VoLUl ts 119 , Nu l teen 52Dsce l reeR 31 , 1997@ Copyright 1997 by the

Ame ric a n C he m ical Soc ie ty

In t roduct ion

This paper evaluates the general i ty with which protein chargeladders can be formed by acylat ing the amino groups onprote ins, and reso lved by cap i l la ry e lect rophores is (CE) . Aprote in charge ladder is a set o f der ivat ives o f a prote in that isseparable by CE into a set of dist inct peaks-the "rungs" of theladder-that dif fer from one another in their values of electro-phoretic mobil i ty. l Each nrng of the charge ladder is composedof a mix ture o f d i f ferent reg io isomer ic der ivat iv 'es o f nat iveprotein that have the same number of modif icat ions and usuall ,v-the same value of charge.l Protein charge ladders are usefur indeternining the values of charge of proteins, l in obtaining values

- 'Co r re :pond ins

au rho r . Te lephone number : (6 t7 i -195-9 -110 .f

t * numbcr : (6 l 7 ) - -195-9S57 . E -ma i l add ress : gw h i res ides@gmL* group h-n rrJ.edu' ,Abrtract

publrsherJ in Adt ' t tnc.e ACS Absrrrrc. / . r , Dccenrber 15, 1997.

tndd]5; : . .1, : i ] : ; ; . '^ 'o le ora ch'*-ee hdder ' sce: creighton' r E ' i \ i t f t t r ' '

, {

IOMALJ OFTHE

A]\4ERINCFIEMICALSOCIETY

Formation of Protein Charge Ladders by Acylation of Amino

Groups on Proteins

Ian J.Sgttol, Janelle R. Anderson, Jinming Gao, Robert G. Chapman,Lyle Isaacs, and George I!I. lVhitesides*

Contribution from the Department of Chemistn' and Chemical Biologt', Hart,ard Uniuersitl'.

l2 Oxford Street, Cambridge, lv[assachusetts 02]38

Receiued Jul; ' 11, 199F

Abstract: The values of charge and electrophoretic mobil i ty of a protein are changed upon acylat ion of i ts cr- and

Lys e-NH3+ groups. Part ial acylat ion of the amino groups of a protein results in a set of derivatives that is often

resolved by capi l lary electrophoresis into a set of dist inct peaks-the "rungs" of a protein charge ladder-that dif fer

incremental ly in the number of residues modif ied. Proteins that have values of lvlW < 50 kD usually form resolved

charge ladders when al lowed to react with acetic anhydride, while proteins that have values of lvlW > 50 kD form

broad unresolved peaks. Resolved charge ladders of proteins that have values of MW > 50 kD may be formed

using acylat ing agents that inroduce several charges upon acylat ion of each oI their Lys e-NH:+ groups. As an

example, L-lactate dehydrogenase (MW : l4l kD) does not form a resolved charge ladder when modif ied rvith

acetic anhydride. When it is acylated with either 1,2,4-benzenetr icarboxyl ic anhydride,3, or i ,2,4,5-benzenetetra-

carboxyl ic dianhydride, 4, however, i t forms charge ladders in which each of the f irst several pairs of adjacent rungs

are separated by approximately 3 or 4 units of charge, respectively. The procedures described in this paper were

used to form resolved charge ladders from 25 proteins dif fering in pI and in NIW.

\J\

of pK" of the N-terminal a-NHl+ group of proteins,a in

evaluating the inf luence of charge on the free energy of bindin-e

of dif ferently charged proteins to a charged l igand.s and in

est imat ing the molecu lar wei -ghts o f pro te ins under non-

denatur ing condi t ions.6

It is convenient to modify the charge of proteins by acylat ing

the e -NH: groups of Lys, s ince Lys is abundant in prote ins ( i ts

f requency of occurrence is 5 .97c, based on 102[ prote ins o f

knorl 'n sequence), reactive, and mostly distr ibuted on the surfa,--e

of rv 'a ter -so lub le prote ins.T lv lod i f ica t ion o f the guanid ino

residues of Arg using a-diketoness'e or amidation of the carbor.v-l

residues of Glu and Asp using a dehydrating agent and aminesl ')

are also useful rvays of modify' ing the surface of a protein. The

ro le o f pos i t ive ly charged e -NH; ' groups of Lys in s tab i l iz ing

the s t ructure o f pro te ins is not *e l l unders tood, a l though Ly 's

gr r )ups l re s l ight ly prefened at the C- terminus of an a-he l i . r ,7

O 1997 Amer ican Chemica l Soc ie t l '50002-7S63t97t023{9- { CCC: S l { 00

12702 J . A r t t . Chem. . ! oc . , l ' o l . l t 9 , N 'o . 52 . l 9g7

pc rhaps t o ba lancc he l i x d i po les . r r An impo r tan t f unc t i on o fL- ls in many prore ins may 'be ro make rhem so lub le . NIod i f ica-t i on o f e -NHr ' - q roups o i Lys on t he su r face o f t he p ro te in ,hor 'e 'er . should a f fc -c t the s tab i l i ry o f the prote in s i_en i f icant lylc 'ss than modi f icar ion o f res idues bur ied ins ide the prote in . wehlve recent ly demonst ra ted th l t c 'ach rung of the char_qe ladderdc. r ived f rom bov ine carbonic anhvdrase I I (BCA I I ) byacc ' t ! l r t i on o f Lys e -NHr - s roups b inds neu t ra l su l f onamidel i c t nds w i r h equa l a t [ i n i t 1 ' . 5 Th i s obse r ' a t i on sucges rs t ha t i nthis protein. Ly's e-NH:- eroups rre nr)t important in determininethe af f in i t l 'o f rhe prote in for neurml l isands, anc l b1 in ferenc lthar chuge on these groups does not determine the contbrmationaround the active.sire. \ \ 'e do not knorv for rv,hich other proteinsth i s conc lus ion w ' i l l ho ld .

Capi l la r l e lec t rophores is is an anal l ' t i ca l technique thatprovides informatit-rn simultaneousl l ' about the values c'rf char-geand coef f ic ienr o f f r ic r ion o f pro te ins in so lu t ion. The e lect ro-pht r re t ic mobi l i t l o I a prote in . as measured b,v ' CE. ,1r .1 . . , ro , ispr \ , )por t iona l to i ts charge ar a s iven l .a lue o f pH. ZqE, andin 'erse ly ' cor re la ted to i l la lcp. i rs coef f ic ient o f f r ic t ion (eql ) . r l - rh *here r l l i s thc ' NI \ \ 'o i the prote in . Cp is a propor t iona l -i t l ' const ln t . and o. is a const rn t that re l r tes to the sh; . rpe o i thep ro te in unde r t he cond i t i ons o f i r s ana l ys i s .

i " c l c e t r o

\ \ 'e define Z..u as the charge of a protein calculated from itsanr ino ac id sequence and anv charged cofactors (bound meta li t rns . prosthet ic -eroups, coenzymes) . and post - t rans la t iona l lyadded residues (char_red suga.rs, phosphare groups), usingstandard values of pK. and the pH of the solut ion. uncertaint iesin the'alues of pK (especial iy for His and for ionizabre residuesw'i th anomalous values of pK.) are suff iciently large thatcalculat ions of Zr.o are inexact. The value of ZgE ref lects anydifferences between the actual and standard values of pK, ofthe charged residues of the protein and includes the contributionsof charge from t ighrly associated counterions.

E lect rophoret ic nrob i l i t ies o f ana ly tes are expressed math-emat ica l ly as the d i f fere nce in ve loc i t ies o f an analyre peak and

( l ) \ \ 'e name each runs of a charse ladder by the degree oImodi f icat ionof the proreins which i t comprises, srar t ing * i rh nar i re protein as rhe zerorh

a neu t ra l ma rke r pc ' ak pe r u : : i o i e l cc t r i cI n eq 2 . L ,n , (m) i . s t he t o t r l i en : i h oF t hethe lc -nsth f rom the in le t o i t ie c . rp i l la rv

, l a s l g . i r , r

f i e l d . s t r ens th (eq 2 ) .cap l l l a r l , L l . r ( n r ) , i s

t o t he de te . t o r and I .

[L?) (*)]( l

( l ) i s r he ' o l t ase app l i ed a ; ro ) s rhe cap i l l . i r y . The p l r r : . e re r .t , ( s ) and r . . ( r ) a re t he t i n res o f n r i s re t i on f o r an rn r l r t e oe . r kx and a neu t ra l ma rke r . r espec t i r e l . v . Equ r r i n r i h .

