Enzymatic Synthesis of Deoxyribonucleic Acid

Full text

THE JOURNAL OF Bro~oo~ca~ CHEMIWRV
Vol. 239, No. 1, January 1964
Printed in U.S.A.
Enzymatic Synthesis of Deoxyribonucleic Acid
XIV. FURTHER PURIFICATION AND PROPERTIES OF DEOXYRIBONUCLEIC ACID
POLYMERASE OF ESCHERICHIA COLZ*
CHARLES
C.
RICHARDSON,? CARL
L.
SCHILDKRAUT,~
H.
VASKEN APOSHIAN,~ AND ARTHUR KORNBERG
From the Department of Biochemistry, Stanford University School of Medicine, Palo Alto, California
(Received for publication, July 18, 1963)
Since the first description of the purification of deoxyribo-
nucleic acid polymerase from Escheriehia coli
(l),
the procedure
has been modified for handling 20-kg quantities of cell paste and
for obtaining a preparation which is physically homogeneous
by several criteria. As a result of the purification, the effects
of known nucleases and of a hitherto unrecognized enzyme on the
priming capacity of DNA have become apparent. This report
describes the modified purification procedure in detail and several
properties of the reaction which were not apparent in studies
with cruder polymerase fractions. Succeeding reports will
describe (a) the properties of an esonuclease which persists in
the most purified polymerase fractions (2), (b) the purification
and characterization of a DNA phosphatase-exonuclease, which
is separated from polymerase in the last step of purification
(3, 4), and (c) the purification of a DNA polymerase from
Bacillus subtilis and a comparison of its properties with t.hose
of the E. coZi polymerase (5).
EXPERIMENTAL PROCEDURE
Materials
Unlabeled deosyribonucleoside triphosphates were purchased
from the California Corporation for Biochemical Research.
32P-Labeled deoxyribonucleoside triphosphates were prepared
as described previously (1). Deoxythymidine-2-r4C purchased
from the New England Nuclear Corporation was enzymatically
phosphorylated to dTTP as described elsewhere (6).
Calf thymus DNA was isolated according to Kay, Simmons,
and Dounce (7). Unlabeled and 3zP-labeled DNA (20 MC per
pmole of phosphate) was isolated from E. coli as described by
Lehman (8). Unlabeled and 3H-labeled DNA (25 PC per pmole
of phosphate) isolated from B. subtilis (SB 19) was a gift from
Dr. Walter Bodmer. DNA from the bacteriophage Xdg was
*This investigation was supported by research grants from
the National Institutes of Health of the United States Public
Health Service.
t Postdoctoral Fellow of the United States Public Health
Service. Present address, Department of Biological Chemistry,
Harvard Medical School,. Boston 15, Massachusetts.
t Postdoctoral Fellow of the National Science Foundation.
Present address, Department of Cell Biology, Albert Einstein
College of Medicine,New York, New York. --
6 This work was done durine the tenure of a Senior Research
Fellowship, United States Public Health Service. Present ad-
dress, Department of Microbiology, Tufts University School of
Medicine, Boston, Massachusetts.
a gift from Dr. Charles Radding. The dAT copolymeri was
prepared by the method of Schachman et al. (9), and the dGdC
homopolymer as described by Radding, Josse, and Kornberg
WY. “Activated” calf thymus DNA was prepared by treat-
ment with crystalline pancreatic DNase (Worthington Bio-
chemical Corporation) as described earlier (11). E. coli soluble
RNA was prepared according to Ofengand, Dieckmann, and
Berg (12). Concent’rations of DNA are expressed as equivalents
of nucleotide phosphorus.
DEAE-cellulose and Whatman phosphocellulose (P-70) were
products of Brown Company and W. and R. Ralston, Ltd.,
respectively, and were processed according to Peterson and Sober
(13). Hydrosylapatite was prepared by the method of Tiselius,
HjertCn, and Levin (14). Streptomycin sulfate was kindly
donated by Merck, Sharp and Dohme Company. Crystallized
bovine plasma albumin was obtained from Armour and Com-
pany. E. coli endonuclease (15) and E. coli exonuclease I (8)
were prepared according to Lehman. Rabbit antiserum to the
E. coli endonuclease was a gift of Dr. I. R. Lehman.
Methods
Assay of Polymerase-The assay measures the conversion of
i4C- or 32P-labeled deosvribonucleoside triphosphate into an
acid-insoluble product.. “Two assays were used for the purifi-
cation of polymerase, a calf thymus DNA-primed assay used in
the early purification procedures (Fractions I through VI.
Table I) and a dAT-primed assay used in the subsequent purifi
cation steps.
Assay A: “Activated” Thymus DNA-primed Assay-The
incubat.ion misture (0.3 ml) contained 20 pmoles of glycine
buffer, pH 9.2, 2 pmoles of MgCh, 0.3 pmole of 2-mercapto-
et,hanol, 40 mpmoles of “activated” thymus DNA, 10 mpmoles
each of dTTP, dCTP, dGTP, and (r-32P-dATP,2 and 0.02 to
0.15 unit of enzyme.3 The incubation period was 30 minutes
at 37”. The reaction was stopped by chilling and the addition
1 The abbreviations used are: dAT copolymer, copolymer of
deoxyadenylate and deoxythymidylate; dGdC, polymer con-
sisting of homopolymers of deoxyguanylate and deoxycytidylate.
2 &2P-dTTP, &2P-dCTP, &2P-dGTP, and dTTP-1% labeled in
carbon 2 of thymine have also been used in routine assays during
the course of enzyme purification, with identical results.
8 Dilutions of enzyme for assay were routinely made in a solution
composed of 0.05 M Tris-HCI, pH 7.5, 0.1 M ammonium sulfate,
0.01 M 2-mercaptoethanol, and bovine plasma albumin, 1 mg per
ml.
222
This is an Open Access article under the CC BY license.

January 1964 Richardson, Schildkraut, Aposhian, and Kombery 223
of 0.5 ml of cold 1 N perchloric acid; the acid-insoluble radio-
activity was determined by the glass filter paper assay described
earlier for a glucosyltransferase assay (16). In the assays
utilizing 32P-labeled triphosphates, the filter paper was dried
on a planchet and the radicactivity was determined in a gas
flow counter; with 14C- or 3H-labeled substrates, the dried glass
filter was placed in a liquid scintillator solution and the radio-
act,ivity was determined in the Packard Tri-Carb liquid scintil-
lation counter. The filter paper assay results were identical
with those obtained by the centrifuge assay described earlier
(1).
Assay B: dA T-primed Assay-The incubation mixture
(0.3 ml) contained 20 pmoles of potassium phosphate buffer,
pH 7.4, 2 pmoles of MgC12, 0.3 pmole of 2-mercaptoethanol,
6 mpmoles of dAT copolymer, 10 mpmoles each of dTTP
and ar-32P-dATP, and 0.02 to 0.20 unit of enzyme.3 The reac-
tion was stopped and the acid-insoluble radioactivity was
determined as described for the thymus DNA-primed assay.
The radioactivity made acid-insoluble in Assay A was propor-
tional to the amount of enzyme added from 0.02 to 0.15 unit.
Thus, with the addition of 5, 10, 20, and 40 pg of Fraction I,
specific activities of 3.7, 3.9, 3.8, and 3.6 were obtained. The
radioactivity made acid-insoluble in Assay B was proportional
to the amount of enzyme added from 0.02 to 0.20 unit. Thus,
with the addition of 0.001, 0.002, 0.005, and 0.01 pg of Fraction
IX, specific activities in the range of 18,000 to 19,200 were
obtained. The control incubation, with enzyme omitted, con-
tained 0.1 to 0.3% of the added radioactivity in either Assay
A or B.
When native DNA was used as primer for the purified poly-
merase, 40 mpmoles of the DNA replaced dAT copolymer and
10 mpmoles each of dCTP and dGTP were added to the Assay
B mixture.
