Endogenous HPRT activity in mycoplasmas isolated from cell cultures

Summary Five mycoplasma species most frequently isolated from cell cultures were tested for the presence of endogenous hypoxanthine phosphoribosyl-transferase (HPRT), activity. All of the five, cultured in cell-free medium, contained variable but significant levels of HPRT. Two strains ofM. hyorhinis exhibited a 13-fold difference in their specific HPRT activity. When infected with any of these mycoplasma species, HPRT-deficient mouse cell mutants rapidly acquired a cell-associated HPRT activity; however, the cells remained sensitive to HAT medium and resistant to 6-thioguanine. On the other hand, normal HPRT-positive cells deliberately infected with the mycoplasmas uniformly became sensitive to HAT medium. The apparent transfer of mycoplasma-specific HPRT activity to HPRT-deficient cells may be used as a sensitive measure of cell infection by these mycoplasma strains. The HPRT activities of mycoplasmas share several common properties so that they can be distinguished easily from the mammalian HPRT isozymes. Compared to the animal cell enzymes, the mycoplasmal HPRT activities are less heat stable, more strongly inhibited by 6-thioguanine, and in general migrate more slowly in electrophoresis at a neutral pH. This work was supported in part by PHS Research Grants 5 R01 GM21014 and 1 P03 GM19100 (Genetics Center Grant to Albert Einstein College of Medicine), and PHS Research Contracts N01 GM 6-2119 and N01-AG-4-2865 (to the Institute for Medical Research), from the National Institute of General Medical Sciences and National Institute on Aging. S. S. is a recipient of a Faculty Research Award from the American Cancer Society.     

Endogenous HPRT activity in mycoplasmas isolated from cell cultures

Five mycoplasma species most frequently isolated from cell cultures were tested for the presence of endogenous hypoxanthine phosphoribosyl-transferase (HPRT) activity. All of the five, cultured in cell-free medium, contained variable but significant levels of HPRT. Two strains of M. hyorhinis exhibited a 13-fold difference in their specific HPRT activity. When infected with any of these mycoplasma species, HPRT-deficient mouse cell mutants rapidly acquired a cell-associated HPRT activity; however, the cells remained sensitive to HAT medium and resistant to 6-thioguanine. On the other hand, normal HPRT-positive cells deliberately infected with the mycoplasmas uniformly became sensitive to HAT medium. The apparent transfer of mycoplasma-specific HPRT activity to HPRT-deficient cells may be used as a sensitive measure of cell infection by these mycoplasma strains. The HPRT activities of mycoplasmas share several common properties so that they can be distinguished easily from the mammalian HPRT isozymes. Compared to the animal cell enzymes, the mycoplasmal HPRT activities are less heat stable, more strongly inhibited by 6-thioguanine, and in general migrate more slowly in electrophoresis at a neutral pH.     

Influence of mycoplasma infection on the incorporation of different precursors into RNA components of tissue culture cells

Seven different tissue culture cells have been cultured with and without mycoplasma (M. hyorhinis) in the presence of various precursors of RNA. Total cellular RNA was isolated and analysed by electrophoresis on polyacrylamide gels. The results obtained with mycoplasma-infected cells can be summarized as follows:

1. 1. When cells are labelled with [8-³H]guanosine or [5-³H]uridine there is some incorporation into host cell 28S and 18S rRNA, but it is less than into mycoplasma 23S and 16S rRNA. [8-³H]guanosine or [5-³H]uridine are also incorporated into host cell and mycoplasma tRNA and mycoplasma 4.7S RNA, but the incorporation into host cell 5S rRNA and low molecular weight RNA components (LMW RNA) is reduced.

2. 2. [5-³H]uracil is not incorporated into host cell RNA but into mycoplasma tRNA, 4.7S RNA, a mycoplasma low molecular weight RNA component M₁ and 23S and 16S rRNA.

3. 3. [³H]methyl groups are incorporated into mycoplasma tRNA, 23S and 16S rRNA, but not into host cell 28S, 18S, 5S rRNA nor into mycoplasma 4.7S RNA.

4. 4. With [³²P]orthophosphate or [³H]adenosine as precursors, the labelling is primarily in the host RNA.

Mycoplasma infection influences the labelling of RNA primarily by an effect on the utilization of the exogenously added radioactive RNA precursors, since the generation time of mycoplasma infected cells is about the same as that of uninfected cells. Mycoplasma infection may completely prevent the identification of LMW RNA components.     

