1

The Involvement of Heparan Sulfate in

FGF1/HS/FGFR1 Signaling Complex

Zhengliang L.Wu1, Lijuan Zhang1, 2, 5, Tomio Yabe3, 6, B. Kuberan1, David L. Beeler1,

Andre Love1, 2, Robert D. Rosenberg1, 2

1 Department of Biology,

3 Department of Chemistry, Massachusetts Institute of Technology, Cambridge,

MA 02139

2 Department of Medicine, Harvard Medical School, BIDMC, Boston, MA 02215

5 Present address: Department of Immunology & Pathology, Washington University School of Medicine, St.

Louis, MO 63110

6Present address: Department of Molecular Developmental Biology, Tokyo Metropolitan Institute for

Neuroscience, Tokyo 183-8526, Japan

Correspondence: rdrrosen@mit.edu (617)-253-8803 Fax (617)-258-6553

Running Title: HEPARAN SULFATE AND FGF1/FGFR1 SIGNALING COMPLEX

Copyright 2003 by The American Society for Biochemistry and Molecular Biology, Inc.

JBC Papers in Press. Published on February 25, 2003 as Manuscript M212590200 by guest on A

pril 16, 2018http://w

ww

.jbc.org/D

ownloaded from

2

Summary

FGF signaling begins with the formation of a ternary complex of FGF, FGFR and

heparan sulfate (HS). Multiple models have been proposed for the ternary complex.

However, major discrepancies exist among those models and none of these models

have evaluated the functional importance of the interacting regions on the HS

chains. To resolve the discrepancies, we measured the size and molar ratio of HS in

the complex and showed that both FGF1 and FGFR1 simultaneously interact with

HS; therefore, a model of 2:2:2 FGF1:HS:FGFR1 was shown to fit the data. Using

genetic and biochemical methods, we generated HSs that were defective in FGF1

and/or FGFR1 binding but could form the signaling ternary complex . Both

genetically and chemically modified HSs were subsequently assessed in a BaF3 cell

mitogenic activity assay. The ability of HS to support the ternary complex formation

was found to be required for FGF1 stimulated cell proliferation. Our data also

proved that specific critical groups and sites on HS support complex formation.

Furthermore, the molar ratio of HS, FGF1 and FGFR1 in the ternary complex was

found to be independent of the size of HS, which indicates that the selected model

can take place on the cell surface proteoglycans. Finally, a mechanism for the

FGF/FGFR signaling complex formation on cell membrane was proposed, where

FGF and FGFR have their own binding sites on HS and a distinct ternary complex

formation site is directly responsible for mitogenic activity.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

3

INTRODUCTION

Heparan sulfate (HS) is a linear and highly sulfated polysaccharide, consisting of 50 to

150 basic disaccharide repeats of uronic acid and D-glucosamine units (1). Sulfation can

occur at 2-O of the uronic acid and 3-O, 6-O and N- of the D-glucosamine and is

catalyzed by a variety of sulfotransferases. Each modification is incomplete, which leads

to sequence variation on HS, and it is very likely that critical sulfate groups determine

the specificity of HS-protein interactions (2). Along the HS chain, the majority of

sulfated residues are clustered in short functional domains separated by relatively less

sulfated oligosaccharide sequences (3). Heparin resembles these functional domains and

is widely used for the functional study of HS. One major function of HS is to interact

with fibroblast growth factors (FGFs) and their receptors (FGFRs) and form

FGF:HS:FGFR signaling complexes (4-7). Defects in HS can cause complete losses of

FGF, Hedgehog and Wingless signaling pathways and lead to severe abnormality in

embryonic development (8,9). The involvement of HS in the FGF molecular signaling

complex suggests that FGF activity and specificity may be modulated by HS and in turn,

by enzymes that synthesize and degrade HS.

FGFs and FGFRs play critical roles in the control of many fundamental

cellular processes, such as cell proliferation, differentiation and migration (10-13). There

are 23 known FGFs and 5 types of FGFRs in humans (14). FGF1 and FGF2 were the first

to be isolated and were called acidic and basic FGF, respectively. Studies performed

primarily on FGF2, have identified a stretch of basic residues in the polypeptide chain as

participating in the heparin-binding site (15). FGFRs belong to a group of receptor

tyrosine kinases and are activated through FGFs and HS induced dimerization (10). The

unspliced form of FGFR contains an intracellular tyrosine kinase domain, a trans-

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

4

membrane region, an extracellular region containing three immunoglobulin (Ig) domains,

a string of acidic residues between the first and second Ig domains (16) and an HS

binding site in the second Ig domain (17). The Ig domain I has been shown to be

dispensable and receptor variants containing only the Ig domain II and III (β form) have

been found to exhibit an equivalent degree of binding to FGFs as the variants containing

all three domains (α form). The Ig domain III can undergo differential splicing and

thereby exhibits IIIb and IIIc forms (16). All receptors show redundant specificity for

ligand binding, i.e. one receptor may bind to several FGFs and one FGF may bind to

more than one receptor (12). FGF1 interacts with almost all FGFRs and FGFR1 is a

common receptor for many ligands ; thus we have chosen FGF1 and FGFR1 to carry out

our studies.

Although the importance of HS in FGF signaling is well documented, the exact

roles of HS in the signaling complex are less well characterized. One key issue concerns

the minimum size of HS in the signaling complex. This size of HS reflects the spatial

arrangement of FGF and FGFR in the complex, thus this parameter is critical for

modeling the FGFR signaling complex. Basically there are three modes of interaction

between multiple proteins and a single chain of HS (Figure 1), designated as the cis, trans

and mix mode. Mix mode contains both cis and trans modes. Different modes of

interaction require HS with different lengths to participate. So far, various models with

different modes of HS-protein interactions and thus different HS length requirements

have been proposed. For examples, hexasaccharide (dp6) was found to be able to link two

FGF1 in a trans mode (18) (Figure 1B, IV); dodecasaccharide (dp12) was thought to be

the minimum length for linking one FGF2 and one FGFR1 in a cis mode (19) (Figure 1A,

I); hexadecasaccharide (dp16) was proposed to be able to fully span a heparin binding

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

5

site created by a 2:2 FGF1:FGFR2 complex in a mix mode (Figure 1C, VI) (20). On the

other hand , the shortest biologically active heparin oligosaccharide has been reported to

be an octasaccharide (dp8) (21), hexasaccharide (dp6) (22,23), trisaccharide, or even

disaccharide (dp2) (24,25). These contradictory findings prompted us to design a more

accurate method to determine the size of HS in FGFR signaling complex.

