94 Bw~kmmu el B .~pk~ua .4~m. 1(/78 119911 q4- I1"~1 ~'4gl pi,c< ,,r 5u!cn,.,: Purqi.hcrx B %, 11167-4f~t8 'q l : '$0t 511

.ID,')',/A ~!! t,~d~3~'ill~tl ~14l;

Evidence for an acid-induced molten-globule stale in interleukin-2: a fluorescence and circular dichroism study

David D r y d e n ~ * and M a l c o l m P. We i r -~

' D~pwolwot ,~; B:,,~he.m~lrl. t mr t',m~ r 4 `'~ ~, .t~lh" tip,,, l~m', (~l?.llerllt~ ' (~ui~ ~. I I Buddmz. Ve,~t~:k. upton TItle ( t ' h' ) and

D~/,o ,:mi'm ./ t,cncsr~ ,, Gla ~o Grotql Real.art h I.Id. Hrcun/.rd. ~lhddh,~. t ¢ I g t

I lr'.ccci'~ed 31 Oct~bct I ttgOI I I~l~,ri~t~d manuscript ret:¢i'.ed 16 Januat,. 1'9911

Ke', ~old',: ]ntcrllsukHi-2: ~'[OllOl globule: Flutpl'e~ccncc; ( ' l rcular dichr~i,,rll

The effect of low pH oa the secondao and tertiao structure o[ the monomedc single-disulphide protein interleukin-2 ILL2} ̀ 'las monitored by fluorescence and circular dichroism saeclroscopy. Be~'een pH 4 and pH 2 there is a gradual hmsening of the tertiao structure as revealed b v changes in ~rosine and tOptophan fluorescence emission, tDptophan flnore~enee ani~trnp), acce~sibilily to the fluorescence quencher acolamide and aromalic circular dichroism. The oleraU melecolar size and secondav, structure content are not significantly changed b) acidilication. These data are consislent ~ilh a 'molten globule' state for IL2 at low pH, in which the h,,drophobic core/secondao struclure is largely intact hut the tertiaD structure is flexible. Similar effects to low pH are seen at sub-denaturing concentrations o| ~-uanidine b)drocbloride. Anal.~sis of fluorescence liletimes and derivative emission spectra of the single tryptophan, Trp-121. shows the existence of t~'o distinct orientations for this side-chain, one of which is affected by a quenching group (the e|fect of which diminishes upon acidificalion) and another ~hieh is essentially unquenched. The identily of the quenching group is unclear but may well be Cys-125. The formation of the molten globule titrates with a pK, of about 2.3: this is unusuall~ low for the acidic groups in proteins and indicates a perturbed pK, of a residue involved in a structurall~ important interaction such as a salt bridge. Candidate residues are Glu-15 or Asp-20, dose to His-16 on the N.terminal helix of IL-2.

lntrt~It~elion

Imerleukin 2 ILL2) is a small monomeric protein of 133 areino acids (molecular ',,,eight 15 500). secreted by T-I}mphoc}tc:,. It promotes proliferation of :'.cli;'atcd "l-helper cells II1 and is al,,c> in,,olved in control of B-cells [2]. T-o, toloxic cells [3]. natural killer cells and I),mphokine actr, ated killer cell> i4.5], q-he high affinity lcceplor for IL2 on human T cells comprises a heavy chain [6] of 70- 75 kDa and a light chain of 55 kDa 17].

• m _ * Pre,,ent addre>,s l)eparlnlen~ of Mol;cu]ar Biolog}. King's fluild-

rag. Mx, field Rt~ad. Edinburgh t£ft9 3.1R t K 4hhrc,.i,~lion~: (;t,(q. guanidlnium ~:hlondc IL2. inlcrlcukln-2: NA~-rA, ,",'-.ICt'l'd t)rosinan]id,.': NAWA. ~-acep, I l~plophanamtde: ANY, ;mihne napth,d~ulph~rlic ;~cid.

(orrc,pondenuc: M.P ~L, em Deparllnem ~t Genetics,. (;lax~ Group Re,,carch t a d . (ireenford R~ad. Grcenf~rd Middlesex, t B 6 0 l t l , t ' K

The amino acid sequence and a low resolution cryslal structure [8] sho,.vs the presence of a single buried tr)ptophan at posilion 121, possibly quenched by Cys- 125 [9] and a disulphide bond between the Iwo remain- mg c3stemes at positions 58 and 105. This disulphide is csscnt;al for maintaining the four helix bundle tertiary structure [10.8]. Mar.y mutations have been studied in an attempt to dcline the residues which bind to the IL2 receptor (summarised in Ref. I1). only some of which have been characterised in terms of structure as well &s activity [9]; it appears that a large proportion of IL2 surhce area is involved in receptor binding.

