Binding interaction studies of selected receptor subpopulations after partial cross linking receptor ligand complexes with a photoactivated heterobifunctional reagent

Binding interaction studies of selected receptor

subpopulations after partial cross-linking

receptor-ligand complexes with a

photoactivated heterobifunctional reagent.

G B Faguet, D Beebe

J Clin Invest.

1986;

78(1)

:67-72.

https://doi.org/10.1172/JCI112575

.

Certain hormonal and nonhormonal binding systems such as the

leukoagglutinin-lymphocyte model exhibit complex receptor-ligand interactions that result in nonlinear

Scatchard plots. Such plots are interpreted as indicating either homogeneous negatively

interacting binding sites or heterogeneous sites with different and fixed affinity. We

assessed the validity of these interpretations in our system by conjugating the ligand to a

photoactivated heterobifunctional agent and cross-linking the conjugate to a subset of

receptors before studying the binding interactions of non-cross-linked sites. Conjugation did

not qualitatively or quantitatively affect the binding properties of the ligand. Cross-linking

was specific, efficient, and stable and had no effect on irrelevant surface receptors.

Cross-linking of only 3% of the total receptors resulted in 50% decreased ligand binding to high

affinity sites consistent with a calculated inactivation of 85% and 2% of high and low affinity

sites, respectively. Such preferential inactivation of high affinity sites in an unequivocal

demonstration of binding sites heterogeneity in this system and shows a clear rejection of

the homogeneous cooperation model.

Research Article

Find the latest version:

http://jci.me/112575/pdf

Binding Interaction Studies

of Selected

Receptor Subpopulations

after

Partial

Cross-linking Receptor-Ligand Complexes

with

a

Photoactivated Heterobifunctional Reagent

GuyB.Faguet*t§and DeborahBeebe§

*Department ofMedicine,andtDepartment ofCell and MolecularBiology,MedicalCollege of Georgia; and§Medicaland ResearchServices, Veterans Administration MedicalCenter, Augusta, Georgia30910

Abstract

Certain hormonal and nonhormonal binding systems suchas the

leukoagglutinin-lymphocyte model exhibit complex receptor-li-gandinteractions that result in nonlinear Scatchard plots. Such plots areinterpreted as indicating either homogeneous negatively interacting binding sites or heterogeneous sites with different andfixed

affinity.

We assessed thevalidity of these interpreta-tions inoursystemby

conjugating

theligandto aphotoactivated heterobifunctional agent and cross-linking the conjugate to a subsetof receptors before studying the binding interactions of non-cross-linked sites. Conjugation did not qualitatively or quantitatively

affect

thebinding properties of the ligand. Cross-linking wasspecific, efficient, and stable and had no effect on irrelevant surface receptors. Cross-linking of only 3% of the total receptors resulted in 50% decreased ligand binding to high affinity sites consistent with a calculated inactivation of 85% and2%of high and lowaffinity sites, respectively. Such preferential in-activation of high affinity sites isanunequivocaldemonstration of binding site heterogeneity in this system and shows a clear rejection of the homogeneous cooperation model.

Introduction

Evidence

suggests thatcomplex

receptor-ligand

interactionsin

certain hormonal and nonhormonal systems reflect

occupancy-dependent affinity (1, 2).

To

date,

such evidence has relied on

demonstrating

enhanced

dissociation of

bound labeled ligand

asreceptor occupancy is

increased

by addition ofan excessof unlabeled

ligand

to the

dissociation mixture

(1,

2).

While

suggestive,

this approach

fails

toprove the existence of

negatively

interacting binding sites

(3-7). Thus, the

existence of

hetero-geneous

binding

sites with different but fixed

affinity

for the

ligand remains

the

preferred

explanation of

such complex

bind-ing interactions. Objective

resolution

of

the

uncertainty regarding

binding

models

underlying complex binding

interactions is hampered by thephenomenological nature

of

receptor-ligand

binding

studies and

by

methodological limitations

that preclude

selectively studying

eachclass of

heterogeneous

binding

sites.If

receptorsbelongingto oneoftwodistinct classes

of

binding sites werecross-linked to the ligand, then data from post-cross-linking

binding

studies would reflect thesecond

non-cross-linked

class of receptor. Suchanapproach

might

provide

an

experimental

Address correspondencetoDr. Faguet, VAMC(11IN), Augusta, GA 30910.

