Reconstitution of endogenous DNA synthesis in isolated nuclei and template activity of chromatin with respect to mode of inhibition by aphidicolin

We succeeded in reconstituting the endogenous nuclear DNA synthesis of the sea urchin. Endogenous DNA synthesis of isolated nuclei was reconstituted by mixing the salt-treated nuclei (chromatin exhibiting essentially no endogenous DNA synthesis) and the salt extract containing DNA polymerase-alpha. DNA synthesis in this reconstitution system showed a level of activity and a mode of inhibition by aphidicolin similar to those of the original isolated nuclei (noncompetitive with respect to dCTP). On the other hand, the inhibitory mode was competitive with respect to dCTP in DNA synthesis in the reconstituted system obtained from the chromatin and purified DNA polymerase-alpha, indicating that some other factor(s) in addition to DNA polymerase-alpha is necessary for the reconstitution with reference to the inhibitory mode of aphidicolin. We also studied the template activity of the chromatin. When chromatin was used as a template, inhibition by aphidicolin of DNA polymerase-alpha was noncompetitive and uncompetitive with respect to the template at high and low concentrations, respectively. Treatment of chromatin with 5 M urea gave urea-treated chromatin (nonhistone protein-deprived chromatin) and the extract (mainly nonhistone protein fraction). Inhibition by aphidicolin of DNA polymerase-alpha was uncompetitive with respect to the urea-treated chromatin. However, when chromatin reconstituted from the urea-treated chromatin and the extract was used as a template, the inhibitory mode by aphidicolin was similar to that with original chromatin, indicating that the nonhistone protein fraction contained factor(s) which modified the inhibitory mode of aphidicolin. Thus, the inhibitory mode of aphidicolin is a useful parameter for monitoring the resolution and reconstitution of endogenous DNA synthesis of isolated nuclei.     

Differential responses of log and stationary phase human fibroblasts to inhibition of DNA repair by aphidicolin

The tetracyclic diterpenoid, aphidicolin, is an effective inhibitor of DNA repair in human cells following ultraviolet irradiation. This inhibition is very efficient in confluent resting cells but not in rapidly cycling cells as measured by (1) analysis of DNA single-strand breaks by alkaline sucrose sedimentation, (2) chromatographic analysis of pyrimidine-dimer removal, and (3) repair replication using CsCl density centrifugation. The inhibition is reversed by deoxycytidine or thymidine but not by deoxyadenosine or deoxyguanosine during the repair period. The data suggest that differences in deoxynucleoside triphosphate pools between cycling and confluent resting cells determine the different efficacies of the agent in these two situations.     

Abrogation of the effects of aphidicolin on NIH3T3 and V79 cells by caffeine

In this communication I show that caffeine (1,3,7-trimethylxanthine) stimulates [³H]thymidine incorporation in aphidicolin-treated V79 and NIH3T3 cells. Flow microfluorometric analysis showed that caffeine, partially or fully, abrogates the cell cycle progression block produced by aphidicolin. Increased cell growth is also observed in cultures treated with both aphidicolin and caffeine compared to cultures treated with aphidicolin only. Microscopic examination of V79 cultures treated with aphidicolin for 8 h showed a marked reduction in the freqeuncy of round mitotic cells, as is expected from a drug which inhibits progression through the cell cycle by inhibiting DNA replication; this effect of aphidicolin was also reduced by caffeine. Biochemical analysis showed that caffeine did not directly interfere with the inhibition of DNA polymerase-α by aphidicolin. Analysis of dNTP pools indicated that caffeine increased the level of dCTP in V79 cells. In aphidicolin-treated V79 cells, the increase in the dCTP level due to exogenous cytidine was almost completely blocked; caffeine also substantially overcame this effect of aphidicolin. These results indicate that caffeine produces its effects on aphidicolin-treated cells by altering the dCTP metabolism.     