A second key issue that remains controversial concerns the stoichiometry of HS

in the FGF/FGFR signaling complex. Intracellular signaling is believed to be initiated

from receptor dimerization and trans-phosphorylation (10,26), but models with different

stoichiometry for HS and FGF have been proposed (Figure 1). For examples, a single HS

chain binds one FGF and two FGFRs (27) (Figure 1A, III); a single HS chain binds two

FGFs, which in turn bind two FGFRs (18) (Figure 1B, IV); one each of FGF, HS and

FGFR first form an FGF:HS:FGFR half complex, two of which then dimerize (28)

(Figure 1C, VII). Because the size of HS can be up to 150 disaccharide repeats, and only

shorter oligosaccharides no more than 7 disaccharide repeats (dp14) were used for

modeling study in most cases, one unresolved question is whether the models established

with shorter oligosaccharides can be extended to the cell surface HS proteoglycans.

HSs from different tissues or developmental stages have different fine

structures (29,30) and can activate or inhibit FGF signaling pathways (31-33). It is

believed that this phenomenon is caused by replacing of critical functional groups during

the synthesis of the HSs (2,32,34). The critical functional groups on HS interacting with

FGFs (19,23,35) or FGFRs (24,36,37) have been investigated previously. For example, 2-

O sulfation at an iduronic acid was found critical for FGF2 binding (38) and 6-O-

sulfation was found important for FGFR1 binding (37,39). But less information about the

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6

relationship between critical groups on HS mediating FGF:HS:FGFR ternary complex to

the groups required for FGF or FGFR binding has been available.

With in vitro modification of HS and a gel mobility shift assay (40), we have

measured the minimum HS size, the molar ratio among FGF1, HS and FGFR1 as a

function of HS size, and showed that both FGF1 and FGFR1 interacts with HS in the

complex . Based upon these results, we suggest that a 2:2:2 FGF1:HS:FGFR1 model best

fits the data. In addition utilizing cell genetic study, we have found that there are different

critical groups at different sites on HS involved in the ternary complex formation and

FGF and FGFR binding interactions. We also find that the ability of HS to form a ternary

complex with FGF1 and FGFR1 is a prerequisite for FGF1 stimulated mitogenic activity.

Based on these data, we propose a mechanism showing how the FGF1/FGFR1 signaling

complex is formed on HS cell surface proteoglycan.

MATERIALS AND METHODS Reagents

Heparin oligosaccharides (dp4 to dp24) were purchased from Iduron (Manchester, UK). Completely

desulfated and N-sulfated heparin sulfate (DSNS), and completely desulfated and N-acetylated heparin

(DSNAc) was from Seikagaku America (Falmouth, MA). FGF1 and FGFR1β (IIIc)/Fc were from R&D

Systems (Minneapolis, MN). PAPS was from Calbiochem (La Jolla, California). 3-OST-1 and 6-OST-1

were prepared as previously described (40). CHOpgsA-745 cell line and 6-O-desulfated heparin (6ODS)

were kind gifts from Dr. Jeffrey D. Esko (University of California, San Diego). Preparation of [35S] labeled

3'-phosphoadenosine 5'- phosphosulfate ([35S]PAPS), radiolabeling of HS, autoradiograph and gel analysis

were the same as previously described (40). Anti-FGF1 antibody was from Santa Cruz Biotechnology

(Santa Cruz, California). CHO-K1 cell line was from ATCC (Manassas, VA). 2-OST deficient cell

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

7

CHOpgsF-17 was prepared as described (41). Fetal bovine serum (FBS) was from Life Technologies

(Rockville, MD). Protein A, Alexa Fluor® 647 conjugate was from Molecular Probe (Eugene, OR)

Heparin Digestion and Disaccharide Analysis

The heparin sample (20µg) was digested with a mixture of heparinase and heparantinase I, II (Seikagaku

Corporation, Tokyo, Japan) at 37oC for 2 hours in 50 µl buffer of 2 mM Ca(Ac)2, 20 mM Na Ac, pH 7.0.

The digestion products were separated with a C18-reversed phase column (IPRP-HPLC) (Vydac, Lake

Forest, CA). The sample was eluted with 2.5%, 6%,10.5%, 18%, 50% acetonitrile in 40 mM NH4H2PO4 and

1 mM tetrabutylammonium dihydrophosphate (Sigma) for 15, 15, 45, 25, 20 minutes respectively, and was

monitored with light absorbance at 232 nm (41).

Expression of FGFR1

The extracellular domain of FGFR1, including Ig domain II and IIIc (from residue 142 to 365) (42,43) was

PCR cloned from a Human Placenta Quik-clone TM cDNA library (Clontech, Palo Alto, California). The

PCR forward primer was GATAACACCAAACCAAACCG and the backward primer was

CCTCTCTTCCAGGGCTTCCA. The PCR product was expressed in a pBAD/TOPO ThioFusion TM

Expression System (Invitrogen, Carsbad, California). The expressed protein was a fusion protein with

thioredoxin at the C terminal and was refolded in 150 mM NaCl, 10 mM Tris pH 8.0, 10% glycerol, 1 mM

L-Cystein.

Binding Reaction and Gel Mobility Shift Assay

For a typical 20 µl binding reaction, 1 µl of FGF1, 1 µl FGFR1 and 0.1 µl of HS (also refer to heparin, if

not specifically stated) were added into appropriate volume of binding buffer. The concentrations of FGF1,

FGFR1 and HS were adjusted as needed before the reaction. The reaction was incubated at 23oC for 20

minutes. The binding buffer contained 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 10 mM of MgCl2

and 1.4 mM KH2PO4 and 12% glycerol. Half of the reaction was loaded onto a native 4.5% polyacrylamide

gel. The gel was run at 6 volt /cm for 70 minutes. After the run, the gel was dried under vacuum and

exposed to phosphor imager plate.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

8

Cell Culture and Mitogenic Assays

Suspension cultures of FGFR1α(IIIc)-expressing BaF3 cells (A generous gift from professor Ornitz, D. M.)

were maintained in AIM V media (Gibco BRL, Rockville, MA), supplement with 5 nM recombinant

mouse IL-3 (R & D Systems). For mitogenic assays, 50 ul of AIM V media (serum free) containing HSs

and FGF-1 at final concentrations of 1 ug/ml and 5 nM respectively, were plated into a 96 well assay plate.

Cells were washed and resuspended in AIM V media (serum free) and 2,500 cells were added to each well

for a total volume of 100 ul. The cells were then incubated at 37 oC with 5% CO2 for 24 hours; afterwards,

100 ul Syto-11 dye (Molecular Probes, Eugene, OR), which was prepared in 10 mM Tris, 1 mM EDTA, 50

mM NaCl, pH 8.0, was added to each well and incubated for 30 minutes at 37 oC. The sample was excited

at 508 nm and the fluorescence emission at 527 nm was then measured using the SpectraMax Gemini XS

(Molecular Devices, Sunnyvale, CA). The data was analyzed with Softmax software and each data

presented was the average of a triplicate determination.