DNA Polymerase U&--A unit of DNA polymerase activity
is defined as the amount causing the incorporation of 10 mpmoles
of total nucleotide into the acid-insoluble product during the
period of incubation. Thus, the radioactivity incorporated in
the calf thymus DNA-primed assay was corrected to give the
total nucleotide incorporation based on a guanosine and cytidine
content of 43% (17) ; the radioactivity incorporated in the
dAT-primed assay was multiplied by 2. The ratio of activities
measured with each assay varies as a result of the decreasing
priming ability of thymus DNA with progressive purification of
the enzyme (see below).
Exonuclease Assays-The nuclease associated with the poly-
merase was assayed as described by Lehman and Richardson
(2). The reaction mixture (0.3 ml) contained 20 pmoles of
glycine buffer, pH 9.2, 2 pmoles of MgC12, 0.3 pmole of 2-mer-
captoethanol, 0.1 Fmole of EDTA, 10 mpmoles of 32P-labeled
dAT copolymer, and 0.05 to 0.30 unit of enzyme. The acid-
soluble radioactivity formed was measured as previously de-
scribed (2). One unit is defined as the amount producing 10
mpmoles of acid-soluble radioactivity in 30 minutes.
The E. coli DNA phosphatase-exonuclease assay and unit of
activity are described elsewhere (3, 4).
Other Methods-Protein was determined by the method of
Lowry et al. (18) after precipitation by cold trichloroacetic acid.
4 This definition of a polymerase unit differs from that presented
previously (1).
TABLE I
Puri&ation of E. coli DNA polymerase*
Fraction No. and stept
I. Extract.
297
20.0
II. Streptomycin.. 139 7.3
III. Autolysis. 119 1.6
IV. Ammonium sulfat’e 106 5.0
V. Acid precipitation. 100 3.5
VI. Ethanol.. 91 4.4
VI. Ethanol. 301 4.4 2,030 30
VII. DEAE-cellulose 227 5.5 7,600 23
VIII. Phosphocellulose 140 0.26 19,000 14
IX. Hydroxylapatite. 129 0.26 18,800 13
-
Unitst Protein
m&‘/ml
Specific
activity
uni1s/mg
pr01eiPI
3.8
18.5
73.2
148.0
200.0
590.0
Yield
%
100
47
40
36
33
30
* We acknowledge the valuable and expert assistance of George
Bugg in the purification of DNA polymerase.
t Steps I through IV were frequently carried out on a smaller
scale, and the fractions were stored until sufficient quantities were
accumulated (see the text).
$ The assay utilizing “activated” calf thymus DNA was used
for Fractions I to VI. The dAT-primed assay was used for the
subsequent purification (Fractions VI to IX). Use of the latter
assay for the crude fractions was not feasible owing to the rapid
destruction of dAT; thus an accurate evaluation of the relative
specific activity of the hydroxylapatite fraction and the crude
extract is not possible.
Deoxypentose was determined by the diphenylamine reaction
of Dische (19). Optical measurements in the ultraviolet region
were made with the Zeiss PMQ II spectrophotometer. Viscosity
measurements were carried out at 37” in Ostwald-type vis-
cometers with shear gradients of about 300 set-1.
RESULTS
Puri$cation of Polymernse
Unless otherwise indicated, all operations were carried out at
O-4”. All centrifugations were at 15,000 x g for 10 minutes.
The purification procedure and results of a typical preparation6
are summarized in Table I.
Preparation of ExtractsE. coli strain B was grown in a
medium containing 1.1% K2HP04, 0.85% KH2P04, 0.6%
Difco yeast extract,, and 1 y. glucose. The growth and harvest-
ing of the cells, in late log phase, was carried out by the Grain
Processing Corporation, Muscatine, Iowa. The cells were
frozen and stored at -20” until used. Partially thawed cells
(450 g, wet weight) were mixed with 300 ml of glycylglycine
buffer, 0.05
M,
pH 7.0, in a large Waring Blendor (5-liter capac-
ity) equipped with a cooling jacket and connected to a Variac.
Slow stirring was begun and after 5 minutes, 1350 g of acid-
washed glass beads (Superbrite, average diameter 200 EL, Min-
nesota Mining and Manufacturing Company) were gradually
added to t,he suspension. When the mixture appeared homoge-
6 The preparation summarized in Table I was carried out on
an 8%kg batch. Each purification step has been carried out on
smaller and larger scales; the volume of fractions, reagents, resin
bed volume of columns (column height kept constant), and gra-
dient volumes were changed proportionately.

224 Enzymatic Synthesis
of
DNA. XIV Vol. 239, X0. 1
neous, stirring was increased to approximately one-third of
maximal speed. After 20 minutes an additional 1200 ml of the
same buffer were added, and the homogenization was continued
for 10 minutes at reduced speed t#o prevent excessive foaming.
During the period of homogenization the temperature was not
permitted to rise above 12”. The beads were then allowed to
settle out and the broken cell suspension was decanted and
saved. An additional 800 ml of buffer were added to the glass
beads, and the residual broken cells were extracted by a lo-
minute homogenization at slow speed. The beads were again
allowed to settle out, and the supernatant fluid was decanted
and combined with the first supernatant fluid to give a final
volume of approximately 2000 ml (Fraction I, Table I). This
fraction may be stored at 0” for at least 2 weeks without loss of
activity.
Streptomycin Precipitation-To 10 liters of extract were added
10 liters of Tris buffer, 0.05 M, pH 7.5, containing 0.001 M EDTA;
then, with constant stirring, 1440 ml of 5% streptomycin sul-
fate6 were added over a 45minute period. After 10 minutes,
the suspension was centrifuged and the supernatant fluid was
discarded. The thick, sticky precipitate was transferred to a
beaker and 1000 ml of potassium phosphate buffer, 0.05 M,
pH 7.4, were added. The precipitate was suspended by slow
mechanical stirring for approximately 12 hours, and the final
volume was adjusted to 2500 ml by the addition of the same
buffer (Fraction II, Table I). Fraction II was stored at 0”
until sufficient quantities were obtained for the subsequent
procedures. There was no detectable loss of activity after
‘storage for 1 month.
AuteZysisEleven lit,ers of Fraction II were made 0.003 M
in MgClz by the addition of 65 ml of 0.5 M MgC12. The suspen-
sion was incubated at 30” for 7 to 12 hours, until 95% of the
ultraviolet-absorbing material at 260 rnp was rendered acid-
soluble? The autolysate was then chilled to 0” and the protein
precipitate that settled out during the digestion was removed by
centrifugation in a Spinco continuous flow refrigerated centri-
fuge at 28,000 x g at a rate of 120 to 130 ml per minute. The
supernatant (Fraction III, Table I) was stored at O”, and no
loss in activity was observed over a l-week period. During the
several hours of nucleic acid digestion at 30”, there was less than
10% loss of polymerase activity. However, if the incubation
was allowed to continue beyond the complete acid solubiliza-
tion of the nucleic acid, there was a further loss in enzymatic
activity, i.e. 10% in 60 minutes.
Am.monium Sulfate Fractionation-To 10 liters of Fraction
III were added 50 ml of 0.2 M EDTA and 50 ml of 0.2 M gluta-
thione. Over a 60-minute period, 3 kg of ammonium sulfate
6The amount of streptomycin sulfate required to precipitate
the nucleic acids and DNA polymerase varied from one prepara-
tion of E. coli extracts to another. Therefore, a trial precipitation
on a small scale was required to determine the optimal amount
of streptomycin sulfate. It was also observed on several occasions
that the 5% streptomycin solution was less effective in precipitat-
ing DNA and the enzyme after storage for 7 to 10 days at
4’.
For
this reason, the streptomycin sulfate solution was prepared just
before its addition to Fraction I.
7 At intervals of 1 hour, l-ml aliquots were removed and cen-
trifuged, and the optical density of the supernatant fluid after
suitable dilution in Tris-HCl buffer, 0.05 M, pH 7.4, was deter-
mined at 260 mp; a portion of the supernatant fluid was precipi-
tated with an equal volume of cold 1 N perchloric acid and the
optical density of the acid-soluble fraction was determined.
were added with stirring, and after 30 minutes at 4”, the pre-
cipitate was removed by centrifugation. To the supernatant
fluid an additional 1.15 kg of ammonium sulfate were added,
with stirring, over a go-minute period and, after 30 minutes at
4”, the precipitate which formed was collected by centrifugation.