Thymidine kinase activity in mouse embryo fibroblast cells infected with murine cytomegalovirus

Thymidine kinase activity was found in whole cell extracts of growing and stationary mouse embryo fibroblast cells after infection with murine cytomegalovirus. Determination of the kinetic constants and heat stability characteristics indicated that the enzyme activity from infected cells was different to that found in uninfected cells in the growth phase. The expression of thymidine kinase activity during virus replication was reflected by the incorporation of (6-³H) thymidine into acid precipitable fractions of infected cell cultures. Preliminary data from kinetic studies showed a reduction in the phosphorylation of thymidine by this enzyme activity in the presence of Acyclovir, a potent inhibitor of herpes virus replication.     

Polyamines in mycoplasmas and in mycoplasma-infected tumour cells.

Three out of four different mycoplasma strains analysed for the polyamine contents contained relatively high concentrations of putrescine, cadaverine, spermidine and spermine. In addition to ornithine decarboxylase (EC 4.1.1.17) activity, the mycoplasmas also exhibited comparable or higher lysine decarboxylase (EC 4.1.1.18) activity fully resistant to the action of 2-difluoromethylornithine, an irreversible inhibitor of eukaryotic ornithine decarboxylase. 2-Difluoromethylornithine did not modify the polyamine pattern of actively growing mycoplasmas. Ehrlich ascites carcinoma cells and L1210 mouse leukemia cells infected with any of the four mycoplasma strains contained, in addition to putrescine, spermidine and spermine, and also easily measurable concentrations of cadaverine; the latter diamine was absent in uninfected cultures. When the infected cells were exposed to difluoromethylornithine, the accumulation of cadaverine was markedly enhanced. The modification of cellular polyamine pattern by mycoplasmas, especially in the presence of inhibitors of eukaryotic ornithine decarboxylase, could conceivably be used as an indicator of mycoplasma infection in cultured animal cells.     

A test for hormonal responsiveness in a mammary epithelial cell line,NMuMg

Summary The NMuMG cell line derived from normal mouse mammary epithelial cells was tested for responsiveness to hormones. The hormones studied included insulin, glucocorticoids (cortisol and dexamethasone), and prolactin. In addition to membrane bound insulin receptors and prolactin receptors, the cells had 2 × 10⁴ cytoplasmic glucocorticoid receptors per cell. Morphological changes were observed in response to hormones. Clusters of cells appeared with greatly increased diameter, and the number of cells per plate was reduced. The rate of DNA synthesis, corrected by cell number, indicates that cell division, and hence cell turnover, was increased by the combination of all three hormones. Insulin greatly enhanced protein synthesis, but glucocorticoid and prolactin did not further increase the rate. The combination of the three hormones produced a change in the synthesis of histones, consistent with the increase in cell turnover. There were substantial responses of enzyme activities to hormonal treatment of the cells. Insulin by itself induced a doubling of the activity of glyceraldehyde phosphate dehydrogenase and perhaps a modest increase in NADH-cytochromec reductase. Lactose synthetase activity showed a three- to fourfold induction of both A and B subunits of the enzyme when the cells were treated with insulin, glucocorticoid, and prolactin, and the effect of the latter two hormones was shown to be additional to that of insulin. This work was supported by Contract N01-CB-43866 from the National Cancer Institute, by Grants GB-38658 from the National Science Foundation and GMS-20338 from the National Institutes of Health, and by the Agricultural Experimental Station at the University of California.     

Chemical carcinogen-mouse mammary tumor virus interactions in cell transformation

Summary We have studied the process of mammary cell transformation in vitro using a single cell clone (Clone 18) from a presumptive epithelial cell line, C57MG, derived from a normal mammary gland; a mouse mammary tumor virus (MMTV) host-range variant (RIII)vp4; and the potent initiating carcinogen 7,12-dimethylbenz(a)anthracene (DMBA). After several serial subcultures, cells treated with virus and then with carcinogen exhibited an altered (transformed) morphology, a dramatic increase in anchorage independence, an increase in multinucleation after exposure to cytochalasin B, an enhanced ability to proliferate in low Ca²⁺ (0.01 mM) medium, and tumorigenicity when inoculated subcutaneously into athymic (nude) mice. Although some of these phenotypic alterations were observed also in cultures treated singly with MMTV or DMBA and in cultures exposed to DMBA before infection with MMTV, enhanced cytochalasin B multinucleation and tumorigenicity were properties observed only in mass cultures of cloned cells first infected with MMTV and then exposed to DMBA. This demonstrates for the first time that exposure of presumptive mammary epithelial cells to MMTV followed by DMBA, but not to either agent alone or to DMBA followed by MMTV, results in malignant transformation of these cells. Support for these studies was provided in part by the National Cancer Institute, Bethesda, Md., Contract N01-CP-01018.     

