Search

The many known eukaryotic DNA polymerases are classified into four families; A, B, X, and Y. Among them, DNA polymerase η, a Y family polymerase, is a low fidelity enzyme that contributes to translesional synthesis and somatic hypermutation. Although a high mutation frequency is observed in immunoglobulin genes, translesional synthesis occurs with a high accuracy. We determined whether the misincorporation rate of DNA polymerase η varies with ambient conditions. It has been reported that DNA polymerase η is unable to exclude water molecules from the active site. This finding suggests that some ions affect hydrogen bond formation at the active site. We focused on the effect of pH and evaluated the misincorporation rate of deoxyguanosine triphosphate (dGTP) opposite template T by DNA polymerase η at various pH levels with a synthetic template-primer. The misincorporation rate of dGTP by DNA polymerase η drastically increased at pH 8.0-9.0 compared with that at pH 6.5-7.5. Kinetic analysis revealed that the Km value for dGTP on the misincorporation opposite template T was markedly affected by pH. However, this drastic change was not seen with the low fidelity DNA polymerase α.

Translesion DNA synthesis (TLS) involves PCNA mono-ubiquitination and TLS DNA polymerases (pols). Recent evidence has shown that the mono-ubiquitination is induced not only by DNA damage but also by other factors that induce stalling of the DNA replication fork. We studied the effect of spontaneous DNA replication errors on PCNA mono-ubiquitination and TLS induction. In the pol1L868F strain, which expressed an error-prone pol alpha, PCNA was spontaneously mono-ubiquitinated. Pol alpha L868F had a rate-limiting step at the extension from mismatched primer termini. Electron microscopic observation showed the accumulation of a single-stranded region at the DNA replication fork in yeast cells. For pol alpha errors, pol zeta participated in a generation of +1 frameshifts. Furthermore, in the pol1L868F strain, UV-induced mutations were lower than in the wild-type and a pol delta mutant strain (pol3-5DV), and deletion of the RAD30 gene (pol eta) suppressed this defect. These data suggest that nucleotide misincorporation by pol alpha induces exposure of single-stranded DNA, PCNA mono-ubiquitination and activates TLS pols.

To elucidate the response to oxidative stress in eukaryotic cells, the effect of an oxidized nucleotide, 8-oxo-2'-deoxyguanosine 5'-triphosphate (8-oxo-dGTP), generated from dGTP with an active oxygen, on DNA synthesis was studied using a cell-free DNA replication system derived from Xenopus egg lysates with a single-stranded DNA template. Amounts of newly synthesized DNA were reduced according to the increasing concentration of 8-oxo-dGTP. Pulse labeling analysis revealed that 8-oxo-dGTP could delay DNA synthesis by reducing the rate of chain elongation. This delay was recovered by addition of a protein kinase inhibitor, staurosporine or bisindolylmaleimide I. These results indicate that a staurosporine- or bisindolylmaleimide I-sensitive protein kinase, such as a protein kinase C family member, may contribute to the delay of DNA synthesis by 8-oxo-dGTP. UV-irradiated single-stranded DNA also caused a delay of DNA synthesis on the undamaged template in the lysates. However, this delay was not recovered by staurosporine or bisindolylmaleimide I. Therefore, the mechanism of delay of DNA synthesis by 8-oxo-dGTP may be different from that by UV lesions. This is the first report that demonstrates an effect of an oxidized nucleotide on DNA replication in eukaryotes.

The mechanism of action of the antitumor nucleoside analog 1-(2-deoxy-2-fluoro-4-thio-beta-D-arabinofuranosyl)cytosine (4'-thio-FAC) was investigated. 4'-Thio-FAC inhibited cellular DNA synthesis, but not RNA and protein syntheses. We observed potent inhibitory action of the triphosphate of 4'-thio-FAC (4'-thio-FACTP) against DNA polymerase alpha, whereas it showed moderate inhibition of DNA polymerase beta and little inhibition of DNA polymerase gamma. The kinetic analysis showed that the inhibition mode of 4'-thio-FACTP towards DNA polymerase alpha was mixed type, implying a chain-terminating effect of 4'-thio-FACTP. The triphosphate of 2'-deoxy-2',2'-difluorocytidine (gemcitabine), a known antitumor nucleoside, did not show potent inhibition of these three DNA polymerases. Thus, the effect of the diphosphate of gemcitabine on ribonucleotide reductase was suggested to be more important for the antitumor action of gemcitabine. From these findings, the main target enzymes of 4'-thio-FAC and gemcitabine appear to be different. We found a synergistic effect of the two drugs in an in vitro model, which supports the above idea.

