Crosslinking of Dam methyltransferase with S-adenosyl-methionine

Volume 2~0, n u m b ¢ r I. 1,17-151 F[~BS t)~)4?g :~; It,l l l,'¢d~r~ttot~ of I~urop~tan llio~h¢.~i~al Socicttet 0014~'~9~1¢~1,tSL;~0 A DONt.* O014~ 9 1 ~01 ~ $

M~felt 1991

Crosslinking of Dam methyltransferase with S-adenosyl-rnethionine C a r o l i n e WenzeP, M a x i m e M o u l a r d t*. A n d e r s L o b n e r - O l e s e # a n d W i l h e l m G u s c h l b a u e r t I Servi¢¢ de BioehhM¢ el d e GOIbtique Mol~cuhffr¢, B¢}t. 142. D~portemenr de Bialojlle Molet'uMtre el C¢lhdaire, Centre tl'Etude~ Nuel~aires de Saday. F.91191 Gif.sur. Vvett¢ Codex. France anti ~Deportment of Mirrohtology, The Technical Vniversio' of Dem~mrk, DK.2800 Lyngby, Denmark

R~¢civcd 30 November 1990; revised version received 16 January 1991 Hil~hly purified DNA,udcnlne nteth)dtranfferasc was irradiated in the presence or different eoncentrmi0nx oF radiolal~lled S.adetlotyl-methiontne tAdoMet) with a ¢onventio~ml Min~ralijht UV.lamp from several mlnute~ u~ to I h while incubatin~ in ice, Incorporation or radioactivity was monitored by electrophoresis of the crossllnk between S-ad©nosyl.methionine and Dam ntcthylase on SDS,polyacr~,'l:tmide lids followq:d by fluo. rolJr~0hv, Cro~slinkin$ reached n maximum in presence of l0 .I~M S.adeno~tl.methiontne: it was inhibited in the ~te~nee of substances which compctttively inhibit mcthyh~tion of DNA by Dam mcthylase, like sincfungin or S.adenosyl-homocystetne, but not in the presence of non.inhibitors like ATP or S.i~obut~,l.adeno~ine, The crosslink obtained was resistant against a wide range of even drastic e:oadition~ commonly used in protein and peptkle chemistry, ProteinL which do not bind S.adcnosyl.mcthioninc, us well as beat inactivated D:~m n~cthtdasc were not photolab~lled, Alter limited proteolysi~ th~ radioactive label appeared only in certain of the pcptides obtained From Western blots carried out with polyclonal antibodies produced a~lainst a synthetic peptidc correspondin$ in its sequence to amino acids 92 ol06 of the D~m methylase, the crosslinkinB of AdoMet could be tcntatiwly mapped at a position after amino acid 106, Dana methylasc" S.Adcnosyl-mcthioninc: Photoctosslinkin$: Sinefun$in', Protcol),sis

1. INTRODUCTION Dam methylase of E. coli recognizes specifically the sequence GATC and transfers a methyl group from AdoMet to the amino group of adenine [1]. Many systems involved in e.g. postreplicative mismatchrepair, replication, transcription of certain genes, transposition and segregation of the chromoson~e are regulated by the state o f methylation of the GATC-site via activating or inhibiting DNA-binding of other proteins [2,3]. Although the biological role of Dam methylase is partially understood, very little is known so far about its structure and mechanism. Dam methylase has a molecular mass of 32 kDa and acts as a monomer. Only one methyl group is transferred to the OATC-site per binding event [4] and the next 3 base pairs flanking the site both at the right and left modulate the rate of methylation [5]. Recently it was shown that AdoMet Correspondence address: W, Gusd~lbauer, Service de Biocl~imie et G6n6tique Mol~culaire, Brit. 142, D~partement de Biologie Molecutaire et Celiulaire, Centre d'Etudes Nuel~aires de Saclay, F-91191 Gif.sur-Yvette Cedes, France

*Present address: CERVI, H6pital de la Piti6, 75013 Paris, France Abbreviations: AdoMet,

S-adetaosyl-methionine; AdoHcy, Sadenosyl-homo_cysteine; SIBA. S-isobutyl-adenosine; CH]AdoMet, 3H~C-S-adenosyl.methionine; [a4ClAdoMet, S-adenosyll3,4)4C]rne thionine; Darn methylase, DNA-adenine-methyltransferase; DTT, dithiothreitol.