Receivedfor publication 7October 1985.

basis for distinguishing homogeneous interacting from

hetero-geneousnoninteracting binding sites.

Using the lymphocyte-leukoagglutinin

(LPHA),'

a

receptor-ligandsystemthatgeneratesbinding data compatible with either homogeneous interacting orwith heterogeneous binding sites with different and fixed affinities,wehave examined the binding interactions ofone of the two assumed receptor classes after selectively cross-linkingthe other receptor class with a

photoac-tivatedheterobifunctionalreagent.Our observationsclearly show that in the LPHA-lymphocyte system, complex binding inter-actions reflect binding site heterogeneity and are inconsistent with the negatively cooperative model.

Methods

Cells

IM9, aB-lymphoblastoid cell line, was grownin RPMI 1640 (Gibco, Grand Island, NY) with 10% fetal calfserum(MABioproducts, Walk-ersville, MD)at37°C ina5%CO2 incubator (FormaScientific, Marietta, OH). Beforebinding assays,cellswere harvestedbycentrifugation at 200gfor 10 min and washed three times with medium. Cell viability assessed by Trypan Blue dye exclusionwasconsistently>95%.

PHA

purification

and iodination

LPHA,apotentmitogenic lectin (8, 9),wasprepared fromasingle batch ofPHA-P(DifcoLaboratories, Inc., Detroit, MI) accordingto a modi-fication of the column chromatography methods of Weber (8),as pre-viously described(9). LPHA wasiodinated (with reductant-free '25Ior

1'3I, 17and20Ci/mgspact,respectively, Amersham Corp., Arlington Heights, IL) bythechloramine-T method of Hunter and Greenwood (10)to aspecific activity of0.1-1

Ci/Ag.

Details ofthe LPHApurification andiodination procedures, and of itsbiological activityhave been pub-lished elsewhere (9, 11, 12).

Insulin iodination

Porcineinsulin,agift from Dr. R. Chance (Eli Lilly & Co., Indianapolis, IN),wasiodinated(reductant-free

3'I,

20

Ci/mg

spact,AmershamCorp.) bythechloramine-T methodof Hunter and Greenwood (10)toaspecific activity of 100-200

Ci/,gg.

LPHA

conjugation

RadioiodinatedLPHAwasconjugatedwiththeheterobifunctional re-agentethylN-5-azido-2-nitrobenzoylaminoacetimidate-HCI (ANB-AI),

agift fromDr. W.S.Allison (University of California,SanDiego, CA), accordingtopublished procedure (13).Briefly, after dialysis against 0.15 MNa2CO3buffer, pH9,overnightat4°C, 0.5 mg of '251-LPHAwas exposedto a 10-fold molar excessofANB-AI and rotated for 8h at room temperature. Thereactionmixture wasfilteredover aG-50 se-phadexcolumn(Pharmacia Fine Chemicals, Piscataway, NJ),and

equil-The

Joural

ofClinicalInvestigation,Inc. Volume78, July 1986,67-72

1.Abbreviations usedinthis paper:ANB-AI, ethyl

ibrated with 0.15 M phosphate buffer, pH 7.4. The pooled radioactive peakcontaining the activated radiolabeled lectin ('25I-LPHA-ANB-AI, 0.1-1

_Ci/;g

spact) was inaliquots and stored at -70°C. All manipu-lations of the activated lectin were carried out in the dark.

Receptor-ligand binding

studies

Binding equilibrium. Increasing concentrations (1.5 x

10'

M to 1.5

X 10-6M) of13 I-LPHA ofconstant specific activity were incubated with fresh cells(106aliquots) or with cell aliquots previously cross-linked with tracerconcentrations (2.65 X 10-8 M) of

125I-LPHA-ANB-AI

in0.5 ml Hanks' buffer containing0.1%ofbovine serum albumin (BSA; Sigma Chemical Co., St. Louis, MO) in polystyrene tubes precoated overnight with 0.5% BSA in Hanks' buffer to minimize nonspecific binding. Con-trolsincluded