Flow Cytometry Analysis

Nearly confluent monolayers of cells in a T-75 flask were detached by adding 10 ml phosphate buffered

saline (PBS) containing 10% FBS and 2 mM EDTA and centrifuged. The cell pellets were placed on ice.

About 1x106 cells was first mixed with 20 µl PBS (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 10 mM

of MgCl2 and 1.4 mM KH2PO4) and 1 µg FGFR1β(IIIc)/Fc, then 4 µg of protein A-Alexa Fluor® 647

conjugate was added. Protein A binds to the Fc region of FGFR1β(IIIc)/Fc. In a separate experiment, 0.5

µg of FGF1 was also added at this point. After 15 minutes of incubation, the cells were washed once with

10 ml of PBS and resuspended in 300 µl PBS containing 10% FBS. Flow cytometry was performed with

FACScan and FACStar instruments (Becton Dickinson).

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

9

RESULTS

Observation of the Ternary Complex of FGF1, HS and FGFR1

Previously, we applied gel mobility shift assay to study antithrombin III/HS interaction

(40). The same method was applied to study FGF/HS/FGFR interactions. A ladder of

defined lengths of oligosaccharides from dp2 to dp24 was first radiolabeled with 6-OST-

1 sulfotransferases (40) and examined on a 15% PAGE gel (Figure 2A). The labeled

ladder was then applied to gel mobility shift assays with FGF1 or FGFR1 (Figure 2B).

FGF1 could bind to tetrasaccharide (dp4) and longer chains, while FGFR1 only showed

significant binding to octasaccharide (dp8) and longer chains. The mobility of the binary

complex FGF1:HS was greater than that of FGFR1:HS. When equivalents of FGF1 and

FGFR1 were mixed with excess dp18 and applied to the assay, a distinct band, with

mobility greater than FGFR1:HS but less than FGF1:HS, was observed (Figure 3A),

suggesting the formation of a ternary complex among FGF1, FGFR1 and HS. This

ternary complex could be observed when the concentrations of FGF1 and FGFR1 were as

low as 128 nM, though either no or less binary complexes of HS:FGFR1 and FGF1:HS

could be observed at this concentration, suggesting the cooperative binding among HS,

FGF and FGFR. Cooperative binding is a characteristic of FGF:HS:FGFR ternary

complex formation (44,45).

To further confirm that the above ternary complex did form at the native state of

the proteins, binding samples were heated before loading onto a gel (Figure 3B). After

the sample was heated at 47 oC, the binary complex of HS:FGFR1 significantly

decreased, whereas the binary complex of FGF1:HS was stable. Heating at 60 oC caused

the disappearance of both of the binary complexes, but not the ternary complex. Heating

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

10

at 100 oC caused the disappearance of all the complexes. This experiment suggests that

the native state of the proteins is required for the ternary complex formation and the order

of the thermal stability of the complexes is FGF1:HS:FGFR1 > FGF1:HS > HS:FGFR1.

To see if the observed ternary complex can form on cells with membrane HS

proteoglycan under similar binding conditions, we carried out flow cytometry assays with

CHO cells and FGFR1β(IIIc)/Fc in the presence or absence of FGF1. FGFR1β(IIIc)/Fc

contained the same functional domains as the FGFR1 used in the gel shift assays, except

for the inclusion of an Fc region of IgG. The binding of FGFR1β(IIIc)/Fc to CHO cells

was monitored (Figure 3C). The HS-deficient CHOpgsA-745 cell showed no binding,

even in the presence of FGF1. On the other hand, HS-expressing wild type CHO-K1 cell

had weak binding, but this binding was greatly enhanced when FGF1 was added. These

experiments suggest that FGFR1β(IIIc)/Fc can bind to the HS proteoglycan on cell

membrane and form a much tighter ternary complex together with FGF1. Interestingly, a

2-O sulfotransferase (2OST) deficient cell CHOpgsF-17 did not bind to FGFR1β(IIIc)/Fc

alone, but also showed a much stronger binding in the presence of FGF1, suggesting the

formation of the ternary complex in the absence of 2-O sulfation (Figure 3C).

The Minimum Oligosaccharide Supporting the Ternary Complex Formation

To determine the minimum oligosaccharide that can initiate the ternary complex with

FGF1 and FGFR1, the labeled ladder (Figure 2A) was applied to a gel shift assay with

both FGF1 and FGFR1 (Figure 4A). The ternary complex could be observed with

tetrasaccharide and larger oligosaccharides, but not with disaccharide.

To confirm the above observation, another gel was visualized with anti-FGF1

antibody (Figure 4B). FGF1 had a low mobility in a native gel electrophoresis. The

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

11

binding of HS to FGF1 added negative charges to FGF1 and the resulting FGF1:HS

complex moved faster than the free FGF. The further addition of FGFR1 slowed

FGF1:HS complex, because of the formation of the ternary complex. The ternary

complexes were observed with tetrasaccharide and longer oligosaccharides, but not with

disaccharide. This experiment confirmed that tetrasaccharide was the minimum HS

required for the formation of a ternary complex of FGF1, HS and FGFR1.

The Molar Ratio of HS in the Ternary Complex

First, we determined the ratio of FGF1 and FGFR1 in the ternary complex. In this

experiment, fixed amount of FGF1, but increasing amount of FGFR1 was added to a

series of binding reactions (Figure 5A). The bands of the ternary complexes were

subjected to densitometry analysis (Figure 5B) and the band density was plotted against

the molar equivalents of FGFR1 to FGF1 (Figure 5C). The slope of the curve declined

sharply when FGFR1 was equivalent to FGF1, suggesting that FGF1 and FGFR1 have a

1:1 ratio in the ternary complex. The continuing increase in band intensity above

equivalence might be caused by a shift of equilibrium, as excess FGFR1 (the reactant)

generated more product (the ternary complex) formation and caused the disappearance of

the FGF1:HS. At the level of 4 equivalents of FGFR1, there was almost no FGF1:HS

observed.

We further measured the molar ratio of HS in the ternary complex. In a series of

binding reactions with the oligosaccharide ladder (Figure 1A), each reaction contained

250 ng oligosaccharide (from dp4 to dp24), 16 pmol of FGF1 and 64 pmol of FGFR1.

The 4:1 molar concentration ratio of FGFR1 to FGF1 overwhelmingly favored the

formation of FGF1:HS:FGFR1, so that almost all the FGF1 was incorporated into the

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

12

ternary complex (Figure 5A). After electrophoresis, the amount of the oligosaccharide

present in the ternary complex in each reaction was calculated by densitometry analysis,

and the molar ratio between the oligosaccharide and FGF1 in the ternary complexes was

then determined (Table 2). It was surprising to find that, independent of the size of HS,

the ratio was consistently near 1:1. Considering the 1:1 ratio between FGF1 and FGFR1,

this result suggests that the ternary complex have a molar ratio of 1:1:1 among FGF1, HS

and FGFR1.

