Multiple forms of DNA polymerase in mouse myeloma

History of DNA polymerase β X-ray crystallography

DNA polymerase β (Pol β) is an essential mammalian enzyme involved in the repair of DNA damage during the base excision repair (BER) pathway. In hopes of faithfully restoring the coding potential to damaged DNA during BER, Pol β first uses a lyase activity to remove the 5'-deoxyribose phosphate moiety from a nicked BER intermediate, followed by a DNA synthesis activity to insert a nucleotide triphosphate into the resultant 1-nucleotide gapped DNA substrate. This DNA synthesis activity of Pol β has served as a model to characterize the molecular steps of the nucleotidyl transferase mechanism used by mammalian DNA polymerases during DNA synthesis. This is in part because Pol β has been extremely amenable to X-ray crystallography, with the first crystal structure of apoenzyme rat Pol β published in 1994 by Dr. Samuel Wilson and colleagues. Since this first structure, the Wilson lab and colleagues have published an astounding 267 structures of Pol β that represent different liganded states, conformations, variants, and reaction intermediates. While many labs have made significant contributions to our understanding of Pol β, the focus of this article is on the long history of the contributions from the Wilson lab. We have chosen to highlight select seminal Pol β structures with emphasis on the overarching contributions each structure has made to the field.

POLB: A new role of DNA polymerase beta in mitochondrial base excision repair

The mitochondrial genome is a matrilineally inherited DNA that encodes numerous essential subunits of the respiratory chain in all metazoans. As such mitochondrial DNA (mtDNA) sequence integrity is vital to organismal survival, but it has a limited cadre of DNA repair activities, primarily base excision repair (BER). We have known that the mtDNA is significantly oxidized by both endogenous and exogenous sources, but this does not lead to the expected preferential formation of transversion mutations, which suggest a robust base excision repair (BER) system. This year, two different groups reported compelling evidence that what was believed to be exclusively nuclear DNA repair polymerase, POLB, is located in the mitochondria and plays a significant role in mitochondrial BER, mtDNA integrity and mitochondrial function. In this commentary, we review the findings and highlight remaining questions for the field.

Inhibition of influenza virus transcription by 2'-deoxy-2'-fluoroguanosine

The nucleoside analog 2'-deoxy-2'-fluoroguanosine (2'-fluorodGuo) is phosphorylated by cellular enzymes and reversibly inhibits influenza virus replication in chick embryo cells within the first 4 h of infection. RNA hybridization studies revealed that primary and secondary transcription of influenza virus RNA were blocked at a compound concentration of 10 microM, but no inhibition of cell protein synthesis was seen even at high compound concentrations (200 microM). In vitro, the triphosphate of 2'-fluorodGuo is a competitive inhibitor of influenza virus transcriptase activity from disrupted virus, with a Ki of 1.0 microM. The cellular polymerases DNA polymerase alpha and RNA polymerase II were only weakly inhibited or were insusceptible to 2'-fluorodGTP. In kinetic studies with the influenza virus transcriptase, 2'-fluorodGTP, in the absence of GTP, blocked elongation of the virus RNA chain. Similarly, by using purified ribonucleoprotein complexes it was found that the addition of a single nucleotide of 2'-fluorodGTP to the virus RNA caused chain termination, which resulted in the blockage of further virus transcription. Furthermore, the specificity for influenza virus transcriptase was confirmed when the transcriptase from partially resistant virus was found to be 10-fold less susceptible to 2'-fluorodGTP (Ki = 13.1 microM).

Comparison of pyrimidine nucleotide synthetic enzymes involved in 5-fluorouracil metabolism between human adenocarcinomas and squamous cell carcinomas

The activities of orotate phosphoribosyltransferase (OPRT), cytidine triphosphate (CTP) synthetase, deoxycytidine monophosphate (dCMP) deaminase, thymidine monophosphate (dTMP) kinase, uridine (Urd) kinase, thymidine (dThd) kinase, Urd and dThd phosphorylases, and DNA polymerase were examined in the eight human lung squamous cell carcinomas and five lung adenocarcinomas, and five tumor-adjacent normal lung tissues. All of these enzymes are involved in pyrimidine nucleotide synthesis. The metabolism of 5-fluorouracil (5-FU) was determined. The levels of these enzymes, except for OPRT, were high in tumor tissues and almost the same between lung squamous cell carcinomas and adenocarcinomas, with no statistical difference. The activities for phosphorylation and degradation of 5-FU were similar in each tissue type of tumor. As 5-FU is incorporated into tumor cells and is metabolized actively to 5-FU nucleotides in squamous cell carcinoma tissues, at almost the same level seen in adenocarcinoma tissues, this drug should have a wide clinical application.