A sulfated glycoglycerolipid, 1-O-(6′-sulfo-α-d-glucopyranosyl)-2,3-di-O-phytanyl-sn-glycerol (KN-208), a derivative of the polar lipid isolated from an archaebacterium, strongly inhibited DNA polymerase (pol) α and pol β in vitro among 5 eukaryotic DNA polymerases (α, β, γ, δ, and ϵ). It also inhibited Escherichia coli DNA polymerase I Klenow fragment (E. coli pol I) and human immunodeficiency virus reverse transcriptase (HIV RT). The mode of inhibition of these polymerases was competitive with the DNA template primer and was non-competitive with the substrate dTTP. KN-208 inhibited pol β most strongly, with a Ki value of 0.05 μM, 10-fold lower than that for pol α (0.5 μM) and 60- or 140-fold lower than that for HIV RT (3 μM) or for E. coli pol I (7 μM), respectively. The loss of sulfate on the 6′-position of glucopyranoside of this compound completely abrogated inhibition. However, the hydrophilic part of KN-208, glucose 6-sulfate alone, showed no inhibition. Other sulfated compounds containing different hydrophobic structures, such as dodecyl sulfate and cholesterol sulfate, exhibited a much weaker inhibition. Our results suggest that the whole molecular structure of KN-208 is required for inhibition. KN-208 was shown to be modestly cytotoxic for the human leukemic cell line K562. Interestingly, a subcytotoxic dose of KN-208 increased the sensitivity of the human leukemic cells to an alkylating agent, methyl methanesulfonate, while it did not potentiate the effects of ultraviolet light or of cisplatin. Int. J. Cancer 76:512–518, 1998.© 1998 Wiley-Liss, Inc.

Sphingosine is a potent inhibitor of a mammalian DNA primase in vitro (Simbulan et al., Biochemistry 33, 9007-9012, 1994). Here we measured the inhibition of DNA primase in vitro by 9 sphingosine-analogues with respect to RNA primer synthesis and DNA primase-dependent DNA synthesis, and their potencies of inhibition in vitro were compared with their in vivo effects on human leukemic cells. Sphingosine, phytosphingosine and N, N-dimethylsphingosine strongly inhibited the activity of purified calf thymus DNA primase, and also inhibited the growth of human leukemic cell line HL-60, exerting strong cytotoxicity. Dihydrosphingosine and cis-sphingosine, which showed more subtle inhibition of DNA primase in vitro, moderately inhibited the cell growth in vivo and caused cell death. In contrast, N-acyl-, N-octyl-, and N-acetylsphingosine (ceramides) showing little inhibition of DNA primase suppressed cell growth only slightly. HL 60 cell was arrested at Go/G1 phase by exogenously added sphingosine. From these results, it is suggested that DNA primase is one of targets of sphingosine, an effector molecule in apoptosis.

We have measured the fidelity of leading and lagging strand DNA replication in HeLa cell extracts. Providing an excess of one dNTP in reactions induces replication errors consistent with misincorporation of that dNTP. With excess dTTP, both substitutions and single-nucleotide frameshifts are induced. Error distribution is nonrandom; reproducible hot spots for a substitution and a frameshift error are observed. Measurements with two vectors having the origin of replication on opposite sides of the mutational target demonstrate that error rates for G.dTTP and C.dTTP mispairs depend on whether the strand is replicated as the leading or lagging strand. Also, the two hot spots are only observed in one origin-target orientation. Replication reactions reconstituted from two fractions derived from extracts are 3-fold less accurate, but the error specificity with excess dTTP is similar to that with extracts. This suggests that the processes responsible for the nonrandom error rates are not lost as a result of fractionation. Furthermore, the reconstituted system is devoid of mismatch repair activity. Thus, mismatch repair is not responsible for the mispair-, site-, and strand-specific differences observed.