At 1o~ pH, Arakawa and Kenney [101, found that terliarv struclure was lost while ,~condary structure was maintained in a mutant IL2. This sort of structural change has been described in, for example, a-laclalbu- rain [12] and fl-lactamase [131 and characterised as a 'molten globule" state with a fairly intact hydrophobic core and flexible tertiary structure [14,15]. We have investigated the conformational changes induced by pH in v, ild DO e IL2 and Iwu mutants a,~d analysed the

f]uore,,cence t i t r i l l i P n ~f t i l e burwo ' /~tt~.~lh,lri rc,lduc i t app,.:ar,, t ha t lo~ pH re'.,uh,, In h+,, ,~f q u c n L ' h i n g ~md

increased r~taii~n of "~'~ ant] ],.r rc',iJu~;, ~,t [1 " ,rod Inulafl~s. ~hiiq the ~v.erall ,c,,~md,~:~ q~uch, e i rmHu- rained.

95

Materials and Mel l~ i s

Wild type IL2 and site directed mutant:, ~,,.erc pre- pared as described previousl~ [9] and stored frozen ill 60% aeetonitrile. 0.1q trifluoroaeetic acid (pH 3.6) until used. Samples were dialvsed exhaustively against 50 mM glycine hydrochlnride. ~-,dium citrate or ~.u..,dium acetate buffers at pH values between 2 and 5 depending on the e~periment. Buffers ,.,,ere bought from BDIt I.td. P,.~le. Dorset. No dependence on the buffer ~,, as found Protein concentration was determined from the absorp- lion at 280 nm assuming an extinction coeffi~:icnt of 9.5.10 ~ M ~cm ~ [11]. pH titration~ ',,,'ere performed by adding small aliquots of 1 M HCI or NaOH with a microliter s)ringe to a stirred solution of protein in a 10- Ill mm cuvette.

Steady state fluorescence and fluoreseencc polarisa- tion measurements were made on a Perkin Elmer MPF3 fluofimeler and fluorescence deri',ative spectra on a Perkin Elmer LS5 fiunrimeter. "the emmsion of protein solutions with an ab~rbance no greater than 0.2 at the excitation wavelength was corrected for backgraund scatter, dilution and. in the quenching experiments. absorbance by acrylamidc [16]. T~rosine fluorescence emission was measured by excitation at both 280 nm and 290 rim. normalisation *ff the 290 nm spectrum to th,: 280 nm based on the emission ratio a: 370 nm. and subtraction of the normali.sed 290 nm spectrum from the 280 nm spectrum. The final tyrosine emission ~,pcc- trum was multiplied by 1.36 to correct tor r¢,i,.%al tyrosinc emission at 290 nm excitation.

Time resolved fluorescence mea~,urements ,sere per- formed with an Edinburgh Instruments 199 T fluorimc- ter fitted with a t t : flast.!amp ~dfich produced a 1.5 ns pulse at 50 kHz. Excitation aas at 295 nm and emission at 340 nm, the bandwidth at both wavelengths was 20 nm. The flash!amp profile was determined ~ith a scattering solution of colloidal silica, and global ilera- tire deconvolution of up to 12 protein fluorescence decay curves determined on a Lommon set of lifetimes for all the decays with ~,ariable intensities of each deca> component 117,18]. Three expon,,ntial decays ,~ere the minimum number of decay components which ga~e a gotM fit to all the data sets.

Circular dichroism (CD) measurements ~erc per- formed on a Jobin Y~on Dichmgtaphe MkV. CD inten- sity is report~M in terms of mean residue ~eight elliptic. icy a~suming a mean residue ~eight of 116.5,

All the spectrometers, had tbermoslated sample c~m-

t GO

i f ~..- ..

\~"

q~iIcd Jl 2q~ lIIll Inlcn,~l~ umt, art j~b~tral~

partments, and all n]ca.~uremcnb ~crc performed at 25°C

Sample,, uved ft~r ( iu ( I unfolding ~ere euuilihrated in each Gu(I c~,ncentrztti~,n c' , ,¢rnlgllt lit r,,Kiin tempera- lure.

