THE JOURNAL OF BIOLOGICAL CHE~~ISTRY Vol. 249, No. 17, Issue of September 10, pp. 568&5676,1974

Printed in U.S.A.

DNA Synthesis on a Double-stranded DNA Template by the T4

Bacteriophage DNA Polymerase and the T4

Gene 32 DNA Unwinding Protein

(Received for publication, February 20, 1974)

NANCY G. NOSSAL

From the Laboratory of Biochemical Pharmacology, National Institute of Arthritis, Metabolism, and Digestive Diseases, National Institutes of Health, Bethesda, Maryland 20014

SUMMARY

T4 DNA polymerase cannot use a nicked duplex DNA molecule as a template-primer for DNA synthesis, apparently because it is unable to displace the 5’ end of the strand paired to the strand serving as the template. Addition of the T4 phage gene 32 DNA unwinding protein (32-protein) facilities strand displacement and allows the T4 DNA polymerase to synthesize DNA using nicked duplex T7 DNA as the tem- plate-primer. The first product of this reaction is a rapidly renaturable DNA which is found to contain a few large branches when examined by electron microscopy. Later in the reaction there is also synthesis of a product containing predominantly alternating residues of dAMP and dTMP.

The DNA polymerase encoded by T4 bacteriophage gene 43 and the DNA unwinding protein encoded by the phage gene 32 (32-protein) are both absolutely required for T4 DNA synthesis in viva (l-4). In vitro the 32.protein lowers the melting tem- perature of double-stranded DNA helices and also facilitates the renaturation of single-stranded DNA, apparently by melting re- gions of intrastrand hydrogen bonding which interfere with the annealing of complementary single strands. Although the 32- protein does not promote substantial denaturation of helical DNA under physiological salt conditions at 37” as judged by hyper- chromic shift measurements (5), the protein does bind to some A-T-rich regions of the helical DNA under these conditions, re- sulting in local areas of denaturation observable by electron microscopy of the glutaraldehyde-fixed DNA 32.protein com- plex (6).

In vitro the T4 DNA polymerase can copy single-stranded re- gions of DNA, using as a primer either the 3’.hydroxy terminus of the opposite strand in the case of duplex DNA partially de- graded with a 3’- to 5’-exonuclease, or the 3’.hydroxyl end of a single strand which has formed a hairpin by annealing with a for- tuitously complementary section of the same strand (7). The 32-protein has been shown by Huberman et al. (8) to stimulate the reaction of the T4 DNA polymerase with each of these types of template. The effect of the 32.protein was greatest under conditions which promoted intrastrand base-pairing (low tem-

perature, high salt), suggesting that the 32-protein was melting regions of intrastrand hydrogen bonding in the single strand that was acting as the template.

While T4 phape DNA replication in tivo requires the copying of a linear duplex DNA, the purified T4 DNA polymerase cannot use an intact or nicked duplex DNA molecule as a template- primer (7, 9). T4 DNA polymerase can bind to duplex DNA, however, since the 3’. to 5’.exonuclease of this enzyme can begin hydrolysis at the 3’.hydroxyl chain end of duplex DNA (10, 11) or at a single-stranded nick in such a duplex (12). Furthermore, the T4 polymerase can repair a gap created by 3’ to 5’ hydrolysis at a nick in duplex DNA (12). It thus appeared that the failure of the T4 DNA polymerase to begin synthesis at a nick or at an end of duplex DNA results from the inability of the phage en zyme to displace the strand which is annealed to the template strand.

The studies reported here were undertaken to determine whether the T4 DNA polymerase could use a duplex template in the presence of the gene 32 DNA unwinding protein, which would be expected to promote the displacement of the strand paired to the strand serving as the template. The present study shows that nicked linear duplex molecules of T7 DNA do serve as a template for the T4 DNA polymerase if the 32-protein is also present. The first product formed is a DNA which is rapidly renaturable and is found to contain a few large branches when examined by electron microscopy. Later in the reaction there is also extensive synthesis of a polymer containing predominantly alternating residues of dAM1 and dTM1’.

MATERIALS AND METHODS

DNA-Unlabeled T7 DNA was prepared as described previ- ously (10). To prepare labeled T7 DNA, Escherichia coli 011’ (13) was grown overnight in M9 medium (13) containing 1.0 I.rg per ml ofthymine, diluted to 1.0 X lo* cells per ml in M9 containing [I%]- or [3H]thymine, grown to 5 X lo* cells per ml and then infected with T7 bacteriophage at a multiplicity of 0.5. T7 DNA was denatured with alkali as described by Studier (14).

E. coli DNA containing either 13Hlthvmine or 13Hlbromouracil .1 - I I

was prepared as described by Hanawalt (15, 16), using a thymine auxotroph, E. coli 70 U3 (from T. Breitman, National Institutes of Health). [3H]Poly[d(A-T)] prepared using E. coli polymerase I was the gift of M. Gellert (National Jnst,itutes of Health). Polymer concentrations are given as that of nucleotide phos- phorus.

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5669

Nucleotides--‘%-, 3H-, and 32P-labeled deoxynucleoside tri- phosphates, obtained from New England Nuclear, and unlabeled nucleotides from various commercial sources, were purified if necessary by DEAE-cellulose chromatography as previously described (10). BrdUTP was prepared by bromination of dUTP (Sigma Chemical Co.) as described for UDP by Howard et al. (17) and purified by DEAE-cellulose chromatography. Note that the correct Brt concentration for this method is 0.92 M rather than 0.92 mM given in the original description.’

Enzymes-32-Protein was purified from E. coli ER 21, a DNA endonuclease I-deficient derivative of E. coli B obtained from J. Eigner (18) (Washingt,on University, St. Louis), which had been infected with bacteriophage T4 (SP 62, am N155 (gene 42)) ob- tained from J. Wiberg (University of Rochester). The SP 62 mutation results in the overproduction of several enzymes which are produced early in infection if DNA synthesis is prevented as it. is in gene 42 mutants which lack dCMP hydroxymethylase (19). This phage was chosen because the cells were also used to obtain other-early enzymes. The SP 62 mutation does not result in the overnroduction of the 32-nrotein (19) or the T4 DNA nolvmerase.2 Thebacteria were grown-and infected in a 300-liter iermentor as previously described (20).