We examined the mechanism of the inhibition of DNA synthesis by a new platinum compound, (-)-(R)-2-aminomethylpyrrolidine(1,1-cyclobutane-dicarboxylato+ ++)-2-platinum(II) monohydrate (DWA-2114R), a derivative of the antitumor drug cis-diamminedichloroplatinum(II) (CDDP), using prokaryotic and eukaryotic DNA polymerases. Preincubating activated DNA with CDDP or DWA-2114R reduced its template activity for prokaryotic and eukaryotic DNA polymerases in a dose-dependent manner. DWA2114R required six times greater drug concentration and two times longer incubation time to show the same decrease of the template activity compared to CDDP. Treatment of primed pUC118 ssDNA templates with the two drugs followed by second-strand synthesis by prokaryotic and eukaryotic DNA polymerases revealed that DWA2114R bound to DNA in a similar manner to CDDP and these adducts blocked DNA elongation by DNA polymerases of eukaryotes as well as of prokaryotes. With these two drugs, the elongations by E. coli DNA polymerase I (Klenow fragment), T7 DNA polymerase and calf thymus DNA polymerase alpha were strongly arrested at guanine-guanine sequences (GG). Stop bands were also observed at adenine-guanine sequences (AG) guanine-adenine-guanine sequences (GAG) and mono-guanine sequence (G). Calf testis DNA polymerase beta was also arrested efficiently at AG, GAG and G, but much more weakly at GG. This pattern was common to DWA2114R and CDDP.

To search a novel factor possibly involved in the regulation of DNA replication for eukaryotic cells, we fractionated the Xenopus egg extract frequently used as a cell free DNA replication system, and obtained a single polypeptide of 50 KDa on SDS-polyacrylamide gel electrophoresis. Data of the amino acid sequence reduced from cloned gene of this protein revealed that this might be a Xenopus homolog of human PA2G4 that was known as a proliferation-associated protein or a cell cycle-regulated protein. Recombinant mouse PA2G4 protein also showed the inhibitory effect on the DNA replication reaction in Xenopus egg extract with the single-stranded DNA template. These results strongly suggest that PA2G4 may have an important role on the regulation of DNA replication in eukaryotic cells.

DNA polymerase eta is recently found as a responsible gene product of xeroderma pigmentosum variant. Differently from other eukaryotic DNA polymerases, such as alpha, beta or gamma, eta polymerase is categorized in Y family. Specific inhibitors for DNA polymerases are useful tools to study the exact role of enzyme in the living cells, however, the inhibitor for DNA polymerase eta has not been developed. We examined the inhibitory effects of several sugar-modified nucleotide analogs on DNA polymerase eta. The arabinonucleotides (araCTP), dideoxynucleotides (ddTTP) and 3'-modified nucleotides (3'-dCTP and 3'-azidothymidine tri-phosphate) did not show any inhibitory effect on DNA polymerase eta. On the other hand, the oxetanocin derivatives those have the oxetane ring as a sugar moiety, OXT-GTP and OXT-ATP, strongly inhibited this polymerase. These results suggest that DNA polymerase eta has a unique recognition mode against the sugar moiety of nucleotide compared with other eukaryotic nucleotide polymerases.

To elucidate the behavior of DNA polymerase eta against the oxidized purine nucleotides, we determined the utilization efficiency of 2-hydroxy-dATP and 8-hydroxy-dGTP by the recombinant yeast DNA polymerase eta using the primer extension assay with the synthetic oligonucleotide template-primers, and compared those by DNA polymerase alpha. Results indicate that DNA polymerase eta incorporates 2-hydroxy-dATP opposite template G in addition to template T and 8-hydroxy-dGTP opposite A in addition to C, respectively. Kinetic analysis revealed that the rate of mutation caused by 2-OH-dATP and 8-OH-dGTP with DNA polymerase eta should be much higher than those with DNA polymerase alpha.

Incorporation of nucleotides opposite the oxidized thymine bases on the template by eukaryotic DNA polymerases alpha and gamma was studied using the synthetic oligonucleotides containing 5-formyluracil or 5-hydroxymethyluracil. Primer extension assay revealed that both polymerases could bypass the oxidized thymine lesion, and incorporated dATP and dGTP opposite 5-formyluracil on the template. On the other hand, dATP was an only nucleotide that was incorporated opposite 5-hydroxymethyluracil by both polymerases as far as examined. These results indicate that 5-formyluracil but not 5-hydroxymethyluracil on template DNA may induce the transition mutation of A:T to G:C in eukaryotes.

To obtain the information for the exact role of replication protein A (RPA) on both eukaryotic DNA replication and repair, the binding preference of RPA purified from Xenopus egg extract against the undamaged and UV-damaged single-stranded DNA was studied by the gel shift assay. Chemically synthesized oligonucleotide containing the pyrimidine(6-4)pyrimidone photoproduct at one site was used as a model of UV-damaged DNA. Results of competition assay and Scatchard plots indicate that RPA preferentially binds to the 6-4 photoproduct oligonucleotide than the undamaged DNA.