Critical Sulfate Groups on HS for the Ternary Complex Formation

We used completely desulfated, N-acetylated (DSNAc), completely desulfated, N-

resulfated (DSNS) and 6-O desulfated (6ODS) heparin as starting materials to investigate

the role of critical groups (Figure 6A). When DSNAc was modified with 6-OST-1, or 6-

OST-1 plus 3-OST-1, it could not bind to either FGF1 or FGFR1 and failed to form the

ternary complex with FGF1 and FGFR1. When DSNS was modified with 6-OST-1 or 3-

OST-1, it still could not bind to either FGF1 or FGFR1 individually, but could initiate the

ternary complex formation. When 6ODS was modified with 3-OST-1, it could not bind to

FGFR1, but could bind to FGF1 and initiate the ternary complex formation. These

experiments suggest that N-sulfation and perhaps to a lesser extent O-sulfation are critical

for the ternary complex formation. On the contrary, 6-O sulfation, which is critical for

FGFR1 binding is not critical for the ternary complex formation.

The chemical compositions of these modified heparins were further validated

through disaccharide analysis (Figure 6B). The major disaccharides found in DSNAc and

DSNS were ∆UA-(1-4)GlcNAc (∆UA, unsaturated uronic acid; GlcN, glucosamine;

GlcNAc, N-acetylated GlcN) and ∆UA-(1-4)GlcNS (GlcNS, N-sulfated GlcN) ,

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

13

respectively. Only very small amounts of ∆UA2S-(1-4)GlcNS (∆UA2S, 2-O sulfated

∆UA) and ∆UA-(1-4)GlcNS6S (GlcNS6S, N- and 6-O sulfated GlcN) were observed in

DSNS, which might arise from non-specific sulfation during the chemical N-sulfation

step. The major disaccharide found in 6ODS was ∆UA2S-(1-4)GlcNS, as a result of the

6-O desulfation of the trisulfated disaccharide, ∆UA2S-(1-4)GlcNS6S. Because of the

presence of a small amount of ∆UA2S-(1-4)GlcNS in DSNS, and ∆UA-(1-4)GlcNS6S in

6ODS, it is hard to reach a definitive conclusion about the importance of 2-O and 6-O

sulfation in the ternary complex formation; nevertheless, since 2-O and 6-O sulfation in

DSNS and 6ODS, respectively, was significantly lower than that in the wild type heparin,

2-O and 6-O sulfation may not be as important in the ternary complex formation as in the

binary complex formation. Interestingly, because the trisulfated disaccharide, ∆UA2S-(1-

4)GlcNS6S was not found in DSNS and 6ODS, yet DSNS/31 and 6ODS/31 could

support the ternary complex formation, it is concluded that the trisulfated disaccharide is

not required for the ternary complex formation.

The Ability of HS to Form a Ternary Complex with FGF1 and FGFR1 but not Binary

Complexes Is Associated with Cell Mitogenic Activity

Ternary complex formation was observed with various modified heparins and heparin

oligosaccharides, but it is more important to know if the formed ternary complexes are

responsible for mitogenic activity. The oligosaccharides and modified heparins were

further tested for their mitogenic activities in an FGFR1α(IIIc)-expressing BaF3 cell

system (21) stimulated with FGF1 (Table 3). FGFR1α(IIIc) contains all three Ig-like

extracellular domains and it showed the same binding characteristics with FGFR1β(IIIc)

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

14

(16). In the assay, disaccharide had background mitogenic activity; tetramer, hexamer

and octamer oligosaccharides had medium mitogenic activities; decasaccharide and

larger oligosaccharides had mitogenic activities comparable to that of heparin. Except for

disaccharide, all other oligosaccharides showed the ability to support the ternary complex

formation (Figure 4). The 6-O and 3-O sulfated DSNAc, which didn’t support the ternary

complex formation (Figure 6A), had almost no activity. On the other hand, 6-O or 3-O

sulfated DSNS and 3-O sulfated 6ODS, which did not support the binary complex with

either FGF1 or FGFR1, but supported ternary complex formation (Figure 6A), showed

significant activities. These data suggest that the abilities of these oligosaccharides and

modified heparins to initiate FGF1:HS:FGFR1 ternary complex but not FGF1:HS and

FGFR1:HS complexes correlate with mitogenic activities.

DISCUSSION

Since Yayon et al. (4) first demonstrated the importance of heparan sulfate proteoglycan

for high affinity FGF2:FGFR1 binding, a large amount of data has been produced,

indicating that HS is essential for FGF signaling (3,5,46). Later, HS was proved to be an

essential component of FGF/FGFR signaling complex (3,10,42), but the roles that HS

plays in the signaling complex formation are still obscure. We have provided a novel

strategy to study the physical parameters of HS as well as the signaling complex as a

whole. This new strategy enabled us to measure the size, molar ratio and molecular

interactions of HS in the ternary complex. It allowed us to study critical groups on HS

involved in the ternary complex formation and FGFR activation. It also allowed us to

correlate the ability of HS to initiate a ternary complex and the mitogenic activity.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

15

Requirements on Ternary Complex Formation Select a 2:2:2 Model Based on our data, we have identified three molecular parameters for choosing

models of the FGF1/FGFR1 signaling complex. First, the minimum HS in the ternary

complex is a tetrasaccharide. Second, the molar ratio among HS, FGF1 and FGFR1 in the

ternary complex is 1:1:1. Third, HS interacts with both FGF1 and FGFR1 in the ternary

complex. These requirements along with the requirement of receptor dimerization (10,26)

are used to evaluate the potential models in Figure 1 (Table 1).

Whereas the requirements on the size and ratio of HS in the complex are obvious,

the requirement that HS interacts with both FGF1 and FGFR1 may not be so obvious.

Because both FGF1 and FGFR1 can bind to HS independently (Figure 2B), with similar

affinity (47,48), it is likely that both of them bind to HS in the ternary complex. If HS

interacts only with FGF1 in the ternary complex, the dissociation of FGF1 from the

polysaccharide will release HS from the ternary complex, thus HS should be liberated

from FGF1:HS and FGF1:HS:FGFR1 at the same temperature, but experiment showed

that HS could only be liberated from the ternary complex at higher temperature (Figure

3B). Again, if HS interacts only with FGF1 in the ternary complex, an HS with no

affinity for FGF1 will not be able to form a ternary complex with FGF1 and FGFR1, but

experiments showed the opposite (Figure 6A). One might argue that FGF1 may

experience a major conformational change upon the association with FGFR1 and thus

have greater affinity towards HS, but this has not been seen in crystallographic studies

(15,18,20,42,49).