The 32-protein was purified by a modification of the method of Alberts and Frey (5). The cells (290 g) were suspended in a buffer containing 1.7 M NaCl, 1 mM EDTA, 1 mM o-mercaptoethanol, and 20 mM Tris-HCl, pH 8.1, broken by sonication, and the nucleic acids removed by polyethylene glycol precipitation (10% final concentration) as described by Alberts and Herrick (21). The supernatant solution was dialyzed and chromatographed on a DNA-cellulose column (100 ml nacked volume) and then on DEAE-cellulose as described (5). Forty per cent df the 32-protein after DEAE-cellulose chromatography-(5 mg) was loaded onto a 17-ml column of hvdroxvlanatite which had been eauilibrated with 20 mM Tris-H&, pg7.i, 0.1 M KCl, 10% glycerol;and 1 mM &mercaptoethanol. A gradient formed from 20 ml of this buffer and 20 ml of the same buffer containing 10% saturated ammonium sulfate was used to elute some contaminating nuclease, and the ammonium sulfate was then removed by washing the column with 20 ml of the equilibration buffer. The 32.protein was then eluted with a gradient formed from 50 ml of 0.04 M and 50 ml of 0.4 M potassium phosphate, pH 7.0, each containing in addition 10% glycerol and 1 mM mercaptoethanol (6). There was a shou- der of nuclease activity overlapping the earliest fractions of 32-protein eluted from this column. Only the fractions from the second half of the 32-protein peak, which were free of DNA exo- nuclease activity measured wit.h denatured, part,ially digested E. coli [“C]DNA and of DNA endonuclease as measured on neutral and alkaline sucrose gradients with T7 DNA were used in the experiments reported here. These fractions gave a single band on disc gel electrophoresis in the presence of sodium dodecyl sulfate (20). 32-Protein concentrations are given in terms of the monomer subunit with a molecular weight of 35,000 (5, 22).

T4 DNA Polymerase-The T4 DNA polymerase is eluted from the DNA-cellulose column used in the 32.protein purification by stepwise elution with 0.15 and 0.4 M NaCl. The pooled polymerase fractions were applied directly to a hydroxylapatite column (20 ml packed column) as described previously (20) and then purified by phosphocellulose chromatography as described previously for the polymerase from a temperature-sensitive mutant (23). The final preparation gives a single band on disc gel electrophoresis in the piesence 07 sodium dodecyl sulfate and is free of endo- nuclease activitv with native or alkali-denatured X or T7 DNA. It had a specifio activity of 15,300 units per mg (20).

Electron Microscop?/-Electron microscopy was generously carried out by Michael Gottesman (National Institutes of Health) using the formamide-spreading technique as described by Davis et al. (28). DNA from the neutral CsCl density gradient of the undenatured reaction mixture shown in Fig. 7C (renaturable DNA, Fraction 20; dAT-rich product, Fractions 25 and 26) was dialyzed as described in the legend of Fig. 6. The DNA was diluted to give a final concentration of 0.1 to 0.2 pg per ml in 0.1 M

Tris-HCI, pH 8.0, 10 mM EDTA, 40 I.rg per ml of cytochrome c, and 40% formamide. In some cases the dialyzed DNA fractions were heated for 3 min at 95” and then cooled quickly by dilution into the solution of Tris-HCl, EDTA, and cyt,ochrome c, followed by the addition of formamide. The DNA solutions were layered on a hypophase of 0.01 M Tris-HCl, pH 8.0, and 1 mM EDTA in 10% formamide. The DNA monolayers were picked up on par- lodion-coated copper grids, multidirectionally shadowed with platinum, observed in a Phillips 200 electron microscope, and photographed on glass plat,es. Molecular lengths were deter- mined with a man measurer as described by Davis el ~2. (28), using circular co\ E DNA molecules for -length calibration. Complexes of the 32-protein and DNA are dissociated by this technique. They can be visualized only if they are fixed with glutaraldehyde before the addition of formamide (6).

Aspergillus Nuclease Sl-Aspergillus nuclease S1 was purified in collaboration with T. Vogel (National Institutes of Health), by t,he method of Vogt (24) through the DEAE-cellulose chroma- tography. A unit silubilizes ld’rg of denatured calf thymus DNA in 10 min at 45”, under the conditions described by Vogt

insoluble polymer was measured by a modification of the method of Furano (25). A glass fiber filter (Whatman GF/C) of 12.5 cm diameter was ruled into a grid of 25 squares of 1.8.cm sides with a No. 1 pencil. The filter was wetted with 10% trichloroacetic acid cont.aining 10e2 M Nar pyrophosphate and then air-dried. Ali- quots of up to 25 ~1 of the reaction mixtures were pipetted onto successive squares. After all the samples had been applied, the filter was washed successively with 56ml of cold 10% trichloro- acetic acid. three times with 50 ml of ice-cold distilled water, and once with i0 ml of 95% ethanol, dried briefly under a heat iamp (10 min), and cut into squares which were counted in a toluene- based scintillation solution (2G). Simultaneous measurement of the incorporation of deoxynucleoside triphosphate int.o polymer and the DNA-dependent formation of deoxynucleotide mono- phosphate from deoxynucleoside triphosphate was determined by thin layer chromatography on polyethyleneimine cellulose (PEI- cellulose) as previously described (10).

Sucrose Den&ty Gradielrt Cenlrijugalion-Neutral gradients were formed from 1.8 ml each of 5 and 20yo sucrose in 0.1 M NaCl, 10 mM Tris-HCl, pH 7.4, 1 mM EDTA, and 0.1% sodium dodecyl sarcosinate. Alkaline gradients were formed from 1.8 ml of 5 and 20% sucrose in 1 M NaCl, 1 mM EDTA, and 0.1% sodium dodecvl sarcosinate, wit.h the pH of each solution adjusted to 12.2 with “NaOH using ‘a low sodium error glass electrode. After centrifuaation for 100 min at 54.000 mm at 10” in the Sninco SW 56 rotor, szmples were collected’ by inserting a 20-plZ disposable pipette through the gradient and then pumping from the bottom of the tube with an LKB peristaltic pump. Fract,ions flowing for equal time intervals were collected onto thick glass fiber filters (Whatman GF/B) which had been prepared as described for the GF/C filters used in the polymerase assay. A grid of 1G squares of 2.2 cm per side absorbed up to 0.2 ml of fluid per square. Acid-insoluble radioactivity was determined as described for the polymerase assay, except that four washes with water were used.