B~ul ls

Tilra.tp.ul of the fhlL,f',.'~,cence c m l s q o n pred~nlinand) due to Trp-I ?1 ,ho~, a large enharcernem m mtens~t> ai 324 nm heh~',~ p]'i 3 (Fig. 1t. There ~as no further change in emi~,~ion aho',e pH 5. Ahhough the full titr;:tion cur~e v, as no z ~blained hel~m, pH 1.5. the pK, of the quenching of feel i~ approx. 2.3. 1-he fractional int,:n~itie~, of t~o of the three fluorescence lifetime cllmptnlem~ r~L;uircd {~ hi the f]u~ re~,~.'ence de~at, s ~f IL2 Lhan~.c dramaltealI', as d~c pH r. h~ered: the longest lifettme. T~ 3.57 +_0.05 n~ (meant + S.D.). in- creases in intensm,' v, hile Ihe sh~riesL L 031 + 0.01 ns. typical of a quenched fluorophore, decrea~s to a n~l i - gible le~d iF ig 2). The intermediate lffet me. T: 1.35 _+ 0.12 ns..,hov, s a ~,light increase in intensib at low pH. A~, ~he pH ",,,-as decreased, a red shift of the emission maximum from 324 mn to 32" nm and an increase in iq|en,,it,, around 350 nm c~n be seen in the first deri',a- m.e ¢mi~,sion ~.pcctra Ir, ot ~,ho~,~n). i-be oxidation of the nearb', C',s-125 m C~sletc acid b 5 th~ acidic pH [19] v, hich has a pK: of 2 [2(I] ~ s ruled out a,, a po~,,,ible

¢

:Z, ~ -

'i e "

!~_ ~ . ?~ ".''. 35 ~C t 5 51

G~

~. dchncd a. ,~ T va [ ~.hc,,.: r = I h, ~ a : prcc~r.,nent~al f~1.r ~nd T the flu,¢c'-,.cn~.c de~.a ~, ;tn~¢

source of thi,, p t l dcp,:ndence h'~ 'he ah,cncc of an; change m i,,oclc,:tric ,ouu,,,,ing gel,, |unpubl ished re,,uh,I.

11.2 ha,, three Is r , ,mes . rc,~du:,, 31.45 and 1t)7. n,me ,H '.quch arc ,.h-,c Io lrp-121 in the terliarx ~,lruclure IX} l',lo'.ff~c f~ttt'~rc',,cc,'lce carl h , mea,,ured indepen- deiill.'., oJ I rp f luore~encc I"., "~uhtractkm due t~. ddfcr- cnl excitation maxima of T~r and " l rp l h e t',n'.,,,mc fluorc,cencc ,pcctrum of 11.2 ha', k , . . a~ 3!~ tam ~ptcal ~1 t~rostnv ~h~ch sht)~, hllle ~ariat~,n m k ...... 1211 l h e cffc~.t ,,f pt] on f luorc~cnce intcnsfl; ,,f 11,2 and 3,-acet;l t'm~,.mam~de INAYA) t,, q~ox~n m [able I N A~& (luc, rc,,cer~ce i~ onl', marglnall ' , :if re, ted h~

IM~ I I I

11~Id¢}

l ' ,r I

-- / / "°I '/

-6O

+;40

i ~ r

2~G 2}~ 280 290 300 310 320

eRVELEI~GTI4 (~J

I:~ 3 \lear/rc,~,hJc v.c~,hl cthpu~.t~', a~ a [unct~.,n ~f pll ,M 2~1 nm. Ihc ~.~! ~aiuc- ,~r¢ 41, luppct ~ur~ct. ~I} !m~ddlc cut, el and 2 16

ackhfi~.,~t,~n ~hcrcas IL2 l~rosine fluore~ence in- creases markcdls het~een pH 4.6 an..~ 2.2. consislen! with the loss of Trp-]21 quenching ~een in this pH range

