Immunological Comparison of Purified DNA Polymerase a from Embryos of Drosophila melanogaster with Forms of the Enzyme Present in Viuo *

Specific antisera to purified DNA polymerase a from embryos of Drosophila melmgaster and to two of the four constituent subunits (a, j3, y, and S) were prepared. These antibodies have revealed the following features of the enzyme. (i) The M, = 148,000 a subunit is very likely derived by in vitro proteolysis from polypeptides with molecular weights of 186,000 and 166,000 that are present in vivo. (ii) The M, = 60,000 fl subunit occurs in rapidly replicating embryos.as both an 85,000and a 60,000-dalton form, but predominantly as a 60,000-dalton form in more slowly replicating cultured cells. (iii) There is no detectable immunologic cross-reactivity between the four subunits. (iv) There is an abundance of antigenic material in embryos that co-migrates with the S subunit of the purified enzyme during polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate.

DNA polymerase a from Drosophila melanogaster embryos is a large, asymmetric 280,000-dalton protein composed of four nonidentical subunits with molecular weights of 148,000 (a), 58,000 (p), 46,000 ( y ) , and 43,000 (6) (1). Of the four subunits, only the isolated a subunit retains catalytic activity (2). It is likely that one or more of the other subunits acts to modify or enhance the activity of the a subunit (3). 'We have prepared specific antibodies against the intact a polymerase and against individual subunits in order to gain a better understanding of the enzyme and its constituent subunits. In addition to the functional studies that such antibodies permit, they provide a useful tool to examine the extent to which proteolysis may have occurred during purification of the enzyme. As demonstrated by Brake1 and Blumenthal (4, 5), proteolysis can be a significant factor in the generation of multiple forms of the enzyme. It is this latter problem that is the principal subject of this paper. We have, in fact, found that significant proteolysis of the a and p and possibly the other subunits has indeed occurred during purification of the enzyme.
Institutes of Health (GM-06196) and the National Science Foundation * This work was supported by research grants from the National . The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Materials
Enzymes and Proteins-DNA polymerase a from D. melanogaster (Oregon R) embryos was purified by a modification of the procedure of Banks et al. (1) and Villani et al. (2). The enzyme so obtained (Fraction VII) was routinely greater than 65% pure as determined by electrophoresis on polyacrylamide gels in the presence of SDS' and had a specific activity of 250,000 to 300,000 units/mg of protein.
Escherichia coli RNA polymerase was purified by the method of Burgess and Jendrisak (6). Bovine pancreatic DNase, E. coli alkaline phosphatase, and E. coli P-galactosidase were obtained from Worthington; rabbit muscle L-lactate dehydrogenase and bovine liver catalase were from Sigma; bovine serum albumin was from Pentex; and protein A from Staphylococcus aureus was obtained from Calbiochem-Behring. Myosin from rabbit muscle was the gift of Dr. James Spudich (Stanford University). KB cell DNA polymerases a and / 3 were gdts from J. Chin and Dr. T. Wang, respectively (Stanford University). E. coli DNA polymerases I and I11 were the g&s of Drs. S. Scherer and P. Burgers, respectively, of this department.
Preparation of Antisera-Antiserum directed against Drosophila DNA polymerase a was kindly prepared by Dr. L. Levine (Brandeis University) using preparations of enzyme which were greater than 60% pure. A male New England rabbit was immunized with 100 pg of enzyme in Freund's complete adjuvant by injection at multiple subcutaneous sites. Four successive booster immunizations, each containing 100 gg of enzyme in Freund's incomplete adjuvant, were administered at 3-week intervals. Bleedings were performed 10 days after each injection.
Subunit-specific antisera were prepared by multiple subcutaneous injections of3-year-old virgin female New Zealand rabbits with either 70 pg of p subunit or 45 pg of a subunit in Freund's complete adjuvant.
Individual subunits were prepared by excising the appropriate COOmassie blue-stained polypeptide from a polyacrylamide gel after eke-trophoresis in the presence of SDS. The resulting polyacrylamide slice was homogenized and mixed with Freund's adjuvant. Five booster injections in Freund's incomplete adjuvant of 70,30, 30, 20, and 20 pg of p subunit and 100,20,30,70,50, and 50 pg of a subunit were given at 3-week intervals. Bleedings were performed 10 days after the last injections.
Purified IgG was prepared by ammonium sulfate precipitation and DEAE-cellulose chromatography (11).
Enzyme Neutralization Reactions-Equal volumes of phosphatebuffered dime and DNA polymerase a (Fraction VII,260 units/ml) in Buffer N (20% glycerol, 2 mg/ml of bovine serum albumin, 40 pg/ ml of leupeptin, 20 mM potassium phosphate (pH 7.1), and 2 m~ pmercaptoethanol) were mixed to give a total volume of 50 pl, and preincubated for 2 h at 20 "C with the indicated amount of either preimmune or immune IgG. Aliquota of this preincubation mixture were then assayed as described (1). In the absence of added IgG, the DNA polymerase retained 1004% of its initial activity under these conditions.
Gel Electrophoresis and Protein Transfers-Polyacrylamide gel electrophoresis in the presence of SDS was performed as described by Laemmli (12). Drosophila embryo extracts were prepared from dechorionated 0-16 h embryos which had been stored at -80 "C by homogenization in the presence of 5% SDS, 10% p-mercaptoethanol, 125 mM Tris-HC1 (pH 6.8), 20% glycerol, 1 m~ phenylmethylsulfonyl fluoride, 10 m~ sodium metabisulfite, 10 pg/ml of leupeptin, 10 pg/ml of pepstatin, and 10 n" EDTA and then heated at 100% for 3 min. Cultured cell extracts were prepared, similarly, from Schneider's cell lime 2 (obtained from Dr. A. Sugino, National Institute of Environmental Health Sciences) which had been grown to 2-4 X lo6 cells/ml in Schneider's complete medium plus 10% fetal calf serum at 25 "C.
Proteins in polyacrylamide gels were transferred to activated aminophenylthioether paper by either blotting (13) or electrophoretic transfer with a Hoeffer Transphor apparatus (14) in 15 mM sodium phosphate (pH 7.0). Activated aminophenylthioether paper was used in place of nitrocellulose to circumvent the problem of nonspecific binding of IgG to some proteins, observed occasionally in nitrocellulose transfers. The transfer efficiency of both high and low molecular weight proteins was routinely monitored with the use of '%labeled marker proteins and autoradiography of the dried polyacrylamide gel after transfer by either blotting or electrophoresis. Transfer by electrophoresis was more efficient than by blotting, especially for large polypeptides. Under the conditions used for electrophoretic transfer, polypeptides of less than 130,000 daltons transferred at nearly 100% efficiency. Myosin transferred at about 30% efficiency. Transfers were probed with purified IgG at a concentration of 20 pg/ml as described (13). They were reprobed after removing bound IgG and protein A by treatment with 2 M guanidine thiocyanate, 0.2 M acetic acid, and 10 m~ dithiothreitol for 20 min at room temperature.
Radwimmune Assay-The radioimmune assay of Crawford and Lane (15) was used to determine the total amount of antigenic material detectable with anti-DNA polymerase a IgG. One unit of antigen is defined as that amount of antigenic material present in Fraction I11 of a polymerase that corresponds to one DNA polymerase unit (1).