' , , . ' ' o

exp ress ions f o r e l ec r ropho r : l i ; n rob i l i r r t eqs I and l i r nJso l ' i ne i o r r . ( eq 3 ) and l / 1 . r eq - l t . r espe . ' t i r e l , , . de i r , . , n : r r : l e rt ha t p l o t s i n t he l / t i n re do rn ln r r e l r ne r r l r r e l . r i eJ i , t / , - ; ! . \ 1 .

f . :

(-t

' ' ' \ - r i 4 r - : \r - l - - l'

L - . 1 \ . 1 1 , /

1 I ' C , l Z , - r ' \

In," L. . ,L, , , \ . t I t /

i - ' . - 16+ l -_16- i ' 1 .( l 9 t C o r t t c z . F . A : . \ r i l t ' L Z " C h u \ ' : \ \ ' h r r c : i r j c s ' C

l e e - l . 6 d . l 7 S , i - t 7 9 t .

t - l r

(- lrI

t \

both _oi these. as-sumpt ions are doubt less incorrect at some present lyun t l c t l ned le r e l o f de ta i l .

( - ' l ) Cao . J . : Comez . F . A . : Haer tc r . R . : \ \ ' h i res ides . C.lcrrr/. Sc'i. L'.5.A. 199{ . 9l . 120-]i - l:0_r0.

( - l t Cao . J . : N l r ks i ch . \1 . : Comez . F . A . : \ \ ' h i t es ides . G .1995 . 67 , _ i09_r - -1100

t 5 r G r o . J . ; I I a n r m e n . \ 1 . : \ \ ' h i r e s i d e s . C_s -.1 7.

a n d c o n s e q u e n t l y t o m o b i l r r r ( e q l ) . w . h i l e t h o s e i n t h e t i m edoma in a re no t . r - \ \ ' e p lo t a l l p r i n r r r y da ta as absorbance l . s- l / t ime (s - r ) ro show ' t rends rn rhe mob i l i r y . o f runss o f p ro te incharge ladders accuntely ' .

E x p e r i m e n t a l S e c t i o n

I la te r ia l s . Ac l l ase I (po rc rne k idner t . a ldo lase (11pe X : rabb i imusc le ) . a l ka l i ne phospharase ( r rpe V I I - \L : bo r ine in res t ina l mucosa) . 'Nan t i -DNP IeE (mouse) . ca rbon i ; anh l ,d rase I I (bov ine ) . ca rbon icanhvdrase U (human), calmodul in r bovine brain) . o.-ch1'morr l ,ps inogen,( ty 'pe I I : bovine pancreas). cyrochrome c (horse hean), ferr i r in ( ty ,pe I : / .horse spleen), p-galactosidase (8. col i ) . g lucose-6-phosphate dehy'dro-genase ( type XI I : rorula yeast) . grorvrh hormone (humant. hemogtobin(bov ine ) , i nsu l i n (bov ' i ne ) , o . - [ac ra lbumin (bov ine m i l k t . L - l ac ta redehydro_eenase ( type XI : rabbir muscte) . m1'o,elobin (horse hean).o ta lbumin (ch icken egg) , p i ' r u r l : k inase ( r ype I I I : rabb i r musc le ) .superox ide d i smutase (bov ine e r r rh roc l res ) , ub iqu i r i n (bo r ine e r l rh -rocy tes ) . sod ium benzoare . and l -e rhy l -3 - (3 -d imerhy . lam inoprop l I tca rbod i im ide (EDACI ue re pu r , ^h rsed f rom S igma (S t . Lou is , N IOt .Bov ine panc rea t i c t r yps in i nh ib r io r . ca rbox l ' pep r idase B (po rc rne

pancreas ) , c rea t ine k inase ( rabb i r musc le ) . deox l , r i bonuc lease I (bov ine

pancreas ) . dex t ranase (Pen ic i l l i t n r s ; r t . hexok inase ( reas t ; . I 1soz1 'me(eg-e wh i te ) . panc re l t i c l i pase (p r r r . - i ne r . papa in (papara l l t ex ) . pe ro r i -dase (horseradish t . phosphol ipase . \ l I c rotu lus aduntt tnte us rr enom).

and nbonuclease A (bor ine pJnCriJ: r r rere purchased f rom \ \ 'onhingron(F reeho ld . NJ t . Hexad imeth r ine b romide (Po ly 'b rene ; . r \ ' - h1 'd roxv

succ in im ide INHS t , I . 2 . .1 -benz : re r r i ca rbox r l i c anhvdr i Je , l ' 1 .+ '5 -

benzene te t raca rbox r l i c d ianh ;C i ' i , 1 : . rnd me l l i t i c ac id r ' "e re pu rch lscd

f rom A ld r i ch ( \ l i l * ' aukee . \ \ ' l r . Su . -c in i c anh l ' d r i de rvas pu rchescd

f rom Eas tman (Roches te r , N I t . L 'ncoa tec l f used s i l i ca cap i l l a r i cs w i t h

an internl l d iametcr of 50 l rn i \ \ ere purchased f l rom P' ' l l r micro

Techno log ies (Phoen ix . AZ t . Cho l rc ac id r ras pu rch rsed f rom L rn '

c rs te r ( \ \ ' i ndham. NF{ r . r \ ' . r \ ' -D rne ih .v l f o rmrmide (D \ lF t \ r r ) i pu r -

c h . r s e d i r o n r E \ t S c i e n c e ( C i b b ' i o r r n ' N J t r \ ' - H l d r o r l s u c c i n i n t i d r I

cho la te ! ! r s p rcpared accorc l i ns l L ) thc p rocedure o i Okeh ' t t ' t e t a l 1

The z r r i t t e r i on 3 -qL r inuc l i d inoprcp rnesu l ton r t c \ ! ' 3s s ) n thc " t zc t l a ' s

p re r i ous lv c lescnber l . rn N ICK Sp in co lumns con t l i n ing C-5 i ) ! ' 3 r ' r t ! ex

ge l r re re pu rch lse t l f ron r Phurm ' rc i . r B io tech (P isc l t l . ' r l l ' NJ r ' The

iH-XttR spectr i . l r rere recorded et Jr-)0 \ lHz on a Brukcr ;1€ctromctdr '

I [ . . ' l , r , r / . Chem. 1997 ' 69 . 216- i - :170

t tS tOk . rh . r r r . \ ' . : Ando . R : Ku i : i l r ke 'T ' Bu ! ! ' C l te rn ' 5 ' ' < " J2n ' 1979

,,(#) ( l )

rune . The n th . rung o f the char -se laddc r o f a p ro re in hav ins , \ ' Lys e -NH:sroups ma' , be composed o[ as manl as ( , \ , | / ( rV-rr t ln i i regio isomericde r i r l t i r es . As an e ramp le , the n in rh runs o i t he charge ladde i o i bov inecarbon ic anhrd rase I [ . an enz rme rha t has i s mod i f r rb le 'L rs e -NH: g roups ,nla- \ compnse up to 18610 resio isomenc der i r r r i res. The assumpt ions madetbr these calcular ions are that a l l Lys e-NH: groups hare rhe sanie reacr iv i r land tha r re rc r ion a r one s i re does no t i n t ' l u ince the reac t i r i r r a t anorhe i :

\1. Pntc' . I ' iar l .

\1 . Ar r r r i . Chem.

(6 t Gao . J . : \ \ ' h i res idcs . G . \1 . . \ ru t ! , C t ten t . 1997 .69 . _57- i_5S0 .(7t Creighron. T. E. Proteins; st rut t r r rer urrd , \ l rs lec. t t [ur propert ies. . \ \ . .

H . F ree mrn and Conrp rn l : Ner r York . t99_ l . o 6 .rS r Takrhashi . K. J. J . Bio l . Ct tent . 1969. 2) . t . 617 I _6 I 79t9 r Robens . C . : Cordora . E . ; \ \ ' h i res i t i es . G . N t . L 'np . ,L i i rhed resu l r s .r l ( ) ' . \ nJe rson . J . R :Cor ron . [ . J , : \ \ ' h r re , i des . c r t t t n fu t t i r f reu l resu l r s .r l l rHo l . \ \ . C . J P r r , ' : . B i t t p tu i . . t / i , / . B rp l . l 9g i . j ; l i j q_ - f , 1 ,( l l r Crossnrrn. P. D. Cupi ! ! t tn Eler . t roplroresir ; T1,r , , , , t . \ pru<.r t te.