Stimulation of human neuroblastoma DNA polymerase alpha and primase activities by a protein factor isolated from rat liver chromatin

Nuclear protein factor type 1 (NPF-1) that simulates IMR-32 primase-associated DNA polymerase alpha 1 and alpha 2 activities has been purified from a high-salt extract of liver chromatin from 6-month-old rats. The final purified factor lacks DNA polymerase alpha, RNA polymerase, and DNA-unwinding or topoisomerase type I activities. The stimulatory activity is destroyed by trypsin (60 min at 37 degrees C), DNase II (60 min at 37 degrees C), and heat treatment (2 min at 68 degrees C). The 125I-labeled NPF-1 does not bind to activated calf thymus DNA or poly(dC). However, it forms a ternary complex with DNA in the presence of DNA polymerase alpha-primase complex (alpha 1 and alpha 2). The ternary complex sediments on sucrose density gradient as a heavier band (11S). The NPF-1 also stimulates (2.5-fold) primase-catalyzed incorporation of GMP and dGMP from the corresponding triphosphates on poly(dC) template even in the presence of a high concentration of alpha-amanitin (400 micrograms/ml). The labeled duplex containing the poly(dC) template, [32P]-GTP, and [3H]dGTP loses 80% of the 32P label and 70% of the 3H label after treatment with 0.3 M KOH and DNase I, respectively. The products were isolated from reaction mixtures incubated with and without NPF-1 and subjected to alkaline sucrose-density-gradient sedimentation analysis. The results suggest that the rate of synthesis of DNA short chains is increased in the presence of NPF-1 without a concomitant increase in the chain length of the newly synthesized products.

DNA polymerase and simian virus 40 infection of resting monkey cells: induction of aphidicolin resistant alpha-polymerase

Tightly confluent monkey cell lines BSC-1 and CV-1 held in stale medium for several days exhibited an extremely low level of thymidine incorporation into cellular DNA. Yet, these cells contained a level of alpha-polymerase equal to about 15% of the level in rapidly dividing cells, and they still were capable of supporting replication of SV40 DNA. SV40 infection and culture in stale medium resulted in a four-fold induction of alpha-polymerase in CV-1 cells, whereas no change in alpha-polymerase level was observed in BSC-1 cells. Characterization of alpha-polymerase partially purified from infected CV-1 cells revealed that 80-90% of the enzyme activity was aphidicolin resistant. SV40 DNA replication in resting CV-1 cells, however, was aphidicolin sensitive. SV40 infection of resting CV-1 cells may induce an aphidicolin-resistant enzyme or lead to a modified alpha-polymerase species.

Evidence in patients with systemic lupus erythematosus of the presence of antibodies against RNA-dependent DNA polymerase of baboon endogenous virus

Immunoglobulin G (IgG) in six out of 30 patients with systemic lupus erythematosus (SLE) strongly inhibited the activity of RNA-dependent DNA polymerase (RDPase) of baboon endogenous virus, M7, while IgG obtained from scleroderma patients, rheumatoid arthritis patients and normal subjects was less reactive. Experiments with anti-human IgG and with IgG F (ab')2-bound immunoaffinity columns indicated that the inhibition of RDPase was antibody-mediated. The RDPase inhibiting activity of SLE IgG was considered not to be due to cross-reactions of anti-nuclear antibodies including anti-DNA, anti-ribonucleoprotein, anti-Sm and anti-SS.B antibodies. SLE IgG preferably inhibited the RDPase activity of baboon endogenous virus and a feline endogenous virus, RD114. These findings support the hypothesis that retrovirus(es) might be involved in SLE.

Immunocytochemical localization of chick DNA polymerases alpha and beta +

An immunofluorescent method using specific antibodies was employed to detect DNA polymerases alpha and beta in chick cells. With monoclonal antibodies produced by four independent hybridoma clones, most of the DNA polymerase alpha was shown to be present in nuclei of cultured chick embryonic cells. With a polyclonal, but highly specific, antibody against DNA polymerase beta, this enzyme was also shown to be present in nuclei. DNA polymerase alpha was detected in proliferating cells before cell contact and in lesser amount in resting cells after cell contact, indicating that its content is closely correlated with cell proliferation. On the other hand, similar amounts of DNA polymerase beta were detected in proliferating and resting cells. Furthermore, DNA polymerase beta was detected in nuclei of most cells, while DNA polymerase alpha was detected only in large round nuclei in seminiferous tubules of chick testis. DNA polymerase alpha is presumably present in cells that are capable of DNA replication, and during the cell cycle it seems to remain in the nuclei during the G1, S, and G2 phases, but to leave from condensed chromatin for the cytoplasm during the mitotic phase.