'3'I-LPHA

binding to fresh cells in the presence of equi-molar concentrations ofANB-AIandequilibrium binding of equimolar concentrations of125I-LPHA-ANB-AIof constant specific activity to fresh cells.After reaching equilibrium (60 min incubation at22°C)thebound and unbound tracer from each incubation were separated by centrifu-gation over a 4:1 mixture of dibutyl phthalate and bis-2-ethylhexyl phthalate (Eastman Kodak Co., Rochester, NY) at 10,000g(Microfuge 12; Beckman Instruments, Inc., Fullerton, CA) for 2 min in conical microcentrifuge tubes (Sarstedt, Inc., Princeton, NJ). The centrifuge tube tips containing the cell pellet were cut off and cell-bound radioactivity of the pellet was determined in a gamma counter (Gamma 4000; Beck-man Instruments, Inc.). Aliquots of cells cross-linked with

'25I-LPHA-ANB-AI but not exposed to

'3'1-LPHA

werecounted to determine the amount andefficiencyof cross-linking. Insomeexperiments binding of

1251-insulin

to IM-9 cells under steady state was determined as described for LPHA except that cells were incubated with

1251-insulin

(1.7X 10-"

M) in the presence of increasing concentrations of native insulin.

Dissociation studies. Cells

(106

aliquots) were exposed to tracer

amounts (2.65 x 10-8 M) of

'251-LPHA

or

125I-LPHA-ANB-AI

in0.1

mlof Hanks' buffer containing0.1%BSA (Hanks'-BSA). After reaching equilibrium, cells were irradiated and unbound tracer ligand was removed bywashing the cells twice with Hanks'-BSA bufferat4°Candreplacing thesupernatantwith50 volumes of Hanks'-BSA buffer alone or in buffer

containing nativeLPHAinconcentrationsequaltothe

'251-LPHA

present

in theoriginal sample. Volume dilution prevents reassociation ofligand

releasedduring thedissociation reaction (1, 2). Addition ofnative LPHA increases the overall occupancy by the ligand and thus enhances tracer

dissociation,asdescribed for this (2) and other systems ( 14). At various

times, duplicate cellaliquots were processed as described above to as-certain receptorligandcomplex dissociation.

Receptor-LPHA

cross-linking

Cross-linking wasinduced as follows: Afterequilibration withtracer concentrations of

1251I-LPHA-ANB-AI,

cellswereplaced ina BSA-pre-coated quartz glass tube andirradiatedfor 5 min.Irradiations were carried out in a Rayonet photochemical reactor (RMR-500; Southern New EnglandUltraviolet Co., Hamden, CT). The apparatus contains four

fluorescenttubeswithanoutput of 975 milliwatts eachat300nm ar-ranged around acylinder.This long wavelength light increases the

reac-tivityof the photo-generated arylnitrene and reduces undesirable pho-tolysisonotheramino acid residues of theconjugatedprotein(15).The

samplesin glass tubesrotateat 5 rpm inacarousel (Merry-go-round, RMR-400;SouthernNewEngland Ultraviolet Co., Hamden, CT),

in-suringuniform irradiation ofallsamples.

Data

analysis

SpecificLPHAandinsulinbindingwascalculatedfromtotalbinding

bysubtractingnonspecific binding,asdetermined bybindingof

radio-ligandin the presence of1,000-foldexcessof nativeligand.Datafrom dualradiolabelexperimentswerecorrectedforspilldownof13'Icounts into the

1251I

window. Data are presentedasaverages ofreplicate exper-imentsasindicated in thefigurelegends.Inall experiments,each data

pointisassessedinduplicate.Standard deviations (not shown)were

gen-erally <5%of the mean. Equilibrium bindingdata were analyzedby

curve-fitting usingthe least-square "ligand" program (16) adaptedto

AppleSoft Basic for theAppleII+ microcomputer(AppleComputer,

Inc.,Cupertino, CA). Fitted parameters are displayed in Scatchard co-ordinates( 17).

Results

Properties

of

the

125I-LPHA-ANB-AI

conjugate.

Fig.

1shows the radioactivity, protein, and ANB-AI profiles of the fractionation eluate ofthe modified

'251I-LPHA.

As shown, nearly all the mod-ified

125I-LPHA

eluted between fractions 11 and 16 with

clear-cut separation between the conjugated and the unconjugated fractions of ANB-AI. The figure inset shows the

absorption

spectraofthe modified and unmodified LPHA. Basedonthe extinction coefficientofANB-AI (E =9 X I03

M`

at318nm), the molecular weight of LPHA (128,000), and the contribution ofANB-AItothe modified LPHA absorbanceat 318 nM, the molar ratio ofreagent tolectincanbe estimated. Reaction

con-ditionswerechosentogenerateconjugates withanANB-AIto

LPHA molar ratio of 1. Becauseonecross-link issufficient, such

aratio reduces undesirable cross-linkingtononreceptor surface moieties often associated withexcessesofheterobifunctional re-agent.