Published by Elsevier Science Publishers B.V.

serves, besides its role as a substrate for the methylation reaction, as an allosteric activator which stimulates Dam-binding to its target sequence GATC [6]. To fulfill these two different roles AdoMet binds with very different affinities to two different binding sites, as shown by tritium-NRM [7]. A widely used technique to undertake studies on the molecular mechanism of protein-substrate interactions is to link the components covalently and irreversibly by UV.irradiation [8,9]. Covalent complexes from methylases and other AdoMet-dependent enzymes with either the natural occurring substrate [10-13] or the photoactivable 8.azido derivative [14-16] have been described in the literature. In two cases it was possible to isolate and characterize regions o f DNA methyltransferases involved in AdoMet binding [16,17]. T o study the mechanism o f AdoMet binding to Dam methylase we have searched for suitable conditions to obtain crosslinks with Dam and its substrate, either radioactively labelled with 3H on the methyl group or with ~4C at positions 3 and 4 in the methionine, with low-dose UV-irradiation to assure only a minimal disturbance of Darn methylase structure during the experiment. We have obtained crosslinks with both substrates which were stable under a wide range o f conditions. Their specificity was proven by experiments taking into account several aspects: heat inactivated Darn methylase does not react with AdoMet; the crosslink reaction is inhibited by competitive inhibitors

147

of Ihc methyluxc rcackm; wurntisn of labclling WRX mchcd in the ~rcdcnce of 10 ,M AdoMet. lm-

mnnalngicnl detection of the protcalyrie Fragments obafter lilnitcd prsteolysis af crox#linkcd Dum mcthylanne was carried out with antibadies procluccd against a synthetic pcptidc corresponding to the amino acid ~eqncnt’c from position 92-166 OF Dam methylasc, This permitted a ramrive rnsppiny of the crosslink rcwztion site nftcr position 106. tain&

2. MATERIALS

AND METHODS

(pDBX)

[IS) 10 > 989’0purity as cstim:~tcd rronr silver-stained SDSpolyacrylamidc geld [ l9,20), Concentralion of purified cnrymc wax dctermincd x~~cctophoton~ctrlcnlly using ma - l.IBS 121) salcula~ctl from tke Trp and Tyr eontenl of Dam mcthylnsc (221. The cnxymr: had a specific activity of about 800 0 U/my 123). ‘H,C.S.adcnosyl-nicthioliinc((“H)hdoMet, IS CPmmul) and “CInbcll~d rrinbow.colot~rctl proiein nrolcculor weight msrkcra wcrc S.ndcnoayl-L[3,J.‘~C]lnetRisnine purchased from Amer&un. ([“C]AdoMct, 59 mCi/mmol) was from C.E.A.. S.adcnesyl. tncthionins (AdoMet) and S~ndenoayl~homocystcillr (AdoHcy) wcrc from Boehringcr Mnnnheim, wincfungilr and S-isobutyl-rdcnanine (SIBA) were gcncrour sifts from Dr M, CiCro (CNRS, Gif-sur-Yvette, France). All protcaxea used wcrc purchased from Uochringcl Mannhcim. 2.2 1 Crosslirlking