The near uhra~iokq circular dichroism ,~pectra (Fig. 3) mc;t~uring Ihe a~}mmelr} of tertiary Mruclure around "lrp-]21. ~how a decrease in intensity as the pH is h~e red Different.e ( I ) ~peclra generated frtlm Fig. 3 ~ho~ a charat|eri~,tic tr~,ptophan signal which decreases

a~ ihc p l l Jails 122] Prev,mq~ publhhed near ultra- ~iolu't CD spcutra of a Cv,,-125 ~ A l a mutant al.~ sho',~ed ~ lu,s ol cll ipti¢il) a~ the pH ho_'ame more acidlc, hut lhe signal-lo-nnive ratio ,xas ve~. poor per- hap~ due I~," a Io~ pro,ein c~,nccnlration 110]. The grad- ual decrease m fluorescence anisotrop} of Trp-121 as lh¢ pH is reduced f f ig . 41 is clearly different from the pH titration of fluore',.cence inlensil~, bul, like Ibe near ultra'.iole! CD, ix r cp roen t ing some gradual change in the tr , .pt~phan en,.ironmenl. [ h e fluorescence ani- ,,otrop~. tleca} lime did not vaD' m any syslematic way v.lth pH and v, as aboul 7 ± 2ns which '~'ould be t y p i ~ l for a spherical prolcin of the molecular weight of IL2

t2~i. At pH 4.6 the t ryptophan i, completel~ buried inside

lilt" protein since H is macces,.~ble to quenching b', iodale ~n,~ [9 I, hov, c,.cr. Ihc p~Iar fluorc,,tence quencher

0 ~'),r /

/ . i

0 ~':t /

O 19,,r / ~ + "

i) lg "I,

0 '3

o se+ / i / /

0 t5+/

llg 4 '~tC+Id~+ qat t + l~tlorc-~-!~tt! +JlIl~+t/+~ ~, a" ~I iun~.llHn +d pH P~,tltatl++~ mas al 295 rm afld t+U++l,+f~ +I I~4 nnl

acolamide can penelra~e the interior of proteins ll6J and at thi~ pH. ac~iamide quenching of Trp-121 ga~e a linear Stern Volmer plot with a quenching constant (,|

0.38 ± 0.02 M :. At pH 22 the quenchieg constam increased to 0+49 ± 0.02 M i indicating a greater degree of sol'cent acce,,,ibilit) al hw, pH. |to~.e~.er, th~.~c quenching constants, at N~th pit v;ducs, are Iota corn-

+.irtmd!', in,J~c,,,~h'~ , " ,~n a l p l l .+ 2 1i+ +" +,Ilov+, ~ t+ c id ] l]ll,+IC~t+<iI~# p,J t .J l l lCtCt- .+- J

IUlit.llw+n ,+l (}u(I ~.~+Ht.cntFAti[in J t pli 4 S ~ lhc cn + h -

,~on intern, it', at 324 nm and an iq rop) al 324 nm or 350 nln (n(+I ++l/+~+++ll) -,h +'~ ,1 ~+.Mual d c q u c n c h m + a n d m- crt++~++c in t r) ,ptophan (lc~b~i~t~ up u~ app++ex 4 h M ( ; + ( ] reprc%'nlm+ a g+ 'adua l h ~ + + e n n l g o f l h c +,IIUctUIC,

~dnd+ r, h + l h , m c d h +̀ ~ompldc u n f i f l d l n ~ |,+ +J r ,md,+n+

cod ,Jt~++¢ 4 5 M a~ ++homn h+ the '+harp change m mlen~sp, al )'+4 nm ;+in,:] the lar+~e red shift m the cmp+~+on ma~tmum m&catlxe of a full~ unlolded pr,+-

~e,n 1231- "l',~o mutant IL2 prolein++ in mhich Phe42 v,a~

changed to a]anmc and tr}ptophan mere a l ~ anal++,+cJ ~.~,'e~r e l a] ~9] h a ~ < rcp,++'ted lhal the tcfhar~+ qlw+lurv of the i+"lulant++ ;~ the ~a,me a- lhe , sM t~.pc 11+2 but lhal the alamne mtJla+++'t is n+{ +ignIflcanll) Iomer dCIV, II~.