Effect ofANB-AIonLPHAbinding. Whetheror not ANB-AI affects the magnitudeoraffinity of LPHAbindingto lym-phocytereceptors was ascertainedinthefollowing experiments. Cellswereequilibrated eitherwith 1.5 X 10-9M to 1.5 X 10-6 M of

'3ll-LPHA

inthepresenceand absence of equimolar

con-centrationsof ANB-AI orwith 1.5X

10-9

Mto 1.5 Xt 10-6 M

'251-LPHA-ANB-AI

conjugateasdescribed under Methods. As shown in Fig. 2 A, binding of LPHAtoitslymphocytereceptors wassubstantially thesame at allligandconcentrations ineach

setof experiments. Replots in Scatchardcoordinates(Fig. 2B) showthatbindingdata from each set ofexperimentsgenerates avirtuallyidentical curvilinear Scatchardplot.Thus,underthese experimental conditions,ANB-AI does not affect themagnitude

orcomplexity of LPHA bindingeither asunconjugated reagent

inthereactionmixture or whenconjugatedto thebinding ligand.

o

.3

C) 310260 500

50

0

6 10 14 18 22 26 30 34

Fraction Number

Figure1.Chromatography of 125I-LPHAconjugatedwith ANB-AIon sephadex G-50. Each tube contained 0.5mlofeffluent. Fractionswere assayed for the presenceof radioiodine (o) inagamma counter; pro-tein(A)at700nmabsorbance; and for presence of ANB-AI(o),at405 nm.Theinset showstheabsorption spectra ofconjugated(upperline)

Dissociation Time

(min)

200

Bound

Figure 2.(A)Binding of'3'I-LPHAand'25I-LPHA-ANB-AItoIM9 cells.Cellswereequilibratedwithincreasingconcentrationsof

`31-LPHA inthepresence(A)and absence(o)of ANB-AI and with

in-creasingconcentrations of 125I-LPHA-ANB-AI(-).Cell-bound radio-activitywasmeasuredinduplicate aliquotsof 106cellsasdescribed in thetext.Receptor-bound ligandisexpressedasbound(logofng 125[-LPHA/106 cells) and is plottedas afunctionof the total ligand added

(zg/106 cells). (B) Replotsthebinding data inScatchardcoordinates. Boundoverfree radioiodinated ligandisexpressedas afunctionof bound(ngX 106cells).

Assessmentofcross-linking. Toassesscross-linking efficiency,

twosetsof 8 X 106 cells were incubated withtracer amounts

(2.65 X 10-8 M)of '251-LPHA or

1251I-LPHA-ANB-AI,

respec-tively.Afterreachingequilibrium, bothsetsweresimultaneously

irradiated as described above, washed twice in Hanks'-BSA

buffer,anddilutedwitha50-foldexcessofHanks'-BSAbuffer

alone, orinbuffercontaining native LPHAat concentrations (2.65 X 10-8M)sufficient tomaintainequilibrium initially in-ducedbythetracer(2). Duplicate aliquotsof106 cellswere

re-moved fromeachset at0, 5, 30, and 60min and processedas

described above to ascertain cell-bound radioactivity. Under these experimental conditions, dissociation ofreceptor-ligand complexes proceeds spontaneously (basal dissociation)oris

ac-celeratedby thepresenceofexcessof nativeligandinthereaction mixture(ligand-enhanced dissociation)asdemonstrated for this (2)and othersystems(1, 14).AsshowninFig.3 A,spontaneous andLPHA-enhanced tracerdissociationwere37 and68%,

re-spectively,at60min.Thesevaluesarecomparabletodissociation experiments performed on nonirradiated lymphocytes (2). In

Figure3.(A)Dissociation of receptor-LPHA complexes.After equili-bratingwithtraceramounts(2.65X Io-0 M/106 cells)of'25I-LPHA (left)or1251I-LPHA-ANB-AI(right),cellswereirradiatedandwashed twicewith buffer.Supernatantswerereplaced bya50-foldexcessof fresh mediumcontaining (o,set 1)ornotcontaining(o,set2)2.65

