A simple and rapid purification method for Escherichia coli DNA polymerase I.

We report a simple, three-step method for the purification of Escherichia coli DNA polymerase I. Its advantages over other procedures are ease and rapidity, the absence of an autolysis or any high speed centrifugation step, and applicability to large quantities of material. In addition, RNA polymerase can be isolated as a by-product. We have applied this method to purify DNA polymerase both from wild type E. coli cells and from cells bearing a lambda prophage carrying the polA gene (Kelley, W.S., Chalmers, K., and Murray, N.E. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5632-5636). This latter source amplifies the amount of DNA polymerase in the cells by at least 10-fold.

We report a simple, three-step method for the purification of Escherichia coli DNA polymerase I. Its advantages over other procedures are ease and rapidity, the absence of an autolysis or any high speed centrifugation step, and applicability to large quantities of material.
In addition, RNA polymerase can be isolated as a by-product.
We have applied this method to purify DNA polymerase both from wild type E. coZi cells and from cells bearing a X prophage carrying the poZA gene (Kelley, W. S., Chalmers, K., and Murray, N. E. (1977) Proc. Natl. Acd Sci. U. S. A. 74, 5632-5636). This latter source amplifies the amount of DNA polymerase in the cells by at least lo-fold.
Several procedures have been published for the purification of DNA polymerase I (l-6). Most employ phosphocellulose or DNA-cellulose for the main purification step. A major problem with these methods is the initial removal of ribosomes and nucleic acids from the cell extracts so that the enzyme will bind to subsequent columns. Richardson et al. (1) accomplished this by a prolonged autolysis step. Disadvantages of this procedure are the high degree of variability, the frequent loss of substantial activity, and the proteolytic modification of the polymerase. Other methods (3-5) utilize high speed centrifugation which severely limits the amount of extract which can be processed conveniently.
Several methods for the purification of RNA polymerase employ polymin P, a basic polymer (polyethyleneimine or aziridine homopolymer), in order to precipitate the enzyme (7, 8). We have found that the DNA polymerase activity remains in solution at concentrations of polymin P which completely precipitate RNA polymerase. After precipitation of the DNA polymerase with ammonium sulfate and desalting, the enzyme is loaded onto a phosphocellulose column. This provides a simple alternative to the autolysis, centrifugation, DEAE-cellulose chromatography, or phase partition steps to remove the nucleic acids.

Enzyme Purification
The following section describes a purification of 200 g of wild type E. coli cells and 25 g of the ApolA lysogen (Tables Purification of E. coli DNA Polymerase I I and II). Indicated volumes refer to the former and values in parentheses to the latter. All steps were performed at 4°C. Preparation of Cell Extract-Cells were lysed by the method of Burgess and Jendrisak (8). Frozen cells were suspended in 800 (85) ml of grinding buffer and were blended at low speed in a Waring Blendor until the solution was at 2-5°C. After 20 min, 10 (1.5) ml of a 4% (w/v) sodium deoxycholate solution were added and the cells were blended 30 s at low speed. After a further 20 min, the cells were blended 30 s at high speed, 500 (120) ml of TGN buffer were added, and the cells were again blended for 30 s at high speed. The cell extract was centrifuged for 45 min at 8000 rpm (7000 X g) in a Sorvall SS 3 centrifuge (GSA rotor) or a Beckman J-21B centrifuge (JA 10 rotor).
Polymin P Treatment and Ammonium Sulfate Precipitation-For the concurrent isolation of RNA and DNA polymerases, the former activity was titrated as described by Burgess and Jendrisak (8). In the case of DNA polymerase alone, a 10% polymin P solution was added to the stirred cell extract to a final concentration of 0.6% (0.06 ml of 10% polymin P added/ml of extract). In either case, the polymin P-treated extracts were spun 45 min at 8000 rpm (7000 X g). The pellet, containing RNA polymerase, was discarded or worked up if desired. The stability of the DNA polymerase activity in the polymin P supernatant was somewhat variable. One batch lost 50% of its activity after 4 days at 4°C whereas several other batches showed no loss in activity after 9 days.
Solid ammonium sulfate was added slowly to the polymin P supernatant to 35% saturation (20 g of solid ammonium sulfate/K@ ml of extract) while stirring. The solution was stirred slowly for 30 min at 4°C and then spun at 8000 rpm for  30 min. The pellet was discarded and the supernatant was brought to 64% saturation (20 g of solid ammonium sulfate added/100 ml of original extract). After stirring for 30 min, the solution was centrifuged as above. The pellet was resuspended in 60 (10) ml of PC04 buffer. The DNA polymerase activity was quite labile at this stage with half of the activity lost after 1 day. Rapid freezing of this fraction rendered it stable for at least '/z year.
Phosphocellulose Chromatography-The ammonium sulfate fraction was desalted by passage through a 600 (150)-ml Sephadex G-50 column equilibrated with PC04 buffer. The activity migrated with the first of two distinct yellow zones. The desalted protein was loaded onto a 320 (50)~ml phosphocellulose column and washed with 3 column volumes of PC04 buffer. It was eluted with a 4 (0.5)-liter linear gradient of PC04 to PC3 buffer at a flow rate of 1.2 ml/min. The DNA polymerase activity eluted at 0.15 to 0.20 M phosphate concentration Fractions with activity greater than 15 units/ml for wild type cells (see Fig. 1) and 300 units/ml for the phageinduced cells (Fig. 2) were pooled. (A small peak of polymerase activity, presumably DNA polymerase II (16, 17) eluted at 0.25 M phosphate (Fig. l).) This fraction of enzyme was fairly stable; 80% of the original activity remained after 4 days at 4°C. The stability seemed to depend on enzyme concentration because a 1:l dilution decayed 2 to 3 times faster. Rapidly frozen samples of this fraction were stable for at least 6 months.
DNA-Sepharose Chromatography-The pooled phosphocellulose fractions were diluted 1:l with PC04 and immediately loaded onto a 47 (47) Tables  I and II). The enzyme was quite stable at 4°C at this stage with no observable loss of activity after several days. Concentration and Storage-The pooled DNA-Sepharose fractions were dialyzed overnight against 20 mM potassium phosphate, pH 6.9 , 0.1 mM EDTA and 5% (v/v) glycerol and then against the same buffer containing 50% (v/v) glycerol. This procedure resulted in a 3-fold concentration.
The enzyme was stored at -20°C with no loss of activity after a year.