To obtain the information for the role of replication protein A (RPA) on the detection of oxidized lesion in the single-stranded DNA, the binding preference of RPA purified from Xenopus egg lysate against the oligonucleotide containing one of three kinds of oxidized thymine residues, 5-formyluracil, 5-hydroxymethyluracil and 5-(l,2-dihydroxyethyl)uracil, was studied by the gel shift assay. Results of competition assay indicate that RPA preferentially binds to the oligonucleotide containing these oxidized thymine residues than the undamaged DNA.

Productive infection and replication of herpesviruses usually occurs in growth-arrested cells, but there has been no direct evidence in the case of Epstein-Barr virus (EBV), since an efficient lytic replication system without external stimuli does not exist for the virus. Expression of the EBV lytic-switch transactivator BZLF1 protein in EBV-negative epithelial tumor cell lines, however, is known to arrest the cell cycle in G 0 /G 1 by induction of the tumor suppressor protein p53 and the cyclin-dependent kinase (CDK) inhibitors p21 WAF-1/CIP-1 and p27 KIP-1 , followed by the accumulation of a hypophosphorylated form of the Rb protein. In order to determine the effect of the onset of lytic viral replication on cellular events in latently EBV-infected B LCLs, a tightly controlled induction system of the EBV lytic-replication program by inducible BZLF1 protein expression was established in B95-8 cells. The induction of lytic replication completely arrested cell cycle progression and cellular DNA replication. Surprisingly, the levels of p53, p21 WAF-1/CIP-1 , and p27 KIP-1 were constant before and after induction of the lytic program, indicating that the cell cycle arrest induced by the lytic program is not mediated through p53 and the CDK inhibitors. Furthermore, although cellular DNA replication was blocked, elevation of cyclin E/A expression and accumulation of hyperphosphorylated forms of Rb protein were observed, a post-G 1 /S phase characteristic of cells. Thus, while the EBV lytic program promoted specific cell cycle-associated activities involved in the progression from G 1 to S phase, it inhibited cellular DNA synthesis. Such cellular conditions appear to especially favor viral lytic replication.

We have studied the contribution of mismatch sequences to the trinucleotide repeat expansion that causes hereditary diseases. Using an oligonucleotide duplex, (CAG)5/(CTG)5, as a template-primer, DNA synthesis was carried out using either Escherichia coli DNA polymerase I (Klenow fragment) or human immunodeficiency virus type I reverse transcriptase (HIV-RT). Both enzymes expanded the repeat sequence longer than 27 nucleotides (nt), beyond the maximum length expected from the template size. The expansion was observed under conditions in which extension occurs either in both strands or in one strand. In contrast, with another template-primer that contains a non-repetitive flanking sequence 5'-upstream of the repetitive sequence, the reaction products were not extended beyond the template size (45 nt) by these DNA polymerases. We then used mismatched template-primers, in which either 1, 2 or 6 non-complementary nucleotides were introduced to the repeat sequence that is flanked by a non-repetitive sequence. In this case, primers were efficiently expanded over the expected length of 45 nt, in a mismatch-dependent manner. One of the primers with six mismatches extended as long as 72 nt. These results imply that the misincorporation of non-complementary deoxyribonucleoside monophosphates (dNMPs) into the repeat sequence makes double-stranded DNA unstable and triggers the slippage and expansion of trinucleotide repeats by forming loops or hairpin structures during DNA synthesis.

Human retinoblastoma (Rb) protein, immunopurified from an extract of recombinant baculovirus infected cells, stimulated 10-100-fold the activity of DNA polymerase alpha from calf thymus or human HeLa cells. Purified Rb protein is composed of two electrophoretically distinguishable forms, i.e., partially phosphorylated and under-phosphorylated forms. Dephosphorylation of Rb protein by protein phosphatase 2A largely diminished its stimulatory effect. On the other hand, a hyperphosphorylated Rb protein, obtained from insect cells overexpressing Rb protein, cyclin E and cyclin-dependent kinase 2 simultaneously, stimulated DNA polymerase alpha more strongly than the singly-expressed Rb protein. These results indicate that the phosphorylation is crucial for the stimulation. Rb protein isolated from human Burkitt lymphoma Raji cells also stimulated DNA polymerase alpha. In contrast, Rb protein did not affect eukaryotic DNA primase or Klenow fragment of Escherichia coli DNA polymerase I. By immunoprecipitation using anti-DNA polymerase alpha antibody, Rb protein in nuclear extract of Raji cells was co-precipitated with DNA polymerase alpha. This result indicates that DNA polymerase alpha exists as a complex containing phosphorylated Rb protein in cells. DNA polymerase alpha specifically bound to a purified Rb protein-immobilized Sepharose column. Rb protein also bound to DNA polymerase alpha trapped to anti-DNA polymerase alpha antibody-Sepharose column, suggesting the direct association of these two proteins. These observations suggest a new function of phosphorylated Rb protein in the regulation of DNA replication.