From the binding assay, we found the minimum HS for FGF1 and FGFR1

binding were tetrasaccharide and octasaccharide respectively (Figure 2B), and the

minimum HS required for ternary complex formation was only a tetrasaccharide (Figure

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

16

4). Based upon these findings, the length of two contiguous FGF1 binding sites on HS

will span at least eight sugar residues, and the length needed for one FGF1 site and one

FGFR1 site will cover at least twelve sugar residues (Figure 1A, Table 1). Obviously, the

cis mode can not explain the findings from the present and past studies(24,25,50) that

shorter oligosaccharides could initiate the ternary complex formation and activate FGF

receptors.

Compare to the cis mode, the trans mode would better explain the requirement of

HS size. In a typical model of the trans mode, two FGFs bind to an oligosaccharide from

the opposite sides; therefore, a tetrasaccharide will be of sufficient in size (Figure 1B,

IV). However, this model does not contain FGFR1/HS interaction and the molar ratio of

HS to FGF1 and FGFR1 is 1:2:2. Model V (Figurer 1B) is excluded, because the

minimum HS in this model is an octasaccharide.

Recently, models with mix mode of interaction have been proposed based on

crystallographic studies (20,28,42, 43). In Schlessinger’s “two-end” model, two

antiparallel oligosaccharides are incorporated and the oligosaccharide can be as short as

hexasaccharide (Figure 1C, VII). The complex is a dimer consisting of two 1:1:1

FGF2:HS:FGFR1 half complexes and is stabilized via both FGFR1/FGFR1 and

HS/FGFR1 contacts. Within each 1:1:1 FGF2:HS:FGFR1 complex, the hexasaccharide

not only makes numerous contacts with both FGF2 and FGFR1, thereby augmenting

FGF2/FGFR1 binding, but also makes contacts with FGFR1 in the neighboring half

complex, thus playing a dual role in the ternary complex formation. Except for the

difference in the minimum size on HS (Table 1), this “two-end” model fits the data

described previously.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

17

Though hexasaccharide was the smallest oligosaccharide shown to be effective in

the “two-end” model (28), we believe that a tetrasaccharide still can make effective

contacts with FGF2 and FGFR1 in the ternary complex. According to Schlessinger’s

crystal structure (28), the first six sugar residues at the non-reducing end of a

decasaccharide make 30 hydrogen bonds, including 9 with FGFR1, 16 with FGF2 in the

same half complex, and 5 with FGFR1 in the adjoining half complex. Among these, all

the contacts to the FGFR1, and 8 out of the 16 contacts to the FGF2 are due to the first

four sugar residues. On the other hand, the first two sugar residues make 12 contacts,

with only one to FGF2, while the next two sugar residues make 10 contacts, with none to

the FGFR1 in the same half complex. This suggests that a tetrasaccharide, but not a

disaccharide, may still fulfill the dual role played by a hexasaccharide in the ternary

complex formation, and thus explains why a tetrasaccharide was the minimum

oligosaccharide capable of initiating the ternary complex formation and possessing

biological activity. Recently tetrasaccharide has also been reported to bind FGF2 and

promote cell growth (23,24,50).

The fact that a tetrasaccharide can not make full contacts with FGF2 and FGFR1

may explain why it is not as active as a hexasaccharide (Table 3), but it still remains

unclear why shorter oligosaccharides like dp 6, dp8 showed lower mitogenic activity than

longer oligosaccharides (Table 3). It is possible that additional HS length helps the

recruitment of FGF1 or FGFR1 to form the ternary complex. The opportunity for each of

FGF1, HS and FGFR1 to encounter each other simultaneously on a cell surface is

probably low. It is more likely that HS, by its abundance, first associates with one of the

proteins (47), FGFR1, for example; then the unoccupied region of HS functions as a

recruiter, where FGF1 can bind and join the complex by proximity. The analysis of the

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

18

binding mechanism and study of critical functional groups on HS further supports this

hypothesis (Figure 7).

Novel Binding Mechanism for HS in the Ternary Complex Involves Different Critical

Functional Groups and Different Binding Site

Plotnikov and Schlessinger’s crystal structures (28,43) show that a new binding grove for

HS forms among two FGFs and two FGFRs. Since the binding environment of HS in the

ternary complex is very different from that on individual FGF or FGFR, it is likely that

the contacts could be made through different set of functional groups on HS, which is

suggested in our experiments. It was previously shown that 2-O sulfation is critical for

FGF1 binding (19,51) and both 2- and 6-O sulfation is critical for FGFR1 binding (Figure

3C, Figure 6A) (52), but N- sulfation is critical for the ternary complex formation (Figure

6). Besides, the N-, 2-O and 6-O trisulfated disaccharide is not found to be important for

the ternary complex formation (Figure 6). Consistently, among the first four sugar

residues of the octasaccharide in Schlessinger’s crystal structure, the N-sulfation and the

backbone structure make 16 out of 22 hydrogen bonds to all the necessary protein

components (28). The observation of different critical groups in ternary complex

formation suggests that by changing critical groups, HS can determine the nature of

FGF1:HS:FGFR1 complex formation.

Different set of critical groups suggests that the site where ternary complex forms

on HS can be different from the site where FGF1 or FGFR1 binds. The fact that enzyme

modified 6ODS and HS on CHOpgsF-17 cells couldn’t bind to FGFR1 but still formed a

ternary complex with FGF1 and FGFR1 (Figure 6A, 3C) suggest that the ternary complex

formation site is different from the site where FGFR1 binds (Figure 7). This is also

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

19

consistent with and explains the previous report that 6-O desulfated heparin showed

mitogenic activity (53). On the other hand, cells with 2OST deficient were able to form a

ternary complex with FGF1 and FGFR1 (Figure 3C) and mount an apparently normal

signaling response to FGF1 (Table 3) and FGF2 (54), though 2-O sulfation is critical for

FGF1 and FGF2 binding (19,35,51). It is also reported that the HS from the culture

medium of the mammary fibroblasts and the myoepithelial-like cells possessed high

affinity to FGF2 but lacked mitogenic activity (55). These evidences suggest that the

ternary complex formation site can be different from the site where FGF1 or FGF2 binds

(Figure 7). The fact that the modified DSNS without the ability to bind either FGF1 or

FGFR1 (Figure 6A) could initiate a ternary complex and possess mitogenic activity

(Table 3) further supports the idea.

A Proposed Mechanism for the Formation of FGF1/FGFR1 Signaling Complex on Cell

Membrane

For all tested oligosaccharides, the 1:1:1 ratio among FGF1, HS and FGFR1 (Table 2)

implies that there is only one site where the ternary complex can form on each HS.

