THE JOUBNAL OF BIOLOGICAL CHEMl8TRY Vol. 249, No. 19, hue of October 10, pp. 6244-6249, 1974 Phted in U.S.A. A Holoenzyme Form of Deoxyribonucleic Acid Polymerase 111 ISOLATION AND PROPERTIES* (Received for publication, April 22, 1974) WILLIAM WICKNER AND ARTHUR KORNBERG From the Department of Biochemistry, Stanford University School of Medicine, Slawford, California 9@05 SUMMARY A new form of DNA polymerase 111 (pol HI), termed holo- enzyme, has been purified to apparent homogeneity from gently lysed Escherichia cofi. Three forms of pol I11 have now been recognized on the bases of their distinctive physical Characteristics and their capacity to utilize a predominantly single-stranded template: (a) pol III, a dimer of 90,000-dalton subunits, which is inactive on single-stranded circular DNA (SS DNA); (b) pol In*, a higher polymer of pol 111, which is active on SS DNA only in the presence of copolymerase III* (cop01 In*), a 77,000-dalton polypeptide; and (c) holo- emyrne, a 330,000-dalton tetramer composed of the two pol I11 subunits and two cop01 III* units, which is distinguished from pol III or pol III* by its activity on SS DNA in the absence of added copol III*. The holoenzyme is dissociated into pol III* and copol III* by chromatography on phospho- cellulose; pol III* is converted to pol I11 by heating, dilution, or aging. Like pol III*, the holoenzyme requires ATP to form an initiation complex with the primer template. DNA polymerase 111 (pol 111) (l), a product of the dnaE genc (2), is inactive in the conversion of 4x174 and 3113 single- stranded viral DNA (SS DNA) to the double-stranded rcplicative form. However, a more complex form of the enzyme, DNA polymerase III* (pol HI*) (3), can catalyze the conversionin the presence of an additional protein, copolymerase III* (copol HI*). Starting with an RNA-primed SS DNA1 template, the first step in this process is the formation of an initiation complex (4) con- sisting of the primed template, spermidine (or DNA-unwinding protein), ATP, and the two proteins, pol III* and ropol III*. In this step, ATP is split into ADP aIid inorganic phosphate. Once the complex is formed, replication of SS DNA to double- stranded replicative form requires neither ATP nor copol HI*. Pol III* cannot be replaced by pol 111 in this reaction. Pol IlI* differs from pol I11 in its physical as well as catalytic properties and appears to be a multimeric form of pol I11 (3). * This work ww supported in part by grants from the National Institutes of Health and from the National Science Foundation. W. Wickner was supported by a National Cystic Fibrosis Research Foundation fellowship and is currently a Fellow of the Mellon Foundation. The abbreviations used are: SS DNA, single-stranded circular DNA; pol I, DNA polymerase I; pol 111, DNA polymerase 111; pol III*, DNA polymerase III*; holoenzyme, DNA polymerase 111 holoenzyme; copol lII*, copolymerase III*. We have now found a third (and presumably more native) form of pol 111. This form, termcd pol I11 holoenzyme, is a tetramer of 330,000 daltoris containing the two pol I11 subunits and two copol III* polypeptides. Replication of SS DKA templates by pol I11 holoeiizyme does not require added copol IIl* and is similar to that catalyzed by mixtures of pol III* and copol HI* in its sensitivity to antibody against copol 11I* and its dependence on ATP. Phosphocellulose chromatography dissociates the holoenzyme into pol HI*, a tetramcr of the dnaE polypeptide, and cop01 III* (3). Pol III* is thermosersitive when isolated from dnaE mutant cells (3). Pol III* can be converted irre- versibly to pol 111 by aging, dilution, or mild hcat treatments. MATERIAL8 AND METHOVS Materials-Materials were obtained as before (3,4). 