The activity of total histone H1 kinases is related to growth and commitment points while the p13suc1-bound kinase activity relates to mitoses in the alga Scenedesmus quadricauda

Under optimal growing conditions, synchronous cultures of the alga Scenedesmus quadricauda underwent three DNA replications and three mitoses during one cell cycle. This resulted in eight daughter cells. By different illumination regimes and temporal addition of cycloheximide, cell divisions resulting in two, four or eight daughter cells per cycle were obtained. The selected cell cycle patterns differed in timing of commitment points, the number of nuclear divisions, and their positioning in the cell cycle. These distinct cell cycle patterns allowed to assess the correlation of histone H1 kinase activity with commitment points and mitoses. The activity of the histone H1 kinases was assayed in cellular protein extracts and after affinity purification using the p13^suc1 protein. The main peaks of kinase activity in the cellular extract were found to correlate with the commitment points. Small histone H1 kinase activity peaks were also found which preceded the nuclear division. Contrary to the histone H1 kinase activity of cellular extracts, the p13^suc1-bound kinase activity preceded the nuclear division, whilst its activity was negligible at the commitment points. Being able to manipulate the timing of commitment points and cell division by manipulating experimental conditions, we could precisely match the commitment points to an as yet unidentified histone H1 kinase activity and mitosis to p13^suc1-bound CDK activity during a particular cell division pattern with overlapping cycles. This provides molecular evidence, that local activation of CDKs regulates distinct events of the cell cycle.     

Elimination of Mycoplasma from various cell cultures

Elimination ofMycoplasma orale-I from chronically infected cell lines was achieved either by treatment with a mixture of antibiotics in a hypotonic solution, or with 10 vol % of anti -M.orale rabbit serum in tissue culture medium. The latter treatment was preferable in most cases, as it was practically harmless to the cells. Inactivation of this antiserum had no effect on its potency. The antibiotic-hypotonic treatment was rather destructive, but to a different degree for the various cell cultures. Both methods were equally useful for the treatment of a monkey kidney cell line contaminated with a mycoplasma strain related toM.hyorhinis. The available anti -M.hyorhinis rabbit serum was toxic for the monkey cells when not inactivated. The potency of the antiserum was rather low and even lower after inactivation. However, prolonged treatment successfully eliminated the mycoplasma. Pre-incubation of the inactivated anti -M.hyorhinis serum with tissue culture medium to which 10% non-inactivated calf serum had been added, favoured the elimination of the mycoplasma.During the treatment of contaminated cell cultures with single antibiotics a strain related toM.hyorhinis became resistant to chlortetracyclin.M.orale- I was found to be resistant to various single antibiotics.We are grateful to Professor Dr. A. Ch. Ruys (University of Amsterdam) and Dr. R. H. Leach (Mycoplasma Reference Laboratory, London) for helpful discussions and for identifying some of our mycoplasma strains; Dr. Leach also for kindly supplying us with his G. D. L.-strain. We thank Dr. H. Cohen and Dr. A. C. Hekker for their criticism and Mr. N. L. M. van Zwetselaar for his accurate technical assistance.     

Specific Labeling of Intracellular Toxoplasma gondii with Uracil*

SYNOPSIS Radioactive uracil was not significantly incorporated into the nucleic acids of human fibroblast cells. Infection of these cells with Toxoplasma gondii resulted in an exponential increase in the rate of uracil incorporation that paralleled the exponential growth of the parasite. One day after infection the rate of uracil incorporation was increased 100-fold. It was established by autoradiography that all of the [³H] uracil was incorporated into the intracellular parasites. A possible explanation for this difference in ability to use uracil is our observation that the specific activity of uridine phosphorylase was 100-fold greater in partially purified parasites than in the host cell.     