The inhibition of the RNA polymerase activity of influenza virus A by pyrophosphate analogues

Substituted methylene diphosphonates are effective inhibitors of the RNA polymerase of influenza A virus causing 50% inhibition of the polymerase activity when they are present in the concentration range 10-85 microM. The inhibitory power of the methylene diphosphonates appears to be related to their ability to chelate with metal ions.

Probable involvement of a glycoconjugate in IMR-32 DNA synthesis: decrease of DNA polymerase alpha 2 activity after tunicamycin treatment

Multiple forms of DNA polymerase alpha activity (alpha 1, alpha 2, and alpha 3) from human neuroblastoma IMR-32 cells untreated or treated with tunicamycin (3 microgram/ml) were separated by DEAE-cellulose column chromatography. Loss of 40--60% of DNA polymerase alpha 2 activity was observed in tunicamycin-treated cells. Ricin 1B, a subunit of intact ricin (Mr, 64,000), was found to be a specific inhibitor of DNA polymerase alpha 2 isolated from control IMR-32 cells. However, DNA polymerase alpha 2 isolated from tunicamycin-treated cells was insensitive to ricin 1B. Heat treatment studies at 50 degrees C showed two completely different inactivation profiles for the DNA polymerase alpha 2 enzymes isolated from the tunicamycin-treated and untreated cells. A probable involvement of a beta-linked galactose-containing carbohydrate chain in the catalytic subunit of DNA polymerase alpha 2 is suggested.

Accessory proteins for DNA polymerase alpha activity with single-strand DNA templates

Three forms of DNA polymerase alpha [DNA nucleotidyltransferase (DNA-directed), EC 2.7.7.7] were partially purified from the combined nuclear extract and postmicrosomal supernatant solution of synchronized HeLa cells. These enzymes, designated DNA polymerases alpha 1, alpha 2, and alpha 3, on the basis of their order of elution from DEAE-Bio-Gel, differ in their abilities to utilize single-strand DNA templates. DNA polymerase alpha 2 has equal catalytic activities with activated and single-strand DNAs as template-primers. DNA polymerase alpha 1 has only partial catalytic activity with single-strand DNA templates, and DNA polymerase alpha 3 is essentially inactive with this template. Successive steps of hydrophobic affinity chromatography and phosphocellulose chromatography of DNA polymerase alpha 2 resolved the polymerase alpha activity and two protein factors (C1 and C2) that are required for its catalytic activity with a DNA template-primer that contains extended single-strand regions. In the absence of the factors, DNA polymerase alpha activity is measurable with activated but not single-strand DNA templates. In the presence of the C1 and C2 factors DNA polymerase alpha activity with single-strand DNA templates is restored to about 75% of the catalytic activity of DNA polymerase alpha 2 with this template.

Preferential utilization of bromodeoxyuridine and iododeoxyuridine triphosphates by DNA polymerase gamma in vitro

Thymidine triphosphate (TTP) and its halogenated analog bromodeoxyuridine triphosphate (BrdUTP) were compared in vitro as substrates for several prokaryotic and eukaryotic DNA polymerases to determine a possible enzymatic preference which might account for the reported finding of nonrandom patterns of incorporation of the analog in eukaryotic cellular DNA as well as help clarify the mechanism for drug-induced activation of latent retroviruses from animal cells. Following nucleotide competition reactions, no discriminatory utilization was detected from a mixture containing equimolar [3H]TTP and [alpha 32P]TTP for any of the polymerases. In contrast, when [3H]BrdUTP was mixed with an equal concentration of [alpha 32P]TTP, it was apparent that eukaryotic DNA polymerase gamma utilized more of the brominated analog triphosphate in preference to the unsubstituted compound. This increased affinity was confirmed by the differences in Km values. Furthermore, the selectivity of polymerase gamma was even more pronounced with the iodinated thymidine analog iododeoxyuridine triphosphate. On the other hand, polymerase gamma failed to discriminate as readily between equal concentrations of alpha 32P-labeled deoxycytidine triphosphate and 125I-labeled iododeoxycytidine triphosphate.