X10-8Mof nativeLPHAper106 cells.Attheintervals shown, dupli-catealiquotsfromeachsetwereprocessedasdescribed in thetext to

ascertain the cell-boundradioconjugated ligand.Dataareexpressedas

percentagesof timezerovalue andplottedas afunction of dissocia-tiontime.(B)Scatchard plots of'31I-LPHAequilibrium bindingto

IM9 cellsbeforecross-linking showingthehigh and low affinity

com-ponentsofthecurve.The fraction of total cross-linkedreceptors(a

=13 ng/106cells) linearly extrapolatestothehighand lowaffinity

componentsintersectingatb=3 ng/106 cellsandc=9ng/106 cells,

respectively. Ratioofboundoverfree'3'I-LPHA (X 100)isplottedas

afunctionofbound '3'I-LPHA (ngX 106cells).

contrast,irradiation of cells equilibrated with

1251I-LPHA-ANB-AIprecluded bothspontaneousand LPHA-enhanced dissocia-tionofreceptor-bound '251I-LPHA-ANB-AI. Asshown,95 and 92% ofreceptor-'251-LPHA-ANB-AI complexes remained

cross-linkedunder basalandLPHA-enhanced conditions, respectively,

after 60 mindissociation.

Binding interactions after cross-linkingoneoftwo(assumed) classesofreceptors. Inthissystem, LPHA-receptor binding

in-teractionsdata generateacurvilinear Scatchardplot (Fig. 3B)

compatiblewithmorethanoneclass of sites with different and fixedaffinity, orwithoneclass of interacting bindingsites.

As-suming heterogeneous sites,theScatchard plotwas

mathemat-icallyresolvedbyweighted nonlinear least-squarescurvefitting

0

co

1-.

Total

'a

4-0

LL

1-co

200

toestimate the equilibrium binding constants (K and R values) of each binding class. Atwo-site model was identified as "best fit" consisting of a small component

(RI

= 3.6 ng/106 cells) of highaffinity sites (K, =2 X 109M-'); and a large component (R2 =388 ng/106 cells) of low affinity receptors (K2 = 5 X 106 M-'). Using dual labeled ligand, we conducted equilibrium binding studies using

'3'I-LPHA

on the assumed class oflow

affinity

sites after

cross-linking

the assumed classof high affinity receptors (with125I-LPHA-ANB-AI) as described under Methods. Cell aliquots equilibrated with

'3'I-LPHA

in the presence or absenceof ANB-AI followed by irradiation and cells equilibrated with

'3'I-LPHA-ANB-AI

without irradiation served as controls. Thefraction of cross-linked receptors (13 ng/ 106 cells or -3% of thetotal) exceeded (by >3:1) the estimated capacity of high

affinity

sites, thus

insuring

that

equilibrium binding

data gen-eratedunder these

conditions reflect

mostly low

affinity

site in-teractions

minimally

contributed to by the class of

high

affinity

sites. This prediction

was

graphically confirmed from

the

inter-section of the line of cross-linked receptors (Fig. 3 B) with the low andhigh affinity components of the Scatchard plot, which showsthat 2 and 85% of the assumed low and high affinity re-ceptors,

respectively,

wereinactivated. The latter

estimate is

fur-ther supported by curve

fitting of

theScatchard plot

following

cross-linking which revealed a 15% remnant active high affinity class(0.56 ng/ 106 cells).

'3'I-LPHA

binding to

cross-linked

cells

(Fig.

4,

solid

squares) generated a

curvilinear

Scatchard plot

qualitatively

comparable to those

derived from

131I-LPHA

bind-ing

to

non-cross-linked

cells

(Fig.

2 B, open squares, open tri-angles, and

solid

circles)

but

with

a

50%

decrease in

binding

to

high

affinity sites.

Insulin

binding

to '25I-LPHA-ANB-AI cross-linked cells. To ascertain the

specificity

of

receptor-ligand

cross-linking

andto assessthe

potential

effect of

cross-linking

onthe cell membrane and on unrelated membrane components, we

performed

the

following experiments:

cell

aliquots

were

equilibrated

with

1251.