Effect of aphidicolin on de novo DNA synthesis, DNA repair and cytotoxicity in gamma-irradiated human fibroblasts. Implications for the enhanced radiosensitivity in ataxia telangiectasia

The antibiotic, aphidicolin, is a potent inhibitor of DNA polymerase alpha and consequently of de novo DNA synthesis in human cells. We report here that in gamma-irradiated normal human cells, aphidicolin (at 5 micrograms/ml and less) had no significant effect on the rate of the rejoining of DNA single strand breaks or rate of removal of DNA lesions assayed as sites sensitive to incising activities present in crude protein extracts of Micrococcus luteus cells. gamma-irradiated human ataxia telangiectasia cells are known to demonstrate enhanced cell killing and exhibit resistance to the inhibiting effects of radiation on DNA synthesis. Under conditions of minimal aphidicolin cytotoxicity but extensive inhibition of de novo DNA synthesis, the radiation responses of neither normal nor ataxia telangiectasia cells were significantly modified by aphidicolin. Firstly, we conclude that human DNA polymerase alpha is not primarily involved in the repair of the two classes of radiogenic DNA lesions examined. Secondly, the radiation hypersensitivity of ataxia telangiectasia cells cannot be explained on the basis of premature replication of damaged cellular DNA resulting from the resistance of de novo DNA synthesis to inhibition by ionizing radiation.     

Effect of aphidicolin on de novo DNA synthesis,DNA repair and cytotoxicity in γ-irradiated human fibroblasts. Implications for the enhanced radiosensitivity in ataxia telangiectasia

The antibiotic, aphidicolin, is a potent inhibitor of DNA polymerase α and consequently of de novo DNA synthesis in human cells. We report here that in γ-irradiated normal human cells, aphidicolin (at 5 μg/ml and less) had no significant effect on the rate of the rejoining of DNA single strand breaks or rate of removal of DNA lesions assayed as sites sensitive to incising activities present in crude protein extracts of Micrococcus luteus cells. γ-irradiated human ataxia telangiectasia cells are known to demonstrate enhanced cell killing and exhibit resistance to the inhibiting effects of radiation on DNA synthesis. Under conditions of minimal aphidicolin cytotoxicity but extensive inhibition of de novo DNA synthesis, the radiation responses of neither normal nor ataxia telangiectasia cells were significantly modified by aphidicolin. Firstly, we conclude that human DNA polymerase α is not primarily involved in the repair of the two classes of radiogenic DNA lesions examined. Secondly, the radiation hypersensitivity of ataxia telangiectasia cells cannot be explained on the basis of premature replication of damaged cellular DNA resulting from the resistance of de novo DNA synthesis to inhibition by ionizing radiation.     

Cell cycle-dependent effects on deoxyribonucleotide and DNA labeling by nucleoside precursors in mammalian cells.

dCTP pools equilibrated to equivalent specific activities in Chinese hamster ovary cells or in nuclei after incubation of cells with radiolabeled nucleosides, indicating that dCTP in nuclei does not constitute a distinct metabolic pool. In the G1 phase, [5-3H]deoxycytidine labeled dCTP to unexpectedly high specific activities. This may explain reports of replication-excluded DNA precursor pools.     

Hydroxyurea: effects on deoxyribonucleotide pool sizes correlated with effects on DNA repair in mammalian cells

We have measured deoxyribonucleotide pool sizes in different cell types: normal human, transformed human (HeLa), and the permanent hamster line CHO-K1. The range of sizes of the four DNA precursor pools in CHO cells is far greater than in human cells. It is a general rule that hydroxyurea causes rapid depletion of pools (except for dTTP) until the pool present in smallest amount is exhausted; this suggests a tight coupling of the pools to DNA replication (the presumed main cause of the depletion). The effect of hydroxyurea on DNA repair after ultraviolet irradiation (namely, a relatively small accumulation of incomplete repair sites blocked at the resynthesis stage) is probably accounted for by the reduced availability of DNA precursors. However, depletion of the dCTP pool is not an adequate explanation for the observed enhancement by hydroxyurea of the inhibitory effect of cytosine arabinoside; we suggest other possible modes of action. Ultraviolet irradiation has only small effects on the levels of deoxyribonucleotides.     

Lack of specific correlation of the deoxycytidine triphosphate pool level with rate of DNA synthesis.

The levels of the four deoxyribonucleoside triphosphate pools and the distribution of cells in the various phases of the cell cycle have been examined in Chinese hamster cells as thymidine, present as a regular constituent in the growth medium, was removed in stages. The results indicate that: 1. Duration of the DNA synthetic phase was lengthened when thymidine was removed from the growth medium. 2.Temporally correlated with lengthening of the DNA synthetic phase upon thymidine removal was a 7-fold increase in level of the dCTP pool, reduction in the dGTP pools, and little or no change in dATP pool. 3.Radioactive labeling procedures indicated that expansion of the dCTP pool could be completely accounted for by increased ribonucleotide reductase activity and that the dTTP pool switched from a largely exogenous thymidine source to endogenous dTTP synthesis as the extracellular thymidine concentration was reduced. 4.Deoxyuridine and thymidine were apparently transported by the same system in Chinese hamster cells, while deoxycytidine was transported by a different system. Although deoxycytidine transport was unaffected by thymidine, phosphorylation of intracellular deoxycytidine compounds to the triphosphate level was stimulated by thymidine. Cytidine transport was not significantly affected by thymidine.     