This precipitate was dissolved in 1.1 liters of potassium phosphate
buffer, 0.02 M, pH 7.2 (Fraction IV, Table I). Fraction IV
has been stored for 3 years at -20” without loss of activity.
Acid Precipitation-Fraction IV (1.1 liters) was dialyzed for
12 hours against 35 liters of sodium acetate buffer, pH 5.55:
ionic strength 0.08, and the resulting precipitate was removed
by centrifugation. The supernatant fluid (Fraction V, Table I)
was then subjected immediately to ethanol fractionation.
Ethanol Fractionation-To 1.1 liters of Fraction V were added,
with constant, rapid stirring, 46 ml of 100% ethanol (-20”)
from a burette over a 60-minute period. The temperature of
the solution was not allowed to rise above -1” after the addi-
tion of approximately 10 ml of ethanol. After 10 minutes at
0”, the precipitate was removed by centrifugation at 0”. To the
supernatant fluid were added, with stirring, 110 ml of 100%
ethanol (-20”) from a burette over a BO-minute period. The
temperature of the solution was gradually decreased until the
final temperature was -4”. The solution was permitted to
warm to 0”, and after sitting for 10 minutes at 0” the precipitate
was collected by centrifugation at 0” and dissolved in 350 ml of
0.02 M K2HP04 containing 0.01 M 2-mercaptoethanol (Fraction
VI, Table I).
DEAE-cellulose Fractionation-A column of DEAE-cellulose
(16 cm2 X 10 cm) was prepared and washed with approximately
5 liters of 0.02 M K2HP04 containing 0.01 M 2-mercaptoethanol
and 0.002 M sodium EDTA, and then was equilibrated with
0.02 M K2HPOI containing 0.01 M 2-mercaptoet,hanol. The
350 ml of Fraction VI (1.5 g of protein) were applied to the
column at the rate of 250 ml per hour. The adsorbent was then
washed with 160 ml of 0.02 M K2HP04 containing 0.01 &r 2-
mercaptoethanol. The protein was then eluted at a flow rat,e of
250 ml per hour with pH 6.5 potassium phosphate buffers con-
taining 0.01 M 2-mercaptoethanol as follows: 160 ml of 0.05 M,
160 ml of 0.1 M, .and finally 160 ml of 0.2 M. The 0.2 M eluate
was collected in IO-ml fractions. Approximately 70% of the
enzyme applied to the adsorbent was obtained in the 0.2 31
eluate. The fractions which contained enzyme of specific
activity greater than 5000 units per mg of protein (approxi-
mately 60% of that applied to the adsorbent) were pooled
(Fraction VII, Table I). The final volume of the pooled Frac-
tion VII was 50 ml.
Phosphocellulose Chromatography-A column of phospho-
cellulose (15 cm2 x 21 cm) was prepared and washed with 5
liters of potassium phosphate buffer, 0.02 M, pH 6.5, containing
0.01 M 2-mercaptoethanol. With the same buffer, 50 ml of
Fraction VII (275 mg of protein) were diluted to 500 ml and
applied to the column with pressure at the rate of 500 ml per
hour. The adsorbent was washed with 160 ml of the above
buffer. A linear gradient of elution was applied with 0.02 M
and 0.3 M potassium phosphate at pH 6.5 as limiting concentra-
tions. The total volume of the gradient was 1500 ml, and 0.01
M 2-mercaptoethanol was present throughout the gradient.
The flow rate was 180 ml per hour, and 20-ml fractions were col-
lected. Of the applied activity, 75% was eluted in a peak
* pH was determined at 25” at an ionic strength of 0.08.

January 1964
Richardson, Schildlcraut, Aposhian, and Kornberg
225
between 2.7 and 3.8 resin bed volumes of effluent (0.17 to 0.25
M
potassium phosphate). The peak fractions containing enzyme
of specific activity ranging from 14,000 to 18,000 units per mg of
protein (65% of that applied .o the adsorbent) were pooled
(Fraction VIII, Table I). The elution of protein was followed
spectrophotometrically at 280 rnp, which permitted the rapid
identification of the enzyme peak, since the only major protein
peak in that region of the chromatogram contained the poly-
merase activity.
Hydroxylapatite
Chromatography-A column of hydroxyl-
apatite (1 cm* x 10 cm) was prepared and washed with 350
ml of potassium phosphate buffer, 0.02
M,
pH 6.5, containing
0.01
M
2-mercaptoethanol. Fraction VIII (20 ml containing
5 mg of protein) was dialyzed against 2 liters of the same buffer
and applied to the adsorbent with pressure at a rate of 20 ml
per hour. The adsorbent was washed with 12 ml of potassium
phosphate buffer, 0.05
M,
pH 6.5, containing 0.01
M
2-mercapto-
ethanol. A linear gradient of elution was applied with 0.05 y
and 0.3
M
potassium phosphate, pH 6.5, as limiting concentra-
tions. The total volume of the gradient was 130 ml, and 0.01
M
2-mercaptoet.hanol was present throughout the gradient.
The flow rate was approximately 5 ml per hour and 2-ml frac-
t#ions were collected. Of the activity applied to the adsorbent,
91 To was eluted in a single, sharp peak between 8.0 and 9.4
resin bed volumes (0.20 to 0.25
M
potassium phosphate) (Fig. 1).
The elution of protein was followed spectrophotometrically at
280 rnp, and the major peak was assayed for polymerase and
protein by the methods described. The specific activity of the
enzyme was constant across the peak, and the fractions of the
entire peak either were pooled and concentrated by pervapora-
tion as described elsewhere (3) or were stored without concentra-
tion (Fraction IX, Table I). In either case the enzyme was
quite stable when stored at 0” in potassium phosphate buffer,
0.15
M,
pH 6.5, containing 0.01
M
2-mercaptoethanol. At a
concentration of 2.5 mg per ml, the hydroxylapatite enzyme,
stored at 0” for 1 year, retained 80% of its initial activity.
Chromatography on hydroxylapatite of from 5 to 100 mg
of protein (Fraction VIII, Table I) has been carried out by
increasing the resin bed volume (column height kept constant)
and gradient volume proportionately.
Homogeneity
qf
Pwijied
Polymerase
Hydroxylapatite Chromatography-Chromatography of the
phosphocellulose fraction on hydroxylapatite gave rise to two
protein peaks (Fig. l), a major peak containing 90% of the
starting protein and a minor one containing approximately 2oJ,
of the initial protein. The major peak was symmetrical and
contained 91% of the polymerase activity applied to the ad-
sorbent. Furthermore, the specific activity of the enzyme,
when assayed with the dAT-primed assay, was constant across
the entire peak, an indication of high protein purity. Rechro-
matography of the protein peak containing the polymerase
activity under identical conditions gave rise to a single, sharp
symmetrical peak that again contained enzyme of constant
specific activity.
Starch
Gel ElectrophoresisElectrophoresis of the purified
polymerase (Fraction VIII) on starch gel at pH 8.8 in Tris-
citrate buffer (20) gave rise to a single migrating protein band.
When subjected to a vertical starch gel electrophoresis for 11
hours at 175 volts and 10 ma, with the bridge solution contain-
I I I I -0
*40 -
t
A
- .30 -
E
&
Czo-
i
_-
c3
‘o.lO-
6
Fm. 1. Hydroxylapatite chromatography of
E. coEi
DNA
polymerase. The phosphocellulose fraction (Fraction VIII)
(5 mg) was adsorbed and eluted from hydroxylapatite as described
in the text. The major protein peak (II) contained the polvmer-
ase activity (dAT-primed as&y) and was separated from the
minor protein neak (I) which contained the DNA nhosnhatase-
exonuclease activity (see the text). Each fraction contained 2.0
ml.
ing 0.30
M
boric acid-O.06 M sodium hydroxide buffer (21),
0.13 mg (0.13 ml) of polymerase migrated approximately 40
mm. After slicing and staining with Amido black,
a
single
band, 4 mm in thickness, was observed.