Factors influencing microbiological assay of cell-culture mycoplasmas

Summary A total of 6432 cell cultures was assayed for mycoplasmas over a 6-year period by aerobic and anaerobic incubation of agar and broth media. Mycoplasmas were detected in 375 cultures (5.8%).M. orale andA. laidlawii accounted for 61.3% of the isolates. Anaerobic incubation detected 98.1% of the isolates; aerobic incubation detected 45.8%. Of factors studied to determine their effect on mycoplasma assay, only two, anaerobic incubation and presence of mycoplasmacidal/static antibiotics, were significant. In separate studies, 86 of 2656 cell cultures (3.2%) were infected with strains ofM. hyorhinis that did not grow on cell-free media. Recommendations are given for microbiological assay of cell-culture mycoplasmas. These studies were supported in part by Contracts N01-AG-4-2865 and N01-AG-8-2117 from the National Institute on Aging and N01-GM-6-2119 from the National Institute of General Medical Sciences.     

Adenosine phosphorylase activity in a mutant Hep-2 cell line contaminated withmycoplasma hyorhinis

Summary Metabolic studies in HEp-2/MP,MIR cells (an adenosine kinase, hypoxanthine phosphoribosyltransferase negative mutant) indicated the presence of adenosine phosphorylase activity. This activity, unknown in established mammalian cell lines, resulted in the glycosidic cleavage of both adenosine and the antiviral drug arabinosyladenine. The activity was observed readily in the presence or absence of the adenosine deaminase inhibitor coformycin. Isopycnic separation of [³H]thymidine-labeled DNA species in CsCl density gradients resulted in the appearance of two distinct peaks. The heavier peak coincided with [¹⁴C]thymidine-labeled marker DNA of human origin, whereas the lighter peak was within the range associated with mycoplasmal DNA. Testing by commercial laboratories confirmed the presence of mycoplasma in HEp-2/MP,MIR cells. The contaminant was identified asMycoplasma hyorhinis, a porcine mycoplasma. Following γ-irradiation (3000 rads) to block cellular mitosis, the mycoplasma-contaminated HEp-2/MP,MIR cells were cocultivated with mycoplasma-free wild-type HEp-2 cells which did not exhibit adenosine phosphorylase activity. Following serial cocultivation in a medium designed to favor the survival of the wild-type cells, adenosine phosphorylase activity was found in the previously uninfected cells. Studies of this nature emphasize the need for investigators to carefully monitor their cell lines for mycoplasma. Presented at the 25th Annual Meeting of the Tissue Culture Association, Philadelphia, Pa., June 1976. This work was supported by Public Health Service Grants DE 02731 from the National Institute of Dental Research and CA 16219 from the National Cancer Institute.     

Aminocentesis cells: Culture,cryogenic storage,isozymes, and finite life in vitro

Summary Forty-six cell cultures established from amniocentesis fluids were preserved in liquid nitrogen and later recovered from the frozen state with little loss of viability as compared to prefreeze viability. Five to 10% glycerol was found to be optimal for preservation in liquid nitrogen, and as few as 5×10⁵ viable cells per frozen ampule could initiate cell growth. Storage in liquid nitrogen did not affect the genetic stability of glucose-6-phosphate dehydrogenase, lactate dehydrogenase, malic dehydrogenase, leucine aminopeptidase, acid phosphatase, or 6-phosphogluconic acid dehydrogenase isozymes of the amnion cultures. These studies were supported by Contract NIH-NIGMS-72-2070, Grant CA-04953-13 from the National Cancer Institute; General Research Support Grant FR-5582 from the National Institutes of Health; and Grant-in-Aid Contract M-43 from the State of New Jersey. Recipient of Research Career Award 5-K3,16, 749 from the National Institutes of Health.     

Induction of macrophage-mediated cytolysis of neoplastic cells by mycoplasmas

Unexpected cytolysis was encountered when nonactivated murine peritoneal macrophages were cultured with [³H]TdR-prelabeled syngeneic or allogeneic tumor cells at a 10:1 ratio. The level of specific cytolysis reached 70% within 48 hr of cocultivation. Similar killing was observed whether the macrophages were derived from untreated, thioglycollate-treated, or germ-free mice. Cytolytic activity was also demonstrated when bone marrow-derived or peritoneal macrophages from 9- and 5-day in vitro cultures, respectively, were employed rather than freshly harvested peritoneal macrophages. Thus, the macrophage-mediated killing was neither the result of in vivo preactivation nor a consequence of the presence of lymphocytes in the assay. Moreover, macrophages derived from different strains caused similar effects. Our study revealed that the neoplastic target cell cultures susceptible to cytolysis by nonactivated macrophages were contaminated with mycoplasma. A mycoplasma was isolated from the supernatant of a culture of the A9HT fibrosarcoma line, identified as Mycoplasma orale, and cultivated. Addition of viable mycoplasma from that isolate to mixed cultures of thioglycollate-elicited macrophages and [³H]TdR-prelabeled mycoplasma-free target cells resulted in specific cytolysis of transformed A9 cells, but not of normal mouse fibroblasts. The level of macrophage-dependent cytolysis correlated with the number of viable mycoplasma cells added and was higher than that attained by activation with LPS at optimal concentration. Similar specific cytolysis was observed with heat-killed mycoplasmas. Our results demonstrate that mycoplasmas may cause selective macrophage-mediated cytolysis of neoplastic but not of normal target cells, perhaps via activation of the macrophages. It is suggested that undetected infection of experimental systems by mycoplasmas may account for some reports on lysis of neoplastic cells by nonactivated macrophages.     