3H-labeled +X174 was a gift of 8. Michal Jazwinski of Stanford. Human hemoglobin was prepared by washing packed erythrocytes in 1 M NaCl solution and then lysing them with 5 volumes of water; debris was removed by centrifugation (20 min at 40,000 x y) and by filtration through DEAE-cellulose in 0.15 M NaCl. Pol I was the Fraction VI1 enzyme of Jovin et al. (5). RNA polymerase was purified by a minor modification of the method of Babinet (6). Other sources were: rabbit muscle lactic acid dehydrogenase from Sigma, sperm whale myoglobin from Mann, and beef liver catalase from P-L Biochemicals. Holoenzyme Assay-Holoenzyme was assayed at 30" in a 25-p1 reaction mixture containing 3 pl of a deoxynucleoside triphosphate mixture (40 mM MgCI2, 0.15 mM [CZ-~~P]~CTP at 100 cpm per pmole, and 0.4 rn~ each of dATP, dGTP, and dTTP); 5 pl of assay buffer (10% sucrose, 50 mM Tris.HC1 (pH 7.5), 50 mM NaCl, 40 mM di- thiothreitol, RNA-primed +XI74 SS DNA (2.0 nmoles of nucleo- tide), prepared as described previously (3); 80 p~ ATP; 4 mM spermidine.Cl and Escherichia coli phospholipids (1 pl of a 4-mg per ml suspension in water). After a 5-niin incubation, acid- insoluble S*P-nucleotide was determined by filtration (7). One unit of holoenzyme is defined as the amount catalyzing the incor- poration of 1 nmole of nucleotide per min at 30". Phospholipid-Phospholipid was extracted from an E. coli H5G0 cell suspension (see below) by the method of Bligh and Dyer (8). After removal of chloroform under a nitrogen stream, the phospholipids were suspended by sonication for 30 s at 0" in 10 mM dithiothreitol. When neutralized with Tris base, the suspension could be stored at 0" for at least 4 months without loss of its abil- ity to support holoenzyme activity. Glycerol Gradient Sedimenlation-Glycerol gradient sedimenta- tion was performed in polydlorner tubes which were boiled for 4 hours in a solution of bovine serum albumin (1 mg per ml) and EDTA (10 mM), washed with water, and then boiled in several changes of 1 mM EDTA. Gradients were composed of 3.6 ml of 35 to 50% glycerol, containing 25 mM imidazole acetate, 20 mM dithiothreitol, and 1 mM EDTA. Sedimentation was at -5" for 9 to 12 hours at 55,000 rpm, in a Beckman SW 56 rotor. Fractions 6244 6245 of 2 drops were rollected through x 21-gauge needle which punc- tured the bottoms of the tubes. Growth uj Cells-E. coli 1.15(iQ wm groan ntr previowly described (3). Cell suspensions (5 X 10'~celIn per ml; 0.2 g of wet pmte per nil) were neutralized with Tris base just hefore freezing in liquid nitrogen and storing in 15-ml aliquots at -20'. rrcitdne 0.028 I4 49 240 440 FUACTIW WUMBER FRKTIW fnwTiw Fir. 1. Chromatography, filtmtion, cmd sedimentation of pol 1x1 holoenzyme. A, 1)EAE-cellulase. Fraction Il (5 ml, 21 tng, 350 units) was applied to R Ill.;A15-cellulose dumn @ X 6 em) in Btiffer C arid eluted with u. linear salt gradient (150 ml, 0 to 0.3 s KaC1, 6-ml fractions). B, gel filtration. Fraction 111 C.24 nil, 3 mg, 180 