LPHA-ANB-AI and in the presence

of

an excess

of native

LPHA. Excesses

of native

LPHAblocked

binding

of

the

conjugated

li-gand

to

specific

receptors

resulting

in a

binding

component

comparable

tothe unsaturable

nonspecific

binding

obtained

from blanks (not

shown).

In

addition, equilibrium

'3'I-insulin

binding

experiments

wereconductedon

IM9

cells

cross-linked

with

1251I

LPHA-ANB-AI.

Insulin

waschosenas asecond

ligand

because the

density ratio

of LPHA toinsulin receptorson

IM9

cellsis

-100:1

(G.

B.Faguet,

unpublished

observations). Thus, binding

interactions of

low

density

receptors would be

appreciably

altered

by

cross-linking

even asmall

fraction of

the

high

density

receptors should the

cross-linking

be

either partly

nonspecific

orshould

it induce allosteric

effects. Fig.

5shows theScatchard

display of

insulin

binding

to fresh IM9 cells andto IM9 cells

pre-cross-linked

with LPHA. The

fraction

of cross-linkedLPHA was cal-culatedat 17

ng/106

cellsor

-80,000

receptors per cell.

Likewise,

it

is

calculated

that

109-3,135 insulin

receptors per cellwere

boundwith

insulin. Thus,

the ratio

of

LPHA

cross-linked

to

insulin-bound receptors over the insulin

concentration

range

assayed

was736:1to26:1.

Insulin

binding

toLPHAcross-linked and

non-cross-linked

IM9 cells was

virtually

identical,

which

strongly

suggests that

LPHA-receptor

cross-linking

hadno

effect

onthe

binding interactions

of non-LPHA receptors. This suggests thatLPHA

cross-linking

was

receptor-specific,

exertedno

mea-surable allosteric

effects,

and didnotaffectthe

complex

inter-actions of irrelevant receptors.

IL

en

Bound

IL

0 100 200 300

Bound

400

Figure 4.Scatchard plotof'3'I-LPHAequilibrium binding to IM9 cellsafter 3% receptorinactivation (m) compared with expected curves forremaining sites assumed homogeneous and negatively interacting (A), orpostulated heterogeneouswithhigh and low components (B).

1

-L0

,--0.5 -m

0 20 40 60

Bound

80 100

Figure 5. Scatchard plots of insulin bindingtofresh and LPHA-cross-linkedIM9 cells. 125I-insulin (1.7X

10-1I

M)wasincubated for 90

minat15°C withtwo setsof 2X 106 cells/ml in thepresenceand ab-senceof various concentrations ofnative insulin.Onesetconsistedof

freshcells (o), the other of cells that had been preincubated with '3'I-LPHAand irradiated (o)asdescribedinthetexttocross-linka frac-tionof LPHAreceptors.Specific '25I-insulinbindingwasdetermined and theboundoverfree ratio of labeled hormone isplottedas a func-tion ofthe total concentration of bound hormone.

B

3- 2-a

U

A

Discussion

Thisreportshowsthataselectfraction of LPHAreceptorscan

becross-linkedtoLPHAbytheheterobifunctional agent ANB-Aland that selective binding interaction studiescanbe conducted

on the non-cross-linked receptor population. One attractive

feature of the heterobifunctionalagentANB-AIisthat itcanbe

activated with 300nm ultraviolet light (13) reducingtoa

min-imum radiation damagetothetarget(13, 18).Thisis ofparticular

importancewhen intact cellsareinvolved. Crucialtothe

inter-pretation ofourresults, and the validity ofourconclusions, is

theassessmentofthespecificityandefficiencyofcross-linking;

the distinction between cross-linked and non-cross-linked

re-ceptors; andtheevaluationofthepossiblecontribution of

bio-logical cross-linking during photolysis (19, 20). The specificity

ofbinding and cross-linking LPHA-ANB-AI to lymphocyte LPHAreceptorsis shown by complete inhibition of LPHA-ANB-Albinding by native LPHA and the factthat LPHAreceptor

cross-linking hadnoeffectonthe binding interactions of

irrel-evantinsulinreceptors. Asignificant degree ofrandomchemical

orbiological cross-linkingwould beexpectedtoqualitativelyor

quantitativelyaffect binding ofirrelevantreceptorsparticularly

atsuchahigh density ofthetargetreceptor(3.4 X 106/cell)and

with a 100:1 target/irrelevant receptors ratio. Irradiation was

determinedtoinduceefficient cross-linkingas -100% of

LPHA-ANB-AIboundto receptors atthe timeof irradiationwas

cross-linked by photolysis. By virtue of the 1:1 LPHA/ANB-AI molar

ratio oftheactivated lectin, this cross-linking efficiencysuggests thatthe ANB-AI attaches LPHAverynearitsbindingsite.