T7 DNA Strand Separation-T7 DNA strand separation by CsCl centrifugation was carried out as described by Englund (27), using poly(U,G) (1:l) obtained from Miles Laboratories. Fractions were collected as described for sucrose gradients.

Nearest Neighbor Analysis-This analvsis was carried out as described by Englund (27).

Stimulation of DNA Synt,hesis by S%Protein (24).

Enzyme Assays-Unless noted ot,herwise, polymerase reactions The DNA unwinding protein stimulates DNA synthesis ca- contained 4 mM MgC12,9 pM EDTA, 3 mM ammonium bicarbonate talyzed by T4 DNA polymerase using single-stranded T7 DNA buffer, pH 8.0, and bovine serum albumin, 60~~ per ml, in addition to the deoxynucleoside triphosphates, DNA, and enzymes given

(Fig. 1.4) as shown previously by Huberman et al. (8). In the

in the individual figures. Radioactivity in trichloroacetic acid- experiment shown the 32.protein to DNA ratio was 0.65, as- suming that one monomer of 32.protein covers 10 nucleotides (5))

1 H. T. Miles, personal communication. and the temperature was 37”. At lower temperatures less DNA 2 Unpublished experiments. was made, but the percentage stimulation by the 32-protein was

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5670

B Native T7 DNA

-32 Proteil * I 2

HOURS

FIG. 1. 32-Protein stimulation of the incorporation of dATP into polymer with alkali-denatured T7 DNA and duplex T7 DNA. Reaction mixtures (25 ~1) contained 1.7 nmoles of T7 DNA, 1.25 pmoles of T4 DNA polymerase, 110 pmoles of 32-protein, and 15 nmoles of each deoxynucleoside triphosphate ([aH]dATP, 4 cpm per pmole) in addition to the components given under “Materials and Methods.” After the indicated times at 37”, 5-J aliquots were precipitated on GF/C filters as described under “Materials and Methods.” 0, omit 32-protein; l , plus 32-protein.

TABLE I

DNA synthesis with T7 DNA requires M-protein

Polymerase activity was measured as described under “Ma- terials and Methods” in reaction mixtures (20 ~1) containing 4 nmoles of each of the four deoxynucleotide triphosphates and DNA and enzymes as given below. 32-Protein was added before the polymerase. Incubation was for 60 min at 37”.

T7 DNA

pmoles

1360 1360 1360 1360 1360 4760 1360

~ 32-Protein

p?W&S

’ 0 2.2 4.5

11 45 45 45

T4 DNA polymerase dATP incorporated

pmolcs pntolcs

2.5 <4 2.5 33 2.5 74 2.5 155 2.5 439 2.5 808 0 <4

greater (data not shown) in agreement with the findings of Huber- man et al. (8).

With duplex T7 DNA as the template-primer, DNA synthesis occurs only if the 32protein is present, in addition to the T4 DNA polymerase (Fig. 1B). DNA synthesis increases with increasing concentration of the 32-protein (Table I) and requires the T4 DNA polymerase (Table I) and all four deoxynucleotide triphos- phates (Table II). The rate of synthesis increased 60% if the 32-protein was added 5 min before the addition of magnesium and the polymerase. There was no further increase by prolong- ing the period before the addition of magnesium and polymerase.

Although no net synthesis of polymer is observed with duplex T7 DNA in the absence of the DNA unwinding protein, there is extensive DNA-dependent conversion of dATP to dAMP (Fig. 2A) as previously reported (10, 11). The 32.protein stimulates this formation of deoxynucleoside monophosphate, in addition to allowing stable incorporation of deoxynucleoside triphosphate into polymer (Fig. 2, B and C).

TABLE II

Deoxynucleotide triphosphate requirement for synthesis with T7 DNA

Polymerase reaction mixtures (20 rl) containing four deoxy- nucleotide triphosphates (12 nmoles each), T7 DNA, 1.7 nmoles, 32.protein, 44 pmoles, and T4 polymerase, 2.5 pmoles, in addition to the components listed under “Materials and Methods,” were incubated for 105 min at 37”.

d’ITP incorporated

Complete. Omit32-protein.......................... Omit dATP, dCTP, dGTP. Omit dGTP, dCTP. Omit dGTF..

fmolcs

524 <2

8 24 32

-i A e -32 Protem

9 E T? DNA=690

I--- pmolesl5p

100

El

-1 I 2

HOURS

C 3cEc&e;o,32

/i

I r I I

I 2

FIG. 2. DNA synthesis and DNA-dependent conversion of dATP to dAMP with duplex T7 DNA. Reaction mixtures (20 ~1) contained 2.8 nmoles of T7 13H]DNA, 44 or 38 pmoles of 32- protein, 2.5 pmoles of T4 DNA polymerase, and 10 nmoles of each deoxynucleotide triphosphate ([r4C]dATP, 2 cpm per pmole) in addition to the components given under “Materials and Meth- ods.” The DNA, deoxynucleoside triphosphates, buffer, and 32-protein were incubated together for 5 min at 37”, before the addition of MgCl and T4 DNA polymerase. At, the times indi- cated, 5-J aliquots were applied on PEI-cellulose and chromato- graphed as described under “Materials and Methods.” The ratio of 32-protein to DNA shown in the figures was calculated assuming that one monomer of 32-protein would bind t.o 10 nucleo- tides (5). There was no degradation of template DNA during the incubation. l , dAMP formation; n , dATP incorporation into new polymer.

Characterization of Product with Duplex T7 DNA

Density Label-The product of the synthesis using duplex T7 [3H]DNA with [3*P]dGTP and unlabeled ISrdUTP is shown in Fig. 3. After equilibrium centrifugation in neutral CsCl, newly synthesized polymer (Fig. 3A) bands at the density of heavy DNA and at intermediate densities between the heavy and light DNA markers. E. coli DNA which has approximately the same GC content as T7 DNA was used as a density marker (Fig. 30)

(29). Most of DNA serving as template-primer bands at a density between that of t.he light and hybrid DNA markers, showing that it has become associated with newly formed product (Fig. 3A). There is no synthesis or change in the banding posi- tion of the template-primer DNA in the absence of either the T4 DNA polymerase (Fig. 3B) or the 32.protein (Fig. 3C). In al- kaline CsCl, some of the product bands at the position of heavy