Sedimentation Pattern,-Schlieren patt,erns taken during the
sedimentation of the purified polymerase (Fraction IX) demon-
strated a single moving boundary (Fig. 2).
Ultraviolet Absorption
Spectrum-The ultraviolet absorption
spectrum was measured on the hydroxylapatite fraction (Frac-
tion IX) in 0.2
M
potassium phosphate buffer, pH 6.5, contain-
ing 0.01
M
2-mercaptoethanol; this solvent was used as a blank.
The absorbances measured were 0.272 at 250 ma, 0.359 at 260
mp, 0.581 at 280 mp, 0.312 at 290 rnl.c, and 0.048 at 300 rnp.
The ratio of absorbance at 280 rnl.c to that at 260 rnM is 1.6 and
suggests that less than 0.3% (by weight) nucleic acid is present.
Amino Acid
Composition-The amino acid composition of the
purified polymerase has been determined and is given in Table
II. These values, compared with those for the “stable” amino
acids of
E. coli
determined by Sueoka (22), show that the amino
FIG. 2. Sedimentation diagram of the purified polvmerase.
Schlieren patterns showing the sedimentation of -polymerase
(Fraction VIII) in a double sector cell 13. 50. and 84 minutes
after reaching full speed. Rotor speed, 59,780 ‘r.p.m.; tempera-
ture, 5.25”; initial concentration, 0.36 g per 100 ml; phase plate
angles, 70”, 65”, and 65’, respectively, in the three pictures; buffer,
0.1 M KCl-0.01 M 2-mercaptoethanol-0.005
M
potassium phosphate,
pH 6.8. We wish to thank Dr. R. L. Baldwin for this analysis.

226 Enzymatic Synthesis of DNA. XIV
Vol. 239, No. 1
TABLE II
Amino acid composition*
These values represent the average of two analyses of one sam-
ple of Fraction IX. The protein was hydrolyzed with twice
distilled 5.7
N
HCI at 105” in sealed evacuated tubes for 48 hours.
Amino acid analyses were carried out with a Beckman/Spinco
model 120 automatic amino acid analyzer. The values listed for
threonine and serine were not corrected for destruction during
acid hydrolysis.
Amino acid
Aspartic acid ............................
Threonine. ..............................
Serine ..................................
Glutamic acid. ..........................
Proline ..................................
Glycine ..................................
Alanine ..................................
Valine ...................................
Methionine.
.............................
Isoleucine
...............................
Leucine .................................
Tyrosine ................................
Phenylalanine
...........................
Lysine ...................................
Histidine ................................
Arginine .................................
-
-
Molar content relative
to histidinet
5.10
2.89
2.18
7.49
3.35
3.59
6.04
3.35
1.53
3.13
6.45
1.95
1.51
3.39
1.00
2.80
* We are indebted to Professor C. Yanofsky for performing this
analysis.
t Cysteine, cystine, and tryptophan were not determined and
were not included in the calculation of relative amino acid con-
tent.
acid composition of polymerase is similar to that of the total
protein of E. coli.
Attempts at Further Purijication-Rechromatography of the
purified polymerase on DEAE-cellulose, phosphocellulose, and
hydroxylapatite under a variety of conditions has not resulted
in any change in the specific activity of the enzyme. Attempts
to crystallize the purified protein from ammonium sulfate have
thus far been unsuccessful.
Exonuclease Activity in Purified Polymerase Preparation
Persistence of Exonucleaae Activity--Present throughout
purification, and at a constant ratio to the polymerase, is an
exonuclease activity
(E.
coli exonuclease II) (Table III); charac-
terization of the latter activity is the subject of a succeeding
report (‘2). When the two activities were measured under
optimal conditions for each, the ratio of the rate of nucleotide
incorporation by polymerase to the rate of hydrolysis by the
exonuclease activity was 2.8 (Table IIIA) ; at lower pH values,
i.e. 7.0, the ratio is considerably increased (Table IIIB).
Examination of the phosphocellulose (Step VIII) and the
hydroxylapatite (Step IX) chromatograms gave no indication
of a separation of the two activities within the polymerase peak
of either chromatogram (Fig. 3). Attempts to purify the poly-
merase further have not altered the ratio of the two activities.
An independent purification of the exonuclease activity, by
means of procedures not common to the polymerase isolation,
resulted in a lOOO-fold purified enzyme having the identical
TABLE
III
Ratio
of
polymerase to exonuclease II during purificiztion
A. Optimal assay conditions were used to measure DNA poly-
merase and exonuclease II. The standard dAT-primed assay
(Assay B) was used to measure DNA polymerase activity. Exo-
nuclease assays were as described in “Methods”’ with the addition
of 5 wmoles of soluble RNA, 0.01 ml of 1:5 dilution of rabbit
antiserum to E. coli exonuclease I, and 0.05 to 0.30 unit of exonu-
clease II. The presence of the RNA and antiserum was necessary
only with Fraction IV and VI and did not affect the exonuclease
assay of the more purified fractions.
Polymerase fraction
Activity Ratio of polymer-
Polymerase ExonIuIc’ease
ase to e~~nuclease
u?&r/n1
IV. Ammonium sulfate. 710 236 3.0
VI. Ethanol. 1,900 705 2.7
VII. DEAE-cellulose. 43,100 16,500 2.6
VIII. Phosphocellulose.. 5,000 1,800 2.8
IX. Hydroxylapatite . 4,900 1,750 2.8
B. Polymerase and exonuclease activities were also measured
at pH 7.4 and 7.0. In these experiments, potassium phosphate
buffer at either pH 7.4 or 7.0 replaced the buffer routinely present
in the standard assay. When exonuclease is assayed at pH 7.4
or 7.0, its activity is decreased.
PH
7.4
7.0
* Fraction IX.
Activity Ratio of polymer-
6.0
10.0
fraction number
Fro. 3. Association of E. coli exonuclease II with polymerase
in the hydroxylapatite chromatogram. The fractions obtained
from chromatography of Fraction VIII on hydroxylapatite (see
Fig. 1) were assayed here for exonuclease II and polymerase
activity in the standard assays (see “Methods”).

January 1964 Richardson, Schildkraut, Aposhian, and Kornberg
227
ratio of polymerase to exonuclease. ilttempts to alter selec-
tively either of the activities with urea or with heat resulted in
inactivation at the same rate [Fig. 4).
Removal of DNA P/wsphatase-Exonuclease-Although chro-
matography of the phosphoceilulose fraction (Fraction VIII)
on hydroxylapatite did not increase the specific activity of the
polymerase (Table I) with respect to either protein or exonu-
clease II activity (Fig. 3), an enzyme was separated from the
polymerase (Peak I, Fig. 1) which produced a striking effect on
the polymerase reaction. The 2’% of the total protein that was
eluted in Peak I of the hydroxylapatite chromatogram is an
enzyme identified as a DNA phosphatase-esonuclease (3, 4),
which increases the ability of a native DNA to serve as primer
in the polymerase system. As described in an earlier report
(23), this activating effect of the new enzyme apparently resides
in its ability to remove inhibitory 3’-phosphoryl groups which
terminate DNA chains (23).
Absence of Endonuclease in Puri$ed Polymerase-When aaP-
labeled native E. coli DNA was incubated with 140 units9 of the
purified enzyme (Fraction IX) at pH 7.4 for 60 minutes in a
standard polymerase reaction mixture lacking deoxynucleoside
triphosphates, there was an 8% drop in viscosity with a simul-
taneous release of 3% of the radioactivity in acid-soluble form
(Fig. 5). These values represented the hydrolysis expected as a
result of the exonuclease II activity discussed above (2). Con-
sistent with this interpretation was the result obtained at pH
7.0; there was a 4% drop in viscosity and the release of 1% of
the DNA as acid-soluble radioactivity. Further evidence that
this slow hydrolysis of native DNA was not due to E. coli
endonuclease (15, 24) was the absence of any inhibition by
soluble RNA (4 mpmoles), the specific rabbit antiserum (0.01
ml), or both together. In addition, assays of the biological
activity of B. subtilis (SB 19) DNA incubated with 140 units of
polymerase under assay conditions showed no detectable loss
of the Trya+ (indole) marker after 60 minutes, and only a 30%
loss after 6 hours of incubation at 37O.10 The assay for biological
activity is the more sensitive test for endonuclease activity
(2) and together with the other data indicates that the observed
hydrolysis of the native DNA was a result of the exonuclease
II activity.