To clarify the molecular mechanism of mismatch formation by DNA polymerase eta, the pH-dependency of misincorporation of dNTP was studied with the synthetic template-primer. Incorporation of dNTP formed Watson-Crick type base pair, such as the incorporation of dATP opposite template T, was slightly affected by pH between 6.5 to 9.0. On the other hand, the misincorporation rate of dGTP opposite template T by DNA polymerase eta was drastically increased according to the increasing pH. Kinetical analysis revealed that this change might be due to the change of Km value for dGTP rather than that of Vmax value. This suggests that the affinity of dGTP on DNA polymerase eta during the mismatch formation with template T should be affected by pH.

1-O-Acetyl-2,5-di-O-p-chlorobenzoyl-3-deoxy-D-ribofuranose (1), derived from the antibiotic cordycepin was coupled with trimethylsilylated derivatives (2a-c) of N4-propionylcytosine, N4-p-toluoyl-5-fluorocytosine and 5-fluorouracil in the presence of trimethylsilyl trifluoromethanesulfonate (TMS-triflate) to give fully acylated nucleosides (3a-b and 3d, respectively). Selective removal of the N4-propionyl group of 3a by treatment with hydrazine hydrate gave 2',5'-di-O-p-chlorobenzoyl-3'-deoxycytidine (4). Deamination of 4 with sodium nitrite in trifluoroacetic acid afforded 2',5'-di-O-p-chlorobenzoyluridine (3c) in good yield. Compounds 3a-d were saponified to give free 3'-deoxycytidine (5a), 5-fluoro-3'-deoxycytidine (5b), 3'-deoxyuridine (5c), and 5-fluoro-3'-deoxyuridine (5d), respectively. These 3'-deoxyribonucleosides (5a-d) were then converted to corresponding 5'-monophosphate and further phosphorylated to the 5'-triphosphates by the phosphoroimidazolidate method. The nucleosides (5a-d) were examined for growth-inhibitory effects on mouse leukemic L5178Y cells, and their IC50 values (microgram/ml) were 1.8, 33, 6.5, and 18, respectively. On the other hand, the antiviral activities of these compounds on a rhabdovirus, infectious hematopoietic necrosis virus (IHNV), were moderate (IC50 = 100-500 micrograms/ml in CHSE-214 cells). The 5'-triphosphates showed remarkable inhibitory effects on DNA polymerase beta and DNA polymerase alpha-primase purified from testes of the cherry salmon, Oncorhynchus masou, but not on common DNA polymerase alpha from same source.

We have determined the fidelity of DNA replication by human cell extracts in reactions containing excess dGTP. Replication errors were scored using two M13 DNA substrates having the replication origin on opposite sides of the lacZ alpha-complementation gene. The data suggest that the average rates for replication errors resulting from G(template), T.dGTP, and A.dGTP mispairs are 25 x 10(-6), 12 x 10(-6), and 3 x 10(-6), respectively. The data also suggest that error rates for both the (+) and (-) strands differ by less than 2-fold when they are replicated either as the leading or lagging strand. This is in contrast to the 33- and 8-fold differences observed earlier for G.dTTP and C.dTTP mispairs on the (+) strand when replicated by the leading or lagging strand complex (Roberts, J. D., Izuta, S., Thomas, D. C., and Kunkel, T. A. (1994) J. Biol. Chem. 269, 1711-1717). Thus, the relative fidelity of the leading and lagging strand replication proteins varies with the mispair and sequence considered. Misincorporation of dGTP preferentially occurs at template positions where dGTP is the next correct nucleotide to be incorporated. This "next nucleotide" effect is characteristic of reduced exonucleolytic proofreading and suggests that these replication errors are normally proofread efficiently. Fidelity measurements performed in the absence or presence of dGMP, an inhibitor of proofreading exonuclease activity, suggest that the leading strand replication complex proofreads some mispairs more efficiently than does the lagging strand replication complex.