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5671

Neutral CsCl Alkolme CsCl

IO 20 30 4q IO 20 30 40 FRACTION

FIG. 3. Neutral and alkaline CsCl equilibrium density gradient centrifugation of the product made from [32P]dGTP and BrdUTP with duplex T7 DNA as the template-primer. The polymerase reaction mixture (63 ~1) contained the constituents listed under “Materials and Methods,” 8.4 nmoles of T7 [3H]DNA (12 cpm per pmole), 15 nmoles each of dATP, dCTP, and [32P]dGTP (46 cpm per pmole), 30 nmoles of BrdUTP, 520 pmoles of 32-protein, and 15 pmoles of T4 DNA polymerase. Magnesium chloride and polymerase were added after 5 min at 37”. After 2 hours at 37”, the reaction was terminated by the addition of 25 ~1 of 0.1 M

EDTA. Chromatography of a 5-11 aliquot on PEI-cellulose thin layer showed that 1.0 nmole of dGTP was converted to polymer, and 2.7 nmoles were converted to dGMP. Neutral CsCl gradients contained 25 ~1 of the reaction mixture, 5.97 ml of a solution of 10 mM NaCl, 1 mM EDTA, and 1 mM Tris-HCl, pH 8.0, and 8.0 g of CsClz (Harshaw, optical grade), giving a refractive index of 1.4036 at 25”. For alkaline CsCl gradients, 25 ~1 of the reaction mixture were added to 0.6 ml of 0.4 M potassium phosphate buffer, pH 12.5.. After 10 min at room temperature, 5.38 ml of a solut,ion of 10 mM NaCl, 1 mM EDTA, and 1 mM Tris-HCl, pH 8.0, and 8.9 g of CsCl were added, giving a refractive index of 1.4090. Heavy (HH) and hybrid (HL) DNA from Escherichia coli grown with bromo[3H]uracil (Schwarz-Mann) and light (LL) DNA from E. coli grown in [zH]thymine prepared and isolated by CsCl equilibrium centrifugation as described by Hanawalt (15, 16) were used as density markers (Panel D). Mineral oil was added to fill the polyallomer tubes, and the solutions centrifuged for 66 hours at 37,000 rpm in the Spinco 50 Ti rotor at 20”. Fractions of 0.15 ml were collected and 0.1.ml aliquots precipitated as de- scribed under “Materials and Methods.” The CsCl concentration of the fractions was determined from the refractive index at 25”.

single-stranded DNA; the remainder band at a density charac-

teristic of single strands with both heavy and light residues. The template-primer DNA bands in the light and mixed heavy and

light regions of the gradient.

The presence in the alkaline gradients of both product and

template label in the position between the heavy and light marker strands is evidence for covalent attachment of at least some of the

product to the template-primer. The apparent absence of tem-

C -Polymerase

D -32 Protein

0 tl 0

1.8 o

8 2 1.7 u

6 % 1.6

o- I s 8= x i5

6 ’

4

2

FRACTION

FIG. 4. Product of the synthesis with duplex T7 DNA as tem- plate annealed with poly(U,G) and centrifuged in a neutral CsCl density gradient. The reaction mixture (200 ~1) contained the constituents listed under “Materials and Methods,” 34 nmoles of T7 [14C]DNA (1.1 cpm per pmole), 50 nmoles each of dATP, dCTP, dGTP, and [3H]dTTP (54 cpm per pmole), 548 pmoles of 32-pro- tein, and 102 pmoles of T4 DNA polymerase. Magnesium chloride and polymerase were added after 5 min at 37”. After 2 hours at 37”, 6.7 nmoles of dTTP had been incorporated into acid-insoluble product. An aliquot (100 ~1) of the reaction mixture was mixed with 10 Imoles of EDTA, 154 nmoles of unlabeled T7 DNA, 150 ~1 of poly (U, G) 1: 1 (1 mg per ml), and 10 ~1 of 30y0 sodium dodecyl sarcosinate in a total volume of 1.2 ml. The solution was heated at 100” for 3 min and cooled quickly in an ice water bath. A solution of saturated CsCl (5.3 ml) buffered with 12 mM Tris-HCl, pH 8.0, was added to give a refractive index at 25” of 1.404. The samples were transferred to polyallomer tubes, covered with mineral oil, and centrifuged at 10” for 34.5 hours at 44,000 rpm in the Spinco 50 Ti rotor. Fractions of 0.2 ml were collected and 50-~1 aliquots of these precipitated with trichloroacetic acid on GF/B filters as described under “Materials and Methods.”

plate label with the product banding in the heavy position does

not indicate de novo initiation of new strands since short template strands arising from uicks near the ends of the T7 DNA could

act as primers without introducing enough radioactive or density

label to be detected under these conditions. Size and Renaturability of Product-The two strands of T7

DNA can be easily separated by equilibrium centrifugation in

neutral CsCl after heat denaturation and annealing with poly- (U ,G) (30). When the reaction mixture from synthesis with

duplex T7 DNA as the template-primer was examined under

these conditions, only a small part of the product banded with the separated T7 strands (Fig. 4A). Most of the product banded at the density of duplex T7 DNA (Peak I, Fig. 4A), T7 DNA

with both helical and single-stranded regions (Fractions 17 to 22), or poly[d(A-T)] (Peak II, Fig. 4A). The T7 [%]DNA

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5672

serving as the template-primer banded as separated T7 strands as well as in the region where duplex T7 DNA or T7 DNA with both duplex and denatured regions would band. This change in the banding pattern of the template-primer is dependent on new DNA synthesis since in the absence of either the polymerase (Fig. 4C) or the 32-protein (Fig. 4D), there was no new synthesis and the template DNA banded as expected as the two separate strands of T7 DNA. The two peaks of product with light density (Peaks Z and ZZ) were also found when the reaction mixture was heated and cooled quickly in the absence of the poly(U , G) (Fig. 4B). The length of the strands containing newly synthesized product was examined by neutral and alkaline sucrose density gradient centrifugation of aliquots of the same reaction mixtures (Fig. 5). Fig. 6 shows sucrose density centrifugation of Peaks Z and ZZ from the CsCl centrifugation of Fig. 4A. The presence of nicks in the duplex T7 DNA used as the template-primer is evident from the alkaline sucrose gradient shown in Fig. 6C.

Since the banding position of DNA in neutral CsCl depends on both the base composition and the extent of single-stranded- ness, the results shown in Fig. 4 suggested that Peak Z was a renaturable DNA with a base composition similar to that of T7, while Peak ZZ was a polymer rich in A and T. This interpreta- tion was supported by the resistance of the products to hydrolysis by the single strand-specific S1 nuclease, and by the base com- position of the isolated peaks.