Differential inhibition of multiple forms of DNA polymerase alpha from IMR-32 human neuroblastoma cells

Three forms of DNA polymerase (pol) alpha from human neuroblastoma IMR-32 were separated by DEAE column chromatography. All sedimented at approximately 7 S in 5-20% continuous sucrose density gradients. All were heat labile, with pol alpha 2 the most (90% inactivated) and pol alpha 3 the least (50% inactivated) sensitive to heating for 5 min at 50 degrees C. pol alpha 1 and alpha 2 efficiently utilized activated calf thymus DNA as template. The most active form, pol alpha 2, used both poly(dA).(dT)12-18 and poly(dT).(dA)12-18 as template at equal rates. Differential inhibition of DNA polymerase alpha activities was examined in the presence of ricin, hemin, and a nonhistone chromatin protein. All three polymerases were inhibited by both ricin (nonreduced) and hemin, with pol alpha 2 the most (80-90%) and pol alpha 3 the least (60%) sensitive in each case. In contrast, only pol alpha 2 and alpha 3 activities were inhibited (80-85%) by rat liver nonhistone chromatin protein.

DNA polymerase of a basidiomycete fungus, Coprinus cinereus

DNA polymerase activity was studied in Coprinus cinereus, a basidiomycete fungus. Only one from of the enzyme could be demonstrated, whether by affinity or ion-exchange chromatography; this enzyme had a molecular weight of 185000 on Sephadex G-200, and was inhibited by mercaptoethanol. Coprinus, a representative of the most advanced type of the filamentous fungi, resembles other eukaryotic micro-organisms in its lack of a mammalian beta-type DNA polymerase. The properties of the polymerase are compared with those of two other fungi, and found to resemble most closely the yeast polymerase A in Mg2+ requirements and template preference.

Further studies on partially purified calf thymus DNA polymerase a

Attempts to prevent the urea conversion of a 200-230,000 molecular weight DNA polymerase alpha to a 150-170,000 molecular weight form by the inclusion of protease inhibitors have not been successful. No other method has been found capable of dissociating a 50-70,000 fragment or subunit from the DNA polymerase subunit. Addition of this 50-70,000 subunit to the polymerase subunit does not aid the binding of the enzyme to DNA, but does have an effect on the utilisation of synthetic template-initiator complexes by the polymerase subunit.

Inhibition of human neuroblastoma DNA polymerase activities by plant lectins and toxins

The effects of concanavalin A and ricin (RCAII, Mr 65,000) on [3H]thymidine incorporation into human neuroblastoma IMR-32 DNA showed reduction of total DNA synthesis to 50% and 70% of control, respectively. Two DNA polymerase (DNA nucleotidyltransferase, EC 2.7.7.7.) activities (alpha and beta) involved in the biosynthesis in vitro of DNA were separated by sucrose density gradient centrifugation from IMR-32 cell homogenate. The DNA polymerase alpha activity was also purified by selective precipitation with polyethylene glycol (Mr 6000) followed by agarose-concanavalin A column chromatography. The activities of both DNA polymerases were examined at various concentrations of mutagenic and nonmutagenic plant agglutinins and the toxin ricin. Concanavalin A and ricin specifically inhibited DNA polymerase alpha activity (activity reduced to 19% and 10%, respectively), whereas DNA polymerase beta activity was inhibited (reduced to 16%) by red kidney bean agglutinin (PHA-P).

DNA polymerase alpha and beta in the California urchin

DNA polymerase alpha and beta were identified in the urchin, Strongylocentrotus purpuratus. The DNA polymerase beta sedimented at 3.4 S, constituted 5% of total DNA polymerase activity, and was resistant to N-ethylmaleimide and high ionic strength. The polymerase alpha sedimented at 6--8 S, was inhibited by N-ethylmalemide or 0.1 M (NH4)2SO4, and was dependent upon glycerol for preservation of activity. Both the polymerases alpha and beta were nuclear associated in embryos. The DNA polymerase alpha was markedly heterogeneous on DEAE-Sephadex ion exchange and showed three modal polymerase species. These polymerase alpha species were indistinguishable by template activity assays but the DNA polymerase associated ribonucleotidyl transferase (Biochemistry 75 : 3106-3113, 1976) was found predominantly with only one of the DNA polymerase alpha species.