Change in Priming Activity of Native DNA as Result
of Pur&ation of Polymerase
Early in the purification of the polymerase, a decrease of
activity was encountered when native DNA was used as primer
(Table IV). Through the ethanol fractionation (Step VI) this
decrease could be circumvented by a limited pretreatment of
the DNA with crystalline pancreatic DNase (see “Methods”),
an endonuclease which cleaves the phosphodiester bond to
produce 3’-hydroxyl and 5’-phosphoryl termini (25). As
purification of the enzyme proceeded, the ability of this “acti-
vated” DNA to direct DNA synthesis also decreased. The
substitution of the dAT copolymer as primer gave the expected
recovery of enzymatic activity at each stage of purification, and
for this reason it was used in the subsequent purification steps
9 Unless otherwise stated, the units of polymerase activity were
determined in Assay B with dAT copolymer as primer.
*O We are grateful to Dr. Walter Bodmer of the Department of
Genetics, Stanford University, for carrying out these experiments.
I I
0 4
Mlnutas at 450cs
12
FIG. 4. Heat inactivation of polymerase and exonuclease II.
Fraction VIII was diluted to a final protein concentration of 0.08
mg per ml in 0.10
M
potassium phosphate buffer, pH 7.35, contain-
ing 0.01 M 2-mercaptoethanol. Incubation was at 45”, and at the
designated times aliquots were removed, diluted with the standard
diluent, and then assayed in the standard dAT-primed assay and
in the standard exonuclease assay (see “Methods”).
1 8o
F 70 -
‘c 60-
2 40-
F‘ZO-
I I I I I
0
10 20
40 50
Time it?minu+es
60
FIG. 5. Assay for nuclease activity in the hydroxylapatite
fraction at pH 7.4. The reaction mixture (3.0 ml) contained
0.2 Fmole of sap-labeled E. coli native DNA, 200 pmoles of potas-
sium phosphate buffer, 20 Fmoles of MgC12, 3 Nmoles of 2-mer-
captoethanol, and 140 units of Fraction IX. Immediately after
the addition of enzyme, 2.0 ml of the reaction mixture were pi-
petted into a viscometer which was maintained in a 37” bath.
The remainder of the mixture was incubated in a tube inserted
into the same bath. Outflow times were measured when indicated.
The total drop in outflow time during the l-hour incubation was
1.0 second. At the times indicated, O.l-ml samples were with-
drawn from the tube and diluted with 0.2 ml of water. Precipita-
tion with perchloric acid and determination of acid-soluble radio-
activity are described under “Methods” for assay of E. coli
exonuclease II.

228 Enzymatic Synthesis of DNA. XIV Vol. 239, No. 1
(see Table I). (Because of the susceptibility of the dAT
polymer to the nucleases present in Fractions I to IV, poor
proportionality was obtained with these cruder fractions and
therefore “activated” thymus DNA was used instead.)
The progressive decrease in the priming capacity of DNA in
the course of purification of the polymerase is largely due to the
removal of endonuclease and the DNA phosphatase-exonuclease.
The former produces 3’-hydroxyl and 5’-phosphoryl end groups,
which have been previously shown to enhance the rate of DNA
synthesis by the purified polymerase (23, 26). The DNA
phosphatase-exonuclease, as mentioned earlier, activates by
removing inhibitory 3’-phosphoryl end groups from the DNA
primer (23).
Properties
of
PuriJied Polymerase
Eject of pH
on
Rate
of
Reaction-Maximal activity was ob-
tained at pH 7.4 in potassium phosphate buffer with either a
native DNA or the synthetic dAT copolymer as primer (Fig.
6). With either primer, the enzymatic activity was approxi-
mately 70% of the optimal value at pH 7.0 and 7.8.
Divalent Metal Requirement-The
purified enzyme requires
added Mg++. In the absence of MgC12, no detectable activity
was observed with native
B. subtilis,
calf thymus DNA, or
dAT copolymer as primers. At pH 7.4 in potassium phosphate
buffer under the conditions of the standard assay, the optimal
Mg* concentration was 7 x 1O-3 M. At 3.3 X lo-’ M and
3.3 X lo-*
M,
30% and 50%, respectively, of maximal activity
was observed. Mn++ can partially fulfill the metal requirement
and also permits the incorporation of ribonucleotides into
a
DNA polymer (27).
Requirement for DNA
Primer-Under the conditions of the
standard assay and without added DNA, there was no detecta-
ble (less than 0.001 mpmole) nucleotide incorporation into an
acid-insoluble product. In the standard assay, 6 mMmoles of
dAT polymer or 40 mpmoles of native DNA
(E. coli, B. subtili-s,
or calf thymus) produced the maximal rate of incorporation
with a given amount of enzyme, 0.1 unit for the dAT-primed
TABLE IV
E;lfect of polymerase purification on primin.g activity
of thymus DNA
Each polymerase fraction was assayed in the standard dAT-
primed assay with either 6 mpmoles of dAT polymer or 40 mp-
moles of native thymus DNA as primer. The rate of nucleotide
incorporation into acid-insoluble product in the dAT-primed
assay is expressed as 100% with each polymerase fraction; the
rate of incorporation in the thymus DNA-primed assay is ex-
pressed relative to the dAT-primed rate (after correction for
the base composition).
Polymerase fraction Activity relative to
dAT-primed asssy
%
III. Autolysate....... . . . . . . 82.0
IV. Ammonium sulfate. 36.0
VII. DEAE-cellulose. . 11.0
VIII. Phosphocellulose 11.0
IX. Hydroxylapatite. . . . . . . . . 2.5*
* The values obtained when this fraction was assayed with
B.
subtilis
and phage Xdg DNAs were 3.0 and 3.1’%, respectively.
90 -
80 -
370 -
E
.-
6
PO -
%
E50 -
Q
?
g40-
s
.530-
.a
t
2
20 -
lo-
0'
6.6 6.8 1.0 7.2 7.4 7.6 7,8 8.0
PH
Fro. 6. pH-activity curve of polymerase. Conditions of the
standard dAT-primed assay were employed with the substitution
of 40 mpmoles of
B. subtilis
DNA and the addition of 10 mfimoles
of dCTP and dGTP when
B. subtilis
DNA was used as primer.
One hundred per cent represents the incorporation of 0.34 mpmole
of nucleotide in the dAT-primed reaction and 0.80 mpmole in the
B. subtilis
DNA-primed reaction. In the dAT-primed assay,
0.09 unit of polymerase was added per assay tube, and in the
B.
subtilis
DNA-primed assay, 3.6 units of enzyme were present.
The pH of each buffer was determined at room temperature and
0.05 M.
assay and 4.0 units for the native DNAs. Heat-denatured
DNA
(B. subtdis
or calf thymus) directs DNA synthesis at 90%
the rate of the unheated, native DNA in the standard assay.
Net &&he& with Purifid
Polymerase-Removal of nucleases
from the polymerase results in an enzyme (Fraction IX) which
carries out net synthesis of native DNA, but at a much slower
rate. However, during the course of replication, very little
hydrolysis of the primer occurs (Fig. 7). In an experiment
with SH-labeled
B. subtilis
DNA as primer, there was a l.bfold
replication” in a lo-hour incubation period during which 90%
of the primer remained in an acid-precipitable form.