The S1 nuclease degrades single-stranded heat-denatured T7 DNA, but not duplex T7 DNA, at both 19” and 37” (Table III). Poly[d(A-T)] is relatively resistant to hydrolysis at 19” where there is extensive self-annealing, but is degraded to a large extent at 37”, which is close to its T,. After heat denaturation, Peak Z is hydrolyzed at both 19” and 37” to an extent which is greater than that of duplex DNA but less than heat-denatured ‘~7 DNA, as expected for a DNA with both helical and single-stranded re- gions. Peak ZZ is similar to poly[d(A-T)] in that it is hydrolyzed to a much greater extent at 37” than at 19”.

Base Composition-Separate reaction mixtures each containing two labeled and two unlabeled deoxynucleoside triphosphatcs and unlabeled T7 DNA were used to determine the base composition of the product. Fig. 7 shows neutral CsCl equilibrium centrifu- gation of the untreated reaction mixtures, and of the reaction mixtures after heat denaturation in the presence of poly(U,G). The undenatured product banding at the same position as duplex T7 DNA contains the four newly incorporated nucleotides in equimolar amoums. In other experiments (data not shown) all of the label from T7 [14C]DNA used as template was also found to band in this position. After heat denaturation in the presence of poly(U , G), there is equimolar incorporation of the four nucleo- tides in bands corresponding to the separated T7 DNA strands. Most of the newly incorporated dCT1’ and dGTP, however, bands with duplex T7 DNA (Peak I) or between the positions

FRACTION

FIG. 5. Neutral and alkaline sucrose density gradient cen- trifugation of the product of the synthesis with a duplex T7 DNA template. Aliquots (25 ~1) of the reaction mixture described in Fig. 4 were mixed with 25 ~1 of 0.1 M EDTA and 50 ~1 of Hz0 before layering on neutral sucrose gradients as described under “Mate- rials and Methods.” Aliquots (25 ~1) of the same reactions were mixed wit.h 25 pl each of 0.1 M EDTA, 2 M NaOH, and 4 M NaCl for 5 min at room temperature before layering on alkaline sucrose gradients.

Neutral Alkaline I I

PeTk I

T7 DNA

.;‘. IO 20 30 IO 20 30

FRACTION

FIG. 6. Neutral and alkaline sucrose density gradient cen- trifugation of the products isolated by CsCl densit,y gradient centrifugation. The fractions comprising Peaks I and II in Fig. 4A were pooled and dialyzed against a solution of 10 mM Tris- HCl, pH 7.5, 50 mM NaCl, and 1 mM EDTA. Samples for the neutral gradients contained 75 ~1 of the dialyzed fraction and 25 ~1 of 0.1 M EDTA. Samples for the alkaline gradients which cont.ained 75 pl of the dialyzed fractions, 25 pl of 0.2 M EDTA, and 50 ~1 each of 2 M NaOH and 4 M NaCl were incubated for 5 min at room temperature before layering on the gradients.

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5673

TABLE III

Renaturability of product measured by resistance to S1 nuclease

DNA was denatured in 0.015 M NaCl (70 pl) by heating for 3 min at 95” and cooling quickly in an ice bath. Hydrolysis by the S1 nuclease at 19” was measured in reaction mixtures (0.1 ml) con- taining either T7 [W]DNA, 500 pmoles, [3H]poly[d(A-T)], 450 pmoles, or the dialyzed fractions (see Fig. 6) from Peaks Z 01’ ZZ in Fig. 4, 5 ~1, in addition to 30 mM sodium acetate, pH 4.6, 0.1 M

NaCl, 1.5 rnM ZnS04, 5yo glycerol, 5 pg of calf thymus DNA (soni- cated, then heated and cooled quickly), and 20 units of S1 nuclease. After 90 min at 19”, the entire reaction mixture was transferred to a GF/B filter, which was treated as described under “Materials and Methods.” Degradation at 37” was measured for 60 min with 12 units of S, nuclease. The per cent solubilization was deter- mined by comparing the recovery of acid-precipitable DNA in the presence and absence of the S1 nuclease.

DNA

Temperature of nuclease treatment

190 I

37”

‘?Zo soluble

Peak I. 6 18 Peak I, heat-denatured. 22 29 Peak II.......................... 8 58 Peak II, heat-denatured. 19 75 T7 DNA. 2 10 T7 DNA, heat-denatured. 99 99 Poly[d(A-T)], heat-denatured.. 26 85

TABLE IV

Nearest neighbor analysis of A-T-rich product

The reaction was carried out and the products isolated by CsCl density gradient centrifugation as described in Fig. 4, except that [a-32P]dATP or [or-32P]dTTP and unlabeled T7 DNA were used. The fractions corresponding to Peak ZZ in Fig. 4 were dialyzed for 48 hours against 0.15 M NaCl, 5 mM Tris-HCl, pH 7.5, 1 mM EDTA, and then for 48 hours against 10 mM NaCl, 5 mM Tris-HCl, pH 7.5. The dialyzed fractions were digested to I’-deoxynucleoside monophosphates with micrococcal nuclease and spleen phospho- diesterase and the nucleotides separated by paper chromatog- raphy using saturated ammonium sulfate-l M sodium acetate- isopropyl alcohol (80:18:2) as described by Englund (27).

Product

Nucleotide incorporated

dAMP dTMP dCMP dGMP

% lotal

[ol-3*P]dATP [or-S2P]dTTP.

occupied by duplex and single-stranded T7 DNA in agreement with the experiment shown in Fig. 4. The lightest peak found both before and after heat denaturation contains predominantly newly incorporated dAT1’ and dTTI’, and very little dCT1’ or dGT1’ (Fig. 7, A ,B, and C). Its banding position is similar with that of poly[d(A-T)] which bands at 1.678 g per cm3 (31). Near- est neighbor analysis of this product (Table IV) shows that it is composed almost entirely of alternating adenine and thymine residues. The formation of this product is a late event, since it is not present after a 15.min reaction (Fig. 70). It is not formed in the absence of the T7 DNA template-primer (data not shown).