Intracellular compartmentation of deoxycytidine nucleotide pools in S phase mouse 3T3 fibroblasts

We labeled mouse 3T3 fibroblasts, synchronized in G0 or S phase, from [3H]cytidine or [3H]deoxycytidine and measured the flow of isotope into and through deoxycytidine nucleotide pools, including the two deoxyliponucleotides dCDP choline and dCDP ethanolamine. Compared to G0 cells, S phase cells had much larger pools with a 20-40-fold faster turnover. The dCTP pool of S phase cells during steady state conditions attained a 6-fold higher specific activity than the pool of G0 cells when labeled from cytidine but a 10-fold lower specific activity when labeled from deoxycytidine. The dCTP pool of G0 cells showed a slow but measurable turnover indicating a limited amount of de novo synthesis also in resting cells. The labeling pattern of dCTP and deoxyliponucleotides of G0 cells was compatible with a simple precursor-product relationship. In S phase cells, however, dCDP choline had a 4-6 times higher specific activity during steady state conditions than dCTP and dCMP when the cells were labeled with [3H]deoxycytidine. We suggest that 3T3 cells contain two distinct intracellular dCTP pools, one labeled preferentially from cytidine and used for DNA replication, the other labeled from deoxycytidine and used for deoxyliponucleotide synthesis. We speculate that the latter pool during S phase may be temporarily sequestered in the cell's membrane fraction before equilibration with the much larger dCTP pool originating in S phase cells from the reduction of CDP.     

Pyrimidine nucleotide and nucleic acid synthesis in human monocytes and macrophages.

In most cell types, the production of deoxynucleotides is tightly coupled to the pace of cell division, and nearly all deoxynucleotides are used for semiconservative DNA synthesis. The capacity of peripheral blood monocytes and macrophages to proliferate is controversial. However, these cells have been reported to produce and release thymidine, which can serve as a precursor or regulator of DNA synthesis by lymphocytes and other cells. To determine to what extent de novo pyrimidine nucleotide synthesis is linked to cell division in peripheral blood monocytes and macrophages, compared to human U937 promonocytes and CEM lymphoblasts, we used a precise precursor-product labeling method. The results showed that in all three cell types, the pace of pyrimidine deoxynucleotide production, and of thymidylate synthesis, was in proportion to the rate of DNA synthesis. The human blood monocytes and macrophages, in contrast to U937 cells, had extraordinarily low deoxyribonucleotide pools (less than 1 pmol/10(6) cells) and synthesized neither thymidylate nor DNA de novo during 7 days culture. Colony-stimulating factors augmented RNA synthesis in monocyte-derived macrophages, and enhanced cell survival, without inducing either DNA or thymidylate synthesis. We conclude that the thymidine released by macrophages derives from dead or dying cells, and not from de novo synthesis.     

Deoxyribonucleotide pools in mouse-fibroblast cell lines with altered ribonucleotide reductase.

Mutant cells lines of 3T6 mouse fibroblasts, resistant to thymidine and deoxyadenosine, have an altered allosteric regulation of the enzyme ribonucleotide reductase (Meuth, M. and Green, H., Cell, 3, 367, 1974). Compared to 3T6, these lines contain larger pools of deoxynucleoside triphosphates, in particular deoxycytidine triphosphate, but show a normal rate of DNA synthesis. Addition of thymidine or deoxyadenosine to 3T6 cells results in large accumulations of the corresponding triphosphates and a dramatic decrease in the dCTP pool, concomitant with inhibition of DNA synthesis. Addition of thymidine to the mutant cell lines also leads to an increase in the dTTP pool but does not result in a depletion of dCTP or inhibition of DNA synthesis. Addition of deoxyadenosine only leads to a small increase of the dATP pool. In general the change in the allosteric regulation of bibonucleotide reductase is reflected in the deoxynucleotide pools.     