unite) was concentrated to 1 ml with ammonium sulfate (see "Materials and Methods") and applied to a column (1.5 X 12 cm) of Rio-Gel A-5m equilibrated with Buffer A. C, glycerol gradient sedimentation. Fraction IV (8.8 ml, 0.4 mg, 120 units) w~as coticetitrated with ammonium sulfate and sedi- mented in a glycerol gradient (see "Materials and Methods"). Aliquutv f2rl) of errch fraction were assayed fur huloenzymc with- out further addition or with addition of either 16 units of cop01 111" or 0.5 unit of pol 111.. The direction of sedimentstion is in- dicated by an atrow in this and subsequent figures. 6246 1 Experiment KO. / Iloloenzyme I FIG 2. Sodium dodervl sulfnte gel electrophuresiv of troloen- ayme. Fraction c` holocnzyrne (IO rg in 100 rl) pas heated for 30 sodium dodccyl ~ulfate-l%~ 8-nrercaptctethn- Tris system of Jovin Y/ a?. (91 m dodecyiuulfrte in tho .go1 arid contained 1274, amylamide and in the reservoir bl 0 257; btsacrylnmidr and were stained with Coom the niethod of Weher and O~born (10) Ilestsir scnnircd at, 580 nni in n (;ilford recording speetrophotomer. Cop01 III*:pol 111 polypeptides __~-I--_-_ _- - __ I Weight ratio 1 Molar ratio M 2 4 6 ! 1 3 3 0.89 1.W 1.00 1.17 0.90 1 . of, ptiorcsi:! (Fig. 2). Two ~)romiirient protein bands were &en at 90,000 aircl 77,000 cialtons. Siirtilar gels sere loaded uith 2, 4, and 6 gg of the purififd holfwrr with C 'ooniami~ the , gels M en ings of the 90,000- and 77,000-dalton praks \+ere rut out and weighed. N'hw the aright ratios of ttir peeks uere corrected for tnolecular wight, nti ecpirnolar ratio for two golypeptidw ws ot)servecl (`I'zrble 11). The 90,000-daltori polypeptidr is prr- suinccf to br the pol 111 (dnaH) polyprptide (hee bt?low and Refs. 2 aid 3) and the 77,000-tlalton polypeptidc, cop1 11 I* (see Yig. 9). l`o dckrmine the molecular wight of holoenzyme, it was filtered through a 1%-Gel A-5rn roltimn and srdimrnted iii glycerol gradients. lioth gcl filtration (Fig. 3) mid g1yecrol yrathertt s4irneritation (Fig. 4) showed that holoenzyrne has a mttlecular weight greater than that of catalase (247,000) but less than that of NKA polymerase (4S0,OOO). Oil the basis of these data and the equirrtolar proportiow of the 90,ocrO- and 77,000- rlnltoii polypeptides (`l'able I I), the holoenzyme is presumed to be larger thmi a dimer (I67,00 daltons) atid srrialler than R hexamer (501,000 daltons), and ttitwforP most probably is a tetranmcr of (pol 111 polyycptide), (copol 111 * polypeptide)2 with a molecular weight of 334,000. Struclurr: oj Pol 111* urd Copol Ill*--Pol 1 lI* has been shown of !)0,000-rlalton subutiita aid to co-chromatograph on gel filtration u ith B-galactmsi `he unexpected breadth of thv pol III* prl filtratioit my rcflcct, the presence of oligoiiiers rrrngiiig from two bunits ur inorr. Glycerol grirditwt sedirneritat ion (Fig. a diineritation cocfficiunt of 7.25 S for pol Ill*. inalecular weights for a spherical protein; they arc roitsistent with air asymmetric molt~tilr of 3fi0,0,000, indimting that pol I1 I * is B tetrtirner of (MtOOO-daltun I)ol!peptides. Copol I I I* has a molwdur weight of approximately 77,000 (Fig. 6) aid is com- pwd of a single 77,000-daltori polypej)tidr (Fig. 7). Conoersion of Pol Ill* to Pol 11/---1%1 Ill* caii be ronverted to `afrnctrt , 11s judged by dretive loss of trrnplatc~a with lorrg single-5trandeti iit