RESULTS
Inhibition of DNA Polymerase a by Specific Antibodies-Antiserum to the Drosophila DNA polymerase a inhibited the enzyme by 85% (Fig. 1); higher concentrations did not result in further inhibition. Under conditions where the antibody produced maximal inhibition of the Drosophila a polymerase, no inhibition of KB cell DNA polymerase a or / 3 , or E. coli DNA polymerase I or I11 was observed.
Antibody directed specifically against the a subunit inhibited the Drosophila enzyme, although not completely, even at high concentrations of IgG. In contrast, the / 3 subunitspecific antiserum produced no inhibition of DNA polymerase activity (Fig. 1). These results are in accord with the previous finding that the a subunit is itself sufficient for catalytic activity and that the / 3 , y, and 6 subunits most likely act to modify the activity of the a subunit, perhaps by increasing the processivity or stability of the enzyme (3).
Specificity of Anti-DNA Polymerase Antibodies-The protein transfer technique described by Renart et al. (13) was used to determine which subunits of the a polymerase were recognized by the various antisera. Purified DNA polymerase was subjected to electrophoresis on polyacrylamide gels containing SDS, transferred to diazotized paper, probed with the antibody to be examined, incubated with '251-labeled S. aureus protein A, and then autoradiographed.
The results of such an experiment using DNA polymerase a antibody is shown in Fig. 2. The major antigenic determinant recognized is the a subunit; the y and 6 subunits reacted to lesser extents. Upon very long exposures a faint band was seen at the /3 position. In addition, a number of other bands were Immunological Comparison of Forms of DNA Polymerase a recognized by the antibody which are most likely minor contaminants present in the a polymerase preparation. The specificity of the antisera prepared against the isolated a and /? subunits is shown in Fig. 3. The a-specific antibody recognized only the a subunit (Fig. 3A). Similarly, the /?specific antibody recognized only the /? subunit (Fig. 3B).
These results indicate that the a and p subunits are distinct from one another and from the y and 6 subunits. They are also consistent with the observation that the polypeptide maps generated by protease digestion of the four subunits are dissimilar (2). A bundance of 6 Subunit-associated Antigenic Material in Drosophila Embryos-Because high molecular weight proteins are transferred inefficiently to diazotized paper by the blotting technique of Renart et al. (13), the a polypeptide was initially indetectable in crude DNA polymerase fractions from Drosophila embryos using antibody directed against the intact polymerase (Fig. 4). However, a large quantity of an immunoreactive polypeptide appeared that had the same mobility on SDS-polyacrylamide gels as the 6 subunit of the purified enzyme. After chromatography on DNA cellulose (l), the DNA polymerase fraction was free of excess 6 subunitassociated antigen. No excess antigen was apparent in extracts of cultured Drosophila cells (data not shown).
As shown by the radioimmune assay, the excess of 6 subunitassociated antigenic material is not retained on DNA cellulose ( Table I) and was in this manner separated from the bulk of the DNA polymerase activity. There was no discrete protein band as judged by Coomassie blue staining that corresponded