. \ , ' r J e n r i c P r c > s : S r n D r e _ r o . C . { , l 9 9 l : p p l i l _ l - r : 'r t _1 r Compron . B . J . :

9 'g r r l f . E . , f . . r , i i / . C t ten t . t 991 . 6 ; . l - s97_160 : .t l - l t E n e e l h t r d r . H . : B e c k . \ \ ' ; K o h r . J . : S c h n r i r t . T . . . 1 , , , , , , , , , - - C l t e n t . , I n t .E '1 . E r t4 l . 199 -1 . - l - ' . 619 -766t l - i t \ o r o r n r . \ 1 . \ ' . . C o b b . K . A ; L r u . J E l c , t . t r t t [ t l t t , r . . , , r , 1 9 9 0 . 1 / .7_l_ i -719.( l 6 r K r r " e r . B . L . : C o h e r r . A . S . ; C u r r n r . r n . . \ J / ( _ 1 1 7 1 t 1 2 t 1 1 t 1 t ( r 1 9 g 9 .j e _ ' . _ i s 5 - 6 t 1 .

C

- l

\ 1

i

Colron

. \n,t ! . Clt t tr t

Formation of Protein Charge [ndders

Chemica t sh i f t s a re repor red in pa r rs pe r m i l l r on dow 'n f i e ld o f

te t ramethy l s i l ane .

Ac .v ta t i on o f Amino Groups o f P ro te in .s . P ro te ins o f N IW < 50

kD were d i sso lved in wa te r a t a concen t ra t i on o f -0 .1 mNl , and l0

vo l % o f 0 .1 N NaOH u 'as added to each so lu t i on to b r ing the pH to- l l . P ro te ins hav in -e va lues o [ N t \ \ ' > 50 kD w 'e re d i sso lved in 100

mr \1 sod ium phospha te . pH 6 .8 a t a concen t ra t i on o [ -0 .1 mNI . F i ve

to 20 equ iv o f ace t i c anhydr ide d i sso lved in d ioxane was added to each

pro te in so lu t i on . an i l t he rexc tan ts were qu ick l y m ixed by vonex ing .

Recc t ions rve re usua l l t ' co rnp le te * i t h in I m inu te . The samp le was

d i l u t e d i n e l e c t r o p h o r e s i s b u i f e r ( 1 5 m l l T r i s - 1 9 2 m N l C l y , p H 8 . a )p r io r to an r l l s i s . Rerc t i ons us in -q re : rgen ts 3 -5 were camed ou t i n a

simi lar manner except th l t the reagents were dissolv 'ed in i t ia l ly ' in D\ lF,

and the p ro te in ch l r -qe ladders were pun t ied on N ICK sp in co lumns

spun at 2000 rpm tor -1 min, pr ior to analysis by capi l l . r ry e lectro-phores is .

Ace t . v la t i on o f Insu l i n . To a suspens ion o I i nsu l i n ( l 00 , i rL , 1 .5ms /mL in d i . s t i l t ed H :O) was added ace t i c anh ld r ide (5 r rL . 100 m} l ) .A f te r 20 m in . rhe re f , c r i on m ix tu re was ad jus ted to pH l ] w i th NrOH(20 r r l . 0 . l N ) and a l l ow 'ed to re rc t r v ' i t h ace t i c anhvdnde (9 l rL , l 0m\ l t . The samp le w 'as d i l u ted in e lec t rophores is bu i f l e r (2 ,5 m lv l T r i s -l 9 l mNI C ly , pH 7 . - l o r 8 . .1 ) p r io r to ana lys i s .

N-H ,vd rox .vsucc in im id . r l Benzoa te . To a m ix tu re o i sod iumb e n z o r t e ( 3 0 0 m _ e , 2 . 1 m m o l ) a n d N H S ( 2 1 9 m - e . 2 . 1 m m o l ) i n l 5 m L

o i DNIF a t 0 cC w ' rs added EDAC (aS0 mg, 2 .5 mmol ) . The m ix tu rewas a l l owed to reach amb ien t tempera tu re and wus s t i ned ove rn igh t .

The solvent was remov'ed in t 'acuo. The residue was extracted wi th

HrO/CH:CI: (3 x 40 mL): the or-eanic layers were combined and dr ied

rv i rh i l {eSO., . The solvent was removed in t 'ucuo, and the * 'h i te sol id

wrs pu r i t - red by co lumn ch ron ta tog rephy us ing l9 : l CH:C l : i l l eOH

i ' i e ld ing 310 mg o f a wh i te so l i d . Th is p roduc t w ts rec ry -s tx l l i zed w i th

CH:C l : /hexenes to g i ve 290 mg; 6 -17c y i c ld o f r \ -h ; -d roxysucc in im idy l

benzoa te as a wh i te so l i d . rH -N lv lR (CDCl r ,400 lv {Hz) 6 2 .92 (s , aH) ,

7 . 5 2 ( m . 2 H ) , 7 . 6 7 ( m , t H ) . 8 . 1 - l ( m , 2 H ) . H R I I S ( F A B * ) c a l c d f o r

,Cr rHr . rN :Or ( i v l * NHr ' ) * 237 .08 '15 , found 237 .0878 .

Benzenehexacarboxyl ic Acid iVlonoanhydr ide (5) . fv le l l i t ic acid(57 mg,0 . l7 mmo l ) was d i sso lved in 5 mL o f D lv lF w i th hea t ing . To

, th is solut ion was added EDAC (58 mg, 0.3 mmol) . The color of the

solut ion turned dark v io let wi th in a 5-min per iod. The react ion mixturewas st i r red for an addi t ional hour at ambient temperature. The cn:de

mi.r ture contain ing the monoanhydr ide of mel l i t ic acid was used in

acylat ion react ions di rect ly . HRiv lS (FAB-) calcd ior Cr:H.rOrr ( iv l - )

323.975-1, found 323.971-1.

Capi l lar-v Electrophoresis (CE). CE exper iments were conducted

on a Beckman P/ACE 5500. Reaction products derived from modifica-t ion of proteins of p[ < 8.4 were analyzed at 25 or 37 oC on an uncoated

capi l lary of fused s i l ica (L,n, : 47 cm. L,r . , = 40 cm. 50 #m internal

d iameter) using 25 mlv{ Tr is-192 mlv l Gly buf fer , pH 8.4 and an appl iedvol tage of l5 or 30 kV; react ion products der ived f rom modi f icat ion

o[ ant i -DNP IeE were analyzed at 25 oC using 500 mlv l 3-quinucl id ino-p ropanesu l [ona te , l 0 m lv { K :SOr , 25 mNI T r i s -192 mNt G ly bu f fe r ,pH 8 . .1and an app l i ed vo l tage o f l 5 kV . The ace ty la ted de r i va t i ves o fi nsu l i n were reso lv 'ed a t 25 oC on an uncoa ted cap i l l a ry o i f used s i l i ca(L. , , = 77 cm, L: . , = 70 cm. 50 l rm internal d iameter) using 25 mlv iT r i s - l 9 l mNl C ly bu f fe r . pH 7 .4 and an app l i ed vo l tase o f 30 kV .Reac t ion p roduc ts f rom mod i f i ca t i on o f p ro te ins o f p l > 8 .4 wereana lvzed a t 25 o r 37 oC in reve rse po lan ty on a Po l lb rene-coa tedcap i l l a ry (L ,o , = .17 cm. L r . , : 40 cm. 50 , r rm in te rna l d iamete r ) , us ing25 m\ { T r i s -192 mNI C ly bu f f l e r , pH 8 .4 and an app l i ed vo l tage o [30k \ ' . : ' )

Results and Discussion

Format ion of Prote in Charge Ladders o f BCA I I Us ingAcylat ing Agents That Introduce DiJTerent ! 'alues of Charge.Acy la t ion o f BCA I I w i rh acet ic anhydr ide, l , conver ts i tspos i t ive ly chuged Lys e-NHr* groups to neut ra l Lys e-NHAc

t l 0 r COr , tn r " . g ' . C - . J . : \ \ ' h i t es idcs ' G i t t ' An r r l ' Chen l9W' 69 '

J. Am. Chent. Soc., Vol. I19, No. 52, 1997 12703

Scheme l . St ructure o f D i f ferent ly Charged Acy la t ingAgents and Thei r Products upon lv lod i f ica t ion o f the Lyse-NH: Groups on a Protein (P-e-NH:)