Interactions of cultured rat synovial and ocular ciliary body cells with two strains of mycoplasma arthritidis

Summary Strains ofMycoplasma arthritidis differ in their ability to cause joint and ocular inflammations. Although the reasons for this difference are not fully understood, pathogenic myćoplasmas commonly require close associations with the cells they damage. Using³H-uridine labeled mycoplasma, we compared cellular interactions of in vitro cultivated rat synovial and ocular ciliary body epithelial cells with two American Type Culture Collection strains ofM. arthritidis shown to differ in their virulence. Radiolabeling assays gave evidence of á stronger retention capability on cultured cells by the more pathogenic strain, 14152. Scanning electron microscopy demonstrated cellular associations with the two strains of mycoplasma, with more of the 14152 adhering to both cell types. Examination by transmission electron microscopy showed evidence of contact between the more virulent 14152 strain and both cell types, but no similar evidence with the comparatively less virulent strain, 19611. The pathogenicity of different strains ofM. arthritidis may vary according to their ability to closely associate with specific target cells involved in the disease process. This work was supported by funding from the National Society for the Prevention of Blindness, Fight for Sight, Retinitis Pigmentosa International, grant EY05415 from the National Institutes of Health, Bethesda, MD, to Thirkill, and, in part, by an unrestricted grant from Research to Prevent Blindness.     

Cytotoxic effect of a T-strain mycoplasma (Ureaplasma urealyticum) on cultured normal human cells (WI-38)

Normal human embryonic lung fibroblasts (WI-38) were infected with Ureaplasma urealyticum, a urea hydrolysing mycoplasma. It was possible to observe reduced rates of multiplication of infected cells and reduced plating efficiency as well as the morphological changes usually associated with mycoplasma infection of animal cells in vitro. The cytotoxic effect on multiplication was sensitive to aureomycin but not penicillin. It was not related to depletion of amino acids or nucleic acid precursors from the cell culture medium but appeared to require that the host cells be growing. Ureaplasmas could not be recovered from cell culture medium after 4 days post infection and their characteristic urease activity could not be demonstrated either in cell culture medium or associated with the cells after the first cell subcultivation. [³H]TdR was incorporated into the nuclei of infected cells and the percent labelled nuclei was reduced compared with uninfected cells. Nuclear labelling indices of infected cells increased as the cells were subcultivated by trypsinization suggesting that the ureaplasmas were removed from the host cell surface by this treatment. In general, the effects of ureaplasmas on WI-38 cells do not appear to be as pronounced as effects of other mycoplasmas on animal cells in culture. It is clear, nonetheless, that the urea hydrolysing mycoplasmas can infect cells in culture and cause discernible effects on the growth and metabolism of these cells.     

Adenosine phosphorylase activity in mycoplasma-free growth media for mammalian cells

Mammalian cells have enzymes that deaminate adenosine to inosine, which can readily be phosphorolysed to hypoxanthine. They do not, however, possess enzymes to form adenine by the cleavage of adenosine. For this reason, the release of adenine from adenosine by mammalian cell cultures has usually been interpreted as indicating the presence of mycoplasma, a frequent microbial contaminant that contains high levels of adenosine phosphorylase. We found that some human lymphoblast cultures free of mycoplasma showed high levels of adenosine cleavage and that this activity resulted from adenosine phosphorylase in the bovine serum used as the culture growth supplement. A survey of 13 serum supplements disclosed that fetal bovine serum (six lots) contains the highest adenosine phosphorylase activity, ranging from 9 to 648 nmol adenine produced per hour per ml serum; newborn calf serum (four lots) has much less activity, ranging from 0 to 5 nmol adenine produced per hour per ml serum; and donor horse serum (three lots) contains no detectable activity. These results suggest that mycoplasma tests dependent on the presence of adenosine phosphorylase or other enzyme activities may give false-positives with cultures containing fetal bovine serum supplements.     