According to Schlessinger’s model (28), this site would be at the non-reducing end of

HS. On the other hand, there could be internal binding sites for FGF1 (48) and FGFR1

inside of the sulfated domains (3). Given the size of the sulfated domains on heparan

sulfate usually covers 12 to 20 sugar residues (1), these sites could be located on different

sulfated domains or located contiguously in the same sulfated domain. Due to the

overwhelming abundance of HS, FGF1 and FGFR1 may first bind to HS individually

(47). Because FGF1 and FGFR1 bind to HS with relative low affinities (44,45) and the

resulting FGF1:HS and FGFR1:HS binary complexes are less stable (Figure 3B), FGF1

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

20

and FGFR1 can undergo binding-releasing cycles rapidly and therefore translocate along

the HS chains. Once an FGF1 and an FGFR1 interacts with each other on HS chain, they

will probably form a half complex at the non-reducing end of the HS. Alternatively, the

FGF1 and FGFR1 can associate with each other in the surrounding solution and the

FGF1:FGFR1 complex then bind to its binding site at the non-reducing end of the HS.

Two half complexes then may associate with each other and form a much more stable

ternary complex (Figure 3B), which allows the receptor trans-phosphorylation (Figure 7).

On the other hand, the binding of FGF1 and FGFR1 to HS can be considered as a

recruiting mechanism to increase the effective concentration and facilitate the association

of FGF1, FGFR1 at the non-reducing ends for the formation of the ternary complex

(Figure 7). In this sense, HS serves as both a “catalyst of molecular encounter” (47) and a

structure component. Shorter oligosaccharides lacking these FGF1 and FGFR1 binding

sites would have less chance to form a ternary complex and thus explain why they

showed less biological activities (Table 3).

Perspective

Overall, we observed a direct link among critical groups on HS, FGF1/FGFR1 signaling

complex formation, and cell growth. The regulatory role of HS in organ development and

cell proliferation has been observed before (8,9,29,30,56), but the mechanism of this

regulation was obscure. It is possible that a cell could regulate its own activity, through

altering the critical groups on HS thus affecting the formation of specific FGFR signaling

complexes. The placement of critical groups on HS must result from the operation of a

set of tightly regulated modification enzymes. Recently, the gene structures for almost all

of these enzymes have been elucidated, and it has been shown that N, 3-O and 6-O

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

21

sulfotransferases exhibit genetic polymorphism and encode distinct isoforms (2,7). These

isoforms differ in substrate specificity and expression pattern, both spatially and

temporally (7), which may cause the specific modification on HS. Further study of these

enzymes is crucial to our understanding of specific modifications on HS and the

regulatory role of HS.

Using chemically and enzymatically modified heparin sulfates and gel mobility

shift assay, we were able to monitor the formation of FGF1/FGFR1 signaling complex

and study the physical parameters of HS in the complex. The results concerning the size,

molar ratio and molecular interactions of HS in the complex and the concept of receptor

dimerization (10,26) allowed us select a 2:2:2 FGF1:HS:FGFR1 signaling model. We

demonstrated that FGF1, FGFR1 and the FGF1:FGFR1 complex can bind to different

sites on HS and only ternary complex formation site possesses mitogenic activity. We

also demonstrated that different critical groups are present in these different HS sites.

Finally, the molar ratio of HS, FGF1 and FGFR1 in the ternary complex was found to be

independent of the size of HS, which provides strong evidence that the formation of the

proposed signaling complex can occur on cell surface HS proteoglycans. This study

demonstrates the structural as well as regulatory roles of HS in the FGF1:HS:FGFR1

complex and can serve as a model for studying other HS molecular interactions.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

22

REFERENCES

1. Bernfield, M., Gotte, M., Park, P. W., Reizes, O., Fitzgerald, M. L., Lincecum, J., and Zako, M.

(1999) Annu Rev Biochem 68, 729-77

2. Rosenberg, R. D., Shworak, N. W., Liu, J., Schwartz, J. J., and Zhang, L. (1997) J Clin Invest

99(9), 2062-70.

3. Gallagher, J. T. (2001) J Clin Invest 108(3), 357-61.

4. Yayon, A., Klagsbrun, M., Esko, J. D., Leder, P., and Ornitz, D. M. (1991) Cell 64(4), 841-8.

5. Rapraeger, A. C., Krufka, A., and Olwin, B. B. (1991) Science 252(5013), 1705-8.

6. Spivak-Kroizman, T., Lemmon, M. A., Dikic, I., Ladbury, J. E., Pinchasi, D., Huang, J., Jaye, M.,

Crumley, G., Schlessinger, J., and Lax, I. (1994) Cell 79(6), 1015-24.

7. Esko, J. D., and Selleck, S. B. (2002) Annu Rev Biochem 71, 435-71

8. Perrimon, N., and Bernfield, M. (2000) Nature 404(6779), 725-8.

9. Lander, A. D., and Selleck, S. B. (2000) J Cell Biol 148(2), 227-32.

10. Schlessinger, J. (2000) Cell 103(2), 211-25.

11. Lemmon, M. A., and Schlessinger, J. (1994) Trends Biochem Sci 19(11), 459-63.

12. Galzie, Z., Kinsella, A. R., and Smith, J. A. (1997) Biochem Cell Biol 75(6), 669-85

13. Naski, M. C., and Ornitz, D. M. (1998) Front Biosci 3, D781-94.

14. Sleeman, M., Fraser, J., McDonald, M., Yuan, S., White, D., Grandison, P., Kumble, K., Watson,

J. D., and Murison, J. G. (2001) Gene 271(2), 171-82.

15. Faham, S., Hileman, R. E., Fromm, J. R., Linhardt, R. J., and Rees, D. C. (1996) Science

271(5252), 1116-20.

16. Givol, D., and Yayon, A. (1992) FASEB J 6(15), 3362-9.

17. Kan, M., Wang, F., Xu, J., Crabb, J. W., Hou, J., and McKeehan, W. L. (1993) Science 259(5103),

1918-21.

18. DiGabriele, A. D., Lax, I., Chen, D. I., Svahn, C. M., Jaye, M., Schlessinger, J., and Hendrickson,

W. A. (1998) Nature 393(6687), 812-7.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

23

19. Guimond, S., Maccarana, M., Olwin, B. B., Lindahl, U., and Rapraeger, A. C. (1993) J Biol Chem

268(32), 23906-14.

20. Stauber, D. J., DiGabriele, A. D., and Hendrickson, W. A. (2000) Proc Natl Acad Sci U S A 97(1),

49-54.

21. Ornitz, D. M., Yayon, A., Flanagan, J. G., Svahn, C. M., Levi, E., and Leder, P. (1992) Mol Cell

Biol 12(1), 240-7.

22. Gambarini, A. G., Miyamoto, C. A., Lima, G. A., Nader, H. B., and Dietrich, C. P. (1993) Mol

Cell Biochem 124(2), 121-9.

23. Zhou, F. Y., Kan, M., Owens, R. T., McKeehan, W. L., Thompson, J. A., Linhardt, R. J., and

Hook, M. (1997) Eur J Cell Biol 73(1), 71-80.