Acylat ing Agent Product AZseg

n nv vl l l l

AoA1

^ lv, l

Yo 2

o\4t'cozx

o. I l l p .e .NH1 '

f'vo 3

n / 1

\z>-{o l l l o\-^x.4,1t \ \

o 4 0

i l ^ ^ .P.€.NH/\-2)r '-u2

t t l'- ^'^\,A^^v 2 v v v 2

I 9o''p.e.rx/#yco2

t t l'o,c-/^cor'a n 'v v 2

U U 2 n

c 0 2 H

co2H5

groups (Scheme l).2r The absolute change in calculated charge,AZ,.q, due to acetylat ion of each of these Lys e-NH;* groupsat pH 8.4 is equal to one unit of charge, based on a standardpK. of 10.7 for each amino group. We observ'e from the chargeladder of BCA II, analyzed at pH 8.4, that each of the f irst fouracetylat ions changes the letectro of the protein by an ap-proximately uniform value (Figure 1). The absolute dif ferencein values of /aerecrro of adjacent rungs, A4.1..1ro, then decreasesfor successive pairs beyond the f irst four. From these results,we assume that the absolute dif ference in values of 7aE of

adjacent rungs, L,Zep, for the first four pairs, is approximately

equal to one unit of charge, i.e., L,ftE=AZr.q. We can therefore

use the approximately equivalent values of LZce for each of

the first four pairs of adjacent rungs of the acetamide chargeladder to construct a "charge ruler" that has increments of LZrq

of one unit of charge.The value of AZs.q for acylat ion of a Lys e-NH3* group on

BCA II is greater when acylat ing agents are used that impart

greater values of negative charge to the protein per modifrca-

t ion. l l We observe, however . that the number o f success i l 'epairs of adjacent rungs of charge ladders of BCA II which have

approximately equal values of Al.1g;11, decreases as the value

of AZ.q increases with reagents 1-5. For example, modif ica-t ion o f BCA I I w i th succ in ic anhy 'dr ide, 2 , resu l ts in a charge

ladder in which each of the f irst two pairs of adjacent rungs are

separated by a value of AZce of approximately two units of

charge, on comparison to the char,ee ruler derived from the

acetamide charge ladder. Similar comparisons show that only

the f irst rung of charge ladders of BCA II formed by acylat ing

wi th 1 ,2 , - l -benzenet r icarbo. ry l ic anhidr ide, 3 , or 1 ,2 ,4 ,5-ben-

zenerdt racarboxy l ic d ianhydnde, { , is separated f rom nat i te

prote in by a va lue of Lke of appro. r imate ly 3 or 4 un i ts o f

(2 t ) The a -NH3 g roup o f BCA I I i s n r tu r r l l y acecy la red .(2 l lThc ca lcu la ted va lues o f AZ , .u fo r acv la r ions ca r r i ed ou t w t th

reagen ts 2 -5 and an r l vzcd us ing pH 8 { runn ing bu f f c r a re based on thcva lucs o i pK . o f t he co r respond ing c r rbox l l i c ac ids : succ in i c (+ .19 : 5 , lS ) ,t r i m e t l i t i c ( 1 . 5 1 ; 3 . 8 1 ; 5 . 2 0 ) . p . v r o m e l I t i c ( 1 . 9 ] : 2 . 3 7 ; + . 1 9 ' 5 . 6 1 t . m c l l i t r c( t . - 1 0 , 2 . t 9 ; 3 . 3 t ; { . 7 S ; 5 . 8 9 ; 6 . 9 6 ) . T h e : e r r l u e s o f p K , * e r e t a k e n f r o mJencks, \ \ ' . P. : Rcgenstein, J. ln Hurulbook t , . f Brr tcherni . i tn, lnd cd. ; Sober"H A . , E d ; C R C P r e s s : C l * e l r n d . O H , 1 9 7 0 , p p J - l 9 l - J - 1 9 - l

oll r

P ' e ' N H \

or l ^

p.".rqx/\-/ 'Yu 2t l

Av

I l 7 0 - I 1 7 , - )

t270.1

T a b l e l .

J. Am. Chem. Soc. , Vo l . I19, iVo. 52, 1997

Panel o f Prote ins Surveyed To Examine the Cenera l i ty o f Form.r t iono i Charge Ladders

C LI D pro tc rn ECNI\\ '

subun i t s " ( kD t "c a p i l l a r l

p [ ' c h a r g e b reage n t ' / zrr , ' L ' t r

Il-'1+

,5

b

7E9

t 0l ll lt \

l ll 5t oT 7t 3l 9l0l rt lj - l

1 l

25i o

2 72S2930- t_1 I

3 l

i n s u l r nb o v i n e p a n c r e a t i c t n p s i n i n h i b i t o rt u h i q u i t i nc r t r t ch rome cr ihonuc lease Ac - l u c t a l b u n r i nl r sc- iz t ntec a l n r o d u l i nm r o g l o b i nhtrnrun grr) lv th hormonep. ip.r ina -ch . r mo t ryps inoeenph t rspho l ipase A2c l rbon ic anh .v d ruse I I t bov ine fc l rbon ic anh- rd rase I I (human)

deor l r i bonuc le l se Ic l rbon ic anhy d r rse \ ' ( mur ine l tsupc rox ide d i smut rsec l rboxvpep t id l se B"o r a l b u m i ndc \ lri.Inuse'pe ru r idase/c re l t i ne k inase(l c l ' l l se Ir l k r l i ne phosphr taser l ucose-6-phosphate dehl drogenasct I

t - - I lc t l te dehl 'dro_lenl-se 'a n t i - D l i P I g Er ldo lasep. ! ru\ f , te k inasep -*u r lactos idaseierr i t in

3 1 . 1 7 5

1 . 2 . r . 1 7

_1. -1 . I1 .2J . + . i l . t

_1. t . L.tr l t l

t ' ) t lr . - . t . 1

l r . 2 r , rt 1 r I

l . - . l . I

1 . 1 5 . l . r3 . + . 1 7 . :

3 . 1 . r . r ll . l t . l . 72 . 7 . 3 23 . 5 . r . l +J . l . - 1 . 1

t . l . t . - 1 9L l . l . 3

I | 1 l a+ . t . i . t J

2.7 . t .+0_3.2.1.2,1

5 7 5 37 .,1 10..18 6 7 .0

l l J t 0 0t --1.7 9 8! a l

t + . - + . 5

l + . 1 1 0 . 91 6 8 + 0I r - .6 6 .322.2 5 .32-1. , i 9 .62 5 1 9 827 .6 _5. I19. i -5 .919 .+ i . 630.-5 -5. ti 0 9 1 |3 1 . _ i 6 . 6_1+._r ,s 9l i . - t _5 II r n r ' )

? ? , ( /

t r n ' i 1t r . v 1 , -

8 6 6 1 r9 1 . 0 - 5 3

106 6 0I 1 5 6 . 1l + 7 8 . 6t_501 5 8 E 7t_ t+ 6 .9.168 5.0179 6.0

1 , 2 . 3 , J , 5I1 . 3

II

)

;t.+

-r

- t A

_5 .5-0 - l

f n

l 8_ _5.9

'1 |

- ; 0

- s 0t t

__i - i

- r :- l _ i

_ 1 1

__l s

_ l tI t

0 9

) e syes

) e syes) e sues)'es) e s)'es,!'e s)'es)'e s! e \ . p r) e :

) 'es\ es, pt '

) e s) c s! e s) c s) e )

) e s) c inono

) e s) c'snononon 0no

t ;

- r 0

-1 rl

-S r-,7 r l

-- i q-6 ( l

- l l

- i 0- l , ) I

- l ( )- t l

- l ( l- l A

- 6 0

t 1- L l

- s .9

, l

I?

,1

' ) !