Detection of a and /? Subunits in Drosophila Embryos and
Cultured Cells in Viuo-Because the purified DNA polymerase a is the product of a substantial number of purification steps, it was important to determine whether modification of the enzyme occurred during this process. We, therefore, examined an extract prepared by directly homogenizing embryos in SDS. To ensure efficient transfer of high molecular weight polypeptides, the transfer was performed by electrophoresis of the proteins to diazotized paper. In addition, amounts of Fractions I through VII. equivalent to 20 units of DNA polymerase, were examined. As shown in Fig. 51 It thus appears that the a and / 3 subunits of purified DNA polymerase a derive from higher molecular weight forms that are present in vivo. T o ensure that we were indeed examining forms found in vivo, extracts were also prepared by direct precipitation of cells with cold trichloroacetic acid followed by resuspension in SDS, and similar results were obtained (data not shown).

DISCUSSION
The use of antibodies specifically directed against the various subunits of the Drosophila DNA polymerase a demonstrates that the M, = 148,000 a subunit is the catalytic core of the enzyme, in accord with the previous observation that this polypeptide has catalytic activity after separation from the p, y, and S subunits (2). Furthermore, studies with these antibodies reaffirm the conclusion reached on the basis of peptide mapping of the isolated subunits that (i) the a and /3 subunits are unrelated to each other, and (ii) that the y and 8 subunits are most likely not derived from the a or p subunits by proteolytic cleavage (2).
We have also found that there is an abundance of antigenic material in Drosophila embryos that co-migrates with the 6 subunit of the purified a polymerase on SDS-polyacrylamide gels. Although we have not yet demonstrated that this material is identical with the S subunit, it is clear that it is not recognized by either a or /3 subunit-specific antibodies. The function of the S subunit and the significance of its abundance in embryos remains to be determined.
Certainly the most striking result presented here is the observation that at least two of the four subunits of the purified DNA polymerase most likely derive from larger polypeptides present in vivo. In the case of the a subunit, it is clear that the 148,000-dalton polypeptide associated with the purified enzyme is generated from larger polypeptide(s) during the purification procedure. Most probably this modification is the result of proteolysis in vivo. However, we do not know whether the M, = 166,000 or the Mr = 185,000 polypeptide, or both, are the functional moieties in vivo. Although the 60,000dalton , L3 subunit of the purified enzyme most likely derives from an 85,000-dalton precursor, the possibility cannot be excluded that the latter species is an immunologically crossreacting polypeptide unrelated to the a polymerase. It is interesting to note, however, that it is much more abundant in rapidly dividing embryonic cells than in the more slowly replicating cultured cells.
It is clear from this and from previous studies (4, 5, 16) that DNA polymerase a is extraordinarily susceptible to proteolysis and possibly other modifications in vitro and that many of the proteolytic fragments retain catalytic activity, at least as judged by their ability to replicate activated DNA. It is highly likely, however, that such proteolysis influences the activity or function of the enzyme. In particular, association of the a and , L3 subunits with accessory proteins may be altered. As an example, the ability of the E. coli dnaB protein to associate with dnaC protein is abolished by brief trypsin digestion of dnaB protein, although there is no effect on its ATPase activity.' Similarly, trypsin treatment of phage T4 gene 32 protein abolishes cooperative binding of the protein to singlestranded DNA, although binding itself is unaffected (17). Purification and characterization of the form($ of DNA polymerase OL that exists in vivo are, therefore, essential prerequisites to the analysis of Drosophila chromosomal DNA replication in vitro.