Sta-bilityofcross-linked complexesisdemonstratedby the minimal spontaneousdissociation (-5%)ofreceptor-boundand irradi-ated LPHA-ANB-AIovera60-min period (compared with 37%

forLPHA), andthe fact thatdissociation ofcross-linked

com-plexeswasnotenhanced byexcessof native ligand.Assessment ofthepotential effect ofcross-linking bindinga fractionof

re-ceptorsonthepost-cross-linkingbinding datamustbeexamined

inthecontextof the opposing hypothetical binding models:

ho-mogeneousinteracting bindingsiteswithoccupancy-dependent

affinityvs.heterogeneous classes ofsiteswith different and fixed

affinity. Cross-linkingofafraction ofbindingsites would have noappreciable effectonbinding interactions of the remaining non-cross-linkedreceptorsifallreceptorsbelongtoasingle class

ofinteracting bindingsites. Indeed, non-cross-linkedreceptors

wouldretain variable affinity dependenton occupancy.

Accord-ingtothismodel, datafrom equilibrium binding studies

con-ductedonnon-cross-linkedreceptorswouldgenerateidentical

curvilinearScatchard plots regardless of whether afraction of

receptorsiscross-linked.Incontrast,inasystemwithtwoclasses

ofnoninteracting bindingsiteswithdifferentand fixed affinity binding, studies performed afterpreferential inactivation ofa

classofsites will reflect interactions of thenon-cross-linkedclass.

Accordingtothis model, the magnitudeand classspecificity of

the inactivation will dictate the extent towhich the pre- and

post-inactivation Scatchard plots differ.Inoursystemwefound

thatinactivation ofonly3%ofthetotalreceptorsby cross-linking,

resulted ina50%decrease insubsequenttracerligandbinding

tohigh affinitysites. Thispreferential inactivation of high affinity

sitesrepresents an unequivocal demonstration of binding site

heterogeneityinthissystemandaclearrejection ofthe

homo-geneousnegatively cooperative binding model. Indeed, graphical

analysis of

the

Scatchard plot

indicates that cross-linking

inac-tivation

of 3%

of

the total receptors

eliminates

85% of the low

capacity-high affinity

receptors

but

only

2%

ofthe

high

capacity-low

affinity

sites (Fig. 3 B),

which

is consistent with a two-class heterogeneous binding site

interpretation

(Fig. 4). It could be argued that the

receptor-LPHA-ANB-AI

complex

dissociates

but that the

conjugate remains cross-linked

by the ANB-AIat asite other than the LPHA receptor.

Should this

be the case,

cross-linking would have no effect on binding

interactions

because, ineffect, all receptors would remain available to the

unconjugated

tracer

ligand.

However, two

conditions

mustbemet

for

thisto

occur: (a)

cross-linking would

havetobe random rather

than

receptor-specific,

and

(b)

the

association

constant of the

uncon-jugated ligand should

be greater than the dissociationrateofthe

conjugated

ligand (concerted

substitution hypothesis).

However, such

alternative

explanation of

our

data is

inconsistent withour

findings

that

cross-linking

is

receptor-specific

and that

conju-gation did

not

measurably affect

thedissociationconstant

for

theligand. We

did

not

examine

and thereforecannotextrapolate

our

conclusions

toother systems

with

complex

binding

inter-actions. However, given the

striking similarities

in

equilibrium

andkinetic

binding

data

derived from

LPHAand

certain

other

ligands

(2, 4, 9, 14),

it

is

unlikely

that our

observations

represent

a

phenomenon restricted

toLPHA receptors.

Acknowledgments

Theauthors thank the reviewer for helpful commentsand M. A. Jones and P.A.Thompson for typing the manuscript.

This workwassupported inpart by the VeteransAdministration.

References

1. DeMeyts, P., J. Roth, D. M. Neville, J. GarvinIII, andM. A. Lesniak. 1973. Insulin interactions with itsreceptors:experimental ev-idencefornegativecooperativity.Biochem. Biophys.Res.Commun. 55:

154-161.