Characterization of the Enzyme
The specific activity with optimally activated poly[d(A-T)] was identical with that previously reported (1, 2). Specific assays for 5' + 3' and 3' + 5'-exonuclease were not performed. When we used potassium phosphate buffer at pH 7.4 (optimal conditions for DNA polymerase I), the total exonuclease activity from Fraction V was about 980 units/ml and the polymerase activity was about 5000 units/ml. Several enzyme purifications (including the data with the lysogenic phage) always showed the same ratio of polymerase to exonuclease activity of 5 to 6 under these conditions.
There was no evidence for endonuclease contamination as tested by incubating 5 units of DNA polymerase I with 1 fig of SV40 DNA Form I for 1 h at 37°C.' The assays for ATPase activity showed less than 3.6 pmol of phosphate released/min/mg of protein. This value was not increased by the addition of single or double-stranded DNA. One single band was observed by gel electrophoresis which migrated identically with a DNA polymerase I sample isolated according to the standard procedure (2).

DISCUSSION
The above purification method for DNA polymerase is capable of yielding homogeneous enzyme in 3 to 4 days. We have applied it to cell masses ranging from 25 to 500 g. It can, in principle, be scaled to handle extremely large amounts of starting material. Table I presents the results of a typical purification.
In several purifications with the wild type strain of E. coli, we obtained the same yields and specific activities. These corresponded quite closely to the recoveries and purities obtained in the other DNA polymerase purifications (1, 2). The yield was significantly lower for the one purification involving the cells carrying the prophage where we found substantially less recovery during both column steps, possibly due to the use of disproportionately large volumes and consequent protein dilution.
We attribute this to the lack of an autolysis step in which the extract is exposed to a prolonged incubation at higher temperatures and to the fact that the protease inhibitor phenylmethylsulfonyl fluoride was included during cell lysis.
Since its introduction by Zillig et al. (7), polymin P has been applied to the purification of a number of enzymes involved in nucleic acid metabolism.
Thus, purification procedures for E. coli and eukaryotic RNA polymerase (8,21), E. cob DNA ligase (22), restriction nucleases (23), DNA gyrase (24,25), and DNA polymerase (6) utilizing this compound have been reported. In all cases, the enzyme in question is precipitated by polymin P and re-extracted with a buffered salt solution. In contrast, our method uses a relatively low concentration of polymin P (0.6%) where few cellular proteins are precipitated (one of the exceptions being RNA polymer-' P. Chambon, personal communication. ase). After precipitation of the supernatant with ammonium sulfate and desalting, the extract can be loaded directly onto a phosphocellulose column. This finding is in accord with Atkinson and Jack (26) who report that 0.37% polymin P removes 99.6% of the DNA and 97.4% of the RNA from a cell extract although somewhat higher concentrations may be required for lysozyme-treated cells. It appears to be possible to isolate other enzymes from the polymin P supernatant.
We have observed a major nuclease activity which elutes from the phosphocellulose column at 0.07 M phosphate, before the DNA polymerase. In addition, as can be seen in Fig. IA, a second peak of DNA polymerase activity elutes at 0.25 M phosphate which is the position at which DNA polymerase II has been reported (16, 17). Finally, we find an exonuclease activity in the DNA-Sepharose flowthrough which we identify as exonuclease III by its mobility in sodium dodecyl sulfate gels. Thus, the method may have general application for treatment of cellular extracts before column chromatography.