24. Ostrovsky, O., Berman, B., Gallagher, J., Mulloy, B., Fernig, D. G., Delehedde, M., and Ron, D.

(2002) J Biol Chem 277(4), 2444-53.

25. Ornitz, D. M., Herr, A. B., Nilsson, M., Westman, J., Svahn, C. M., and Waksman, G. (1995)

Science 268(5209), 432-6.

26. Heldin, C. H. (1995) Cell 80(2), 213-23.

27. Springer, B. A., Pantoliano, M. W., Barbera, F. A., Gunyuzlu, P. L., Thompson, L. D., Herblin,

W. F., Rosenfeld, S. A., and Book, G. W. (1994) J Biol Chem 269(43), 26879-84.

28. Schlessinger, J., Plotnikov, A. N., Ibrahimi, O. A., Eliseenkova, A. V., Yeh, B. K., Yayon, A.,

Linhardt, R. J., and Mohammadi, M. (2000) Mol Cell 6(3), 743-50.

29. Allen, B. L., Filla, M. S., and Rapraeger, A. C. (2001) J Cell Biol 155(5), 845-58.

30. Lindahl, U., Kusche-Gullberg, M., and Kjellen, L. (1998) J Biol Chem 273(39), 24979-82.

31. Zhang, Z., Coomans, C., and David, G. (2001) J Biol Chem 276(45), 41921-9.

32. Pye, D. A., Vives, R. R., Hyde, P., and Gallagher, J. T. (2000) Glycobiology 10(11), 1183-92.

33. Guimond, S. E., and Turnbull, J. E. (1999) Curr Biol 9(22), 1343-6.

34. Berry, D., Kwan, C. P., Shriver, Z., Venkataraman, G., and Sasisekharan, R. (2001) Faseb J 15(8),

1422-4.

35. Turnbull, J. E., Fernig, D. G., Ke, Y., Wilkinson, M. C., and Gallagher, J. T. (1992) J Biol Chem

267(15), 10337-41.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

24

36. Loo, B. M., Kreuger, J., Jalkanen, M., Lindahl, U., and Salmivirta, M. (2001) J Biol Chem

276(20), 16868-76.

37. McKeehan, W. L., Wu, X., and Kan, M. (1999) J Biol Chem 274(31), 21511-4.

38. Maccarana, M., Casu, B., and Lindahl, U. (1993) J Biol Chem 268(32), 23898-905.

39. Pye, D. A., Vives, R. R., Turnbull, J. E., Hyde, P., and Gallagher, J. T. (1998) J Biol Chem

273(36), 22936-42.

40. Wu, Z. L., Zhang, L., Beeler, D. L., Kuberan, B., and Rosenberg, R. D. (2002) FASEB J 16(6),

539-45.

41. Zhang, L., Lawrence, R., Schwartz, J. J., Bai, X., Wei, G., Esko, J. D., and Rosenberg, R. D.

(2001) J Biol Chem 276(31), 28806-13.

42. Pellegrini, L., Burke, D. F., von Delft, F., Mulloy, B., and Blundell, T. L. (2000) Nature

407(6807), 1029-34.

43. Plotnikov, A. N., Schlessinger, J., Hubbard, S. R., and Mohammadi, M. (1999) Cell 98(5), 641-50.

44. Nugent, M. A., and Edelman, E. R. (1992) Biochemistry 31(37), 8876-83.

45. Pantoliano, M. W., Horlick, R. A., Springer, B. A., Van Dyk, D. E., Tobery, T., Wetmore, D. R.,

Lear, J. D., Nahapetian, A. T., Bradley, J. D., and Sisk, W. P. (1994) Biochemistry 33(34), 10229-

48.

46. Lin, X., Buff, E. M., Perrimon, N., and Michelson, A. M. (1999) Development 126(17), 3715-23.

47. Powell, A. K., Fernig, D. G., and Turnbull, J. E. (2002) J Biol Chem 277(32), 28554-63.

48. Mach, H., Volkin, D. B., Burke, C. J., Middaugh, C. R., Linhardt, R. J., Fromm, J. R., Loganathan,

D., and Mattsson, L. (1993) Biochemistry 32(20), 5480-9.

49. Plotnikov, A. N., Hubbard, S. R., Schlessinger, J., and Mohammadi, M. (2000) Cell 101(4), 413-

24.

50. Delehedde, M., Lyon, M., Gallagher, J. T., Rudland, P. S., and Fernig, D. G. (2002) Biochem J 9

51. Ishihara, M., Kariya, Y., Kikuchi, H., Minamisawa, T., and Yoshida, K. (1997) J Biochem (Tokyo)

121(2), 345-9.

52. Lundin, L., Larsson, H., Kreuger, J., Kanda, S., Lindahl, U., Salmivirta, M., and Claesson-Welsh,

L. (2000) J Biol Chem 275(32), 24653-60.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

25

53. Kariya, Y., Kyogashima, M., Suzuki, K., Isomura, T., Sakamoto, T., Horie, K., Ishihara, M.,

Takano, R., Kamei, K., and Hara, S. (2000) J Biol Chem 275(34), 25949-58.

54. Merry, C. L., Bullock, S. L., Swan, D. C., Backen, A. C., Lyon, M., Beddington, R. S., Wilson, V.

A., and Gallagher, J. T. (2001) J Biol Chem 276(38), 35429-34.

55. Rahmoune, H., Chen, H. L., Gallagher, J. T., Rudland, P. S., and Fernig, D. G. (1998) J Biol Chem

273(13), 7303-10.

56. Brickman, Y. G., Ford, M. D., Gallagher, J. T., Nurcombe, V., Bartlett, P. F., and Turnbull, J. E.

(1998) J Biol Chem 273(8), 4350-9.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

26

FIGURE LEGEND

Figure 1. Various Models of FGFR Signaling Complexes with Three Modes of Protein/HS Interaction

(A) Three models with cis mode of interaction. The yellow balls represent FGFs; the blue bars represent

FGFRs; the red rods represent HS. (B) Two models with trans mode of interaction. (C) Three models with

mix mode of interaction. In model VII, two chains of HS are incorporated.

Figure 2. Oligosaccharide Labeling and Binding Assay with FGF1 and FGFR1

(A) An oligosaccharide ladder from dp2 to dp24 was radiolabeled with 6-OST-1 and separated with a 15%

PAGE gel. The dp2 was prepared from digestion of a 6-OST-1 labeled heparin. The size of each

oligosaccharide is indicated. (B) Measuring the minimum FGF1 and FGFR1 binding HS. The

oligosaccharide ladder was used to do the binding assay with FGF1 or FGFR1. Each lane contained

roughly 25 ng of oligosaccharide and 31 pmol (50 ng) of FGF1 or 24 pmol (100 ng) of FGFR1. F1, FGF1;

R1, FGFR1.