Electron Jlicroscopy-The products of the reaction with T7

DNA, the T4 polymerase, and the 3!-protein were examined in the electron microscope by Michael Gottesman after fractionation

I .6

i

: 52

if’ 0

IO 20 30 IO 20 30 FRACTION

FIG. 7. CsCl equilibrium density gradient centrifugation of the product of the incorporation of each of the four deoxynucleo- side triphosphates with duplex T7 DNA template. The four reaction mixtures (75 ~1) contained the constituents shown under “Materials and Methods,” 13.4 nmoles of T7 DNA, 205 pmoles of 32.protein, 38 pmoles of T4 polymerase, and 20 nmoles of each deoxynucleoside triphosphate including [3H]dTTP (19 cpm per pmole) and either [W]dGTP (6.3 cpm per pmole) (Panel A), [W]dATP (5.9 cpm per pmole) (Panel B), or [W]dCTP (8.4 cpm per pmole) (Panels C and D), as indicated in the figure. Mag- nesium chloride and polymerase were added after 5 min at, 37”. After 120 min at 37”, 3.2 nmoles of dGTP and 5.9 nmoles of dTTP had been incorporated in Reaction A; 5.5 nmolcs of dATP and 5.6 nmoles of dTTP in Reaclion B; and 3.0 nmoles of dCTP and 6.5 nmoles of dTTP in Reacliorb C. After 15 min 1.0 nmole of dCTP and 1.3 nmoles of dTTP had been incorporat,ed in Reaction D. An aliquot (25~1) of each reaction mixture was denatured with poly(U,G) as described in the legend to Fig. 4. Another aliquot (25 ~1) was mixed with 10 pmoles of EDTA and 10 ~1 of 30yo sodium dodecyl sarcosinate in a total volume of 1.2 ml. Saturated CsCl (5.3 ml) was added to both the native and denatured samples (refractive index 1.404), which were centrifuged at, 10” for 34.5 hours at 44,000 in t.he Spinco 50 Ti rotor. l , incorporation into polymer of [aH]dTTP; 0, incorporation into polymer of [“Cl- dGTP, dATP, or dCTP. Arrows indicate position of separated T7 strands as shown in Panel E. Z and ZZ designate the peaks corresponding to t,hose in Fig. 4.

of the reaction mixture by centrifugation in neutral <Xl. Frac- tion 20 from the centrifugation of the undenatured reaction mix- ture shown in Fig. 7C, which contained 60% T7 DNA template and 40’% newly synthesized DNA with all four nucleotides, was a mixture of linear double-stranded molecules, and linear mole- cules cont.aining a few long branches (Figs. 8 and 9A). Of a total of 56 such molecules traced, 12 contained one branch, 1 contained two, and 2 contained three branches, with an average branch length of 2.0 pm. The ability of this fraction to reanneal

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5674

A Product - Peok I

m Branched

8. T7 DNA Template

2 4 6 8 IO 12 14 16 I8 20 22 24 26 28

LENGTH (pm)

FIG. 8. Length distribution of the product of the synthesis using dunlex T7 DNA with T4 DNA nolvmerase and 32nrotein and of the T7 DNA template. Electron micrographs of renatura- ble DNA (Fraction 20 from Fig. 7C (native)) and of duplex T7 DNA were made and measured as described under “Materials and Methods.”

with itself is shown clearly in electron micrographs made after heat denaturation and rapid cooling of the DNA. Under condi- tions where the template ‘I’7 DNA was quantitatively converted to single strands, 45 out of 99 molecules of denatured Pea/c I were single strands, 27 were linear double strands (hairpins), 10 were linear molecules with single- and double-stranded regions, and 17 were complex branched structures such as those shown in Fig. 9, C and D. The lightest peak from the same gradient (Fractions 25 and 26) which contained predominantly newly synthesized poly[d(A-T)], was highly branched when examined before or after heating (Fig. 9B). Partial denaturation in 82% formamide has been used to identify easily melted, A-T-rich, regions in duplex DNA (32). Under these conditions, T7 DNA and Frac- tion 20 had small regions of denaturation, while the product in Fractions 25 and 26 was entirely single-stranded, as expected from its base composition (not shown).

DISCUSSION

The T4 DNA polymerase is one of several proteins which are absolutely required for T4 phage DNA synthesis (l-3). While replication in vivo requires the copying of the linear duplex phage DNA, the purified polymerase itself is unable to use such a tem- plate in vitro (7, 9). This defect is not due to an inability of the enzyme to bind to double-stranded DNA, since the T4 enzyme catalyzes extensive hydrolysis and resynthesis at the 3’.hydroxyl ends and nicks of duplex DNA (10-12, 27). The polymerase synthesizes DNA by the addition of nucleotides to the 3’-hydroxyl end of the chain serving as the primer. Therefore, synthesis on a duplex template requires that the 5’-hydroxyl chain end of the

strand hydrogen bonded to the strand serving as template be removed either by physical displacement or by enzymatic deg- radation. The T4 DNA polymerase lacks a 5’. to 3’-exonuclease which could hydrolyze the opposing chain (12, 33). It is ap- parently also unable to physically displace this strand, since the enzyme cannot begin synthesis at the end or at an internal nick in a duplex DNA (7, 9). The phage enzyme differs in this re- spect from E. coli polymerase I which can both remove the 5’ end of the complementary chain with its 5’- to 3’.exonuclease activity, and physically displace it (9, 34). The 5’.exonuclease is not required for synthesis at a nick, since polymerase I from which the 5’.exonuclease activity has been removed by limited proteolysis (9) or by mutation (35) can still begin synthesis at an internal nick in duplex DNA.

The DNA unwinding protein encoded by the T4 phage gene 32 is apparently able to perform strand displacement both in viva and in vitro. In vivo, it has been found that in the absence of the 32.protein there are fewer branched recombinant DNA molecules whose formation requires displacement of part of one strand of each of the two duplex DNAs and annealing of com- plementary regions of the 2 recombining molecules (36-39). Furthermore, the 32.protein may also be temporarily displacing the strand complementary to the template strand at each grow- ing point during DNA synthesis (5), since phage DNA replica- tion in vivo requires that the 32-protein be present throughout the period when DNA is being made (4) in quantities such that there are approximately 170 monomer units of 32.protein per growing point on the phage DNA (5,40). In vitro the 32.protein binds to A-T-rich regions of duplex DNA, as shown by local areas of denaturation in electron micrographs of the glutaralde- hyde-fixed DNA 32.protein complexes (6).