Early nuclear events in lymphocyte proliferation : The role of DNA strand break repair and ADP ribosylation

The previously reported extensive DNA strand breakage in resting murine splenic lymphocytes is not an artifact of the extraction or assay procedure. The benzamide inhibitors of poly(ADP ribose) synthetase (pADPRS), such as 5-methoxybenzamide (MBA), had been shown to block the strand break repair occurring within 2 h of activation of splenic lymphocytes by the mitogen concanavalin A (conA); the inhibitors also blocked early events in proliferation, such as blast formation, as well as entry into S phase. Inhibitors of pADPRS blocked lymphocyte proliferation by inhibiting the activity of this enzyme, rather than by non-specific effects. Aphidicolin, an inhibitor of alpha-polymerase, also prevented DNA strand break repair in conA-stimulated cells but, unlike MBA, did not prevent blast formation. DNA strand breaks accumulated in the presence of MBA at the same linear rate (300-400/h) in both resting and conA-treated cells. We and others had hypothesized that this accumulation was due to a continuous production of strand breaks in lymphocytes, leading to their accumulation in presence of repair inhibitors. However, incubation of the cells with aphidicolin at concentrations that inhibited repair did not result in any increase in strand breaks. The hypothesis of continuous cycling of breaks is incorrect; accumulation of breaks was due to some indirect effect of MBA, such as a possible disinhibition of an ADP-ribosylation-sensitive endonuclease described in other cell types. All of the early stages of lymphocyte proliferation, including blast transformation (but not DNA synthesis) require ADP ribosylation. Repair of DNA strand breaks is not a precondition for blast formation, though experiments involving the combined effects of MBA and aphidicolin showed that repair of the breaks is essential in order for the cells to replicate their DNA. Our data are consistent with a model suggesting that DNA strand breaks introduced into differentiated cells act as an additional safety-catch mechanism that restrains them from replicating their genetic material but not from undergoing the early stages of proliferation.     

DNA polymerase activity in a repair-deficient human cell line

A human low-density-lipoprotein (LDL) receptor-deficient diploid fibroblast cell line (GM1915) was determined to be short patch competent (DNA polymerase-beta) and long patch deficient (DNA polymerase-alpha) for DNA excision repair. Analysis of DNA from GM1915 cells or from WI38 control cells, following treatment with a mutagen known to initiate long patch excision repair, showed that GM1915 cells exhibited decreased resynthesis of oligonucleotide segments excised during repair. When cells deficient in DNA polymerase-alpha activity were permeabilized to permit LDL entry, repair synthesis immediately increased. These data suggest that DNA polymerase-alpha is not activated by mutagen treatment in GM1915 cells and that introduction of LDL into the cells results in activation of the enzyme.     

Compartmentation of dCTP pools. Evidence from deoxyliponucleotide synthesis

The nucleotide fraction of cultured 3T6 and 3T3 mouse fibroblasts contains deoxy-CDP choline and deoxy-CDP ethanolamine as well as the corresponding riboliponucleotides. In permeabilized cells both deoxyliponucleotides were formed from dCTP. In intact cells they could be labeled from [5-3H] deoxycytidine or cytidine via transformation of the nucleosides to dCTP. Their turnover was slow compared to that of dCTP. When rapidly growing 3T3 cells were labeled during 90 min from deoxycytidine the specific activity of dCDP choline was 2.4 times higher than that of dCTP while after labeling from cytidine both nucleotides (and CTP) reached the same specific activity under steady state conditions. Also dCDP ethanolamine was labeled more rapidly from deoxycytidine than from cytidine. Our results suggest that the deoxyliponucleotides were synthesized from a dCTP pool that was labeled preferentially from deoxycytidine. Earlier work (Nicander, B., and Reichard, P. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 1347-1351) had demonstrated synthesis of DNA from a dCTP pool labeled preferentially from cytidine. Taken together our results suggest that deoxyliponucleotides and DNA are synthesized from separate dCTP pools.     

Qualitative differences between replicative and repair synthesis of DNA in normal and transformed mouse cells as measured by precursor discrimination

Inhibitors of DNA polymerase alpha such as aphidicolin (APC) or 1-beta-D-arabinofuranosyl-cytosine (araC) cause DNA-strand breaks to accumulate after UV-irradiation, at sites where repair resynthesis is inhibited. Transformed cells accumulate fewer such breaks than normal cells do; this may be due to differences in the extent, or the nature, of excision-repair synthesis in transformed and in normal cells. We have looked for differences in the nature of repair synthesis, comparing the labelling of DNA by deoxycytidine (dC) and araC through UV-induced repair in normal and transformed mouse cells. We have made parallel determinations of precursor discrimination in replicative synthesis, and find that normal cells discriminate better against araC in replicative synthesis than do transformed cells. But repair synthesis discriminates against araC less than normal replicative synthesis does, to a similar extent in both cell types. Thus, there are qualitative differences between the DNA polymerases engaged in UV excision repair and replication in normal and transformed mouse cells; but there is no evidence for a predominantly araC-insensitive repair synthesis in transformed cells, such as might account for the difference in break accumulation.     