The heterogeneity of cytoplasmic deoxyribonucleic acid polymerase from Physarum polycephalum

Cytoplasmic DNA polymerase (DNA deoxynucleotidyltransferase, EC 2.7.7.7) was partially purified from Physarum polycephalum. The first step of the purification procedure utilized the fact that the enzyme on gel filtration behaves in anomalous fashion. The second step was either ion-exchange chromatography or sucrose-density-gradient centrifugation. The partially purified DNA polymerase was heterogeneous and at least four species with different sedimentation coefficients (5.5S, 7.2S, 8.6S and 11.5S) were detected. Calculated molecular weights indicated a tendency for stoicheiometric polypeptide aggregation, accompanied by an alteration of the three-dimensional structure froma compact spheroid to a more open elliptical form. Sodium dodecyl sulphate/polyacrylamide-gel electrophoresis and computed molecular weights suggest an active protomer in the range of 113000 daltons; all data pertain to I 0.045, which was maintained during the whole procedure.

Studies on the purification and properties of a 6.8-S DNA polymerase activity found in calf-thymus DNA polymerase-alpha fraction

The heterogeneity of calf thymus DNA polymerase-alpha has been further investigated. In particular, an enzyme (enzyme D) which exhibits higher activity on poly(dA) . (dT)10 (A:T = 20:1) compared with that on activated DNA, has been further purified and its properties compared with two other activities of the DNA polymerase-alpha fraction (enzymes A1 and C) which do not show a preference for poly(dA) . (dT)10 over activated DNA. As with A1 and C, enzyme D was shown to have many of the characteristic properties of DNA polymerase-alpha in that it is an acidic protein as judged by its binding to DEAE-cellulose, has a molecular weight of about 140000, does not use a poly (A) . (dT)10 template-initiator complex and is inhibited by N-ethylmaleimide. It exhibits anomalous gel filtration behaviour on Sepharose 6B and it binds relatively weakly to DNA-cellulose compared with DNA polymerase-beta. The extreme sensitivity of enzyme D to inhibtion by N-ethylmaleimide distinguishes it from A1 and C, as does its elution position from a DEAE-cellulose column. On the other hand enzymes C and D are readily inactivated by heating at 45 degrees C unlike enzyme A1. The possible interrelationships of the multiple activities of calf thymus DNA polymerase-alpha are discussed.

DNA polymerase activities in differentiating mouse neuroblastoma N-18 cells

The activities of two DNA polymerases (DNA nucleotidyltransferases) were characterized in mouse neuroblastoma clone N-18 on the basis of their apparent molecular weights (determined by sucrose density gradient centrifugation: polymerase-alpha, 7.5-8 S; polymerase-beta, 3-4 S) and relative inhibition by sulfhydryl-blocking agents. N-Ethylmaleimide (10 mM) and iodoacetamide (1.5 mM) inhibited DNA polymerase-alpha activity completely, whereas only 35-40% inhibition was observed for DNA polymerase-beta under similar conditions. DNA polymerase-alpha activity was reduced 50-70% in N-18 cells that had been induced to differentiate by 4 micro M bromodeoxyuridine, and the low molecular weight DNA polymerase-beta activity remain unchanged. With activated calf thymus DNA as template, only DNA polymerase-alpha activity was stimulated in the presence of added ribonucleotides and purified Escherichia coli RNA polymerase.

Novikoff hepatoma deoxyribonucleic acid polymerase. Identification of a stimulatory protein bound to the beta-polymerase

The Novikoff hepatoma DNA polymerase-beta sediments as a 7.3S form in crude extracts but during purification sediments as a 4.1S form (after diethylaminoethyl-Sephadex chromatography) or as a 3.3S form (after DNA-cellulose chromatography). If 0.25 M ammonium sulfate or 0.5 M NaCl is included in the sucrose gradients, the 7.3S form sediments at 3.3 S; after removal of the salt, it sediments again at 7.3 S, indicating the reversibility of the aggregation phenomenon. By careful adjustment of ionic strength in the gradient, four distinct and reproducible forms of the enzyme sedimenting at 7.3, 5.8, 4.1, and 3.3 S can be generated. The isoelectric point of the DNA polymerase also changes during purification; the 7.3S form has a pI of 7.5, while the 4.1S form isoelectrically focuses at a pH of 8.5. During DNA-cellulose chromatography, the Novikoff beta-polymerase is separated from a stimulatory factor designated as Novikoff factor IV. Factor IV is a protein as shown by its sensitivity to protease and resistance to nucleases. It is responsible for converting the 3.3S enzyme to the 4.1S form since the 3.3S homogeneous DNA polymerase-beta sediments at 4.1 S in the presence of factor IV. Factor IV confers stability to the polymerase in low ionic strength buffers as well as stability to heat denaturation. Factor IV has the ability to increase the activity of the 3.3S homogeneous polymerase by about fourfold.