/ , 2 , 3 , { . 51 , 1 , 5I t r -

/ , 3 , {1 , 2 , 3 , 4 , 5-j

J , J . J/ i

l t <

1 . 1 . 5

1 r- I I t

I i t 1

'See SwISS-PROT dunbase. The inlemet address fo. SWISS-PROT dalabase: htlpJlexpasy.hcuge.ch,/JproLlsprot-top.html. N{olecular ueighrslre given tbr the aggregale of all subunits. Values of ltlw are calculated from the sequences of the pioteini rnd include th€ iuw of any boundmetal. heme group, phosphale residue or carbohydrate residue. Values of pl and .4.". the values of Z*q of Dati!e proteins. were calculated from thcsequences of native proreins lsing the fol lowing standard values of pK. for rhe amino acid residues: Arg = l : : Lys = 10.7; Tyr = 10.21 Cys 19.1: c(-NHr' = 7..1: His = 6.5; a-COOH = .1.9; Glu = 4.4: Asp = 4.0. These values are midpoints of lhe range of values of pKn given in CreighlonlT. E. Proreins' Stucturcs and Llolecular Propenies: W. H. Freeman and Company: New York. 19931 p 6. ' Capillary charge' indicates thecharge on the surface o[ lhe capillaries. An uncoated capill!ry is negatively chcrg.d. and a Polybrene-coated cepil]3ry is posilively charged. 'Success

in tb.mrlion ofcharge ladders. The designalion'pr" indicates thrt the rungs of rhe charge lrdder were onl]- prniilly resolved. r' Rergents thirt weresuccessful in generaling chcrge ladders are shown in bold. In cases where a charge ladder could no! be resol!.'d. the reagents shown in itrlics arclhore thrt r1ere (ried. ' Values of .48. the values of ZcE of narive proreins. were derermined by erlrapolxrion ro rhe -!-intercept in plots of1ri,".,,,, vsnAZ,.r, trsing the rungs of lowest orerall chorge in the corresponding charge ladders. These values we.e obrrired using 25 mill Tris-l9l miV Olyrunning bufter, pH 8J. 25 "C. /The value of pl of bovine caibonic anhydiase I[ were txken from the folloliing: Deu$ch. H. F.lnr' J. Eiochu|1987 '19 101-l13 Holmes. R s Euf J Biochen l9?7 , 78. 5 t I -520. ! The value of 4" was estimated usins l5 mlvl Tris- 192 mll cly runninSbuffcr. pt l 3.,1. at37'C. iThe valueof iul \ o f carboxy pepridase B was raken from rhe foi lorving: Fotk. J. E.: Piez. K. A.;Canoll . $' . R.lClrdncr'! . A. J. Biol. Chen. 1960. 2J 5,2212-2177. The value'of j l of carbo xype ptid rse B was raken fro-m rhe fol lowing: Sasaki, [ . ; Cotoh. H:Yamrmo(o.R.: Hrsegarva. H.; Yrmahita. J. i Horio. T. ! . Biochent. '1979, gf. tS:Z-tS+9.'The values of i l lw and pl-of dextranase were !ak.n from thcfol lo\! ing: Susiura, A.: f to.A.; Ogiso,T.: Kato, K.; Asano. H.Br'achim. Bioph\s. Acta 1973,309.357-362.'Tne value of lvl \ l ofperorid:r\c wn\taken from thi follorving: Welin-der. K. Eur. J. Biochen. fCZC. CO, .lg:-S0:. The value of pl of peroridrse was taken from tbe folk)*inslNlrehl). A. In tr lethotls- in Enltnolog;-, Colowick, S., Kaplan. N.. Eds.: Academic Press: NeJ Yor{. 1955: \ 'ol .2. pp 801-813. Thc vrh'tc of

{u of peroxidase was nor calculated due to rhe ambigui(! in the number o[ charged carbohldrate residues on the native protein r Thc vrluc\ of

{u of these proreins we.e esrimated using charge ladders formed using reagent 3. I The srrte of rggreg ron oi glucose-6_phosphlte dch}droScniN'fr'o:m Caadr,i r (rilir w;N uken trom the iollo*:ing: Engel, H. J.: Doischie. rV.: Albeni, iil.: Oo.[t. C. F.- Brochim. bophtl A.t!] 196l'' lql

509 -516 .

char-se. respectivel) ' . The f irst rung of a char_se ladder of BCAII made usin-s benzenehexacarboK)' l ic acid monoanhydride, 5,how'ever, is separated from native protein by a I 'alue of LZceof on l ,v ' .1 .6 un i ts o f chuge.

Thc ' obsen 'ed decreases in va lues of A, t r .1 . ;166 for success ivepai rs o f adacent run_ss o[ charge ladders o f BCA I I may bedue to decreases in l 'alues of AZqE and/or to increases in thecoef i ic ient o f f r ic t ion o f the prote in rv i rh increas ing number o [nrod i f ic r r t ions. Decreases in va lues oI AZce ma] ' be due to ( i )

ch lnses in the \ ' r lues o f pK. and af f in i t ies tor ions o f theion iz . rb lc groups on the prote in as the mrgni tude of i ts chr rge

incrc . r 'es l rnd/or to ( i i ) ch lnges in the d is t r ibut ion o f ions

response to increases in i ts charse ( " ion re larat ion") ' : l Changcs

in va lues of the coef f ic ient o f fnc t ion may be atmbutcd to ( i )

contr ibutions oI mass and I 'olume from the conjugrtcd.rcyl l t ing

agent and/or to ( i i ) confon 'n l t iona l chanses resu l t ing f rom

increasin-gly uniavorable electrostat ic interacdons on thc surfacc

of the prote in w i t f r increas ing number o f ac} le t i665 ( thc Process

of "e lect rosta t ic denalurat ion") . : *we ev'aluat.d t ; ; ; ;nr.,u' t ion of m35s and 'olume from

acylat ing agents r iS to rf t . coeft-rcient of fr ict ioo of BCA II

oi". i i . i t" ; i l ;protein rr i th rergenrs thxt inuoduced dif ferent

r o n . J . A . S t o p h t t . J . l g t n . 7 - l\ - - ' r . 1 l l l \ u l l . J . . 1 . r \

l - i - i - t l r ) r D . . t f L t - t t Q l a i r i 6 , / \

{ l - l ) A r i r r m ' I : \ t r e r ' \ " P ' S ; h c 1 t e r ' A J | n l C h c n t l 9 3 l ' : - ' o ' /

Formation of Protein Charge Ladders J. Am. Chem. Soc., Vol. I19, No. 52, 1997 12705

lJ2rcg

1 0

M."q

I

o oAoA

\ \ ^ ^ , ,ff7vwtn

o . t I)-\.,i

n

\\\V

an

> { . 1 5I

o. {.20a

?

0.0 0 .5 1 .0 r .5 2 .0 2 .5 3 .0- lA lmc (1 t r r r - t ;

Figure l . Formation o[ charge ladders of BCA II at pH 12 usingacylat ing agents I -5. The scale (n) below each electropherogram

' indicates the number of modif ied Lys e-NH1 groups on rhe prorein.rThe nAZ*o scale at the top of rhe electropherograms has increments ofone unit of charge. The two isozymes of BCA II dif fer by one unit ofcharge. Gln in the minor isozyme (pl = 5.4), labcled with a f i l ledtr iangle (r), is replaced by Arg in the major isozyme (pl = 5.9) aslabeled with a f i [ed square (I). The neurral marker, p-merhoxyb€nzylalcohol, is indicared by a f i l led circle (O). lvlodif icat ion of BCA II byacetic anhydride yields derivatives of the major isozyme that overlapwith those o[ the minor isozyme. Nlodif icat ion of BCA II by acylaringagents 2-5 dist inguished rhe two isozymes of BCA II. The asterisksindicate impurit ies from the protein sample. The - l / t ime scale ar rhebottom of the set of electropherograms is proport ional to the mobil i r iesof the peaks in each electropherogram relat ive to the neurral marker.The electropherograms were col lected ar 25 oC using 25 mlvl Tris-192mlvl Gly, pH 8.4 and an applied voltage oi l5 kV.

values of mass and volume but only changed the value of Zsrqof the protein by one unit of charge per modif icat ion. Acylat ionof BCA II with N-hydroxysuccinimidyl benzoate and iodoaceticanhydride contr ibute 104 and l&+ D of mass per acylat ion,respectively. We observe that the values of A491..6o forsuccessive pairs of adjacent rungs of the charge ladders formedwith these reagents are approximately the same as those of theconesponding pairs from the acetamide charge ladder (Figure2) ; when the larger cho l ic ac id res idue (389 D) is con jugaredto BCA [I, however, only the values of A4g1..1r. of the f irst threepa.irs oi adjacent rungs of the resulting charge ladder are similar.These results suggest that the derivatives of BCA II thatconstitute the rungs of the acetamide and iodoacetamide chargeladders do not hal]e sienificant concributions of mass and volumefrom their respecrive;onjugared acylet ing agents. Any dif fer-ences in the coeft lcient of fr ict ion among the derivatives thatmake uP successive rungs of each ladder mxy then be attr ibuted

Figure 2. Plot of the mobi l i t ies of the nrngs of charge ladders of BCA

II formed w i th acet ic anhydr iCe ( x ) . iodoacet ic anhydr ide ( t r ) .