As previously reported (23, 26), pretreatment of the native
DNA with either
E.
coli endonuclease, pancreatic DNase, or
E.
coli DNA phosphatase-exonuclease markedly increases the
rate of DNA synthesis (2- to 30-fold). Also, by removing the
11 The number of replications is defined as the ratio of moles of
synthesized DNA phosphorus to the number of moles of phos-
phorus in the primer DNA.

January 1964
Richardson, Schildkraut, Aposhian, and Kornberg 229
Release OF H3
Minutes 0
FIG. 7. Measurement of degradation of SH-B. subtilis DNA
during the course of replication by polymerase. The incubation
mixture (2.0 ml) contained 120 pmoles of potassium phosphate
buffer, pH 7.0, 12 pmoles of MgC12, 1.8 pmoles of 2-mercapto-
ethanol, 0.2 pmole each of dATP, dCTP, dTTP and &*P-dGTP,
60 mpmoles of aHlabeled native B. subtilis DNA, and 160 units
of Fraction IX (Assay B). At the times indicated, O.l-ml samples
were withdrawn from the tube and diluted with 0.2 ml of water.
The acid-insoluble SzP and 3H radioactivity were determined as
described in “Methods” for the polymerase assay.
3’-phosphoryl termini of native B. subtilis DNA, it has been
possible to obtain a 5-fold net synthesis product with 85% of
the primer undegraded (23). Further studies relevant to the
net synthesis of native and heat-denatured B. subtilis DNA,
the fate of the primer, and the characterization of the product of
the reaction are under current study.
Heat-denatured DNA as Primer-Earlier reports (28, 29)
from this laboratory have indicated that with some preparations
of the DNA polymerase from E. coli heat-denatured DNA is the
preferred primer. However, the purified polymerase (Fraction
IX) utilizes heat-denatured and native DNA at approximately
equal rates (1.9 and 1.5 mpmoles per pg of protein in 30 minutes,
respectively). With either DNA template, extensive synthesis
of the order of a B-fold replication is observed in a 6-hour period.
This ability to replicate double stranded DNA provides a strik-
ing difference between the E. coli enzyme and the DNA poly-
merase induced by infection of E. coli (26) with bacteriophage
T2. The latter is able to replicate single stranded DNA at a
much faster rate and to a far greater extent than double stranded
DNA (11). In this regard the T2 phage polymerase is similar
to the calf thymus polymerase described by Bollum (30).
(EAT and dGdC Synthesis, de Novo and Primed-The
hydroxylapatite fraction of polymerase carries out synthesis
de novo of dAT as described for cruder polymerase preparations
(9). However, for a given amount of enzyme the lag period is
extended. Thus, it required 100 units of the purified enzyme
compared to 25 units of the DEAE-cellulose fraction (Fraction
VII) to produce dAT after a lag period of approximately 4
hours. The presence of dAT polymer in the reaction mixture,
as described earlier (9), permitted the synthesis of dAT polymer
without a lag period. If either a primed or unprimed dAT
synthesis was carried out at pH 7.0, there was only a very slow
degradation of the product after the completion of the reaction
(Fig. 8), in contrast to the rapid degradation reported earlier
with the less purified enzyme fractions (9) ; 75% of the deoxy-
nucleotides initially present were utilized for synthesis of dAT.
The influence of enzyme purification on the primed synthesis
of dGdC is even more striking. With 140 units of the phos-
phocellulose enzyme (Fraction VIII) in a reaction mixture con-
taining dGdC primer, no synthesis was observed during a 5-hour
observation period. However, the addition of 1 unit of E. coli
Oh. I I I I I
I I
I I
,\”
0
100
200 300 400 500
MinutQs
FIG. 8. Primed net synthesis of dAT copolymer. The reaction
mixture (1.0 ml) contained 0.33 pmole each of dATP and dTTP,
6 -moles of dAT polymer, 6.6 Lcmoles of MgC12, 66 pmoles of
potassium phosphate buffer, pH 7.0, and 200 units of Fraction IX.
Absorbance was measured in the Zeiss PM& II spectrophotometer
in a 0.7-ml glass-stoppered cuvette with a light path of 0.2 cm.
The reaction was carried out at 37”.
$0
8
c-4 2
x4
Q‘ 6
68
E
CO/i
QndOfwClQase add
.c 10
$) 12
0 14
i? 16
g 15
n 0 50 100 150 200
.x
Minutss
FIG. 9. Effect of E. coli endonuclease on primed dGdC syn-
thesis with the phosphocellulose fraction of polymerase. The
reaction mixture (0.5 ml) contained 30 firmoles of potassium phos-
phate buffer, pH 7.35,3 amoles of MgC12,0.5 rmole each of dGTP
and dCTP, 10 wmoles of dGdC polymer, and 140 units of
polymerase (Fraction VIII). E. coli endonuclease (1 unit)
was present as indicated. Incubations were performed in quartz
cuvettes with quartz inserts, and absorbance was measured in
the Zeiss PM& II spectrophotometer. The reactions were car-
ried out at 37”.

230 Enzymatic Synthesis of DNA. XIV Vol. 239, Ko. 1
TABLE
V
Reversal of polymerase reaction: pyrophosphorolysis
of 32P-E.
coli
DNA*
The incubation mixtures contained (in 6.0 ml) : 0.07
M
potassium
phosphate buffer, pH 6.5, 0.007
M
MgCI,, 0.001
M
2-mercapto-
ethanol, 0.0017
M
inorganic pyrophosphate, 0.4 pmole of
3WE.
coli
DNA (4 X lo6 c.p.m. per rmole), and 115 pg of hydroxylapatite
fraction of polymerase. After incubation at 37” for 2 hours (Ex-
periment 1) or 4 hours (Experiment 2), 0.2 ml of calf thymus his-
tone IIb (5 mg per ml) (32) was added to precipitate the DNA
quantitatively. To the supernatant fluid, after centrifugation,
were added 5 rmoles of each of the four deoxynucleoside triphos-
phates as markers, and the mixture was adsorbed to a column of
Dowex l-chloride (10 cm X 1 cm*). The column was washed with
0.01
M
HCl; dCTP was eluted wit.h 0.01
M
HCl and 0.05
M
KCl;
dATP and dGTP were eluted in that order with 0.01
M
HCl and
0.1
M
KCl; and dTTP was eluted with 0.01
M
HCl and 0.2
M
KCl.
Experiment
1
2
dATP
%
18
18
Recoveriest
dCTP dGTP dTTP Total
-___
% % % %
19
17
19
73
21 19 19
77
* We acknowledge the skillful assistance of LeRoy Bertsch in
these experiments.
t Recovery of 3*P-containing material applied to the column.
As determined by optical density measurements, between 70 and
90% of each deoxynucleoside triphosphate applied to the column
was recovered in the eluate.
endonuclease to the reaction mixture resulted in the prompt
synthesis of dGdC polymer (Fig. 9).
Reversal of Reaction--Pyrophosphorolytic cleavage of DNA
by DNA polymerase has been inferred largely from the observed
exchange of 32PP with deosynucleoside triphosphates, a reaction
which is absolutely dependent on DNA (31). In the absence of
deoxynucleoside triphosphates, a small amount of 32PP was
converted to a No&adsorbable form. This finding suggested
a slow or limited pyruphosphorolysis of the added calf thymus
DNA. The following experiment was designed to determine
this point. With 32P-labeled native
E.
coli DNA, pyrophosphate,
and the purified polymerase, the release of four 32P-labeled
deoxynucleoside triphosphates and in the molar proportions
expected from
E.
coli was readily demonstrated (Table V).
In the two experiments, 6.9 and 8.90/e of the DNA was released
after 2- and 4-hour incubation periods, respectively. Of the
radioactivity released, 73 to 77% was identified chromato-
graphically as deoxynucleoside triphosphates (Table V). The
low pH at which the incubations were carried out was favorable
for pyrophosphorolysis and minimized the hydrolytic action
by exonuclease; only 8% of the radioactivity released from the
DNA was susceptible to the action of semen phosphomono-
esterase.