The 32-protein increases the rate at which the T4 DNA polym- erase copies single-stranded DNA and repairs gaps created at the 3’ end of duplex DNA under conditions which suggest that it acts by interfering with intrastrand hydrogen bonding in the strand serving as the template (8). The 32-protein does not stimulate E. coli polymerases I (8), II (41), or III (41), which may indicate a specific interaction of the 32.protein and the T4 DNA polymerase. However, the possibility that this specificity is due to different template requirements for the host and phage enzymes has not yet been completely ruled out.

The studies reported here show that the 32-protein does en- able the T4 polymerase to use a double-helical template (Fig. 1). The products of the reaction are a rapidly renaturable DNA and a polymer consisting primarily of alternating adenylate and thymidylate residues. Most of the T7 DNA molecules present in the reaction do function as a template-primer, as shown by the density shift of labeled T7 DNA template-primer to the hybrid region of a neutral CsCl density gradient after synthesis with ISrdUTP (Fig. 3). In this experiment new product was measured by the incorporation of radioactive dGTI’, so that the DNA, but not the A-T-rich product, was followed. 111 alkaline CsCl, part of the product bands between the heavy and light single-stranded markers, suggesting covalent attachment of at least some of the product to the primer. The absence of tem- plate label in the product that bands with the heavy DNA in alkaline CsCl does not indicate de novo chain initiation since the heavy product may be covalently attached to primer strands too short to contribute radioactive or density labels. The experi- ments reported do not exclude the possibility that there is some chain initiation iu this system. However, the T4 DNA polym- erase cannot initiate chains using poly(dT) as template (42). Furthermore, there was no effect on the synthesis with duplex

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5675

I.. . .

“-.

FIG. 9. Electron micrographs of the products of synthesis with duplex T7 DNA. The samples from Fig. 7C (native) were prepared and photographed as described under “Materials and Methods.” All samples are shown at the same magnification. A, renaturable DNA (Fraction 20); B, d(A-T)-rich product (Frac-

T7 DNA of E. coli RNA polymerase, or the ribonucleoside tri- phosphates (data not shown), or both, although the synthesis of RNA may be involved in the initiation of T4 DNA synthesis in viva (43).

The rapid renaturability of the DNA product of the reaction with T4 DNA polymerase and the 32-protein was shown by its banding position in neutral CsCl after heat denaturation (Fig. 4), its resistance to S1 nuclease after heat denaturation (Table III), and its branched appearance in electron micrographs before denaturation (Figs. 8 and 9A), as well as its appearance as branched and hairpin molecules in electron micrographs made after heating and rapid cooling of the DNA (Fig. 9, C and D). Such renaturable structures result when the polymerase copies either the newly synthesized product or the displaced strand rather than the strand originally serving as the template, as shown in Fig. 10. Similar products are formed by E. coli polym- erase I using either linear duplex T7 DNA containing many internal nicks (34, 44) or double-circular DNA containing nicks (9, 35). Detailed schemes for generating such molecules have been presented previously (9, 44). The importance of copying the displaced strand in generating these renaturable structures

tions 25 and 26); C and D, renaturable DNA (Fraction 20), heat- denatured and rapidly cooled. The intact circles in A are co1 E DNA added as a length standard. The arrows point to the ends of a molecule the length of T7 DNA.

J

+--===I__J/___t 2 FIG. 10. Scheme for the formation of rapidly renaturable DNA

with a nicked linear duplex DNA template. Heavy line indicates newly synthesized DNA.

is suggested by the observation that a mutant E. coli polymerase I which is unable to hydrolyze the displaced strand, gives a more highly branched product than the wild type enzyme (35).

The product containing predominantly alternating adenine

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

5676

and thymine residues was characterized by its banding position in neutral CsCl before or after heat denaturation (Figs. 4 and 7), base composition (Fig. 7), greater resistance to S1 nuclease diges- tion at 19” versus 37’ (Table III), nearest neighbor analysis (Table IV), and the appearance of highly branched molecules in electron micrographs made before or after heat denaturation (Fig. 9) which were converted to single strands in 82’$$ forma- mide. This product contains small, but significant amounts of dGMP and dCMP (Table IV) and it is not made in the ab- sence of dGTP or dCTP (Table II), or in the absence of the 32. protein (Fig. 4) or the T7 DNA template (data not shown). It probably results from the creation of a poly[d(A-T)] template by continuous strand slippage (45, 46) after the copying of a region of the T7 DNA with a short stretch of alternating ade- nylate and thymidylate residues. The poly[d(A-T)] product appears late in the reaction (Fig. 7, C and D), and the quantity of this product compared with the renaturable DNA varied from experiment to experiment, and with different preparations of the 32-protein.

The strand slippage required for poly[d(A-T)] synthesis may be facilitated by the 32.protein. In the absence of this un- winding protein T4 DNA polymerase does not synthesize poly- [d(A-T)] de ROUO, in contrast to E. co2i polymerase I and micro- coccal DNA polymerase (7). When poly[d(A-T)] is used as the template-primer, the phage enzyme makes less than a single copy before both the template and new product are hydrolyzed by the 3’-exonuclease associated with the T4 DNA polymerase (7). The addition of the 32.protein to the T4 DNA polymerase greatly increases both the rate and extent of synthesis with a poly[d(A-T)] template.3

Although the 32-protein allows the T4 polymerase to use a duplex template, the product is not a complete copy of the tem- plate. At some point after synthesis begins, the polymerase stops copying the strand complementary to the primer strand and begins instead to copy either the primer strand itself or the displaced strand, producing a rapidly renaturable hairpin struc- ture (Fig. 10). Presumably this is prevented in oioo by one or more of the other protein factors required for T4 DNA replica- tion.

7. GOULIAN, M., LUCAS, Z. J., AND KORNBERG, A. (1968) J. Biol. Chem. 243, 627-638

8. HUBERMAN, J. A., KORNBERG, A., AND ALBERTS, B. M. (1971) J. Mol. Biol. 62, 39-52

9. MASAMUNE, Y., AND RICHARDSON, C. C. (1971) J. Biol. Chem. 246, 2692-2701

10. HERSHFIELD, M. S., AND NOSSAL, N. G. (1972) J. Biol. Chem. 247.3393-3404

11. ENGLUND, P. T. (1971) J. Biol. Chem. 246, 3269-3276 12. MASAMUNE, Y., FLEISCHMAN, R. A., AND RICHARDSON, C. C.

(1971) J. Biol. Chem. 246, 2680-2691 13. STUDIER, F. W. (1969) Virology 39, 562-574 14. STUDIER, F. W. (1969) J. Mol. Biol. 41, 189-197 15. HANAWALT, P. C. (1967) Methods Enzymol. 12A, 702-708 16. STEIN, G. H., AND HANAWALT, P. C. (1972) J. Mol. Biol. 64,

393407 17. HOWARD, F. B., FRAZIER, J., AND MILES, II. T. (1969) J.