Eukaryotic DNA polymerase alpha pauses at some sites on the natural DNA template of M13mp2. Terminal misincorporations of dA or dG, in place of dT, by DNA polymerase alpha have been reported to be within one of the pause sites, pause site II (positions 6269 and 6270 (Fry, M., and Loeb, L.A. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 763-767)). The DNA products arrested within pause site II (position 6270) were separated, annealed with synthetic templates, and further elongated by DNA polymerase alpha. It was confirmed that a considerable amount of terminal misincorporation of dG in place of dT occurred at this position. When M13mp2 DNA was coated with various amounts of replication protein-A (RP-A), however, DNA polymerase alpha was able to overcome the pause site II, whereas pause bands at other sites barely decreased. In contrast, Escherichia coli single-stranded DNA-binding protein did not specifically abolish the arrested band at pause site II, though it generally suppressed the reaction. Since RP-A hardly increased the elongation frequency from the primer carrying a 3‘-mismatched terminal deoxynucleotide, the reduction of arrested products by RP-A may be attributed to the change in the incorporation mode from noncomplementary to complementary deoxynucleotides within pause site II and may not be due to the reinitiation from the mismatched 3‘-terminal deoxynucleotide. To confirm this, we amplified the reaction products at pause site III by means of a polymerase chain reaction method and showed that the complementary strand to pause site II, which was elongated in the presence of RP-A, did not carry any detectable misinsertion. Therefore, the errorprone step of the DNA synthesis catalyzed by DNA polymerase alpha may be readily avoided by RP-A.

The direct effect of the eukaryotic nuclear DNA-binding protein poly(ADP-ribose) polymerase on the activity of DNA polymerase alpha was investigated. Homogenously purified poly(ADP-ribose) polymerase (5 to 10 micrograms/ml) stimulated the activity of immunoaffinity-purified calf or human DNA polymerase alpha by about 6 to 60-fold in a dose-dependent manner. It had no effect on the activities of DNA polymerase beta, DNA polymerase gamma, and primase, indicating that its effect is specific for DNA polymerase alpha. Apparently, poly(ADP-ribosyl)ation of DNA polymerase alpha was not necessary for the stimulation. The stimulatory activity is due to poly(ADP-ribose) polymerase itself since it was immunoprecipitated with a monoclonal antibody directed against poly(ADP-ribose) polymerase. Kinetic analysis showed that, in the presence of poly(ADP-ribose) polymerase, the saturation curve for DNA template primer became sigmoidal; at very low concentrations of DNA, it rather inhibited the reaction in competition with template DNA, while, at higher DNA doses, it greatly stimulated the reaction by increasing the Vmax of the reaction. By the automodification of poly(ADP-ribose) polymerase, however, both the inhibition at low DNA concentration and the stimulation at high DNA doses were largely lost. Furthermore, stimulation by poly(ADP-ribose) polymerase could not be attributed to its DNA-binding function alone since its fragment, containing only the DNA-binding domain, could not exert full stimulatory effect on DNA polymerase, as of the intact enzyme. Poly(ADP-ribose) polymerase is co-immunoprecipitated with DNA polymerase alpha, using anti-DNA polymerase alpha antibody, clearly showing that poly(ADP-ribose) polymerase may be physically associated with DNA polymerase alpha. In a crude extract of calf thymus, a part of poly(ADP-ribose) polymerase activity existed in a 400-kDa, as well as, a larger 700-kDa complex containing DNA polymerase alpha, suggesting the existence in vivo of a complex of these two enzymes.

Eight kinds of 5-substituted 1-β-D-arabinofuranosyluracil 5 '-triphosphates ((E) - (3-nitrostyryl) (6), (E) - (3-aminostyryl) (9), (E) - (4-nitrostyryl) (7), (E) - (4-aminostyryl) (10), (E) -styryl (8), phenethyl (11), (RS) - (3-azido-2-hydroxypropyl) (17), and (RS) - (3-amino-2-hydroxypropyl) (18) derivatives) were synthesized. Among these analogs, araUTPs bearing strongly hydrophobic styryl groups at the 5-position (6-10) were shown to have selective and strong inhibitory action on deoxyribonucleic acid (DNA) polymerase α purified from cherry salmon (Oncorhynchus masou) testes. The 5-azidopropyl derivative (17) also inhibited this polymerase. The compounds with a nitro and an amino group on the 5-styryl substituent showed essentially the same activity, but the 5-phenethyl derivative (11) and the 5-aminopropyl derivative (18) showed greatly reduced inhibitory action. On the other hand, in the case of DNA polymerase β, all the analogs showed similar inhibitory effects. The influence of hydrophobic and steric effects of substituents at the 5-position of araUTP on DNA polymerases α and β are discussed.