Figure 3. Observation of FGF1:HS:FGFR1 Ternary Complex

(A) Complex formation and its concentration dependence. Excess dp18 was used for the reactions and the

concentrations of FGF1 and FGFR1 in the binding reactions are indicated. F1:HS, binary complex of FGF1

and HS; HS:R1, binary complex of HS and FGFR1; F1:HS:R1, ternary complex of FGF1, HS and FGFR1.

(B) Thermal stability of the binary and ternary complexes. Around 0.8 µM each of FGF1, FGFR1 and dp18

were used in all the binding reactions. The heating temperature and the duration (in minutes) for each

reaction are indicated. (C) Ternary complex formation on CHO cells assayed with flow cytometry. The

cells were sequentially sorted without any protein, with FGFR1β(IIIc)/Fc, and with both FGF1 and

FGFR1β(IIIc)/Fc. Left panel, HS-deficient CHOpgsA-745 cell. Middle panel, HS-expressing CHO-K1 cell.

Right panel, a CHO cell expressing 2-O sulfation negative HS. The fluorescence tagged FGFR1β(IIIc)/Fc

was monitored. F1, FGF1; R1/Fc, FGFR1β(IIIc)/Fc.

Figure 4. Minimal HS in the Ternary Complex of FGF1:HS:FGFR1

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

27

(A) Binding assay with the oligosaccharide ladder in the presence of FGF1 and/or FGFR1. Each lane

contained 100 ng oligosaccharide and 16 pmol each of FGF1 and/or FGFR1. FGF1:HS:FGFR1 is visible

from dp4 to dp6. (B) Western blot of another gel with anti-FGF1. In order to see FGF1:HS, limited FGFR1

was applied to each binding reaction. Lane 1 and 2 contained no HS.

Figure 5. Molar Ratio between FGF1 and FGFR1 in the Ternary Complex

(A) Gel shift assay with excess HS (dp12). Lane 2 to 9, constant concentrations of FGF1 (0.8 µM), but

increasing concentrations of FGFR1 were applied. The molar concentration ratios (R1/F1) were indicated

on the top. No protein was in lane 1 and only FGFR1 (3.2 µM) was added in lane 10. (B) Densitometry

analysis of A. (C) The band intensity of the ternary complex was plotted against the molar equivalents of

FGFR1 to FGF1.

Figure 6. Effect of Critical Sulfate Groups on the Formation of the Ternary Complex

(A) 3-OST-1 and/or 6-OST-1 modified heparin oligosaccharides and chemically modified heparin were

shifted with FGF1 and /or FGFR1. ~/31, modified with 3-OST-1; ~/61, modified with 6-OST-1; ~/31,61,

modified with 3-OST-1 and 6-OST-1; F1, FGF1; R1, FGFR1. (B) Disaccharide analysis of the heparin and

chemically modified heparins. a, heparin; b, DSNAc; c, DSNS, d, 6ODS

Figure 7. A Proposed Mechanism for the Formation of the FGF1/FGFR1 Signaling Complex

In inactive state, FGF1 and FGFR1 bind to their primary binding sites on the HS chain. The binding sites

can be contiguous in a same sulfated domain or separated in different sulfated domains, depending on the

size of the sulfated domains. During the activation, one FGF1 binds to one FGFR1 on the same chain of HS

and form a half complex at the non-reducing ends. Two of these half complexes then associate to an active

molecular signaling complex.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

28

TABLES

Table 1. Selecting FGFR1 Molecular Signaling Models with HS Size and Stoichiometry

Models in Figure 1 I II III IV V

VI VII

VIII

Minimum HS sizea dp 12 dp 8 dp 20 dp 4 dp8 dp8 dp 4 dp8

Molar ratio of FGF1, HS, FGFR1

1:1:1 2:1:2 1:1:2 2:1:2 1:1:1 2:1:2 2:2:2

2:1:2

HS/FGFR1 interaction

yes yes yes no yes yes yes yes/nob

Receptor dimerization

no yes yes yes no yes yes yes

Fit no no no no no no yes no

aMinimum HS is derived from either literature or calculation whichever is smaller. Calculation is estimated

according that the minimum FGF1 binding HS is a tetrasaccharide and minimum FGFR1 biding HS is an

octasaccharide (Figure 2B)

bOne FGFR1 interacts with HS, the other does not.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

29

Table 2. The Molar Ratio of Oligosaccharide to FGF1 in the Ternary Complex

HS size 4 6 8 10 12 14 16 18 20 24

HS (pmol) 16.7 12.5 14.6 16.6 15.5 15.2 19.5 14.1 14.4 13.8

FGF1/HS 1.04 0.78 0.90 1.03 0.96 0.95 1.21 0.88 0.90 0.86

Note: Each binding reaction contained 16 pmol of FGF1, 64 pmol of FGFR1 and 250 ng of oligosaccharide.

The molecular weight of oligosaccharides was calculated based on the molecular weight of the fully sulfated

disaccharide, which is 577.

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

30

Table 3. Sulfation versus complex formation and mitogenic activity

HS Sulfation aF1:HS aHS:R1 aF1:HS:R1 bMitogenic Activity

Dp2 2, 3, 6-O, N- - - - 0.07± 0.08

Dp4 2, 3, 6-O, N- + - + 0.35 ± 0.15

Dp6 2, 3, 6-O, N- + - + 0.46 ± 0.14

Dp8 2, 3, 6-O, N- + + + 0.51 ± 0.12

Dp10 2, 3, 6-O, N- + + + 0.75 ± 0.20

Dp12 2, 3, 6-O, N- + + + 0.83 ± 0.19

Dp14 2, 3, 6-O, N- + + + 0.95 ± 0.22

Dp16 2, 3, 6-O, N- + + + 0.97 ± 0.22

DSNAc/61 6-O- - - - 0.03 ± 0.05

DSNAc/61,31 3, 6-O- - - - 0.04 ± 0.05

DSNS/61 6-O, N- - - + 0.75 ± 0.25

DSNS/31 3-O, N- - - + 0.56 ± 0.17

6ODS/31 2, 3-O, N- + - + 0.66 ± 0.20

HSpgsF-17c 3, 6-O, N- - - + 0.72 ± 0.14

aThe formation of the corresponding complex is indicated by plus and minus signs.

bRelative values to the mitogenic activities of heparin.

cThe HS prepared from mutant cell CHOpgsF-17

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from

and Robert D. RosenbergZhengliang L. Wu, Lijuan Zhang, Tomio Yabe, B. Kuberan, David L. Beeler, Andre Love

The involvement of heparan sulfate in FGF1/HS/FGFR1 signaling complex

published online February 25, 2003J. Biol. Chem. 

  10.1074/jbc.M212590200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

by guest on April 16, 2018

http://ww

w.jbc.org/

Dow

nloaded from