of ith at4 filtration profile tn that of ti\ Ihta from gel filtration and glycerol (1 , 3) intlivate a inalccular wcigtit for pol thmcr of the !40X),000-claltoii dnuE poly- These tlttta woulri giw quite difft Cntaly/ic* Projmfim oj Pol Ill Ilolom2ym~ /loloenzyme Recpires A TY4ur earlier studics of pol I I I * had shown an ATl' rrquircwcnt For thia enzyme when assayrd with ~opol Ill* on priinw tcinplates ~ith hrg, single-straded rtyions (4). `I`hih ATP rrcluircwrrit specifit- for the formation of e put I I I * iriitiation t*oinpIci, MI ewnt which prrc-eded HiYA chain c,longution, qwrniidittr or DN,l unninrling proteiii, ATP, and platc wcw dl found iwciwwy for formation of pol I I1 * initiation roirtples. Ihriiig this r( iori, ATI' MBS rleavd to ADP atid 15, 1% hich rcinaiued bound to thv roniples. Holocwaymc has also Ixcn found to IR clcpcritlerit OII ATP for DNA synthesis on primer templ~tes with lor~g, sirigle-stranded rrgioiili (`Fable 1111, but, like 1101 I1 I* and pol Ilf (I, 4), it doe5 not require ATP when filling short gap5 (datu not shown). Cop01 Ilf * FU~ZC~~QM in iloloenzyn~e--Unlikp pol 111 *, holo- tot require additional copol IfI* for syirthasis on I long gaps (Table IV); this activity of thr holo- wtwr, wnsitivr to nritilx-dy ngniiist, cop01 I I1 * (Table V), as might be rspe.ctett from thr prewure of cop01 lII* polypeptide8 in the holoenzyme molecule (Figs. 2 aiid 7). 1'01 I li*, copcil I I1*, a J)N.\-lhditig ag in (%%> optimd rate) or by other proteins ( < 10% optinial rate). Further studiw will be neceswy to detcrmine the rolr of phospholipids in this reaction. Other kinetic p8,rameters of holoertzynie were tfetcrmined (Fig. 10). lkspite its lubility and camples structure, holoenzyme can synthesize DNA at a liiieur rate for at least 5 niin (Fig. 10.4) and this synthesis is propcrrtionnl to added enzyme (Fig. 10B). Like pol 111, itzj activity is optimal at a very low Yalt concentration 6247 7 I FRACTION TRACI ION FIG. 3 (left). Gel filtration of holoenzyme. Pol 111 holoenzyme (50 units), SH-leucine-labeled +)i171 phage (3,000 cpm), polymerase holoenzyme (0.05 me), catalase (2 mg), W[ (Y2,OOO cpm), arid human hemoglobin (0.6 mg) were mixed (sample volume 0.3 ml) and applied to a Bio-Gel A-5m colttmn (1 X 12 em) equilibrated with 30% glycerol, 0.5 mg of bovine serum albumin per ml, 0.05 M Tris.CI (pH 7.5) 2Q mx dithiothreitol, 0.1 I animo- nium acetate, snd 1 mM EDTA at 4". Fractions (0.2 ml) were assayed for T, RNA polymerage (ll), catrrlase (A,&), hemoglobin (A ,ao), and DX.4 polymerase I iI ttoltrenaynie (m described under "Mnterials and Methods") One unit on the ordtnale scale hw the following equivalents: 95174, 1,ooO cpm of `H (0); RNA polymerase, incorporation nf 0.25 prnole of ar-id-insnluhle ribo- nucleotide (0 ); I)NA polymerawl* 111 lioloeitzyme, incorporatiuri of 10 pmoles of acid-insoliiblc nucleotide (or 0.01 unit HI^ defined under "Mnterids and Methods") (e 1; catdrcse, Ales; of 0.05 (A); hemoglobin, A13(L of 0.0% (shaded hetagon.?); and ATP, 1,000 cyrn of SH (hezagotls). FIG. 4 (cenler). Glycerol gradient sedimentation of holorn- ayme. Woluenzyme (50 units), catnlnse (0.35 nrg), and herrtoglo- I,,,,, IW.Mo lr g 77wo 1 c .ACTIC DEHYDROGENASE w P g 50,MO E d o 51oi52025~~~)1~~ FRACTION FIG. 6 (left). Glyeernl gradient sedimentation of ropol IIl*. Copol III* (10 pg), lactic acid