Culture of adult rat lung cells: Benzo(a)pyrene metabolism and mutagenesis

Summary A method is described for obtaining and culturing large numbers of lung cells from normal adult male rats. The lungs were perfused in situ to remove blood cells and then perfused via the trachea with a trypsin-collagenase solution to initiate tissue digestion. The tissue was further digested in the enzyme solution and approximately 2×10⁸ viable lung cells were obtained per animal. Primary cultures contained a mixed cell population. Through eight subcultures about 70% of the cell population possessed an epithelial-like morphology, whereas the remaining 30% was fibroblast-like. Three clones of epithelial-like cells were isolated at the fourth subculture. The mass culture lung cells and the epithelial-like clone that was studied retained a normal karyotype and did not grow in soft agar. Both the mass culture cells and the epithelial clone metabolized the lung carcinogen benzo(a)pyrene (BP) to water-soluble products. Furthermore, the mass culture lung cells metabolized BP to intermediate(s) which mutated Chinese hamster V79 cells from ouabain sensitivity to ouabain resistance. These lung cell cultures have potential use in cell transformation, mutation and carcinogen metabolism studies. Visiting scientist from Hungary. This research was supported by Grant 5 R01 CA20022 and Public Health Service Contract N01 CP33278 from the Division of Cancer Cause and Prevention, National Cancer Institute, National Institutes of Health.     

Rapid detection of mycoplasma contamination in cell cultures using sybr green-based real-time polymerase chain reaction

Summary We have developed a simple method for rapid detection of mycoplasma contamination in cell cultures using SYBR Green-based real-time polymerase chain reaction (PCR). To detect eight common contaminant mollicutes, including Mycoplasma (M. arginini, M. fermentans, M. orale, M. hyorhinis, M. hominis, M. salivarium, M. pirum) and Acholeplasma laidlawii, four primers were prepared based on the 23S rRNA regions. Using these primers and a minimum of 100 fg of mycoplasma genomic DNA, the 23S rRNA regions of these eight mycoplasma species were consistently amplified by real-time PCR. In contrast, no specific specific amplification product was observed using DNA templates prepared from various mammalian cell lines. Frozen and cultured samples of several cell lines were tested for mycoplasma contamination to evaluated the utility of this method. Of 25 samples that tested positive for mycoplasma by Hoechst staining, which requires two passages of cell cultures started from frozen samples, mycoplasma was detected by real-time PCR in 24 samples of cell extracts prepared directly from frozen samples. When cultured samples were used for this assay, the accuracy of the diagnoses was further improved. Thus, this technique, which is simple, rapid, and sensitive enough for practical application, in suitable for handling many samples and for routine screening for mycoplasma contamination of cell cultures.     

Isolation and Characterization of a Bacillus subtilis Mutant with a Defective N-Glycosidase Activity for Uracil-Containing Deoxyribonucleic Acid

Crude cell extracts of Bacillus subtilis 168T exhibit enzyme activity capable of releasing free uracil from phage PBS1 deoxyribonucleic acid (DNA) in the presence of ethylenediaminetetraacetate. By measuring the enzyme activity in 300 clones that emanated from mutagenized cells, we obtained a mutant strain that did not show this N-glycosidase activity. The mutant strain, designated as TKJ6901 (urg-1) exhibited no physiological abnormalities. We observed the intracellular action of the enzyme by following the fate of uracil-containing DNA in cells from wild-type and mutant cultures. When infection with phage PBS1 was allowed in the presence of chloramphenicol, extensive degradation of phage DNA was observed only in the wild-type cells. When bromouracil residues were converted to uracil residues by ultraviolet light irradiation in the presence of cysteamine, the DNA was extensively fragmented in the wild-type cells. These single-strand breaks were rejoined upon postirradiation incubation. In contrast, such fragmentation of the DNA was not observed in the mutant cells, indicating that the uracil residues were not removed from the DNA. This demonstrated that the N-glycosidase activity was involved in the excision of uracil in DNA. A transformation assay with four types of recipient strains with combinations of N-glycosidase and DNA polymerase I deficiencies indicated that DNA polymerase I was involved in the later steps of this base excision repair pathway initiated by the action of the N-glycosidase.