Selection and characterization of mutant S49 T-lymphoma cell lines resistant to phosphonoformic acid: evidence for inhibition of ribonucleotide reductase

Phosphonoformic acid (PFA) and its congener phosphonoacetic acid (PAA) are inhibitors of viral replication whose mechanism of action appears to be the inhibition of viral DNA polymerase. These drugs inhibit mammalian DNA polymerase to a lesser extent. We sought to characterize the effects of phonoformic acid on mammalian cells by examining mutants of S49 cells (a mouse T-lymphoma line), which were selected by virtue of their resistance to phosphonoformic acid. The 11 mutant lines that were resistant to growth inhibition by 3 mM PFA had a range of growth rates, cell cycle distribution abnormalities, and resistance to the inhibitory effects of thymidine, acycloguanosine (acyclovir), aphidicolin, deoxyadenosine, and novobiocin. Most mutant lines had pools of ribonucleoside triphosphates and deoxyribonucleoside triphosphates similar to those of wild-type S49 cells. However, one line (PFA 3-9) had a greatly elevated dCTP pool. When this mutant line was further characterized, no apparent defect in DNA polymerase alpha activity was seen, but an increased ribonucleotide reductase activity, as assayed by CDP reduction in permeabilized cells, was observed. The CDP reductase activity in the PFA 3-9 cells decreased to wild-type control levels, and the CDP reductase activity of wild-type cells was also greatly reduced when PFA (2-3 mM) was added to permeabilized cells during the enzyme assay. These results demonstrate that PFA can directly inhibit ribonucleotide reductase activity in permeabilized cells. In addition, when PFA was added to exponentially growing cultures of either wild-type or PFA 3-9 mutant cells, the drug caused an arrest in S phase of the cell cycle and a decrease in all four deoxyribonucleotide pools, with the most dramatic decrease in the dCTP pools. The reduction in the dCTP pool level could be reversed by addition of exogenous deoxycytidine, but this reversed PFA toxicity only marginally. These observations suggest that PFA is an inhibitor of mammalian ribonucleotide reductase and that partial resistance to PFA can be effected by mutation to increased CDP reductase activity resulting in a large dCTP pool. This mutation results in less than twofold resistance to PFA, suggesting that other sites of inhibition coexist.     

Human monocytes undergo excessive apoptosis following temozolomide activating the ATM/ATR pathway while dendritic cells and macrophages are resistant

Immunodeficiency is a severe therapy-limiting side effect of anticancer chemotherapy resulting from sensitivity of immunocompetent cells to DNA damaging agents. A central role in the immune system is played by monocytes that differentiate into macrophages and dendritic cells (DCs). In this study we compared human monocytes isolated from peripheral blood and cytokine matured macrophages and DCs derived from them and assessed the mechanism of toxicity of the DNA methylating anticancer drug temozolomide (TMZ) in these cell populations. We observed that monocytes, but not DCs and macrophages, were highly sensitive to the killing effect of TMZ. Studies on DNA damage and repair revealed that the initial DNA incision was efficient in monocytes while the re-ligation step of base excision repair (BER) can not be accomplished, resulting in an accumulation of DNA single-strand breaks (SSBs). Furthermore, monocytes accumulated DNA double-strand breaks (DSBs) following TMZ treatment, while DCs and macrophages were able to repair DSBs. Monocytes lack the DNA repair proteins XRCC1, ligase IIIα and PARP-1 whose expression is restored during differentiation into macrophages and DCs following treatment with GM-CSF and GM-CSF plus IL-4, respectively. These proteins play a key role both in BER and DSB repair by B-NHEJ, which explains the accumulation of DNA breaks in monocytes following TMZ treatment. Although TMZ provoked an upregulation of XRCC1 and ligase IIIα, BER was not enhanced likely because PARP-1 was not upregulated. Accordingly, inhibition of PARP-1 did not sensitize monocytes, but monocyte-derived DCs in which strong PARP activation was observed. TMZ induced in monocytes the DNA damage response pathways ATM-Chk2 and ATR-Chk1 resulting in p53 activation. Finally, upon activation of the Fas-receptor and the mitochondrial pathway apoptosis was executed in a caspase-dependent manner. The downregulation of DNA repair in monocytes, resulting in their selective killing by TMZ, might impact on the immune response during cancer chemotherapy.     