dehydrogenase (250 pg), hemogloltirt (`200 pg), and myoglobin (100 pg) were mixed (in 200 4) and sedi- mented through a glycerol gradient (15 to 30% glycerol, 50 mM Tris-Cf (plI 7.5), 20 SnM dithiothreitol, and 1 m~ EDTA; the gradient wns 3.6 ml in a polyallomer tube in an SW 5(3 rotor. Sedi- mentation wns for 21 hours at 55,000 rpm at 4'. Fifty-one single drop fractions were collected through a 21-gauge needle from the bottom of the gradient and assayed for lactic acid dehydrogenase (Also, Fractions 2 to 8)' hemoglobin (Also, Fractious 20 to 361, myoglobin (Adlo, Fractions 38 to 431, arid copol 111' (asyayod tis described (3)). FIG. 7 (center). Sodium dodecyl sulfate gel electrophoresis of copol HI*. Copol III* (5 rg) w8s boiled in 1% sodium dodecyl sulfate-1% 8-mereaptoethanol €or 1 niin and subjected to electro- FUACIION bin (0.50 mg) were sedimented in a glycerol gradient (susnple vol- ume 0 15 mi, 3 6 ml, 35 to 50% glycerol gradient). Conditions and assay of holoenzyme sedimentation are described under "Ma- terials and Rlet8hoda." One unit oil t,he ordinate wale has the following equivalents: holoenzyme, irtcorporntion of 10 pmoles of acid-insoluble nucleotide or 0.01 unit as defined under "Material8 and Methods" (e); hemoglobin, Acto of 0.2 (A); and catalase, A,@* of 0.025 (0). FI~. 5 (right). Glycerol gradient sedimentation of pol III*. Glycerol gradients were as described under "Naterids and Illeth- o&." Pol Ill* (17 units), cata1ase (4.0 mg), and henioglohin (0.5 rng] in 0.2 nil wre Inyered on one gradient; %second gradient, run in the same rotor, tiad N sanipfc (0.2 nil) of hemoglobin (0.5 rug) urid pol 1 ((1 pg of Fructivn 5 (8)). The peak of ho~noglobirr was ut I`rachn 31 in each gradient; resufts from the two grirdicnts are plotted togethcr for simplicity. One unit on the ordrnak malt hw the following equivalents: hemoglobin, of 0.2 (0); pol I, incorporation of !20 pmoles of acid-insoluble nucleotide (A 1; catalase, At06 of 0.0% (a); iind pol IlI*, incorporation of 2 pmoles of acid-insoluhle nticleotide (0 ), 02 04 OB 08 @I TEMPLATE. I I phnresis in a 107; acryIamide-0.25% bisncrylnniide slab gel (Beck- man Microzone Cell, model It-101) for 3 hours at 50 ma ut room temperature. The gel arid reservoir buffers of Vifiuela et ul. (12) were used. h-klead proteios (J, U, h3, E, arid V) were subjected to oloctrophoresis in parallel as molecular weight standards. We thank Ur. dhersood Casjens, of this department, for help with this techaique. FIG. 8 (i-ight). Heat treatment of pol 111`. Pol HI+ (Fraction V) in 30% glycerol-50 znM Tris+Cl (pH 7.5), 50 mv NaCI, `20 nu% dithiothreitol, and 1 tnu EDTA (3) was heated nt 39" for the times indicated and LtsYayed on activated calf thynius DNA (1) or CJn RNA-primed +S SS UNA (3). Standard activity values (I@%) were 10 prnoles of nucleotide incorporated in 10 min at 30' on &E- tivated calf thymus DNA and 8 pmoleR of nucleotide incorpo- rated in 10 min at 30" on Ith'A-primed (PX SS DNA. 6248 TABLE 111 I€oloenzyme is A TP-dcpendenl Holoenzyme (0.035 unit) or pol IIK' (0.044 unit with saturating cop01 111,) was assayed on (dT)o-(dA)lroe &Y described (3), with or without 80 ~BI ATP. Fom of polymerase I Addition of ATP I DXA syathcsis _1_~----- . pnrd65 0.1 72 0.2 58 T.