The determination of binding constants for binding between carbohydrate ligands and certain proteins

CarbohydrateResearch, 33 (1974) 377-382 0 ElsevlerScientic Pubbsbmg Company, Amsterdam - Prmted m Belgum

Note

The determination between MICHAELE

of binding

carbohydrate JOLLEYANDCORNELISP

ligands

constants

for

and certain

binding proteins

J GLWDEIKANS

NatronalInstrtuteof Arthritrs,Metabohsm, and Digestme Dtseases, NatIonalInstmtes of Health, Bethesda, Maryland 20014 (U S A ) (Ftecelved December Ilth, 1973, accepted December 19th, 1973)

For some time, we have been mvestlgatmg a group of seven lmmunoglobuhn A myeloma protems whch have bmdmg speclfic~ty for polysacchandes contammg sequences of P-D-( I -+hnked D-galactopyranosyl residues ‘-’ These globuhns preclpltate with larch arabmogalactanl, gum ghattl’, and mammahan-lung gala&an4 We have found that the fluorescence mtensltles of some of these proteins (J-539, M-601, X-24, and X-44) are altered upon bmdmg w&h haptens, and, by utrllzmg tbs property, we have determmed the bmdmg constants (K, values) of two of these proteins (J-539 and X-24) mth (1+6)-/?-D-gala&o-trlose and -tetraose3 We have now found that the method is such that K, values of lo* can readily be measured, ths IS lo* to lo3 times more sensltlve than the method of eqmhbrmm dlalysls6, the foremost alternatlve for the determmatlon of assoaatlon constants When proteins are n-radiated at 280 nm, they fluoresce wth an ennsslon maxunum at 330-350 nm This fluorescence (fl IS due mamly to tryptophany1 residues, and tyrosyl residues appear to contnbute very httle to the phenomenon7 Energy em&ted by tryptophan 1x1the excited state can be transferred to bound hgands by proteins m a radIationless process’ d the hgand contams aromatlc or mtro groups The quantitative change m fluorescence observed by us on bmdmg snnple carbohydrates [transparent to ultraviolet (u v ) radlatlon] to protems IS unusual, and these changes in mtenslty can be ratlonahzed as follows Any change m the mlcro-enwonment of fluorescmg tryptophanyl residues will produce a change m the overall fluorescent propertles of the protem, be It m the Intensity or fluorescence maxImum Such a change may be induced by changes m (a) conformation, either large or subtle, (b) pH, (c) temperature, (d) solvatxon, etc Such changes may also be brought about m certam protems when the protem bmds to small moIecuIar moletres Thus, the intensity

of

Iysozyme

IS mcreased

upon

bmdmg

ohgosacchandes

derwed

from

Yeast enolase etiblts a slmllar phenomenon on bmdmg magnesnnnl’ In addition to our four an&(l++D-galactopyranan IgA protems, certam myeloma nnmunoglobuhns havmg antlphosphorylchohne speci&ty have mdependently been found 10 Increase their fluorescence-mtenslty upon bmdmg phoschltmg

NOTE

378 c

sf I”