A ' -hydroxysucc in imidy I benzoate ( r ) , or rV-hydroxysucc in imidyl cholate(O; as a funct ion of nM*. Some of the more heavi ly acylated rungs

of the charge ladders formed wiu\ the l / -hydroxysuccin imidyl esters

of bcnzoic or chol ic acid could not be resolved. The electropherogramswcre col lected at 25 oC using 25 mI{ Tr is-192 mfvl Gly, pH 8.4 andan app l i ed vo l tage o f l 5 kV .

o'o ]',|'-" i& '''lo

4'o

Figure 3. Acetylat ion of the ar:r ino groups of myoglobin at pH l2and 9. At pH 12, acetylat ion occurs selectively on the Lys e-NH2 groupsof the protein. At pH 9, modif icauon is not select ive. and acetylat ionoccurs on both the o.- and Lys e-\H: groups. Coinject ion experimentsestabl ished the identi ty of the derirat ives of the protein. The scales (n.and o. * n,) bclow each electropherogram indicate the number ofmodif ied Lys e-NH3 groups on the native protein and a-NH1 acetylatedprotein. respectively. The f i l led circle (O1 indicates the neutral marker,p-methoxybenzyl alcoho[. The electropherograms were col lected at 37oC us ing 25 miv t Tr is -192 ml l Gl .v , pH 8.3 e lecr rophores is buf fer .

to dif ferences in their conforination due to the process o["electrostat ic denaturation". lr We conclude from these studiesthat decreases in values of &r,1..6o for the successive pairs ofadjacent rungs of the charge ladders of BCA II formed withreagents I and 2 are caused by' electrosradc effects that in-fluencevalues of AZce and./or the coefr ' ic ient of fr ict ion of the protein.The values of Al l .r . . . ,ro of the plrs of adjacent rungs of charge

o o

o\c#o)'\

rcct lc rnhydr ldc

lodorcc l lc rnhydr ldc

N-hydroryrucc ln lmldy l b .nro. t r

N. hydro ryr ucc ln lm I dy l cholc tc

il(14 . ,

F o* o

Ag ̂A e ^

I V ^^ n - 6^ n - o^ *

^ *F ( n

h -h f-l

t < t t -r< l{

^ CO"H

)\-o\)co,xo l l l

F^\ico,xo co2H

Acstylat ion at pH 12

12706 J. Am' Chent' Soc" l 'ol ' I 19' r \o' 52' 1997 Colton

BA

i I l,1ill,1r1nl,r,u,t,r,',lr SuoaroxideDismurassI I _ ,t ll illlll/UVl/Vt/tJVrvur\ ".,ioliil r, *or

_JLJUU,Uv.u', ,S.,,_

a t

n 4 1 2 1 0

.0.5 o.o 0.5 ' l .0o ' 1 l l lme (10 ' l s ' 1 ;

till l P a i a r nilr r ( P l 3 i 2 ' l k D )! - n- l \ / \

I t / \ l \ r 1 |l l n / v v \ / ,l v " u \

- / - r t l

n { 3 2 1 0

o I - ,n, 'J ioto-, r ' ' ) t

5 2 . 0

_lI 'i l LysczT-a Il l ( p t ro 9 1 . : ( l i l=r , i liL-ny-,n-,tt -/Ul.-;1-;1;-i

Myoglobin( p l 6 . 8 , 1 8 k D )

l | | t I l l r r l f

r l l l l r n

o .o 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0' tAlme ( tOr s ' t1

ladders formed with reagents 3-5, however, may in addit ion

be inf luenced by dif ferences in the coeff icient of fr ict ion of the

proteins that consti tute successive rungs of each charge ladder

due to appreciable contr ibutions of mass and volume from the

conju-eated acylat ing agents.

Select ive Acety la t ion o f the e-Amino Groups of Lys a t

High Values of pH. The o'-NH-',* group on the N-terminus of

a protein has a lower value of pK. (PK, - 1-9) than do the

. - N H r * g r o u p s o f L y s ( P K . , - l l ) ; t h e n u c l e o p h i l i c i t y o f t h ec-NH: group is also lorver than that of Lys e-NH1 groups'a'2o

Acy la t ion o f an ammonium group (RNH:*) occurs on ly by

reaction of i ts neutral fonn (RNHI) rvi th the acylat ing agent.

Since a neutral e-NH: group is more basic and more nucleophil ic

than a neutral c-Nll: group, reacrion at a hi,eh value of pH where

both amino groups are deprotonated rv' i l l occur selectively on

the e-NH: group. At low'er values oi pH, w'here the a-NH3+ is

par t ia l ty deprotonated but the e-NHt* is essent ia l ly ent i re ly

protonated, i t is poss ib le to ach ie l 'e se lect ive acy la t ion on the

c-NH3 pos i t ion.

F igure 3 shorvs e lect rophero-srams co l lec ted at pH 8 '3 o f

acety: lat ion reactions of myoglobin carr ied out at pH 12 and 9'

At pH 8.3, the value of AZcE from native protein is greater for

a prote in that is acety la ted on one of i ts Lys e-NH;* groups

thrn for one that has i ts cr -NH;+ group acety la ted ' due to

cl i f terences in the vl lues of pK" (and. consc'quently ' of fract ional

protontt ion and a! 'erage chrrrge) of rhe nv'o residues' Part ial

[ , i l i l11rfi l i l ,r,r,rr aLacra,bumin

_-.,|

8' 14 kD)

il|nl A [m"""r

^- - , vV\ / \

- 0 . 5 0 . 0 0 . 5 1 . 5 2 . 0 2 . 5

o . l i t t m o ( 1 0 ' l 3 ' 1 )

L l l ln 3 2 1 0

. 0 . 5 0 . 0 0 . 5 1 . 0'1 l t lmo (10 ' l e ' t1

Figure 4. Represenrorive charge ladders of proteins fonred by selective acetylation of L)_s €-NHl grouPJ a! PH l: The neutral mrrker'

p-merhoxybenzyl alcohol, is indicared by the fil'led circles (a) and oative proteins are indicated by the filled squrres (I) The number of acet!'lated

Lys €-NH: groups (n) is indicareO Uetor" eaci "i."i.ptr*og."..

t.ql Charge ladders of proleins hlving valu.j ot'pl < 8 -{ analyzed using uncoated I

capillari.s. (B) charse ladders ot proreins;;;i"e '"li'.' "e

pr ' i 't' anatvlzed.*ine P"iio':i:--"i"1:,1::l'l!:: l:;t::T '""'' on fte bottom

of'rhe electiopherogiams is proponional to rhe mobilities of rhe peaks in each elec!rcpherogrxm relllive lo the neutrirl muKer'

acetylat ion of myo-elobin at pH l2 resulted in a charge ladder

in which each of the f irst ferv pairs of adjacent rungs were

separated by a value of AZce of approximately one unit o[

charge. Acety la t ion car r ied out a t pH 9. resu l ted in two

overlapping charge ladders; the rungs of the minor charge laddcr

had values of , /{ .1.. ,ro that overlapped with those of the rungs o[

the charge ladder rnra. a t pH l l - Acetv la t ion o f myoglob in a t

pH 7 gave results similar to those oUtaineA from thc mocliflcatit:n

at pH 9. A co in jec i ion o f react ion mix tures f rom pH l2 and 9

confirmed that two Ji" int ' charge ladders werc fornrcd by

acety la t ion o f success i l ,e e-NH: groups at pH 9: onc dcr ivcd

from natir ' . rnyogotin' and the other dcrivcd from myoglr:bin

acetylated on the N-terminus The results also confirmcrl thlt

the e-NH3 grouPs " ; ; ; t i ; t

myogobin arc sc tcc t ivc ly rc ty l r tcd

at pH l '2. The dif ference bctwccn thc rwo chrrgc h<l '

].,,-,pp," -, T1*. ilin ;1, "J.'H'ff" frfr' J;?ii!:il:i..!-:ilffi il: ;: 1t'ti jlTp'fr#tobia irapproximrtetv so*';:;;;1;;' 'cd (P& = I'o lx Gty eNHr')'

S uccessfu I E xam p t'-r o f j! n-c.'YO^ o( Proldnr Wc

. * pro,.a,r:,i.:ill* j;;ro,ru!ffff"i:i;lilXl"'T.I;.';fu;'"-ir a sG or 6' Fvcior * cstudied contrin tuf ipft subuniu' 'fftrl

hna'hG 1'qryl'

;l;. :,; ; : :::lni :y,i : *H'ffi 'Xffi-1A shor ' , 's sc! 'c13l s

6

Carboxypeptidase I(pl 5 9, 3.t k0)

___-__

i_.