Biol. Chem. 244, 1291-1302 18. EIGNER, J., AND BLOCK, S. (1968) J. Viral. 2, 320-326 19. WIBERG, J. S., MENDELSOHN, S., WARNER, V., HERCULES, K.,

ALDRICH, C., AND MUNRO, J. L. (1973) J. Viral. 12, 775-792 20. NOSSAL, N. G., AND HERSHFIELD, M. S. (1971) J. Biol. Chem.

246,5414-5426 21. ALBF,RTS, B. M., AND HERRICK, G. (1971) Methods Enzymol.

2lD, 198-217 22. CARROLL, R., NEET, K., AND GOLDTHWAIT, D. A. (1972) Proc.

Nat. Acad. Sci. U. S. A. 69, 2741-2744 23. HERSHFIELD, M. S. (1973) J. Biol. Chem. 246, 1417-1423 24. VOGT, V. M. (1973) Eur. J. Biochem. 33, 192-200 25. FURA~O, A. V. (1971) Anal. Biochem. 43, 639-640 26. NOSSAL, N. G., AND SINGER, M. F. (1968) J. Biol. Chem. 243,

913-922 27. ENGLUND, P. T. (1972) J. Mol. Biol. 66, 209-224 28. DAVIS. R. W.. SIMON. M.. AND DAVIDSON. N. (1971) Methods

Enz‘jmol. 21: 413-428 ’ ,

29. SHAPIRO, H. S. (1968) in Handbook of Biochemistry (SOBER, H. A., ed) 1st Ed, pp. H-33, H-49, The Chemical Rubber Co., Cleveland, Ohio

30. SUMMERS, W. C., AND SZYBALSKI, W. (1969) Virology 34, 9-16 31. ERIKSON, R. L., AND SZYBALSKI, W. (1964) Virology 22,111-123 32. WOLFSON, J., DRESSLER, D., AND MAGAZIN, M. (1971) Proc.

Nat. Acad. Sci. U. S. A. 69, 499-504 33. COZZARELLI, N. R., KELLY, R. B., AND KORNBERG, A. (1969)

J. Mol. Biol. 46, 513-531 34. KELLY, R. B., COZZARELLI, N. R., DEUTSCHER, M. P., LEH-

MAN, I. R., AND KORNBERG, A. (1970) J. Biol. Chem. 246, 3946

35. HEIJNEKER. H. L.. ELLENS. D. J.. TJEERDE. R. H., GLICKMAN,

Acknowledgments-I would like to thank Doctors John Wiberg B., VAN DORP, V., AND ‘POUW~LS, P. H’. (1973) Mol. Gen:

and Joseph Eigner for the phage and bacteria used in this study, Genet. 124, 83-96

36. TOMIZAWA, J., ANRAKU, N., AND IWANA, Y. (1966) J. Mol. Mr. Douglas Johnson and Mr. David Rogerson, Jr., for growing Biol. 21, 247-253

the bacteria, and Dr. Michael Gottesman for taking the electron 37. KOZINSKI, A., AND FELGENHAUER, Z. Z. (1967) J. Viral. 1.

micrographs. 1193-1202

38. BROKER, T. R., AND LEHMAN, I. R. (1971) J. Mol. Viol. 60, REFERENCES

1. EPSTEIN, R. H., BOLLE, A., STEINBERG, C. M., KELLEN- BERGER, E., BOY DE LA TOUR, E., CHEVALLEY, R., EDGAR, R. S., SUSMAN, M., DENHARDT, G. H., AND LIELAUSIS, A. (1963) Cold Spring Harbor Symp. Quant. Biol. 28, 376-394

2. DEWAARD, A., PAUL, A. V., AND LEHMAN, I. R. (1965) Proc. Nat. Acad. Sci. U. S. A. 64, 1241-1248

3. WARNER, H. R., AND BARNES, J. E. (1966) Virology 28, 10&107 4. ALBERTS, B. M., AMODIO, F. J., JENKINS, M., BUTMANN, E. I).,

AND FERRIS, F. L. (1968) Cold Spring Harbor Symp. Quant. Biol. 33, 289-305

5. ALBERTS, B. M., AND FREY, L. (1970) Nature 227, 1313-1318 6. DELIUS, H., MANTELL, N. J., AND ALBERTS, B. (1972) J.

Mol. Biol. 67, 341360

39. 40. 41.

42.

43.

44.

45.

46.

131-149 BROKER, T. R. (1973) J. Mol. Biol. 81, 1-16 SINHA, N.. AND SNUSTAD, D. P. (1971) J. Mol. Biol. 62,267-271 SIGAL,’ N.; DELIUS, H., XORNBERG, T., GEFTER, M. L., AND

ALBERTS. B. (1972) Proc. Nat. Acad. Sci. U. S. A. 69.3537- 3541 ’ . ’

HUANG, W. M., AND LEHMAN, I. R. (1972) J. Biol. Chem. 247, 31393146

SPEYER, J. F., CHAO, J., AND CHAO, L. (1972) J. Viral. 10, 902-908

SCHILDKRAUT, C. L., RICHARDSON, C. C., AND KORNBF,RG, A. (1964) J. Mol. Biol. 9, 2445

KORNBERG, A., BERTSCH, L. I,., JACKSON, J. F., AND KHORANA, H. G. (1964) Proc. Nat. Acad. Sci. U. S. A. 61, 315-323

OLIVERA, B. M., SCHEFFLER, I. E., AND LEHMAN, I. R. (1968) J. Mol. Biol. 36, 275-285 s F. D. Gillin and N. G. Nossal, unpublished experiments.

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from

Nancy G. NossalDNA Polymerase and the T4 Gene 32 DNA Unwinding Protein

DNA Synthesis on a Double-stranded DNA Template by the T4 Bacteriophage

1974, 249:5668-5676.J. Biol. Chem. 

  http://www.jbc.org/content/249/17/5668Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/249/17/5668.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on July 18, 2018http://w

ww

.jbc.org/D

ownloaded from