E ______--______ AFmax _________

s E x 0 8 E s C Sk 2

___

b Fig

Ltgand added

1 TypIca change m fluorescence mtenslty when a hgand IS added to a myeloma IgA

phorylcholine “*“_ Some proteins do not show a change m fluorescence on bmdmg wtth hgand, but, for those that do, the bmdmg constant may be determmed by the procedure reported here The method ISwell suited to carbohydrate-bmdmg proteins, as carbohydrates do not absorb m the u v or vrstble spectrum, and hence cannot Interfere wtth, or contnbute to, observed fluorescence Fig 1 shows a typlcal curve obtamed on addmg vanous proportions of hgand to a protein The maxunal fluorescence (Um,) ISobtained by addmg small proportions of solid hgand to the protem untd no further change occurs We assume that the relative increase m fluorescence mtenaty IS drrectly proportronal to the fraction of avatlable sites occupred by ligand T!XS assumption is vahd only If all sites have the same afhmty for hgand, and If the bmdmg of hgand at one site does not affect the bmdmg of hgand at another site Therefore, at any point b on the curve (see Ftg l), we can say that 5 (the fraction of available sites occupied by hgand) equals a/Mm, Knowing the protein concentration and the number of sttes available per molecule (which may be determmed by repeatmg the experiment at a Merent concentratron of protern’ 3), the concentratron (c) of free hgand may be determmed c = pgand] - V[total sites] A Scatchard plot (V/c uer.susij) of the data’ wrll yield a lme whose slope equals -K,_ As an example, the changes 111 fluorescence mtenslty due to the bmdmg of methyl or-r.-arabmopyranostde [methyl 5-nor(hydroxymethyl)-/3-n-galactopyranostde] to the pepsin fragment (Fab’) of J-539 antr-galactan IgA mmmnoglobuhn are shown m Fig 2 and Table I, and the corresponding Scatchard plot IS shown m Fig 3 The fact that a stratght hne IS obtained IS indicative of the correctness of our assumpttons The method has a hmitatron, m that, rf no change m fluorescence intensrty IS observed on bmdmg wtth hgand, It IS useless However, from the fact that four out of seven proteins m our anti-galactan series show the phenomenon, tt appears that these bmdmg-related fluorescence changes are quote common The advantages of thts method over the foremost alternative method for determinatron of assoaatron

NOTE

AF

379

c 0

--e--m--

4--e-

pax____‘_-‘__‘___---------

I I

I 2 Hapten concentration

I 3 (x10”

MI

FIN 2 The change m fluorescence mtenslty of J-539 lmmunoglobuhn A (Fab’) upon the addltlon of methyl a-L-arabmopyranoside

constants (namely, eqtuhbnum dialyses) are many (I) The method IS simple and quick, reqmrmg only one hour for measurement of a & value (2) It requires only small quantities of material, even for protein-hgand systems havmg low K, values (For a system with a K, as low as 102, the amount of protein and ligand needed are 1 nmole and 75 moles, respectively ) (3) It IS accurate and very reproducible (4) Radroactlve derrvatrves of the hgand need not be prepared, so that haptens remam unchanged (5) It IS versatile, there IS no restriction on the K, value measured (6) The K, values of whole antigen w&h Fab’ fragments of armbodies may be measured, and tlus has hitherto been nnposslble. (7) The protein may be of small molecular weight, and need not be nondralyzable E‘xeERlMENTAL

The fluorescence titrations were performed according to the general method used for quenchmg-t&rations s. A Perkm-Elmer MPF-3 fluorescence spectrophoto-

380

Fig 3 Scatchard plot of i;lc tarsus I for unmunoglobuhn pyranoslde (see text)

NOTE

J-539 (Fab’)

and methyl a-r-arabmo-

meter was used, It was fitted with a thermostated, sample-cell assembly, through whch water at 35 0” was circulated For the best results, the protem solutions [havmg an absorbance (A) of A280nm t0 051 were centnfuged to remove suspended dust particles The protem solution (1 5 ml) was added to each of two cells (10 x 10 x 45 mm), one bemg used for hgand addltlon (test cell), and the other as the reference cell A magnetic stnrmg-bar (6 x 1 x 1 mm) was placed m the test cell After thermal eqmhbration (several mmutes), the difference m the mtensltles of the fluorescence of the test and reference solutions at 340 nm (excltatlon at 295 run, band wxdth, 2 nm) was determmed several times, and these differences were averaged The mtenslty of fluorescence of a buffer blank was also determmed Llgand soIufion* was added to the test solution m small ahquots (3-45 ~1) by means of a syrmge mlcroburet (model number SB2, Micro-Metnc Instrument Co, Cleveland, 0h10) The rmxture was stirred m the ceti while hgand was added, and the cell was returned to the MPF-3 instrument, where It was ieft for several mmutes to eqmhbrate The mtenslties of fluorescence of the test and reference solutions were then determmed several times, *Two determmahons were performed for each compound For the prehmmary measurement, an arbitrary hgand concentration was employed, dependmg on the K, value expected-the higher the bmdmg constant, the less concentrated the hgand solution Havmg thus obtamed a prelumnary bmdmg-constant, the concentration of the hgand solution was so adjusted that the protem would be half-saturated after the addition of ~20 ,ul of &and soIutlon to the test cell In tbs way, a good &stnbuoon of pomts could be obtamed.