,mpbin@ lhat Phc-a2 i'. a hinding site r~idue, lhc+r l"~'h,J,.i,+ur in dupli~.ate cq'~nment,+ uf tho~ d~crlh,cd abo+.c mac, ~er} ,+nnlar to the: '~ild t},pe, except +thal the

higher q u a n t u m ++old ol Trp-42 J9] almost oh~un.~ lh< titration of the cmi,~aon ,+f lrp+121 l-here ma+, a blue ++hfl't in t h e cm+~+~a+n m a x i m u m tff th+: 1:42'+/ fliUlatll ~I~,IIi b+~t) IiIIl io ~ nl~i ~t~ lh~ pll m~s Io+~'e~ed due m

dequenching of the hu~i~xl Trp-121 Three flu(~tevcence lifetime+ q+ll h t tM the fluorescence deca), of tht- m,atanth, though m I)'PJ F42W mutant th,:y mere differ- enl from the mild type and +.'~mnc,t no~ he a+~%~ial~d +,ith a part+::ular t r}ptophan. ~t pH 4.,5. the 1ifettm~ ~nd fr+ctional intcn,it+e, of F+t2A ~ere 0.29 z 0.01 n~ (0.72~. l.L~ : O.lt~ m, ((i.22~ and ~.03 2 Off5 m, (0.05) and o,~ F42V+ were 0 29 : I~01 i0.33). 2.75 = o.12 ((I 3~). 5 9V~ , 0 , (~ . ( 0 . 2 c ; ) Th' . : a r o m a t ~ ( I ) . ,p~.~-trum o f F 4 2 ~ . .

<F~g 6+ , h u m ' , unl+~ a , h g h I d c~ r ca ,~ : m c}hpilCtI} m & -

'7 Ice+

i

3 i

Bill

!Q

{ ,++ ,+ I +P

+ i + ~ . J "

\ . +~ :+ / ~ ,z- +. • "L< '.

C <+,t ++0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

++ ( . If+ +' ~,:ar+++I+,+,,+ +++ tlu~,It'+-~+I+~.c ~+,+Jr+tm+!~r- a~ ,+ fur.~t~.+n +, ~ u . d~m,~::.~+an+ ~cr~.cnl+,oJ++n d~ pl I 4 +~¢ ~++j c+ .Vah,,~ M 2 <+¢ nrn ta} .'~m:++t~<~+ at

9~

8o! 7

-80

-120 [

-ISD !

!

J -2°~s~ 2-T~-~7o 2eo 2~o 3oo 31o 3~o

VIaVELENGTH if, m) Fig. ,5 Mean re,.iduc '.xeight ellipticity of the Phe-42 ~ Ala (upper curve) and Phe-42 ~ TTp (lowe[ curve) mutant-.. The pH was 421 and

4.33. re~pectiveb..

caring that it has essentially the same structure as the wild type. The ellipticity of F42W is substantially lower due to a large negative contribution from Yrp-42 re- vealed in the difference spectrum between F42W and the wild type. Lowering the pH frmn the value of 4.6 in Fig 6 showed a drop in ellipticity in a manner analo- gous to the change in the wild type.

Discussion

The pH titration of the fluorescence intensity, polarisation and aromatic circular dichroism of IL2 reveals a Iocalised titration of a charged group which results directly or indirectly in quenching of the fluores- cence of Trp-121 at high pH. and a more extensive gradual change in the protein structure.

The titration of T~ and T~ matching the titration of the fluorescence emission and the virtual independence of T~ on pH implies that there are at least two confor- mations of Trp-121, one that is affected by I~tration of the quencher, the other, represented by T~, is protected by distance, orientation or other side chain residues from the quencher. The conformation represented by 7~ at low pH and T I at high pH should have a different

emission spectrum from the 7~, component. This is con- si~tent with the observed red shift in emission which is often a~sociated with a more polar tryptophan environ- ment [24]. The multiple lifetimes of tryptophan have been attributed to different rotameric conformations of the indole side chain [25] combined with effects of a nearby quenching group [26]. The two conformations observed will be in equilibrium with each other perhaps via a ring flipping mechanism similar to that observed for phenylanine and tyrosine, although the large size of the indole ~ide chain would make this a rare event [27]. A simple displacement of Trp-121 could also account for the postulated conformations but a displacement of the quenching group alone wo,ld not lead ~o the effec~.s observed on the polarisation and circular dichroism. The identity of the quenching group is unknown though Cys-125 one turn above Trp-121 on helix F, and prob- ably aiso buried or partly buried, has been suggested [9]. There do not appear to be any acidic groups close to Trp-12': it thus appears that a global conformational change dependent on protonation of acidic groups re- mote from "frp-121, indirectly results in dequenching due to movement of a quenching group, probably Cys- 125.

Further evidence for a global conformational change on acidification comes from tyrosine measurements. Tyrosine is strongly quenched by hydrogen bonding of the phenolic hydroxyl to amide groups 121.28] which could be supplied by the protein backbone or side chains such as Asn-30, next to Tyr-31, or Gin-106 next to Tyr-107. An overall loosening in the structure could weaken such bonds and thus relieve quenching, or re- duce energy transfer to Trp-121.