\iiiIx IV Iiolosnzywie -ES no& slimitlaled ?q cop01 111' Pol 111* (0.006 unit) and holoenzyme (0.W unit) were msayed on activated calf thymus DNA, on RNA-primed &X 85 DNA, or on (dT),. (dA)tjoo, rn previouafy dewrihod (3). Copol fII* (0.1 pg) was added to the amay where indicated. 1 1 DNA smthesLI on temalater nf TATl1.E b` Wobentprte ix serrsilire 10 antibody uguirtsi cop2 III* Holoenzyme (0.007 unit) was mmyed on RNA-primed &i SP DNA iia described under "Ivloterinls and Methods." Where indicated, holoertzyme WNS first mixed at 0' with an exce~y (%I fig) of anti-ropol III* (3). Enzyme assays on activated cdf thymus DNA were rniidurted accorciing to the mehd of Korn- berg urd Gelier (1). RNA-primed #X SS DNA . . . . RXA-primed #X SS DNA . . . . . Activated cnff thymus DNA.. . ~. . Activated calf thyrnue DNA.. . . . .I + BALO2.s 33 4 32 70 (Pig. 1OC) and within a broad pII range centered iiear 7.5 (Fig. IOU). UIScUB8I(xr` Ttirough studies of the rrpliccttiori of single-strattded AI 13 ui~d a174 viral DNA, we have observed two riels forms of I)NA polymerase 111, physicalIy arid fuiictionally distkt. from that described originally (I). The new forms have thP capacity of pol 111 to utilize a gapped duplex template but can, in addition, replicate a11 extensively single-straxided DNA. The form de- scribed in this report (Iioloenzyme) is probably newest, to the form of the polymerare that is functional in iso. It represents all of the pol 111 identified in a gentle lysate of cells and is a tetra- meric protein composed of the two 90,000-dalton polypeptides of pdymerase mid two 77,000-dalt~n polypeptides uf a copokymrrase (Pig. 11). The holoenzyme can be resolved into these eompo- '100 I I I t I 80 ii 5 t - 0 B' -/* I 0.2 0.4 0.6 0.8 1 .o 1.2 Fro. 9, Phospholipid dependence of hotoenzynie activity. Ho- loenzyme (0.W2 tinit) urus a~~iyed as described urider "Materials and Methods," except that the phospholipid concentration was varied. Lipid was total Escherichia colt lipid extract (0 ), E. colt pho~phatidyletltancil~nine (0 1, or ox broiti diphosphntidylglyc- era1 (a). The latter two lipids were generous gifts of Ifr. E. P Kennedy, of Harvard Univereity. i 0 0 LIPID (rna/ml) 20 iieiits by dwomatctgraphy on pho~phorellulose. The "core" polymerme avtivity [mviou~ly iwlatotl us tin oligonirr (probably a tetranwr) of the pol 111 pdppcpticlcs nw culltvl po1ymcra.w I I1 * (3) ; the copulymerase titis1 for 1101 I I1* activity on sin& stsrantfs \v*s culled topolymcrarp Ill*. Our uw or the term holoet~zyrnc for the cornphx of pol I I I * untl c.oyolymrrrise 111 * is based 011 the analogy with thr cornplt*s of thc core polynrerahe anti IT subunit which constitutes WSA polymerase (13). Like the u subunit, wpdymerasc 11 I * wws ni the inifsl htqp of foriniiig a rornples with the template but appears to be dispensable during tho rqdiratioti itself (4). ltrceiitly, Iluncitz and Wickmr (1 4) haw rrportetl the partial purification of a 150,000-dalton protein, termrd Ftrctor 11, wttivh nib DPU'A polymerase 111 to catalyze nded templates in the presence of sperm ing prot~in, und wpoi 1X1*. TMortuiintely, thc.ir data do not permit an evaluation of whether they usud pol I1 I, pol 111 *, or a mixture of hth. These two forms of pdymerasc could be &s- tingui4xd by gel filtration (3). Our preparations of pol 111 * aid the holoenzyme have been purified 40,000- aid 1 B,M)O-fold, respectively, to tqqmrent homogerwity , without our uncovertng a requircineiit