1997 t2707Formation of Protein Charge Indders

' 1 l t l m c ( 1 0 ' : 1 ' 1 1

Figure 5. Elecrropherograms of thc react ion products f rom modi f ica-t ion of crear ine k inase ( iv lw = 87 kDl ar pH 6.8 using acylar ing agents3-5. The scale of the - l / r ime axis is proporr ionar to the mobir i t ies ofpeaks in each electropherogram relat ive to the neutra l marker, p-methoxybenzyl a lcohol (O). The Fr l led square ( I ) is nar ive crear inekinase. The electrophero-erams were col lected at 25 oc using 25 mlv lTr is-192 mtvl Gly, pH 8.a and an appl ied vohage of 30 kV.

formed from proreins having values of pl < 8.4, the pH of therunning buffer.

Proteins having values of pl > 8.4 have a net posit ive chargeat pH 8.4 and tend to st ick to the negatively charged surface ofuncoated capi l lar ies. To overcome this problem, we usedPolybrene-coated capillaries to reverse the charges on the surfaceof the capi l lary. creating a posit ively charged surface reversesthe direct ion of elecrroosmotic f low (EOD.zs Analysis ofsamples must therefore be done in reverse polari ty and resultsin the grouping of rungs of charge ladders in decreasins orderof modif icat ion (Figure 48).

We assume that each rung of a charge ladder comprisesproteins that have approximately the same values of ke ndcoeff icient of fr ict ion.26 We have observed, however, that themore heavi ly acylated rungs of some charge ladders aresuperimposed on a broader background. We do not presentlyunderstand the nature of this broad heterogeneous peak; i tscharacterization wil l be the subject of later work.

Charge Ladders of High II I IV Proteins with Acylat ingAgents 3-5. Proteins that have values of lvlW >50 kD do notform resolved charge ladders when modihed wirh acetic

. (25i Elecrroosmot ic t lorr (EOF) is the bulk t low of solut ion in a capi l lary

(nat resul ts f rom the appl icat ion oI an elecrr ic f ie ld. The value of EOF ist v "p i c r l l y exp ressed . r ' i h . r ime requ i red fo r a ncu t r r l mr rke r to m ig ra tet toT jh! in ler of a crpi l l l ry to rhe detector w' indow.

, ( :6) The poor pc.rk shape for rhe rungs of some of the cherge ladders

shown in Figure iB n. ' " r - be due to s l ighc di f fcrences in the vclues o( keand/or coef l ic ienr o i f r icr ion for the di f ferent regio isomcrs that makc upt h e s e r u n g s .

J. Am. Chem. Soc . l 'o l . I19, lvo. 52.

- 1 , " t i m e ( 1 0 ' 3 s ' t ;

2 . 52 .01 . 51 . 00.5

Ai i l i l r ' 1rtt i l l l l lr l l lr/r ri, i . lli;':lilt/ilirlr/irl,/ f l l l ' , ; 'u ' I7! / ' t t t t t t 'l , * f i . i t

' u ' \ * ^

- J V Y " i I r r r r l r r r r r r r r r r r r n- ' , 2 2 - # n..15;,' 5 10 15

nVr"q- 5 0 5 1 0 1 5

4

n \ ? s e q

. l 0

Figure 6. (A) The values o[ AZcs berween each o[ rhe f irsr four parrsof adjacent rungs of rhe aceramide charge ladder of BCA II areappro.ximately uniform and can be used as a "charge ruler" to est imrre4r.Th. f i t led square (I ; is narive BCA II. The scale of rhe - l l t imearis is proporrional to the mobil i r ies of peaks in rhe electropherogramrelat ive to the neutral marker, p-methoxybenzyl alcohot (Ot. Theasterisks indicate impuriries from rhe protein sample. The electrophero-gram was col lected at 25 oC usine 25 mNI Tris-192 mlvt Gly, pH 8.a.(B) The value of 7r, o( -3.- l oi nariv,e BCA II is est imated b.vplott ing the mobil i r ies of rhe f irsr t ive rungs of rhe aceramide chargeladder as a function of nLZ^ (LZ,,q = one unit of charge) andextrapolat ing to the x-intercepr. The assigned errors in measuredelectrophoretic mobil i t ies are -5%.

anhydride. The lack of resolut ion is due to the low value ofthe ratio of terms Lkeliltr = l/fu[a (eq l) that accounts for theseparation between adjacent rungs of the charge ladder. In orderto overcome this problem. w'e tested the abi l i ty of acylat ingagents 3-5 to yield resolved charge ladders on modif icat ion ofhigh lvfW proteins. Nlodifications performed with these reagentsresult in higher values of LZSE|i l ta = 3/NI" to 6/, lF betweeneach of the f irst several pairs of adjacent rungs of thecorresponding charge ladders-and consequently better resolu-t ion of these products by CE. \ ! 'e were able to form resolv'edcharge ladders of creatine kjnase ( lvtw : 87 kD) (Figure 5).g lucose-6-phosphate dehydrogenase (NlW : l l5 kD) , andL- lac ta te dehydrogenase ( \1 \ \ ' : 147 kD) by us ins reagenrs

Est imat ing the Charge of Prote ins in Solu t ion by CEUsing Prote in Charge Ladders . Esr imat ions of the charge ofproteins over a renge of values of pH have been performed usingmethods based on iso ion ic po in ts lT and Donnrn potenr i t l

B

FI-l-Li- n0

. \

( r / ) N o z f , k r , Y . : T l n f o r d . C , \ l e t h t t d , t E n . - v r t t t l . 1 9 6 7 , 1 , 7 1 5 - 1 . 3 1 .

12708 J . Arn. Chern. Soc. , \ 'o l . I19. lv 'o . 52, 1997

{ .5 0 .0 0 .5 1 .0 1 .5 2 .0' 1 l t i m e ( 1 0 ' r 5 ' t ;

nlZseq

0 5 1 0 1 5 2 0

2 . 5 3 . 0

2 5

Figure 7. (A) Electropherograms of the charge ladder of L-lactatedehydro-eenase (NIW : 117 kD) formed at pH 6.8 using acylatingagents 3 and 4. Each unit in n (the number of modihed Lys e-NH3+residues) conesponds to values of M,.o of approximately 3 and 4 unitsof char-ee, respectively. The neutral marker, p-methoxybenzyl alcohol,is indicated by the f i l led circles (Ot. The electrophoresis buffer is 25mjl l Tris-192 mil t Gly, pH 8.-1. The - l l t ime scale on the bottom ofthe f i ,gure applies to both the electropherograms. (B) A plot of themobil i t ies of the run-es of the char-ee ladder of L-lactate dehydrogenasevs nt5,2,,, used to estimate the value of tru o[ -1.5.

measurements. ls '2e A major drarvback of these procedures,

however , is that they requi re s ign i f icant amounts ( l -10 mg)of pure proteins. We have developed a method based on chargeladders that is used to estimate the charge of proteins in solut ionof d i f ferent va lues of pH us ing on ly pe quant i t ies o f pro te ins; lpure samples are not required in this method provided that anyimpurit ies can be identi t ' ied and separated from the rungs ofthe charge ladder.