1503 1506 1509 1512 1515 1520 1530 1550 1580

I 002 1 004 1 006 I 008 I 010 1 013 1 020 1033 1 053

II 5 19 5 80 5 81 4 81 15 82 1 82 5 82 5 81 4 17 65 79 8 81 0 82 05 82 6 83 2 84 15 85 2 85 I

gorrected)

4 OS 62 14 8 45 90 96 10 5s 11 6 12 1

(corrected)

AF

0 5362 0 6123 0 6522 0 6957 0 7645 0 8406 0 8768

0 2935 0 4493

V

209 6 320 9 383 0 437 4 465 8 496 9 546 1 600 4 626 3

Bound ligand w4

1115 2,225 3,331 4,433 5531 7 349 10,951 18018 28 280

Total Iigmd hw

“F, = 73 6, concentratton of protem (C,,) = 714 nt& AF,,,,, = 13 8, concenttatlon of stock solution of llgand (CL) = 558 5 mM

3 G 9 12 15 20 30 SO 80

added ti0

Total fluoresceme

Correchorr

faclor

Total

vohe

TITRATION OF J-539 FAD’ WITH METHYL tl-L-ARABINOPYRANOSIDf

TABLE I

1 115 2 224 3 331 4 433 5 530 7 349 IO 951 I8 017 28 280

Free &and WO

263 46 201.96 160.96 138 15 11793 94 66 69 81 46 66 31.00

‘G/C

382

NO-IX

and the drfferences averaged as before Thrs procedure was repeated ten tmes, untrl a total volume of 0 15 ml of hgand solution had been added Small quantltres of sohd hgand were then added to the test solution until an increase m fluorescence no longer occurred, tlus point was taken as bemg the maximum increase in fluorescence mtenslty of the rmmunoglobuhn due to the additions of hgand, z e , when all available sites on the protein were occupred by hgand (see Fig 2). The increase m the intensity of fluorescence due to the addition of each ahquot of hgand was corrected for dilutron of the sample (see Table I) As may be seen from Table I, the concentration of free bgand at any trme IS essentrally identical to the amount of hgand added, due to the fact that, here, the bmdmg constant IS extremely low (see columns 9 and 10 m Table I) For higher bmdmg-constants, the concentratron of free hgand differs substantrally from the total concentration of hgand The values for 7/c were plotted uer.su.sC (see Fig 3), and association constants were determmed from the slopes of these Scatchard plots by the method of least squares (K, =4 03( f 0 03) x 102) REFERENCES 1 M POTI-ER, E B MUSHINSKI, AND C P J GLAUDEhwNS, J Immunol, 108 (1972) 295-300 2 M POITER AND C P J GLAmEhfANS, Methods Enzymol, 28 (1972) 388-395 3 M E JOLLEY, S RUDIKOFF, M POTTER, AND C P J GLAIJDEMANS, Blochenzlstry, 12 (1973) 3039-3044 4 C P J GLA~EMANS, M E JOLLEY, AND M PO-R, Carbohyd Res, 30 (1973) 409413 5 S RUDIKO~,E B MUSHINSKI,M POI-IZR,~ P-J GLAumzhf_ws,~~~ M E JOLLEY, J Exp Med, 138 (1973) 1095-1106 6 A H SEHON, Methods Immunol Immunochem , 3 (1971) 383-394 7 F W J TEALE, Blochem J, 76 (1960) 381-388 8 3 E MCGIJIGAN, Metho& Immrmol Immunochem , 3 (1971) 395-404 9 S S LEHRER AND G D FASMAN, Brochem BIophy>Res Commun, 23 (1966) 133-138 10 J M B-R AND B WEFJER,J B1o1 Chem , 241 (1966) 2550-2557 11 R POLLET AND H ,248 (1973) 5443-5447 12 R PoiLm, personal commumcatIon 13 C J HAr_Fhf,m AND T NISHIDA,Blochemrstry, il (1972) 3493-3498