esperitnetrts

Indicated amounts of ‘H-labcllcd

or non-labcllcd AdoMet and its nnnlogucs were pipettcd into Eppundorf-tubes, MCI was rcmovcd by evnporfition in a Speed vat, This treatment doer not affect ~hc stnbility of AdoMct as proved by Dam mcthylnse activity tests carried out with dried substmtc, HaSOn was neutralized with BaCOj, and Bas01 removed by centrifugalion prior to lyophilization. Dam methylasc solution was added to the dried material and samples were preincubatcd on ice for IO min (sample volume was always 20 jtl). They were spotted into microtiter plates precooled on ice, Irradiation with UV-light wascarried out for the itldicated times placing a Mineralight handlamp (4 W, 254 nm) directly on the plate, while keeping the temperature of the irradiated samples always around 0°C. Buffer conditions were 50 mM K/Na.phosphate pH 7.6, 100 mM NaCI, 2 mM EDTA, 2 mM DTT, and S% glycerol. DTT is known IO hinder crosslink by its radical scavenging capacity, but did not affect the reaction in this case as verified by varying DTT concentrations. 2,3. Gel electroyhoresis, jluorogruphy Gel electrophoresis was carried out according to Laemrnli [I91 on 12.5 and 15% polyacrylamids gels, respectively. Gels were Coomassie-blue stained destained as usual, Gels of radioactively labelled enzyme were either stained/destained or fixed in 25% isobutylalcohol/lO% acetic acid prior to impregnation with fluorographic reagent (Amplify; Amersham). Dried gels were exposed to Kodak Xomat S films with oneCromex HI plus intensifying screen for 2-5 days. 2.4. Limited proteolysis The AdoMet/Dam methylase complex was subjected to enzymatic digestion under limited conditions. Buffer conditions wereas described under crosslink experiments, except for trypsin where an elevation of phosphate buffer concentration to 200 mM led to a remarkable stabilization of the generated fragment. Proteases were added to indicated enzyme/substrate ratios and reactions were carried out at 4OC for the indicated time intervals. Higher temperatures resulted in fast degradation of Dam methylase without leading to a reproducible pattern of proteolysls products on the SDS-gel indicating a destabiliza-

148

Protein3

wcrb

blcrtlcd

Tcmidry teturntrd Tri+HCI, subn~rd

from SKX~pt~lyttrryl~mid~ grli (hfillipere) using rhc Millipsrc

canto SDE clc~trul~lotting syi&dm (hfillipore). The mrnrbrrnc: WISI; with 3% bovine serum nlbunrinc in TRS-buffer (50 rnhl IS0 mM NnCI, pt-f 7.51, warhcd far I min in TlfS nntl ip1. for 2 h with polyclorr#l rrrtibodics (prtxlucad ttgainst n ryn-

polyvinyl.tlill~torid~memhr~~~~

Writ pcptidc with the rcqumsc YQFREEFNKSQDPFR ecrrrcsptlnding lo lhc ?rcyucncc 92-106 041Dam mtthylnrc, suprJlicd by &of, M. Marinur, Univ. tlr hlsuxnuliur*r~~s) in TBS plus 0.05% twcca 20,

Artcr thrcs 5.nrin washer in TBS, the membrane wax incubated for 2 h with alkaline plro~phntase cunjugrtsd swine irrrarunuglob~rlinr to rubbit imerrlnoglob:rlin~ (Dakopatn, Cupo~~lra~c~i). After three udditional waxhcs IIIC nrsmbranc wad atninctl with 5-broma&chluroa

Mntlulyl pl~srapl~~~c nnd N iwclbluc i~~rwnlium xcribcd by nlakc ct ill. [24].

3. RESULTS

AND

crxeniiully

nr tic.

DISCUSSION

3. I. S~~ecificirv of the crosdink The-yield o~cr&slink obtained

depended on the time

inrcrval of irradiation in a linear manner up to one hour, while longer irradiation did not lead co further labclling (Fig. 1). Under the conditions used in these experiments Dam methylase lost only half of its activity in one hour of irradiation (not shown). Irradiation induccd crosslinks between AdoMet and Dam methylase occurred only when both components were present in the solution in intact form. If either Dam mcthylase or AdoMet were irradiated prior to addition of the other component no labelling was observed (not shown). Other proteins which have no affinity for AdoMet, like phosphorylase b, bovine serum albumin, ovalbumin, carbonic anhydrase, soybean trypsin inhibitor, and CYlactalbumin were not labelled, nor was Dam methylasc inactivated by heating at 37°C for 15 min prior to addition of AdoMet and irradiation (not shown). In the presence of lOO-fold excess of unlabelled AdoMet the yield of radioactive crosslink obtained was drastically reduced while the presence of the same concentration of

