SDS- appears that the 110-kDa WGA-bindingglycopro-
polyacrylamidegel. Moreover,whenintactWGA- teinisanAF-bindingprotein andthatAFdoesnot
1 was applied to the column of Sephacryl S-400, recognize GlcNAc but recognizes a-mannose in
WGA-1 passed through the gel. These results the 110-kDa WGA-binding glycoprotein. When
suggestthatWGA-1 isaverylargemoleculethatis we used a ConA-sepharose column to purify AF-
composed of subunits. Therefore, it is probalbe binding protein, we could not obtain the active
that the four WGA-binding plycoproteins are fraction by the elution with 0.5 M a-methyl-D-
asociatedwithone anothethroughdisulfidebonds. mannoside. Therefore, we used WGA-agarose
We demonstrated previously that addition of instead of ConA-sepharose to purify AF-binding
WGA
to the culture medium of embryos caused protein, even if AF may recognize a-mannose.
the dissociation of the embryos [12]. In addition, In our experiments in vivo, when either the
the aggregation ofcellscausedbyAFfailstooccur trypsinized fragments ofthe 110-kDa glycoprotein
after treatment ofthe surface ofcells with trypsin or those of the other three glycoproteins were
[11]. These results suggest that WGA-binding added to dissociated embryonic cells Simula-