DISCUSSION
By several criteria, including chromatography, sedimentation,
and starch gel electrophoresis, the extensively purified poly-
merase had appeared to be a homogeneous protein. However,
further chromatography on hydroxylapatite was rewarded by
the separation of an enzyme, the DNA phosphatase-exonuclease,
whose identificat,ion has provided additional insights into the
polymerase reaction. Still present in the purified polymerase,
however, is an exonuclease activity, called exonuclease II, which
is described in the following report (2). This nuclease, exhibit-
ing its optimal activ-ity at pH 9.0, hydrolyzes DNA at a very
slow rate in the neutral pH range in which polymerase activity
is maximal. Thus far all our attempts to separate these two
activities have been unsuccessful. This raises the question
whether a single protein with dual activities is responsible for
polymerizing deoxymononucleotides and for the stepwise hy-
drolysis of these residues. What has become clear is that this
esonuclease function is not an essential property of all DNA-
synthesizing enzymes, since the purified
B. subtilis
DNA poly-
merase (5) shows no significant exonuclease activity.12
Assuming that the enzyme is near homogeneity and that the
molecular weight is 100,000 (see “Appendix”), an estimate can
be made of the number of DNA polymerase molecules per
E.
coli cell. The number is approximately 400 molecules of DNA
polymerase per bacterial cell; in terms of DNA mass, this
represents about 1 molecule of polymerase per 2 X lo7 gram
molecular weight of DNA.
Previous studies have demonstrated that the exact replica-
tion of DNA depends on the complementary base pairing of the
deoxynucleotidyl substrate residues with those in the DNA
template (26, 29). Current work is directed toward the mech-
anism of initiation of replication and the fate of the DNA
template during the course of replication. By using a template
labeled with r5N, *H and 3H, substrates labeled with 32P, and
various physicochemical methods, it appears that a hybrid of
template and product is first formed. Upon heating or alkaline
denaturation, the product can be separated completely from
the primer but differs from naturally occurring DNA in the ease
with which it renatures (23, 26). Further studies are required
to establish the basis for this unusual behavior of the DNA
product.
Tests for the biological activity of DNA synthesized in vitro
have been described elsewhere (23). With the purified poly-
merase, free of endonuclease activity, an extensive replication
(8fold) of the primer could be accomplished with little loss in
genetic activit,y (30% decrease). However, no genetic activity
was found in newly synthesized material isolated by CsCl
density gradient centrifugation which could not be accounted
for by associated primer atoms. Earlier studies (29) with heat-
denatured Dh’A as a primer showed large increases in biological
activity concomitant, with DNA synthesis and have been con-
firmed more recently with the purified polymerase. However,
the final level of activity obtained is always less than that of the
unheated control and very likely resides in hybrid molecules.
Recently, Litman and Szybalski (33) have used heat-denatured
bromouracil-labeled B. subtilis DNA as primer and a less puri-
fied fraction of
E.
coli polymerase, and have claimed that the
restored transforming activity resides in newly synthesized
DNA. This finding requires confirmation by more rigid proof
that primer molecules are in fact absent from the DNA product.
The polymerase reaction depends critically on the nature
of the DNA primer. As demonstrated earlier (23, 26), the
introduction of new end groups into the DNA primer molecule
can either stimulate or lower priming activity, depending on
12 It should be pointed out that nucleotide removed by exo-
nuclease action during the course of polymerization would be ex-
pected to be replaced without residual effects provided the supply
of deoxynucleoside triphosphate remains sufficient.

January 1964
Richardson, Schildkraut, Aposhian, and Kornberg
231
whether such groups are 3’-hydroxyl or 3’-phosphoryl end groups,
respectively. Thus the great variability in DNA primer which
has been observed depends at least in part on the presence of
such end groups, which may be altered by specific endonucleases
or phosphatases or by mechanical shearing. For these reasons,
further studies are indicated to define the presence of such
termini in DNA, and these would employ homogeneous and
perhaps smaller and better defined DNA molecules as primers.
Since the yield of purified enzyme from
E.
coli (approximately
7.5 mg per kg of cell paste) is so small, a survey of other organisms
and animal tissues has been made. Extracts from these other
sources were found to contain lower levels of polymerase activity
than those in
E.
coli. Extracts of
B.
subtilis, an organism low
in nuclease activity, contained only one-fourth the DNA poly-
merase activity found in
E.
coli. Furthermore, purification of
the
B. subtilis
enzyme (5) presented considerable difficulties in
removing the associated nucleic acid. Thus, with the avail-
ability of large quantities of
E.
coli cells, a purification procedure
designed to handle kilogram quantities of cells, and the relative
stability of the enzyme,
E.
coli still remains a favorable source
of DNA polymerase.
SUMMARY
1. A procedure for the purification of deoxyribonucleic acid
polymerase from
Escherichia
coli adapted for kilogram quantities
of cell paste yields a preparation which is homogeneous by sedi-
mentation, chromatography, and starch gel electrophoresis.
Chromatography on hydroxylapatite in the final step of purifi-
cation separates a minor component which has been identified
as a new enzyme, a DNA phosphatase-exonuclease (3, 4).
2. The most purified DNA polymerase fraction is relatively
feeble compared to cruder polymerase fractions in replicating
native (or heated) DNA. This may be attributed to the re-
moval of endonuclease and of the DNA phosphatase-exonuclease.
The former, by introducing additional terminal 3’-hydroxyl
groups, and the latter, by removing terminal 3’-phosphoryl
residues (which block such 3’-hydroxyl groups), activate the
priming capacity of DNA many fold.
3. Another exonuclease activity persists in the most purified
polymerase fractions in a ratio that has not been altered by
fractionation procedures, heating, or urea treatments. However,
this activity does not interfere seriously with DNA synthesis.
Under optimal conditions (pH 7.0) for replication of DNA,
only slight destruction of primer occurs, as judged by release of
nucleotides from 3H-labeled primer or by loss of genetic activity.
4. Reversal of the reaction has been demonstrated directly
by cleavage of 32P-labeled
E.
coZi DNA by inorganic pyrophos-
phate to yield the four deoxynucleoside triphosphates in proper
molar proportions.
APPENDIX: MOLECULAR WEIGHT AND SEDIMENTATION COEFFICIENT
OF DEOXYRIBONUCLEIC ACID POLYMERASE
R. L. BALDWIN
From the Department of Biochemistry, Stanford University School of Medicine, Palo Alto, California
The finding of a single boundary in sedimentation velocity to sedimentation equilibrium in a short column of solution
patterns (Fig. 2) indicated the deoxyribonucleic acid poly- (2.5 mm) at 7447 r.p.m., nearly identical results were obtained
merase to be sufhciently homogeneous with respect to molecular from photographs taken at 23 and 343 hours. The data, plotted
weight to warrant its measurement. Preliminary measure- by Method II of Van Holde and Baldwin (36), are shown in
ments of the dependence of s on c were made with a phospho- Fig. 11. The presence of some curvature in this plot indicates
cellulose preparation (Fraction VIII) and then extended to the heterogeneity. The value of 1.3 x lo5 for the molecular weight,
hydroxylapatite preparation (Fraction IX) ; a single boundary
M,,
is significantly higher than that found by combining s and
was found in each case and the values of s obtained fell on the
same curve of s uersus c (Fig. 10). At zero concentrat,ion,
s20,w was estimated to be 5.64 x lo-l3 second. The diffusion
coefficient of this preparation was measured in a synthetic
boundary cell (double sector, capillary type) at a rotor speed of
7447 r.p.m.; the average concentration of the two solutions
forming the boundary was 0.22 g per 100 ml. The value found
for @zo,~ was 5.02 X 10-v cm2 see-I. The partial specific volume
was computed from the amino acid composition (omitting
cysteine and tryptophan, whose values were not available) to be
0.742, by the procedure outlined by Cohn and Edsall (34).
Calculation of the molecular weight by the Svedberg equation
(taking s at c = 0.22 g per 100 ml, and neglecting a possible
activity coefficient correction) gave
M = 1.00 x
105. When
s20,w is plotted against log
M
for a series of globular proteins,
chiefly enzymes (35), the results for DNA polymerase fall as
close to the smooth curve as do most of the other data, showing
that these values of s and
M
are self-consistent. FIG. 10. The dependence of .s on c for DNA polymerase. 0
refers to a phosphocellulose preparation (Fraction VIII); n
When the hydroxylapatite enzyme fraction was centrifuged refers to a hydroxylapatite preparation (Fraction IX).