5

min irradation 10 30 60

90

Fig. 1. Time dependence of crosslink reaction. IO PM [3H]AdoMet and 5,7 FM Dam was irradiated for various time intervals as indicated.

yield of crurnlinks dcpcndcd upsn the time si’ irmdinrion and rhe AdoMet conccntrution in a manner Aimilar to the one abxcrvcd using AdoMet a#-lnbellcd in the methyl ~raup (Fig, 4). The lower signal an the fi’ilmin Fig. 4 is due ta a specific uctivity used in these experimcntx about 250 times less compared with that of [‘HJAdoMel, These experiments establish the covalent binding of probably the entire AdoMct molecule to the enzyme. The crosslinkcd product obtained remained stable under conditions commonly used for peptide purification (precipitation with trichloritcetic acid, presence of hydrophobic solvents like acctonitrile) as well as prolonged incubntion with protenses in several buffers. Boiling of samples for 3 min in sample buffer f 191prior to loading them onto SS-gels did not affect the intensiry of the band observed in fluorogram compared with a non-boiled sample (Fig. 5, fluorogrnm, lane 4).

ATP did not &‘cct the crcsslink reacticxl (not thswn), To

study

the

clcpcndcnec

of

the croralinkv

on

PM) of Dam methylase was irradiared in the presence of various concentrarions of [JH]AdoMet from 3,4-N PM (Fig. 2). With less than 10 PM [%]hdoMct the yield of crosslink depended on substrnte concentrations, while higher concentrations did nor lead to further labelling, This in. dicates that cavnlcnt attachment of the substrate to the enzyme depended upon the formatian of a spesific eornplcx, Complex formation was inhibited by sincfungin and AdoHcy. Both substances are competitive inhibitors for the mcthylase reaction [6], while SIBA, a specific inhibitor of certain eucaryotic cytosine-methyltransfcrascs [25] had no influence on the kinetics of DNA methylation by Dam mcthylasc [6] and showed only very little influence on the yield of crosslink formation (Fig. 3). At the present state we can not decide whether crosslinking of AdoMet cakes place in the catalytic or in the allostcric site rf the enzyme [6,7]. To demonstratr .:!at radioactive labclling was due to the covalent atta\ ?r.cnt of the whole AdoMet molecule and not the result L, methylation of Dam rncthylasc by AdoMet, experiments were carried out with AdoMes ‘*C-labelled in positions 3 and 4 of methionine. The [2H]ArloMct

cancermarion,

110 pmoI(5.7

yM slnofungin 0

Fig,

3. Crosslinking

10

20

100 1000

I I10

For a preliminary mapping of the crosslink site, erosslinks of Dam methylase with AdoMet and [‘H]AdoMet were subjected to limited proteolysis with proteases of several specificities. Analysis was either by fluorography or immunological detection in a Western blot, carried out with polyclonal antibodies produced against a synthetic peptide corresponding to a large part of region II [27] of the amino acid sequence of Dam methylase (see legend of Fig. 5). The limited digestion by various proteases suggests that Dam methylase possesses a domain structure where the larger fragment is about 20-21 kDa. A second, 12 kDa fragment is visible on the Coomassie-stained gel in the slastase and papain digestion. The same pattern was observed in thermolysin digestion (not shown). Tryptic digestitin led to only one fragment of about 2.9kDa on the Coomassie-stained SDS-gel as well as on

t

pM StBA 20

100

1000

I10

FMAdotlcy 20

100

1000

in the presence of specific inhibitors for rnethylation, Dam methylase (5.7 $4) and [%]AdoMet 60 min in the presence of indicated amounts of competitive methylase inhibitors.

(10pM) were irradiated

for

149

the more sensiiive fluorography and Western blot. Papain digestion gave rise to two pepcides of CXL 21 and 19 kDa, which contain the AdoMet label and the antigenie site, Cleavage on either terminus is possible. Elascase digestion yielded 3 pepcides of ca 22, 19 and 17 kDa which contained the AdoMet label, but only the first reacted with the antibody. The 1% and 17-kDa pepcides must derive from the 22 kDa domain. Whether the 22 kDa pepcide contained the C-terminal (from amino acid ca $0 to 298) or the N-terminal (from amino a) bD

Coomaeaie

12

otein

3

b)

4

5

6

543

acid 1 ta fa 20(I), or neither (c-g, from amino acid 50-240) is presently unknown. A cut of 3 to 5 kDa to yield the 1% and 17.kDa peptides, respectively, however, eliminated the ancigenic site, It is thus probablc chat chc 22-kDa pcpcide is the C-terminal part and therefore, the Adokiec binding site is downstream of the antibody recognition site, i.e. afccr amino acid 106. Several peptides were observed in chc VS digestion. Only the largest one (27 kDa) contained both the AdoMct label and the antigenie site. Since the next blot

Weet62rn

2

1

c)

kD

12

Fluoaogrem

3

4

b

6

7

8

200 92 94 69

69

41

30

Fig. 5. Limited proteolysib of crosslinked enzyme with several proteases. Dam methylase crosslinked with [%]AdoMet (Fluorogram) or AdoMet (Coomassie-stained gel and Western blot) were digested with several proteases under limiting conditions. (a) Coomassic stain: lane 1, protein molecular weight markers; lane 2, Dam methylase; lane 3, tryptic digestion, enzyme/Dam = 1:lOO (w/w), 4T, I2 h; lane 4, V8 digestion, enzyme/Dam = I :20(w/w), 4”C, 36 h; !; ne 5, elastase digestion, enzyme/Dam = I : 100(w/w), 4”C, 12 h; lane 6, papain digestion, enzyme/Dam = I:100 (w/w), 4”C, 30 min. (Buffer conditions see section 2. (b) Western blot: poiyclonai antibodies used for this experiment were produced against a synthetic peptide corresponding to the sequence 92-106 (YQFREEFNKSQDPFR) of Dam mcthylase. Lane 1, Dam methylase; lane 2, tryptic digestion; lane 3, V8 digestion; lane 4, elastase digestion; lane 5, papain digestion. (c) Fluorography: lane 1, “‘C-labeHed rainbow-coloured protein molecular weight markers: lane 2, crosslinked Dam methyl?.se; lane 3, crosslink, incubated at 4T for 24 h; lane 4, crosslink, incubated and boiled in sample buffer for 3 min; lane 5, tryptic digestion; lane 4, V8 digestion; lane 7, elastase digestion; lane 8. papain digestion.

150

xmallsr peptidc (63 k!3~) did nc)t srpp~ar on the Wcstcrn blot, 1~1 contained the r#t!iorstivt! !&cl, it appenrr that thsre # a cat wirlrirr the nntibedy reccqnition sitetthe Cterminal gcptidc from position lW to 278 ban about 23 kI3a). The AdaMct lnbcllcd 17-kDu pcptidc iscderived frsm fhc ZJ=kDa frngmcnt. Similar rrrgumit?nta8s for rllc clastwse reaction nupgest that the AdoMet-lt~bclled site is located downstrcnm from amino acid lCJ6, i.e. between amino aeicla166 and ca 248, tt is noreworthy rhar this fragment eenrninr the ‘IWO regions If! end IV, llighly conserved in fill adeninc mcttlyltranafertexs (26,273. Kegian IV hns been euygested [29) to be the possible AdoMet-binding site. The present data eviclcntly de not allow furrlrer conclusions about the excjet location of chc AdoMct binding site, Purification of labcllec! peptides genrratcd by limited nnd complctc proteolysis, ns we!! IS their chnracterizntisn by N-terminal aquencing and fast atom bombardment spcctreseopy are under way. We arc also investigating the possibility to obtain crosslinks by irradiation with 8n excimer laser. nrs;no,v/crlblrr,,otlc WCWC~FIWZ~UI I$ Drs K.H. Hiilsmann and U. Hahn (Ucrlin) for supplying the overproducing strain MBtOI (pBQX). WC thank Prof. M. Marinus (Worccslcr, MA) fur the kind gif1 of Dam antibodies and 1% G&o @if) for sincfungin and SIBA, C.W. wns supported by Twining OrtInt ST2-0034 from rhc European Community.

REFERENCES [I) [2] [S] (4) [s]

Hatlman, S., Brooks, J.E. and Masurcknr, M. (1978) J. Mol. Biol. 126, 367-380. Mcsscr, W. and Noycr-Wcidncr, M. (1988) Cell 54, 127-135. Barras, F. nnd Marinus, M.G, (1989) Trends Gcnct, S, 139-143. Herman, GE. and Modrich, P. (1982) J. Biol. Chcm, 257, 2605-2612. Bergcrat, A., Kricbardis, A, andGuschlbaucr, W.(l989) 3. Biol. Chcm, 264, 4064-4070,

Dulban, P,. #randolin. G, Boulny, F., Hrbppc, .I. and V&xi&, PC (19013) Bitxhcmislry 27, ~141-5149. Yu, P.H. (19%3) f%.xhim, Iiiophyx. &~a 7.12, 517-534, P@larcr, M.A., Aloms\ny, S., Valcra, I., Marin, CAU, D. and hlaio, 3-M. (1984) Rioshcm. J. 223, 61-66. Hurst, J.H,, Billin&cy, M.t. and Lovznberg, W, (1982) Winohem. IIiophfa, Rcr. Commun. 122, 4%SO@. Siam, S. and FrIcdmRn, S. (1990) J. Dial. Chem. 36S,J27R-4283. Kalscr, I.I., Klrdiantx, D&t., loran Kirk, E.A. and Haley, B.8, (1913) J. 13iol. Chcm. 2%. 1747-1721. &m&r, J., Jcncks, R,A,,Xhani, S, and M,ruhrwr, RaG. (198)) J. Dial, Chcm. 261, 7697-77OO, Reich, ND. and Evrrcu. E.A. (199O) J, Viol, Chcm. 265, 13929”8934. Friedman, X., pcrsannl conrmunisalian. Hillsmann, K.&l,, Quass, R., BruncrI. H.-P., Georarlis, Y. and Hahn, 11. (1991) Gent, in press. Lacmmli, U.K. (1970) Nature 227, 680-6#J. Mcrril, C.R., Goldman, B,, Sedman, S.A. and Ebcrt, M.H. (1981) Scicnco 211, 1437-1431. Canror, C.R. and Schimmcl, P.R. (1980) Biophyaical Chemiary Part II, pp. 38O-381, W,il. Frcemnn, San Francisco. Brooks, J.E., Blumcnlhnl, R,M. and Gingeras, T,R. (1983) Nucleic Acids Ret. I 1. 837-8.51. Gcicr, G.E. and Mddrich, P. (1979) .I, Bial. Chem. 254, 1408-1413. Rlake, M.S., Johnston, K,H., Russ&Jones, C.S. and Gotschlich, EC, (1984) Anal. Uioehem, 136, 175-179, Blanchard, P., Dodic, N., Fourrcy, J-L,, G&c, RI,, Lawrence, B,, Mnlinn, H., Pnolamonacci, P., Vedcl, M,, Ternpete, C., Robert-G&o, M. and Ledcrer, E. (1986) in: Biological mcthylalion and druy design (Uorchard, R.T., Crevcliny, C.R. and Ucland, P.M. cds) pp. 435-446, NUITI~IKJ Press, Clifton, New York* Lnustcr. R., Kricbardis, A, arid Guschlbaucr, W. (1987) FEDS Lctt. 220, 167-176. Guschlbaucr, W. (1988) Gcnc 74, 211-214.