The values of L,ke are approximately uniform for the pairsof adjacent rungs of lorvest overal l charge of a char_ge ladderand are used as a char_qe ruler to est imate the value of Zce of

nat ive prote in , { r . r For example, as pre l ' ious ly descr ibed, thevalue of LZgs at pH 8.4, for each o[ the f irst four pairs ofadjacent rungs of the acetamide charge hdder of BCA II, isassumed to be appror imate l l ' equal to one un i t o f charge, the

l 'a lue o f AZ, .o (F igure 6) . To est imr te the va lue of {u o fBCA II at pH 8.-1, rve f irst plot the mobil i t ies of the f irst f iverungs as a funct ion o f r rAZ, .u . rvhere r r is the number o f

acetv la t ions. These d l t r are then f i t b l the method of l inear

( lS tFord . C . L . . \ \ ' i nzo r . D J . B r , , t ' l t i r t t . g i l t l t i t . r ' s . . . t r ' / r r 1931 , 70J ,109-I l ] .

t l g rO j teg , C . ' L t rnd ' rh l . P . \ \ ' o lg . r ' t . N l . B r r ' c l r i r r t . B i r tp l t . t i . , { c ' t r r 1989 .t2 t7 l . -117-_ ' i l _ ' 1 .

t . 1.' t1;i_ 5

t ? . ;u -^ ' i l l - - : t 5

lagl:*i t rzr / 10

r r '16

.s0 -40 -30

_-:t

-1 0 0 1 0

t t a q

F i g u r t S . C o m p e r i s o n o i i r l u : : . . i Z . l i o t p r o t e i n s t o t h e c o r r c -

s p o n . i r n r r a l u e s o f 4 . q . T h e r r l u e s o i 4 u f o r c r e r t i n e k i n l s e ( l - l t .

g luco 'e -6 -phospha te deh l d roe :n . ] :3 r l 6 t . and L - l l c t l t e , Jeh r d rosenrse

( l7 l r i e r -e es t im l ted us ing ch . r rg : l r JJc rs made us ins re f , sen t 3 . The

va lues o [ 2 . . f o r a l l o the r p ro re : ; ] s r i p resen ted in th i s t - t gu re were

es r im l red f ron r the i r respec t i r c r c : l : n ide ch r rge ladders . Thc d l shed

d iegon l l l r ne i s added as a r i su r l a rJ ro rep resen t a l : l r e l r t i on be t r "een' , ' a lues o i 4 r enJ Z , ' "u . Pcaks a r : nun tbe red accord ins to the i r ID

numbers i r t rn t Tab le l .

least squares to a l ine rvhich is def ined by eq 5. In th is equl t ion,

/ r l r . . , ,o refers to the mobi l i ty of the nth rung of the cherse

ladder, p3r. .oo is the mobi l i ty of nat ive protein ' -Cp/, \ ld is the

slope of the l ine, and nLZr* is approximately equal to the

di f ference between values of 4.e Md 4r, the charge of the nth

rung of the charge ladder (eq 6). Exrapolation to the .r-intercept

(where p}..,,o - #:i:.,," : 0 and 4.e:48 - 0, nm _ neuual

marker) of the fit l ine gives the value of 7r, of BCA II of-3.3 (eq 7). Values of 1e \r 'ere similarly calculated for high

lvlW proteins, such as L-lactate dehydro-eenase (Fi-eure 7) using

acylating agents 3-5, where M,.o is equal to approximately

3-6 uni ts of charge.

A

uJ ^^O ( ) ' Z UN

B 0.00i \ -t \ \

t \{ \' in I \ \

-C-3- ; o .os - z i . - r . s \

{ ' 4-l 1 \c { l \

st \o r a ai - 0 . 1 0 ;

\ rg ' i \t = " j \

t \ .o . r s l \ t

) '

C,lrlr..,,o - -\{uM,.o) + /,3,..,,o (5)

nMrrr= 7 i , - 1 , (6)

o! l . t ' . r , n

f - ' _ - - : :l - f ' t r - t 7 \

/ L p I

t+tLtl^/

( 7 )

Colron

l l l l l r ;

Figure 8 compares the va lue s o f Z i7 o i d i f fc rcnt Pro lcrn ' l to

the con'esponding , '0,r. , of Z', ,0' the "u'alues of Z-q of n'rt ivc

proteins. We obserl 'ed good igretmcnt bct* 'ccn vrlucr of

{ , and 4.0 rvhen?; i ; ; , - '9,^,1," 'vr lucs o( zn <-10'

horvever, rve found ir 'h 'u'otute velues of & t*crc'ovetc' t t i '

meted b1'the .ntgl i toot ' nrt*"d \\ 'e essumc thrt t tr values

of AZcE for the p' i"-of adjacent nlngs of lorrcrt ot<nl l chuge

o f a c h' g e I ad d e r ;' ;'#;'::'lt':: "J.' l.f'ffig; 1 "; ; l r . ; i i Z,,u' , i t thPc" r ' f , lues ot Arc'r '

thrt o f 1\Z,er,,h. n ;;' :'' ::'i t:,lll} :J ffi:l'I;vatue of ziu of calmodulin is t: lf,:

o( L?<t of cxh o( thc

charge when trie assume that f ic^vatlt;,;; ch:rge laddcr

i;;;;;." pairs of e..l jrcent rungs ol rts f

?ormation of Protein Charqe Indders

p l

Figure 9. Summary oI the resulrs of experiments designed ro formprotein charge laddcrs. The values of pl of the proreins were plotredagainst their values of lvlw. Proreins from which we were unable toform charge ladders are indicared by f i l led circles (Ot. proteins fromwhich we successful ly formed charge ladders are represenred by opencircles (O) i f rhey were analyzed using uncoated capittaries or by opensquares (C) i f they were analyzed using Polybrene-corred capi l lar ies.Peaks are numbered according to rheir ID numbers from Table L

is appro.r imately equal to one unit of charge, the correspondingvalue of AZr.o; i ts value of 4.0, however, is -25 units ofcharge. We obtained better agreement between the values of

4, md tro of native calmodulin when we assumed that each. value of ATce was equal to 0.6A2r.o, i.e., 0.6 of a unit of charge,' to estimate the value of tcE. This low value of Lfts for even

. -the first several pairs of adjacent rungs of the charge ladder of' tcalmodulin is reasonable based on the masnitude of the value

of 4.q of calmodulin.

Conclusions

The combination of charge ladders of proteins and CE is anew biophysical tool that can be used to estimate the values of

J , Am. Chem. Soc. , l 'o l . I19, lv 'o . 52, 1997 l2 l ( ) ' t

charge oI pro te ins and to exr lore va lues of pK. , . ion ic assoc iat ion, in teract ion o f charged l rgands wi th prote ins, and a rangco f o the r phenomena i nvo l r i ng cha rged spec ies i n te rac t i ng w i r l rpro te ins that are not a lways access ib le by o ther merhods. Th i , ,w'ork demonsfates that charge ladders can be made from a broarlarray of proteins dif fering in \ l \ \ 'and pl; among the 32 proreinr

we e, \amined, 25 y ie lded re so lved charge ladders (F igure 9 ,1\! 'e conclude that charge ladCers can be formed rel iably fronrprote ins that hav 'e va lues o i \ lW < 50 kD. Prote ins that havcI 'alues of l ! l \ ! ' > 50 kD ma1 form charge ladders when al lorvetl

to react * ' i th reagents 3-5, *h ich impar t approx imate ly -3 ,- .1 , and -6 un i ts o f negat ive charge, respect ive ly . for eac l r

e-NH3+ group acy' lated. A iew of the high lvl \ \ ' proteins rharw'e examined did not forrn resolved charge ladders, e! 'en wirhreegents 3-5, instead broad unresol l 'ed peaks w'ere observed.l ' ,

\ l 'e have not currently identi l ied the reasons for these fai lures

In the ceses * here charge laC,jers of proteins w'ere not formed,the value of the rat io of terms, L.Z1E/,1[d, betrveen dif .ferent acl lated rungs of these charge ledders mav hal 'e to be >

6/,11u to obtain suff icient sepuation by CE to resolve a charsc

ladder .

Acknowledgment. lvlurine carbonic anhydrase ! 'was a git ifrom Dr. David Christianson (L?enn). This work was supported

by the National Inst i tute of Health Grants Givl5l559 anc]Giv130367. The NNIR faci l i t res at Harvard were supported byNIH Grants l -S l0-RR04870-01 and CHE-881.1019. FAB- iv tSwere obtained by Dr. A. Ty' ler at the Harvard University N1assSpectrometry Faci l i ty supponed by Grants from NSF (CHE-

9020043) and NIH (SIO-RR06716) .

JA9723-r9 l

(301 We did not inc lude cenain commercia l proreins that gave ambiguousresul ts upon acylat ion. Nat i l 'e herokinase and pancreacic l ipase f romWorthington and hemoglobin f rorn Sigma revealed mul t ip le pea-ks by CE.The impunt ies present in these sanples made assignment of the some ofthe rungs oI the charge ladders of rhese proteins di f f jcul t due to thcot 'er lapping of peaks.

:3=

310 324

300

)< zz }29. J ^ ^

2ooo - t "g23

M W = 5 0 k 0- z-rd zo0 .^-

- - -Oiz-rY^)

t1?"t i tao c s15

r ,5] ,2t n v

A - C6 C t u t o C t

oo s

J 2'|