for a protein such as Faotor I I. C`ormivably, such a fnetor could function in the conversion of poi III to pol Ifl', The catalytic activities of the several forms of pol ill iit thc replirrtf ion of the E. coli chromosome have not \wen clarified. In our studies of the structure of these enzymes during catalysis (0, polymerase. aid copolymerasc units form a complex on the primer ternplate during initiation. Still uncertain are the st& chiometry of the proteins and their spatial arrangement in this 6249 m I" Nhcl ImMI - DU FIG. 10. Properties of holoenzyme. A, holoenzyme (0.12 unit) ww assayed in a 250-pi reaction mixture as described under "hla- terials and Methods." Aliquots (%MI) were withdrawn for deter- mination of acid-insolnhte nucleotide. B, holoenzyme was as- sayed for 5 min at 30" irr 25-4 reaction mixtures as described under "hfaterials and Metho&." C, holoenzyme (0,0!! unit) was as- aoyed as desctihcd under "Materiala and Methods," with addi- tional NaUl as indicated. D, holoenzyme (0.05 wit) was ~ssnycd as described under "l\laterixls and Methods," but with 10 mM Tris-CI (pH 7.5) replaced by 50 miii TrisPCf or imidazoIe.Ct at the indicated pll, complex and whether they Function as holoenzyme or in some other form of the pol 1II*-copol III* guir. Even though all of the polymerase 111 appears as pol III holoenzyme in the gentle lysate, this complex may be associated with additional repiica- tion component,s in the d. The futictional featturfs of holoenzyme and of poi I I I * compltk- ineliteti with copol Ill* are similar, inc~luding aa absolute dr- peridence on ATP and a sensitivity to anti-eopol HE*. An irn- HOLOENZYME PHOSPHOCELLULOSE CHROMATOGRAPHY COWL rn* a pwm FIG. 11. Hypotheticnl forms of polymerase 111. Pol represents the pol 111 (dnaE) polypeptide, arid copol represents the cop1 ITI* polypeptide. portant distinrtion is a drpcncirnrr of holoenzymc on the prm ence of phospholipids not consistently observed nith pol 1II*. The lipid factor may provide some structural substitute for the complex in which holoenzyme riorrnally fiiida itsell` in the rcfl. It EFEKENCEH 1. KORN~ERG. T.. ASD GKFTI~R. M. T,. {I9721 J. Rid. Ohem. 247. ., &?6$,-537d ' 2. GI:FTEK, M. L., HIKOIA, Y., Korrxmrto, T., WECHSLEH, J. A., AND BARNOKX. C, (1971) Proc. ,Val. dcad. Sci. U. S. A. 68. 31 50-3 153 (1973) Proc. Nnl. Ararl. Rci. U. S. A. 70, 1764-1767 3. WICGNLR, w., biEKMAS, H., C;ICIDI:K, K., hND klORNE%ERIi, a. 4. WXCKNER, W., AND KORKBERO, A. (1073) Proc. ?la[. Acad. Sei. 5. Jovrx, T. &I., ESDLLIHD, P. T., AXD Bcxrsca, L. I,. (1969) J. 6. BARINET, C. (1967) 13rcichern. Biophys. Res. Commun. 26, 639- 7. JOYIN, T. M., ENULUWO, 1'. T,, -\SD KMINREIIG, A. (1969) J. 8. BLIOH, E. G., ALD UYEK, W. J. (1959) C`ataad. J. Bzoehern. 9. .JOYIN, T. hX., CHRA~UIACH, A., AXD NAU(:HTON, hf. h. (1BFi) 10. WEBHI<, K., AZNU OBIKIHN, 5%. (1969) J. Biol. Cltertz. 244, 4400- 11. Cs.inin~n~ix, M., ~ND Bmc, P. (1062) Proc. Nnt. dcad. Sei. 12. VI?~C`YLA, E., ALGHANALI, I D., AND OCHOA, 8. (1967) Eur. J. 13. BURWS, R. R., TRAYEKS, A. A., DUNN, J. J., AND BAUTZ, 14. HunwiTz, J., AND WICKNER, S. (1954) Proc. Nat. Acad. Sc-r. I,', S. A. 70, 3679-3683 Biol. Cheni. %44, 2996 -3008 644 Rial. Chern. 244. 30WMO18 Phgsid. 37, 911-914 Anal. Biocherrt. 9, 351-369 5412 U. S. A. 48, 81-94 Btochem. 1, 3-11 E. K. F. (1969) Nature 221, 43-40, U. S. A. 71,6-10