2.Faguet, G.B.1979. Leukoagglutinin bindingtohuman

lympho-cytes:experimentalsupportfornegative cooperativity.Am.J.

Physiol.

237(3):E207-E213.

3.Cuatrecasas, P., andM. A. Hollenberg. 1975.Bindingof insulin and other hormonestonon-receptormaterials: saturabilityspecificity

and apparentnegative cooperativity.Biochem.Biophys.Res.Commun. 62:31-41.

4. Pollet, R. M., M. L. Standaert, andB. A. Haase. 1977. Insulin binding to human

lymphocyte

receptor. Evaluation of the negative cooperativitymodel.J.Biol. Chem. 252:5828-5834.

5.Verries,B.,G. Fayet, and S. Lissitzky. 1974.Thyrotropin-binding

properties of isolated thyroid cells and theirpurifiedplasma membranes. Relation ofthyrotropin-specific bindingtoadenylate-cyclase activation. Eur. J.Biochem. 42:355-365.

6. Limbird, L. E., P. DeMeyts, andR.J. Lefkowitz. 1975. Beta-adrenergicreceptors:evidence fornegative cooperativity. Biochem.

Bio-phys.Res. Commun.64:1160-1168.

7.Sandvig, K., Olsnes, andA.Pihl. 1978.Interactionsbetween abrus lectins andsephadex particles

possessing

immobilizeddesialylated fetuin. Eur.J.Biochem. 88:307-313.

8.Weber,T. H.1969. Isolationandcharacterization ofa

lymphocyte

stimulating leukoagglutininfromredkidneybeans. J.Clin.Lab.Invest.

14(Suppl. III): 1-80.

9.Faguet, G.B.1977.Mechanismsof lymphocyte activation.Binding kinetics ofphytohemagglutininto human

lymphocytes.

J.Biol. Chem.

10. Hunter, W. M., and F. C. Greenwood. 1962. Preparation of

iodine-'3'labeled human growth hormone of high specific activity. Nature (Lond.). 194:495-496.

11. Faguet, G. B. 1979. Mechanisms oflymphocyteactivation. The roleof suppressor cells in the proliferative responses of chronic lymphatic leukemialymphocytes. J. Clin. Invest. 63:67-74.

12. Faguet, G. B. 1976. Mechanisms of lymphocyte activation. II. Negative cooperativity in the binding of phytohemagglutinin (PHA)to its lymphocyte receptors. In Leukocyte Membrane Determination Reg-ulating ImmuneReactivity. V. P.Eijsvoogle,D. Ross, and W.P. Zeig-lemaker,editors. Academic Press, Inc., New York. 65-72.

13.Lewis, R. V., M. F. Roberts,E. A.Dennis, and W. S. Allison. 1977. Photoactivated heterobifunctional cross-linking reagents which demonstrate theaggregationstateof phospholipase A2biochemistry. 16: 5650-5654.

14.DeMeyts,P. 1976.Cooperativeproperties of hormone receptors incell membranes. J.Supramol.Struc. 4:241-258.

15.Vladimirov, Y. A.,D. J.Roshchupkin,and E. E.Fesenko. 1970.

Photochemical reactions in amino acid residuesandinactivation of

en-zymesduring,uv irradiation:areview.Photochem. Photobiol. 11:227-246.

16.Munson,P.J.,andD.Rodbard. 1980.Aversatilecomputerized approachforcharacterizationofligand-bindingsystems. Anal. Biochem.

107:220-239.

17.Scatchard, G. 1949. The attraction of proteins for small molecules ofions. NY Acad. Sci. 51:660-672.

18. Fleet, G. W., J.R.Knowles, and R. R. Porter. 1972. Theantibody bindingsite. Labelling ofaspecific antibody against thephoto-precursor

ofanarylnitrene. Biochem.J. 128:499-508.

19.Scandella,C.J.,P.Devaux,and H. M. McConnell. 1972.Rapid

lateraldiffusion of phospholipids in rabbit sarcoplasmic reticulum. Proc. Natl. Acad. Sci. USA. 69:2056-2060.

20. Schlessinger, J.,D.E. Koppel, D. Axelrod,K.Jacobson,W. W.

Webb,and E.L.Elson. 1976.Lateraltransport of cellmembranes: