Processive DNA Synthesis by DNA Polymerase I 1 Mediated by DNA Polymerase I 11 Accessory Proteins *

From the §Department of Biological Sciences, Molecular Biology Section, University of Southern California, Los Angeles, California 90089-1340, the (IHoward Hughes Medical Institute Department of Microbiology, Cornell University Medical College, New York, New York 10021, the IlDivision of Biochemistry and Molecular Biology, the University of California, Berkeley, California 94720, the **Department of Molecular Genetics, Center for Research and Development, Barcelona, Spain 08034, and the $$Department of Biological Chemistry and the Molecular Biology Institute, UCLA School of Medicine, Los Angeles, Californk 90024

An interesting property of the Escherichia coli DNA polymerase I1 is the stimulation in DNA synthesis mediated by the DNA polymerase I11 accessory proteins fl,y complex. In this paper we have studied the basis for the stimulation in pol I1 activity and have concluded that these accessory proteins stimulate pol I1 activity by increasing the processivity of the enzyme between 160-and 600-fold. As is the case with pol 111, processive synthesis by pol I1 requires both B,y complex and SSB protein. Whereas the intrinsic velocity of synthesis by pol I1 is 20-30 nucleotides per s with or without the accessory proteins, the processivity of pol I1 is increased from approximately five nucleotides to greater than 1600 nucleotides incorporated per template binding event. The effect of the accessory proteins on the rate of replication is far greater on pol I11 than on pol 11; pol I11 holoenzyme is able to complete replication of circular single-stranded M13 DNA in less than 20 s, whereas pol I1 in the presence of the y complex and B requires approximately 6 min. We have investigated the effect of B,y complex proteins on bypass of a site-specific abasic lesion by E. coli DNA polymerases I, 11, and 111. All three polymerases are extremely inefficient at bypass of the abasic lesion. We find limited bypass by pol I with no change upon addition of accessory proteins. pol I1 also shows limited bypass of the abasic site, dependent on the presence of fl,y complex and SSB. pol I11 shows no significant bypass of the abasic site with or without B,y complex.
Escherichia coli contains three DNA polymerases: pol I,' pol 11, and pol I11 (for a review, see Ref. 1). pol I11 is the main replicative polymerase (2, 3), and the functional holoenzyme contains at least 10 subunits (4). The polymerase function resides in the a subunit (5) and the 3' to 5' proofreading exonuclease is localized in the c subunit (6). These two proteins, along with a third subunit, 0, comprise the pol I11 core (7), an enzyme with relatively low synthetic processivity, synthesizing approximately 11 nucleotides per template binding event (8). The accessory proteins, / 3 and the five-protein y complex subassembly, form a P-sliding clamp on duplex DNA (9) which tethers the pol I11 core to the template thus converting pol I11 core into a highly processive enzyme capable of replicating long (Xi000 nucleotides) SSB coated singlestranded circular DNA viral templates in less than 20 s (4,8,10,11). pol I is a single-subunit enzyme (103 kDa) possessing polymerase as well as 3'45'-and 5'+3'-exonuclease activities (1). Since cells containing polA deletions are viable, pol I is not required for chromosome replication in E. coli (12).
However, under normal growth conditions, the 5'+3'-exonuclease of pol I acts on the lagging strand during replication to excise RNA primers from Okazaki fragments (1). pol I has also been shown to play an important role in repair of UV- Bonner et al. (22) showed that DNA pol I1 activity is increased 7-fold in cells induced for the SOS response and that the polymerase is capable of limited insertion and bypass of abasic lesions in uitro. The structural gene for pol I1 has recently been shown to be the same as the d i d gene (13,23), which is a member of the SOS regulon controlled directly by Lex A repressor (24). These results clearly demonstrate that pol I1 is an SOS-induced polymerase and likely participates in some aspect of error free or error prone DNA synthesis.
Evidence from earlier studies has suggested that the / 3 and y subunits of the pol I11 holoenzyme can stimulate DNA synthesis by pol TI (25). Recently, Hughes et al. (26) have shown with highly purified proteins that pol 11-dependent DNA synthesis is stimulated by the @,y complex accessory proteins of pol I11 holoenzyme. Since these accessory proteins are known to stimulate pol I11 activity by increasing the processivity of the enzyme (4,27,28), they might affect pol 11 activity by a similar mechanism. If so, addition of /3,y complex, perhaps in the presence of other SOS-induced proteins, such as UmuC and UmuD (15,16) may stimulate bypass of blocking lesions in DNA. In the work reported here, we have characterized the effect of /3,y complex accessory proteins on the processivity of pol 11. We have also compared the bypass of abasic lesions by all three E. coli DNA polymerases.

EXPERIMENTAL PROCEDURES
Materiaks-Unlabeled and labeled nucleotides were purchased from ICN Biochemicals. DNA oligonucleotides were synthesized on an Applied Biosystems 381A DNA synthesizer. The abasic site containing oligonucleotides were synthesized as described (29). M13 mp18 ssDNA uniquely primed with a synthetic DNA 30-mer was prepared as described (9). E. coli DNA polymerase I1 was prepared by a modification of a published procedure (22) from an overproducing strain' where phosphocellulose peak fractions were applied to a diethylaminoethyl cellulose (Whatman DE52) column followed by chromatography on an AGATP type 2 column (Pharmacia LKB Biotechnology Inc.). Other replication proteins were purified as described a (5), e (61, ac complex (30), y complex (31), pol I11 holoenzyme (32), and SSB (33). Concentration of proteins was determined by the method of Bradford (34) using bovine serum albumin as a standard. Terminal deoxynucleotidyltransferase (TdT) was purchased from Boehringer Mannheim, and T4 polynucleotide kinase was purchased from United States Biochemical.
Terminal Deoxynucleotidyltransferase Reactions-Oligonucleotides were extended in TdT buffer (40 pl volume) containing 2.5 p M oligonucleotide, 2.5 mM dNTP (0.6 mM each dATP, dCTP, dGTP, dTTP), and 100 units of TdT. Reactions were incubated at 37 "C for 2 h, followed by addition of another 100 units of TdT and an additional 1-h incubation. Reactions were then heated 10 min at 70 "C and cooled on ice. Extended oligonucleotides were phenol/ chloroform-extracted and ethanol-precipitated in the presence of sodium acetate as described (35). DNA concentrations were checked by absorbance at 260 nm. Extended oligonucleotides were 5' endlabeled using [y-3'P]ATP and T4 polynucleotide kinase (35) and electrophoresed on 8% denaturing polyacrylamide gels. Oligonucleotides were extended from 60 or 71 nucleotides to an average length of 150 nucleotides.
Polymerase Assays on Synthetic Oligonucleotides-15-mer DNA primers were 5' end-labeled using [y-3'P]ATP and T 4 polynucleotide kinase (35) and annealed to DNA oligonucleotides in annealing buffer by heating to 100 "C, followed by slow cooling (at least 2 h) to room temperature. Primers (15-mers) were annealed to templates 56 nucleotides from the template 5' end. Thus, extension of the primer to the end of the template yields products 71 nucleotides long. Polymerase reactions were in buffer A. Preinitiation complexes were formed in 10-pl volumes using 14 nM primed oligonucleotide (as 150-mer), 130 nM SSB (as tetramer), 0.5 mM ATP, 60 pM dCTP, 60 p M dGTP, 11 nM ( 3 (as dimer), 7 nM y complex, and either 38 nM at complex or pol I1 and incubated 5 min at 37 "C. Reactions were initiated by addition of dATP and dTTP to final concentrations of 60 p M each and terminated with the addition of 10 pl of 20 mM EDTA in 95% formamide. Samples of the polymerase reaction mixtures were heatdenatured at 100 "C for 5 min, cooled on ice, and loaded onto 10% polyacrylamide gels containing 8 M urea. Electrophoresis was performed at 2000 V for 2 h to resolve extended primers. Autoradiograms of gels were made by overlaying medical x-ray film (Kodak GPB-1) with an intensifying screen and exposing overnight at -70 "C. Alternatively, gels were exposed to phosphorimager cassettes overnight and scanned on a phosphorimager (Molecular Dynamics). Reactions on synthetic oligonucleotides containing a single abasic site were performed as described above except that 50 nM ac complex or pol I1 were used on 50 nM SSB-coated primed oligonucleotide.
Polymerase  ) were added on ice, and replication was initiated upon adding dNTPs to a final concentration of 60 p~ each. Reactions were carried out at 37 "C for 10 min and quenched with formamide/EDTA. Products were separated on 10% acrylamide-urea gels and visualized by autoradiography of dried gels. Only the products of replication from the 3' P end-labeled primer can be visualized on autoradiograms. The major replication products detectable by this experimental system are: (i) replication block, replication products that stop opposite the base before the abasic site (85 nucleotides); (ii) misincorporation products, replication products that terminate opposite the abasic site (86 nucleotides); (iii) bypass products, replication products that bypass the abasic site and terminate at the end of the template (116 nucleotides).

RESULTS
Stimulation of pol 11 by pol IIZ Accessory Proteins Using Synthetic Template-Primer Molecules-To examine the effect of the B,y complex accessory proteins on the processivity of pol 11, we have used both a synthetic primer-template system and uniquely primed single-stranded M13 DNA. The primertemplate region required to bind and hold onto the /3,y complex is quite large; approximately 36 bases upstream and downstream of the primer 3' terminus are req~ired.~ To satisfy these length requirements, we synthesized a 71-mer oligonucleotide and extended it with terminal deoxynucleotidyltransferase to produce average template lengths of 150 nucleotides. Extended templates were then annealed to a specific 5'-32Pend-labeled primer (15 nucleotides long), preincubated with SSB protein, @ and/or y complex accessory proteins, and pol 11. DNA synthesis was initiated upon addition of dNTP and denatured reaction products were resolved on 10% polyacrylamide gels, and primer extension was visualized either by autoradiography or phosphorimaging.
The effect of the pol I11 accessory proteins on synthesis by pol I1 is shown in Fig. 1. There were no full length DNA chains synthesized by pol I1 alone or pol I1 in the presence of either the @ or y complex accessory proteins added separately (Fig. 1, lanes 1-3). In these reactions there is a pattern of products ranging from one to 15 nucleotides added to the primer. However, addition of both @ and the y complex resulted in full length synthesis of the oligonucleotide within the shortest incubation period examined (10 s; Fig. 1, lane 4 ) . A primer-template reaction without added polymerase is shown to verify the lack of polymerase contamination in the accessory proteins (Fig. 1, lane 5 ) . The effect of @,y complex on the at subunits of pol I11 is shown for comparison (Fig. 1,  lanes 6 and 7). Only low molecular weight products are produced by pol I1 alone suggesting that the polymerase acts distributively or is very slow in elongation. The addition of both y complex and @, which forms a sliding clamp on the primed DNA, conferred onto pol I1 a rate greater than five nucleotides per s. A possible explanation of the data of Fig. 1 indicates that, as with pol 111, the effect of added @,y complex is to increase the affinity of pol I1 for the DNA primertemplate. No stimulation of pol I1 was observed using 71-mer primer-templates that were not extended by terminal transferase (data not shown), consistent with the stimulation being conferred by the accessory proteins since there is a minimum length of DNA required for the assembly of the accessory protein Effect of @,y Complex on pol 1 1 Using M13 ssDNA Templates-A study of the @ and y complex accessory proteins on either pol I1 or the at subunits of pol I11 on short synthetic templates (Fig. 1) demonstrated that the pol I11 accessory proteins stimulate both polymerases (Fig. 1, lanes 4 and 7). To investigate the differences in the degree to which @ and the y complex can stimulate pol I1 compared to pol 111, we used the longer template molecule, M13 mp18 single-stranded (ss) DNA primed with a synthetic DNA 30-mer in either the presence or absence of SSB. Products of DNA synthesis were analyzed in native agarose gels ( Fig. 2A. and alkaline agarose gels (Fig. 2B), and the total amount of nucleotide incorporated was also quantitated (Fig. 2C). In the absence of @,y complex, or in the presence of either @ or the y complex, the at polymerase showed no detectable product (Fig. 2, lanes 1, 2,  and 3). In the presence of @ and y complex the a t polymerase completed synthesis of some of the 7.2-kb M13 mp18 templates within 5 min (Fig. 2, A and B, lane 4 ) . pol I1 alone extended the primer on M13 mp18 slightly ( Fig. 2A, lane 9); the product length was below the 500-base resolution limit of the alkaline gel (Fig. 2B, lane 9). The extent of synthesis by pol I1 was unaffected by the presence of either @ or the y complex, or both @ and the y complex (Fig. 2, A-C, lanes 10-12).
Effect of SSB Protein on the @,y Complex Stimulation of pol 11-The presence of SSB protein is required to achieve highly processive synthesis by the pol I11 holoenzyme on ssDNA (8). The at polymerase was much more efficient at synthesis of the M13 mp18 template in the presence of SSB and its accessory proteins than in the absence of SSB (compare lanes 8 and 4 in Fig. 2). In order to determine the effect of SSB on pol I1 synthesis stimulated by @,y complex, we repeated the assay described for Fig. 2 in the presence of SSB. pol I1 alone was stimulated approximately 8-fold by SSB (compare lanes  9 and 13, Fig. 2C). Synthesis by pol I1 in the presence of SSB was unaffected by either @ or y complex, wherein the primer was extended approximately 1.5-2 kb (Fig. 2, A and B, lanes  13-15). However, in the combined presence of @,y complex, and SSB (Fig. 2, lane 16), pol I1 completely replicdted the M13 mp18 template. Hence, in the presence of SSB the accessory proteins had a profound effect on pol 11, but in the absence of SSB the accessory proteins had little effect on pol 11.

Accessory Proteins Do Not Increase the Intrinsic Velocity of
pol 11-To determine the rate of pol I1 replication in the presence and absence of accessory proteins, kinetic experiments were performed on uniquely primed M13 mp18 ssDNA "coated)) with SSB. An autoradiogram of a 0.8% agarose gel of the replication time courses is shown in Fig. 3. Three levels of pol I1 enzyme were used in these assays: 27 fmol (1 nM, 0.81 polymerase to primed circle), 270 fmol (10 nM, 81 polymerase to primed circle), 3500 fmol(l40 nM, 1001 polymerase to primed circle).
With approximately stoichiometric levels of pol I1 alone, little synthesis was detected before 6 min, and no full length product was observed (Fig. 3A). Addition of @,y complex, however, resulted in accumulation of full length replicative form I1 between 4 and 6 min (Fig. 3A). A somewhat different response was observed with the 8-fold molar excess of pol I1 over template (Fig. 3B). Without the accessory proteins, full length replicative form I1 was produced by 12 min. The  addition of the accessory proteins resulted in the appearance of full length replicative form I1 between 2 and 4 min. Yet another result was obtained with the huge molar excess of pol I1 (Fig. 3C). Now the pol I1 replicated the template to full length between 2 and 4 min whether the accessory proteins were present or not, indicating that the accessory proteins did not increase the intrinsic velocity of synthesis by pol 11. The accessory protein stimulation observed at low levels of pol I1 is likely due to an increase in processivity but may be explained by an increased rate of association with the primer template. In the next experiment (Fig. 4), it is shown that the B, r complex do increase the processivity of pol 11. However, the B, r complex must not provide pol I1 with sufficient processivity to allow it to completely replicate the M13 mp18 template in a single binding event. If pol I1 would not have dissociated from the accessory proteins for a full round of M13 mp18 synthesis, then 27-fmol (1 nM) pol I1 would have formed full length M13 mp18 DNA in the same amount of time as seen for 270-fmol (10 nM) pol 11.
The stimulation by @,r complex on the rates of DNA synthesis by pol I1 was quantified by measuring the kinetics of deoxyribonucleotide incorporation (Fig. 3). The rate of incorporation depended on the concentration of pol I1 when measured in the presence or absence of @,r complex. In the absence of the accessory proteins, the template was replicated a t a rate of 10 nucleotides per s with 270 fmol (10 nM) of pol II; with 27 fmol (1 nM) of pol 11, replication was about 10-fold slower. Addition of accessory proteins resulted in at least a 3fold stimulation of the rate of incorporation with 10 nM pol 11, to a value of 30 nucleotides per s. At 1 nM pol 11, inclusion of accessory proteins produced a 20-fold stimulation, from 1 to 20 nucleotides/s. Processivity of pol 1 1 in the Presence and Absence of Accessory Proteins-To measure the processivity of pol I1 in the absence of accessory proteins and SSB, an experiment using excess challenge DNA was performed using stoichiometric pol I1 with the extended oligonucleotide (-150-mer) primed with a 32P end-labeled DNA 15-mer (Fig. 4A). pol I1 was preincubated with the primed template and replication initiated by the addition of 10-fold excess unlabeled challenge DNA and unlabeled nucleotides. For reaction times between 30 s and 3 min following addition of challenge DNA, product distribution ranged from one to 20 nucleotides. The processivity of pol I1 alone is about five nucleotides as determined by computing the weighted average of integrated band intensities (Fig. 4A, 0.5-and 1-min reactions).
The processivity of pol I1 on SSB "coated" primed M13 mp18 ssDNA containing the accessory protein clamp was measured directly in an excess challenge DNA processivity assay (Fig. 4B). In this processivity experiment, stoichiometric pol I1 along with B,r complex were preincubated with M13 mp18 ssDNA primed with a 32P end-labeled DNA 30-mer and "coated" with SSB. Replication was initiated upon addition of a 4-fold excess of challenge primed (unlabeled) M13 mp18 ssDNA which also contained a preinitiation complex formed using the @,r complex accessory proteins. Hence, as pol I1 dissociates from the 32P end-labeled primer it will become trapped on the excess unlabeled challenge primed template. At various times the extension reaction was sampled and the  length of primer extension was analyzed in an alkaline agarose gel. The analysis shows at least some pol I1 remains associated with the initial 32P primed template for over 2 min. The length distribution of products finally stabilized by 3 min. Densitometric analysis of the smear of products identified the peak at approximately 1.6 kb with half-maximal intensity on either side of the peak at approximately 0.75 and 3.0 kb (not shown).
Hence, the processivity of pol I1 in the presence of p,r complex and SSB is approximately 1.6 kb. A 3-kb product synthesized over 2 to 3 min indicates an intrinsic velocity for pol I1 between 20 and 30 nucleotides per s. This agrees with the velocity determined in Fig. 3. A control analysis using the pol I11 ae complex, which is known to be processive over the length of this template, is shown in the three rapid time points on the right side of the gel.
An experiment identical to Fig. 4 was performed in which SSB was present but the accessory proteins were omitted. No products greater than 200 nucleotides were observed in the alkaline gel even after 22 min of replication (data not shown). Therefore, under these conditions, pol 11, in the absence of accessory proteins, gives no significant signal in the agarose gel. Analysis in a denaturing polyacrylamide gel of the pol I1 extension products in the absence of accessory proteins showed the products increased in size steadily throughout the time course and reached a length distribution of 50-180 nucleotides after 22 min of replication (data not shown). All the templates had been extended, consistent with the distributive action of pol I1 in the absence of accessory proteins. This analysis of pol I1 processivity on primed M13 mp18 was not extended to the case of p,r complex in the absence of SSB since the accessory proteins do not significantly stimulate pol I1 in the absence of SSB (see Fig. 2).
Effect of Accessory Proteins on Nucleotide Incorporation and Bypass of Abasit Sites by pol 1, 11, and 111-pol I11 and pol I1 have been implicated in SOS-induced error prone DNA repair synthesis. Since the accessory proteins enhance processivity of both polymerases, it was important to determine whether those proteins could affect the ability of the polymerases to insert nucleotides opposite a well-defined abasic (apurinic/ apyrimidinic) template lesion and/or stimulate extension be- and at were assayed with and without the addition of 0 ,~ complex for insertion and bypass at the abasic site. The results are shown in Fig. 5. pol I1 exhibited slight misincorporation of a nucleotide opposite the abasic site with little extension beyond the abasic site (Fig. 5, lanes 5 and 6 ) . Addition of accessory proteins had little effect on misinsertion; however, bypass of the abasic site increased from barely detectable to about 2% bypass (Fig. 5, lane 8). No p ,~ complex effect was seen in the absence of SSB (data not shown). In contrast, pol I11 exhibited no detectable misincorporation or bypass with or without the accessory proteins (Fig. 5, lanes 1-4). The synthetic oligonucleotide template is not optimal for analyzing the role of processivity subunits because the replication blocking lesion is so close the primer (15 nucleotides). To avoid this difficulty, the 60-mer oligonucleotide containing a single abasic site was ligated into linear 4x174 DNA, allowing us to investigate translesion replication when the blocking lesion is more distant from the primer terminus. The assay used to measure polymerase incorporation and bypass of abasic sites with the longer template is shown in Fig. 6A. A single abasic site was introduced a t a unique location. The 4x174 DNA was primed 86 nucleotides upstream from the abasic lesion with a 32P end-labeled 20-mer oligonucleotide; replication by the various polymerases and accessory proteins was initiated as described under "Experimental Procedures,'' and the products were resolved on denaturing gels. The size of the replication products indicates whether synthesis is completely blocked at the abasic site (85 nucleotide product), if a nucleotide can be incorporated opposite the abasic site (86 nucleotide product), or if the lesion can be bypassed (116 nucleotide product). However, if the misincorporation opposite the abasic site is very limited, the resolution of this band from the intense band before the lesion is very difficult.
The results of the various polymerase reactions are shown in Fig. 6B. Bypass of the abasic lesion was very inefficient for all three DNA polymerases, and so the gel is overexposed to show the differences. Because of the extremely limited misincorporation, we were unable to make a reliable comparison of this aspect of the system among the polymerases. The extent of bypass was estimated from a densitometer tracing of this exposure (for bypass) compared to a lower exposure in which the relative amount of blocked replication product could be measured. Significant lesion bypass (approximately 3%) was observed for pol I (Fig. 6B, lane I). There was no effect of B,-y complex on pol I-dependent bypass at the abasic site ( l a n e 2). A barely detectable bypass of the abasic lesion was observed for pol I1 alone (Fig. 6B, lane 3). Addition of SSB protein inhibited both replication and bypass by pol I1 (Fig. 6B, lane 4 ) , an effect also seen on synthetic oligonucleotides (data not shown). The accessory proteins alone had little effect on bypass of the abasic site catalyzed by pol I1 (Fig. 6B, lanes 5 and 6 ) . In contrast, addition of SSB, B,-y complex increased replication up to and exceeding the abasic lesion (approximately 2%) (Fig. 6B, lanes 7 and 8). A third pol I11 accessory protein, 7 , which increases the processivity as tetramer) in 141 pl of buffer A containing 0.5 mM ATP and 60 p M each dCTP and dGTP. Replication was initiated upon addition of an excess DNA challenge mixture consisting of 5.9 nM M13 mp18 ssDNA primed with the same DNA 30-mer (but unlabeled) which also contained a preinitiation complex upon a 6-min incubation at 37 "C with SSB (1.2 p M as tetramer), 0 (11 nM as dimer), and y complex (2 nM) in 159 pl of buffer A containing 60 p~ each dCTP and dGTP, and 120 p~ each dATP and dTTP. At the times indicated, 42-pl aliquots were removed and quenched with 40 p1 of 1% sodium dodecyl sulfate, 40 mM EDTA. Reactions were analyzed by electrophoresis at 40 V for 20 h in a 1.2% alkaline agarose gel. The gel was neutralized, dried, and exposed to x-ray film. The three lanes on the right side of the gel are time points from a control reaction which was performed in a similar manner except pol I11 ac complex replaced pol 11. The positions of the size standards visualized by ethidium bromide staining induced fluorescence after neutralization of the alkaline gel are marked on the right side of the gel. Only the products of replication from the end-labeled primer can be visualized on the autoradiograms of denaturing acrylamide gels. Replication blocked at a base before the abasic site gives a product of 85 bases, misincorporation opposite the abasic site gives a product of 86 bases, and the complete bypass product is 116 bases. B, translesion replication reactions using the three E. coli DNA polymerases were performed as described under "Experimental Procedures." Lanes from left to right: pol I (Klenow fragment); pol I (Klenow) + @ + y complex; pol 11; pol I1 + Ssb; pol I1 + @ + y complex; pol I1 + @ + y complex + T; pol I1 + Ssb + @ + y complex; pol I1 + Ssb + @ + y complex + T; ac subassembly of pol HI; a 6 complex + Ssb+ @ + y complex; pol 111 holoenzyme (naturally purified).
of pol I11 core (28), was assayed with pol I1 using the abasic site template (Fig. 6B, lanes 6 and 8). Addition of 7 had no effect on bypass of the abasic site. In addition, three different assemblies of pol I11 were assayed on the abasic lesion-containing template: the complete 10-subunit holoenzyme (HE, lanes 11 and 12), the two subunit a e (lane 9 ) , and the eightsubunit ac,B,y complex (lane 10). Of the enzymes tested, the three forms of pol I11 showed the most restricted bypass at the abasic lesion (less than 0.5%).  (26) showed that pol I1 activity was markedly enhanced by the presence of B,y complex.
With highly purified DNA polymerase 11, we have investigated the functional interaction of pol I1 with the sliding clamp formed by the P,y complex accessory proteins of the pol I11 holoenzyme, using as templates synthetic oligonucleotides and long single-stranded viral DNA with and without unique abasic lesions. We have concluded that the primary effect of P,y complex is to increase pol I1 processivity between 150-and 600-fold. Similarities in the effects of B,y complex on pol I1 and pol I11 include: requirement for SSB protein; a minimum template-primer length requirement for assembly of the active complex; optimized stimulation in activity at approximately stoichiometric levels of P,y complex with each polymerase. However, there is also an important difference in that the accessory proteins enhance pol I11 processivity to a far greater extent than pol 11. The pol I11 complex is able to catalyze complete replication of M13 mp18 ssDNA circles within about 20 s, whereas the pol II-B,y complex cannot complete a full round of M13 mp18 synthesis without dissociation and even a t saturation requires between 2 and 4 min to complete the circular template (Fig. 3). Under assay conditions in which pol I1 is approximately stoichiometric with primer, pol I1 alone incorporates nucleotides at a rate of approximately 1 per s. This rate is increased 20-fold in the presence of a stoichiometric amount of /3,y complex and SSB. This correlates well with the observed processivity for the pol II/B/-y complex (Fig. 4). A t 20 nucleotides per s, a residence time of 2-3 min predicts a product of 2.4-3.6 kb, well within the 0.75-3-kb processivity measured in Fig. 4.
The faster optimized rate of synthesis for pol I11 compared to pol I1 may reflect differences in the biological functions of the two enzymes; pol I11 for rapid chromosomal replication, and pol I1 perhaps for specialized "limited extent" synthesis accompanying repair. The stoichiometry required to achieve optimal stimulation of pol I1 by /3,y complex suggests that the accessory protein-pol I1 interactions are specific, and therefore, likely to be biologically important. For comparison, we find that pol I and bacteriophage T4 polymerase also exhibit more rapid DNA synthesis in the presence of /3,y complex, but the degree of stimulation is small and occurs only at extreme substoichiometric levels of the polymerase (-50 mol of complex per mol of pol I or T4 polymerase)? There are no obvious conserved amino acid sequences or domains that might imply a structural relationship between pol I1 and the a subunit of pol 111. Thus, interactions of pol I1 and pol I11 a subunit with P,y complex are likely to involve external noncatalytic regions containing common tertiary structural elements.

Activity of DNA Polymerases at a Replication-blocking Le-
sion-Although pol I1 has been shown to incorporate nucleotides opposite abasic sites in vitro, it extends these inefficiently (22). Since it has been suggested that factors which increase polymerase processivity may play a role in translesion replication (37)(38)(39)(40), we investigated the effect of pol I11 accessory proteins on bypass by pol I, pol 11, and pol I11 of one particular lesion, an abasic site. Addition of accessory proteins to E. coli polymerases I and I11 had little or no effect on lesion bypass, whereas addition to pol I1 resulted in increased lesion bypass. Misincorporation and bypass were markedly less for pol I11 than for pol 11 (Figs. 5 and 6). It is important to note that abasic site bypass is extremely inefficient for all polymerases. Since genetic data implicate pol I11 in lesion bypass and pol I11 holoenzyme has extraordinary processivity, factors other than (or in addition to) increased processivity are likely to be required for translesion replication by pol 111.
Although recent observations suggest a specialized role for pol I1 in one or more DNA repair pathways requiring SOS induction, pathways requiring intervention of pol I1 remain to be elucidated. We have recently constructed a new insertion mutation in the pol I1 gene which removes both transcription and translation start signals and more than 50% of the 5' portion of the gene.' In preliminary experiments, we find that this mutant appears to have a normal growth rate. Thus, it now appears virtually certain that pol I1 is not required for replication of E. coli; however, the observation that the activity of pol I1 is stimulated significantly by stoichiometric levels of the pol I11 accessory proteins suggests that pol I1 might utilize these interactions in uiuo. The availability of mutants should allow investigations into the possible role of pol I1 in DNA repair synthesis, either cooperatively with or as a backup to pol I and 111.