Greater freedom of movement of Trp-121 at lower pH would increase the rate of interconversion of the tryptophan conformations. This is revealed in the de- pendance of fluorescence anisotropy and aromatic cir- cular dichroism as a function of pH shown in Figs. 3 and 4. T[-~ese show that the rigidity and asymmetry of the environments of aromatic residues decreases almost linearl 3 with decreasing pH.

There is no apparent gross change in size of the protein, due to partial unfolding, since the anisotropy decay time remains typical of a 15000 molecular weight protein. Neither is there a significant change in solvent or acrylamide accessibility to Trp-121. The gradual loo- sening of the protein structure as the pH is dropped. may be similar to the acid expansion observed in some other proteins [15]. This expanded state has been found to be equivalent to metastable folding intermediates caused by low denaturatant concentrations.

Baum [121 concluded that the acid expanded state or u-lactalbumin maintained secondary structure around a hydrophobic core with slowly fluctuating tertiary struc- ture. Other workers showed this acid expansion causing several effects on fluorescence such as a red shift,

99

minimal increases in t ryptophan exposure to quenchers. loss of aromatic circular dichroism but not of backbone circular dichroism, no significant change in radius of the protein and the exposure of hydrophobic patches causing aggregation and increased binding of the fluo- rophore ANS. Most of these effects have been observed in this study of IL2, and Arakawa and Kenney [10] have reported the increase in ANS fluorescence at low pH. The p H effect is also seen by N M R observations of Hi', residues in IL2 [29]; His-55 is buried at pH 4.1 but becomes exposed to solvent below pH 2.9. The same group also observed that His-16 was perturbed by ad- jacent negatively charged group(s). His-16 is part of helix A and is therefore adjacent both to Glu-15 and Asp-20. one turn up the helix [81; it is probable that pK, values of at least one of these residues will be signifi- cantly lowered by proximity to the protonated im- idazole. Groups such as these with anomalously low pK~ values could be involved in the acid-induced un- folding seen below pH 4.

To conclude, these results imply that at pH less than 4, IL2 possess an acid expanded state similar to the "molten globule" state in a- lactalbumin [12]. A possible hydrophobic core for this molten globule could be formed from the three aromat ic groups, Phe- l lT. Trp- 121 and Phe-124 on helix F and Phe-44 on helix B which forms the interface between the two helixes [9,8 I. The disulphide link between Cys-58 and Cys-125 has also been shown to be essential for maintenance of the protein structure [101. The aromatic groups on helix F. spaced one helix turn apart , could stack face to edge to face as found in some high resolution structures 1301. An increase in slow fluctuations in this structure would affect the anisotropv and circular dichroism in the manner observed without allowing rapid access of solvent or fluorescence quenchers. In F42W a negative C D signal for the new Trp-42 is still observed at low pH suggesting that it is still in an environment with :-igr~'fi- cant tertiary structure; the solvent exposed side of helix B around position 42 would appear thus to be a stable structure even at p H 2.

Alternative packing arrangements of the aromatic residues in the hydrophobic core could give r i ~ to two or more conformations of Trp-121. one of wh;ch is affected by a specific quenching group coupled to an acid-induced expznsion with a pK, of 2.3. It is not possible to identify ",.r,y amino acid with this p K . in the vicinity of Trp-121; however, Cys-125 prcvk~asly sug- gested as the quencher 191 since it is one turn up helix F from Trp-121, could interact with the N H group of the indole ring, a crucial position with respect, to the fluo- rescence emission of t ryptophan [261. It is likely that the groupts) t i lrating with p K , of 2,3 are invoked in the overall integrity of the tertiary structure and their proto- nation leads to a global loosening of the tertiary fold. Asp and Glu usually have pK, of about 4,5 but proxim-

it3 to a positi',el3 charged residue could markedly re- duc t their pK, : candidate residues arc Glu-15 and Asp-20. both close to His-t6 which itself is known to interact wi th a negatixely charged group 1291. Posilive identificatilm of the group'.: involved in acid-induced expansion would require hi~hcr resolution analysis.

Acknowledgements

We wish to thank Dr, A. Slade for helpful discus- sions ano the SERC and EEC for their support, Dr, RS. Baines for protein purification and Prof. R, Pain for his constant advice and support.