Elevation of dCTP pools in xeroderma pigmentosum variant human fibroblasts alters the effects of DNA repair arrest by arabinofuranosyl cytosine

DNA excision repair inhibition by arabinofuranosyl cytosine (ara-C) or by ara-C/hydroxyurea (HU) was measured in log phase and confluent cultures of normal and xeroderma pigmentosium (XP)-variant human fibroblasts following insult by ultraviolet (UV) light (20 J/m²). Repair inhibition was determined by measuring the accumulation of DNA single-strand breaks/10⁸ daltons following cell culture exposure to ara-C or ara-C/HU in a series of 3 hr. pulses up ro 24 hr. after UV insult. Both normal and XP-variant derived cells showed a wide range of sensitivity to ara-C in log phase cells (0.2–9.4 breaks/10⁸ daltons DNA), although strand break accumulation was constant for each specific cell line. The same cells were more sensitive to ara-C/HU with a 2–14 fold increase in DNA strand breaks depending upon the cell line assayed. In confluent cultures of normal cells, maximum sensitivity to ara-C and ara-C/HU was achieved with similar levels of repair inhibition observed (16.1 and 16.5 breaks/10⁸ daltons, respectively). The same level of repair inhibition was observed in confulent XP-variants receiving ara-C/HU, but was reduced by 62–68% in cells treated with ara-C alone. Ara-C repair arrest was more rapidly reversed by competing concentrations of exogenous deoxycytidine (dCyd) in XP-variant compared to normal cells, especially in confluent cell cultures. In ara-C/HU treated cells, the level of dCyd reversal was reduced in the XP-variant when compared to cells exposed to ara-C alone. However, the same addition of HU had relatively little effect on dCyd reversal in normal cells. The measurements of dNTP levels indicate an elevated level of intracellular deoxycytosine triphosphate in XP-variant vs normal cells. The implications of these results are discussed as they relate to possible excision repair anomalies in the XP-variant.Abbreviations ara-C arabinofuranosul cytosine - dCTP deoxycytosine triphosphate - dCyd deoxycytidine - dNTP deoxynucleoside triphosphate - dT thymidine - HU hydroxyurea - XP xeroderma pigmentosium This research was sponsored jointly by the National Cancer Institute under Interagency Agreement #40-5-63, and the Office of Health and Environment Research, U. S. Department of Energy, under Contract W-7405-eng-26 with the Union Carbide Corporation.     

Evidence for the involvement of substrate cycles in the regulation of deoxyribonucleoside triphosphate pools in 3T6 cells

Pool sizes of deoxyribonucleoside triphosphates (dNTPs) in cultured cells are tightly regulated by i.al., the allosteric control of ribonucleotide reductase. We now determine the in situ activity of this enzyme from the turnover of the deoxycytidine triphosphate (dCTP) pool in rapidly growing 3T6 mouse fibroblasts, as well as in cells whose DNA replication was inhibited by aphidicolin or amethopterin, by following under steady state conditions the path of isotope from [5-3H]cytidine into nucleotides, DNA, and deoxynucleosides excreted into the medium. In normal cells as much as 28% of the dCDP synthesized was excreted as deoxynucleoside (mostly deoxyuridine), leading to an accumulation of deoxyuridine in the medium. Inhibition with amethopterin slightly increased ribonucleotide reductase activity, while aphidicolin halved the activity of this enzyme (and thymidylate synthase). In both instances all dCDP synthesized was degraded and excreted as nucleosides. This continued synthesis and turnover in the absence of DNA synthesis is in contrast to the earlier found inhibition of dCTP (and dTTP) turnover when hydroxyurea, an inhibitor of ribonucleotide reductase, was used to block DNA synthesis. To explain our results, we propose that substrate cycles between deoxyribonucleosides and their monophosphates, involving the activities of kinases and phosphatases, participate in the regulation of pool sizes. Within the cycles, a block of the reductase activates net phosphorylation, while inhibition of DNA polymerase stimulates degradation.