Studies on DNA alpha-polymerase of mouse myeloma: partial purification and comparison of three molecular forms of the enzyme

Activity of DNA alpha-polymerase in extracts from MOPC-104E was not associated with a single protein molecule, but with several molecular species that differed in isoelectric point. The three most abundant of these enzyme species were first separated from other DNA polymerases and then resolved from each other by repeated chromatography on diethylaminoethylcellulose columns. Next, with the use of glycerol gradient centrifugation and DNA-cellulose column chromatography, the three species were further purified to a state representing more than 5000-fold purification over the crude extract. These three highly purified enzyme species exhibited very similar catalytic properties, and the main activity of each species sedimented at the same rate (6-7S) in glycerol gradients containing 0.5 M KCl. Analysis of the polypeptide content of each species revealed that polypeptides of about 150 000 and 60 000 daltons cofractionated with the DNA polymerase activity. The multiple alpha-polymerase species did not appear to result from in vitro proteolytic cleavage, since multiple species were observed in extracts prepared under several different types of conditions, including the presence of the protease inhibitors, phenylmethanesulfonyl fluoride, or trasylol. The three species were recovered in about the same relative amounts from both the nuclear and cytoplasmic fractions of MOPC-104E, and it appeared that multiple species of alpha-polymerase were also present in extracts from fetal bovine liver.

Evidence that DNA polymerase-alpha of calf thymus contains a subunit of molecular weight 155000

Small samples of the 8-S species of enzymes (A1 and A2) occurring in the DNA polymerase-alpha fraction of calf thymus, have been extensively purified using non-denaturing (normal) polyacrylamide gel electrophoresis. When peak fractions of activity on normal gels were subjected to dodecylsulphate-polyacrylamide gel electrophoresis, a polypeptide at 155000 correlated with polymerase activity. Samples of the 7.3-S (C) enzyme prepared from A2 by treatment with 2.4 M urea or isolated directly without exposure to urea, also showed the presence of a 155000-Mr polypeptide. It is concluded that the 7.3-S (C) enzyme, of previously estimated molecular weight 155000-170000, is a single polypeptide and that the 8-S enzymes A1 and A2 contain an additional subunit of 50000-70000 molecular weight.

DNA polymerases of rat liver. Partial characterization and effect of various inhibitors

Three distinct DNA-dependent DNA polymerase activities have been partially purified from normal rat liver. Soluble activities are separable into two distinct fractions (P1 and P2) by phosphocellulose chromatography. A low-molecular-weight DNA polymerase was isolated from purified nuclei. The enzymes were characterized according to chromatographic and sedimentation behavior, enzymological properties, and response to various inhibitors. The results indicate that fraction P1 corresponds to the high-molecular-weight enzyme and suggest that polymerase P2 may be derived from partial dissociation of the high-molecular-weight enzyme. The molecular weight of polymerase P1 was estimated to be about 250 000 by Sephadex column chromatography. Both fraction P2 and nuclear DNA polymerase appeared to be low-molecular-weight enzymes. However, the molecular size of these activities was apparently different. The estimated molecular weights of nuclear and P2 enzyme are about 40 000 and 25 000, respectively. As with the nuclear enzyme, polymerase P2 (but not P1) appeared to be free of detectable exonuclease activity. All of these polymerases showed a marked preference for initiated polydeoxyribonucleotide templates. The rat liver polymerases differed in their ability to use poly[d(A-T)-A1 primer-template, as is shown by the ratios of their activity with this synthetic polymer to that with activated DNA: 0.5, 2.75, and 1.34 for P1, P2, and nuclear polymerase, respectively. Denatured DNA was a poor template for both enzymes P1 and P2, but it was inert as template for the nuclear enzyme. Although each of these polymerases required all four deoxynucleoside triphosphates for maximal activity, they catalyzed a high rate of synthesis in the absence of one or more deoxynucleoside triphosphates. Such a 'limited' synthesis was much more extensive for polymerase P2 and nuclear enzyme than for P1 was the most sensitive of the three to sulphydryl reagents, ehtidium bromide, heparin, and single-stranded DNA. The responses of P2 and nuclear enzymes to various inhibitors were very similar. However, these two enzymes respond differently to heat and high ionic strength.