232 Enzymatic Synthesis of DNA. XIV
Vol. 239, No. 1
0
(&La) qm.;~oo t-n,.
.3
Fm. 11. Sedimentation equilibrium of DNA polymerase.
Plotted according to Method II of Van Holde and Baldwin (36),
where r is the radial distance in the cell and c is the concentration
at
T
(c at the meniscus is denoted by c.). The dashed line refers
to the initial slope of the plot.
D. This result is expected for a preparation with some in-
homogeneity, since the molecular weight measurements by this
method (M,) places a greater value on species of higher molec-
ular weight. The initial slope of the line in Fig. 11 gives a value
for M of 1.10 X 106. (However, the weight average molecular
weight (M,,,) could not be measured from this experiment for
technical reasons.)
Further studies of sedimentation velocity patterns at different
pH values and salt concentrations are indicated at this point to
determine whether or not aggregation phenomena are responsi-
ble for the observed physical heterogeneity.
EXPERIMENTAL PROCEDURE
The buffer was 0.1
M
KCI, 0.01
M
2-mercaptoethanol, and
0.005
M
potassium phosphate, pH 6.8, at 25”. Measurements
of ?he viscosity and density at 25” gave q/qu, = 1.0037 and
P/Pw =
1.005~ (the subscript w refers to distilled water) ; these
ratios were taken to be constant, independent of temperature.
The sedimentation measurements were made within 2” of 6”
and converted to 20” with water as solvent, in the usual way,
by assuming that sq/(l -
@p)
and
Dq/T
are constants. Con-
centrations were computed from the schlieren patterns by using
the known optical constants for the ultracentrifuge and taking
dn/d.c = 1.90 x 10-S for c in grams per 100 ml. A Spinco
model E ultracentrifuge, with temperature control and schlieren
optics, was used.
25.
SINSHEIMER,
R. L., J. Biol. Chem., 206,445 (1954).
26.
RICHARDSON,
C. C.,
S~HILDKRAUT,
C. L.,
AND KORNBERG,
A.,
Cold Spring Harbor symposia on quantitative biology, Vol. 88,
Long Island Biological Association, Cold Spring Harbor,
New York, in press.
27.
BERG,
P.,
FANCHER,
H.,
AND CHAMBERLIN,
M., in H.
VOGEL
(Editor), Symposium on informational macromolecules,
Academic Press. Inc., New York, 1963, p. 467.
23. KORNBERG,
A., Sc>ence; 131, 1563 (iSSO).- -
29. KORNBERG.
A.. Enzumatic sunthesis of DNA. Ciba lectures in
microbiai b&hem&y, Jo”hn Wiley and Sons, New York,
1961.
30. BOLLUM, F. J., J. Biol. Chem., 236, 2399 (1969).
31.
BESSMAN,
M. J.,
LEHMAN,
I. R.,
SIMMS,
E. S.,
AND KORNBERG,
A., J. Biol. Chem., 233, 171 (1958).
32.
SATAKE,
K.,
RASMUSSEN,
P. S.,
AND LUCK,
J. M., J. Biol.
Chem., 236, 2801 (1960).
REFERENCES
1. LEHMAN,
I. R.,
BESSMAN,
M. J.,
SIMMS,
E. S.,
AND KORNBERQ,
A., J. Biol. Chem., 233, 163 (1958).
2. LEHMAN,
I. R.,
AND RICHARDSON,
C. C., J. Biol. Chem., 239.
233 (1964).
33.
LITMAN,
R. M.,
AND SZYBALSKI,
W., Biochem. and Biophys.
Research Communs., 10, 473 (1963).
34.
COHN,
E. J.,
AND EDSALL,
J. T., Proteins, amino acids, and pep-
tides as ions and dipolar ions, Reinhold Publishing Corpora-
tion, New York, 1943, p. 371.
35.
EDSALL, J.
T., in H.
NEURATH AND
K.
BAILEY
(Editors), The
proteins, Vol. I, Part B, Academic Press, Inc., New York,
1953, p. 634.
3.
RICHARDSON,
C. C.,
AND KORNBERG,
A., J. Biol. Chem., 239, 36.
VAN HOLDE,
K. E.,
AND BALDWIN,
R. L., J. Phys. Chem., 62,
242 (1964). 734 (1958).
4.
RICHARDSON,
C. C.,
LEHMAN,
I. R.,
AND KORNBERG,
A., J.
Biol. Chem., 239, 251 (1964).
5.
OHAZAKI,
T.,
AND KORNBERG,
A., J. Biol. Chem., 239, 259
(1964).
6.
OKAZAKI,
R.,
AND KORNBERG,
A., J. Biol. Chem., 239, 269
(1964).
7.
KAY, E. R. M., SIMMONS,
N. S.,
AND DOUNCE,
A. L., J. Am.
&em. SOL, 74.1724 (1952).
8.
LEHMAN,
I. R., J. Biol. Chem., 236, 1479 (1960).
9.
S~HACHMAN,
H. K.,
ADLER,
J.,
RADDING,
C. M.,
LEHMAN,
I. R.,
AND KORNBERG,
A., J. Biol. Chem., 236,3242 (1966).
10.
RADDING.
C. M..
JOSSE.
J..
AND KORNBERG.
A.. J. Biol. Chem..
237, 2Sii9 (1962). ’ ’ , ,
11.
APOSHIAN,
H. V.,
AND KORNBERG,
A., J. Biol. Chem., 237,
519 (1962).
12.
OFENGAND,
E. J.,
DIECKMANN,
M.,
AND BERG,
P., J. Biol.
Chem., 236, 1741 (1961).
13.
PETERSON,
E. A.,
AND SOBER,
H. A., J. Am. Chem. Sot., 76,
751 (1956).
14.
TISELIUS,
A.,
HJERTEN,
S., and
LEVIN, 6.,
Arch. Biochem.
Biophys., 66, 132 (1956).
15.
LEHMAN,
I. R., Roussos, G. G.,
AND PRATT,
E. A., J. Biol.
Chem., 237, 819 (1962).
16.
JOSSE, J., AND KORNBERG,
A., J. Biol. Chem., 237, 1968 (1962).
17.
CHARGAFF,
E., in E.
CHARGAFF AND
J. N.
DAVIDSON
(Editors),
The nucleic acids, Vol. 1, Academic Press, Inc., New York,
1955, p. 307.
18.
LOWRY,
0. H.,
ROSEBROUGH,
N. J.,
FARR,
A. L.,
AND RANDALL,
R. J.. J. Biol. Chem.. 193, 265 (1951).
19.
DISCH;,
Z., in E.
CHA~GAFF AND
J.
fi. DAVIDSON
(Editors),
The nucleic acids. Vol. 1. Academic Press. Inc.. New York.
1955, p. 285. ’ ’ , ,
20.
SMITHIES,
O., in C. B.
ANFINSEN,
JR., M. L.
ANSON,
K.
BAILEY,
AND J.
T.
EDSALL
(Editors), Advances in protein chemistry,
Vol. 14, Academic Press, Inc., New York, 1959, p. 65.
21.
POULIK, M.
D., Nature, 189, 1477 (1957).
22.
SUEOKA,
N., PTOC. Natl. Acad. Sci. U. S., 47, 1141 (1961).
23.
RICHARDSON,
C. C.,
SWILDKRAUT,
C. L.,
APOSHIAN,
H. V.,
KORNBERG,
A.,
BODMER,
W.,
AND LEDERBERG,
J., in H.
VOGEL
(Editor), Symposium on informational macromole-
cules, Academic Press, Inc., New York, 1963, p. 13.
24. LEHMAN,
I. R., Roussos, G. G.,
AND PRATT,
E. A., J. Biol.
Chem., 237,829 (1962).