References

1 Robb. R.J. Ma~.cr, P.C and Garlic'k, R. (1985) J. Immunol, Mtthod~ 81, 31-42.

2 Pike. BL.. Raubio.chelL A and No,-~h G I V 119841 Pn~. Na i l Acad St i [ S A 81 791g-7921.

3 Ezard. R.. Corlhe~). P.. Nabholz. M.. Lov, enthal. J.~ Za¢ch, P.. Plaetlnck+ G and MacDonald. H 11985) J. Immunot. 134. 1644- 1652.

4 Hinney, C.S, Kutiba~,~}sht. K. Kern, D.E and Gilhs, S. 119811 Nalur¢ d.on&ml 201. ~35-338

5 Mule. J.J. Shu, S. and Rosenlxrg. SA 119~51 J Immunol 135, 646-652

6 Hatak~yama. M. l~udo. M.. M01alrlolo, S.. Kom~, T. Doi. T. Miyata. 1., Mi}asaka. M. and laniguchi. T (19891 Science 244. 551-556

7 Leonard. s.~.J.. Depper. J.M., Kamehisa. M Kronke. M., Peffer. N ,I., S~.etlik. PB, Sullivan, M and Gleen. W.C. 11985t Science 2~0. 633-639.

8 Brar, dbuher. B.J.. Ik'~me. T.. Kcnney. W (.. and McKa 5. D.B, H9871Sciet~c¢ 238. 1707 1709.

9 Weir. M.P.. Chaphn, MA.. Wallace, D.M.. Dykes, C.W. and Itobdcn, A.N. 119881 Btochernisl~ 27, 6883-6892.

tl) ~raka,xa. 1 and Kcnne 3, W.C, t1988j Int. I Pepl Prot Res. 31. 468 ,I.73

I1 Cohen. F L Ko:.,cm P.A,. Ktmtz. ID_ Ep~.tem. L.B, Ciardelli, T L and Smith. K.A i1986i Science 234. 349 352.

12 Baum. J., l~['~orl, (..M. E~ans. PA and Hurtle}. C 11989) Biocheml:.tr) 28, 7-13.

13 Gov.. Y. and Ftnk. A.L t198tll Bi~hemi',t~ 28. 94'~ 952 lZt Dolgll',h. DA. Gllmanshin. RJ . Brazhinkov. E.V.. Bychkova.

V E, Scm*mtn~v.. G V Venypuninov. S,Y. and Pntsyn. OB. (19gl) FEB:~ |.ett 136. 311 315

15 Ku,.:ajima. K i19891 Proteins 6, 8%-1u3. 16 Eftink. M.R :nd (ihiron, CA. !l~.~Sl} Anal. 8iochem I1,1, 199-

227 17 l]irch. D,J.S. and Imhof, R.E. (1985~ Anal. Instrum. 1,4, 293-329 18 Knutson, J.R.. Beechem, J. rand Brand, L 119831 Chem. Phys. Lett.

102, 501-507. 19 Crc~ghton. T.E tlt1831 Protein:, Struclure and M lecular Princi-

ples. Freeman. Ne',,. York. 20 Andrev,~. S. ~nd ~hfmdt, C.L.A, 11927i J Biol Chem. 73. 651-

654. 21 Lako'ai,cz, J.R 119831 Principle~ of Fluore~-er, ce Sp~,clro,a:opy.

P!¢num Prc~,s, Ne~ York. 22 Stri~.k/ord. E.H. 119741CRC Crit Re', Bioehem. 2. 113-175. 23 Kronman. MJ and ttt~lme~. LG, (19711 Photochem. PhotobJoL

la. 113 12,4.

2++ +',+.+,JB ~.++ ,mi Ra'+ncr D " . I +ig+I)) J ,~.m ( h + m "+,,, 1+~

_~+ '~++,l+J¢r {. +t+~}}(.} RU ~. i++,+ph~ l++ I ?+:

R J and kd, : lh+~h | '1. ',:d, b. pp aA.2 4,~; ".l,~r~:cl l)..'kkcr N..-~

2'~ Mlx,mu I I Su,'ukl, l~+l. Ak~s!,:, .+';. l :ur~' .a.M. [~t~!D. I . lhr~m~.ml.J ~ . N~gJ~htma. N. Ni~hlmura. Y and Arat.J.Y. Ilgq~h Prot ln~2 X. 335 336