Incorporation Of Deoxyuridine Research Articles

Abstract IA022: Uracil DNA glycosylase activity limits deoxyuridine contamination in genomic DNA and is essential for the type-1 interferon response in cells treated with ATR kinase inhibitors

Abstract Most chemotherapies target DNA replication, and their efficacy depends on the DNA damage response that integrates DNA repair with the cell cycle. Widely used standard-of-care antimetabolites inhibit thymidylate synthase and increase the incorporation of deoxyuridine (dU) into DNA by polymerases. Contamination of dU in DNA is limited by the DNA damage response kinase ATR that induces cell cycle checkpoints and reduces the rate of DNA replication. ATR inhibitors (ATRi) induce origin firing across active replicons and cause ribonuclease reductase degradation in otherwise unperturbed cells. This increases both the amount of DNA replication and the amount of free dUTP in cells. Thus, ATRi increase the incorporation of dU into DNA by polymerases. Since ATRi also inhibit cell cycle checkpoints, ATRi induce more dU contamination than antimetabolites. ATRi-induced dU contamination in DNA is associated with an innate immune response. We showed this with the simple observation that ATRi-induced dU contamination and IFN-α/β expression is reversed by low doses of thymidine. Here we show that ATRi-induced IFN-α/β response is dependent on uracil DNA glycosylase (UNG) which removes dU from DNA. Our data are consistent with a model in which UNG-dependent base excision repair (BER) removes dU from DNA, ultimately generating cytoplasmic dsDNA that induces IFN-α/β. We propose that ATRi-induced dU contamination contributes to dose-limiting leukocytopenia and inflammation in the clinic and CD8+ T cell dependent anti-tumor responses in mouse models of cancer treated with radiation therapy. Citation Format: Pinakin Pandya, Frank P. Vendetti, Joseph A. Ghoubaira, Sudipta Pathak, Joshua J. Deppas, Reyna E. Jones, Yunqi Zhang, Daniel Ivanov, Raquel Buj, Katherine M. Aird, Jan H. Beumer, Robert W. Sobol, Christopher J. Bakkenist. Uracil DNA glycosylase activity limits deoxyuridine contamination in genomic DNA and is essential for the type-1 interferon response in cells treated with ATR kinase inhibitors [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: DNA Damage Repair: From Basic Science to Future Clinical Application; 2024 Jan 9-11; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2024;84(1 Suppl):Abstract nr IA022.

Constitutive stable DNA replication in Escherichia coli cells lacking type 1A topoisomerase activity

Type 1A topoisomerases (topos) are ubiquitous enzymes involved in supercoiling regulation and in the maintenance of genome stability. Escherichia coli possesses two type 1A enzymes, topo I (topA) and topo III (topB). Cells lacking both enzymes form very long filaments and have severe chromosome segregation and growth defects. We previously found that RNase HI overproduction or a dnaT::aph mutation could significantly correct these phenotypes. This leads us to hypothesize that they were related to unregulated replication originating from R-loops, i.e. constitutive stable DNA replication (cSDR). cSDR, first observed in rnhA (RNase HI) mutants, is characterized by its persistence for several hours following protein synthesis inhibition and by its requirement for primosome components, including DnaT. Here, to visualize and measure cSDR, the incorporation of the nucleotide analog ethynyl deoxyuridine (EdU) during replication in E. coli cells pre-treated with protein synthesis inhibitors, was revealed by “click” labeling with Alexa Fluor® 488 in fixed cells, and flow cytometry analysis. cSDR was detected in rnhA mutants, but not in wild-type strains, and the number of cells undergoing cSDR was significantly reduced by the introduction of the dnaT::aph mutation. cSDR was also found in topA, double topA topB but not in topB null cells. This result is consistent with the established function of topo I in the inhibition of R-loop formation. Moreover, our finding that topB rnhA mutants are perfectly viable demonstrates that topo III is not uniquely required during cSDR. Thus, either topo I or III can provide the type 1A topo activity that is specifically required during cSDR to allow chromosome segregation.

A High-Throughput Continuous Assay for Screening and Characterization of Inhibitors of HIV Reverse-Transcriptase DNA Polymerase Activity

The authors have devised a continuous fluorescence-based assay to measure HIV reverse transcriptase (RT) polymerase activity for both high-throughput screening (HTS) and mechanistic characterization of inhibitors. The designed substrate is composed of a recessed DNA primer annealed to a DNA template that is labeled at the 5'-terminus with a donor fluorophore (AlexaFluor 488). RT-catalyzed incorporation of an acceptor-labeled deoxyuridine (dUTP-AlexaFluor 555) at the 3'-terminus of the fully extended DNA primer juxtaposes donor and acceptor fluorophores, resulting in robust fluorescence resonance energy transfer that can be monitored kinetically in real time. The assay is sensitive, permitting the use of low enzyme concentrations (<0.5 nM), and can be miniaturized for use in 384-well HTS mode. The authors further show that this assay is capable of evaluating inhibitor mechanism of action by confirming the binding mechanism of a set of nonnucleoside RT inhibitors. Given the versatility and the lack of requirement for costly platforms or radioactivity, this assay may serve to accelerate and streamline the discovery and characterization process for future antiviral agents.

Three DNA Polymerases, Recruited by Different Mechanisms, Carry Out NER Repair Synthesis in Human Cells

Nucleotide excision repair (NER) is the most versatile DNA repair system that deals with the major UV photoproducts in DNA, as well as many other DNA adducts. The early steps of NER are well understood, whereas the later steps of repair synthesis and ligation are not. In particular, which polymerases are definitely involved in repair synthesis and how they are recruited to the damaged sites has not yet been established. We report that, in human fibroblasts, approximately half of the repair synthesis requires both pol kappa and pol delta, and both polymerases can be recovered in the same repair complexes. Pol kappa is recruited to repair sites by ubiquitinated PCNA and XRCC1 and pol delta by the classical replication factor complex RFC1-RFC, together with a polymerase accessory factor, p66, and unmodified PCNA. The remaining repair synthesis is dependent on pol epsilon, recruitment of which is dependent on the alternative clamp loader CTF18-RFC.

Polypurine Hairpins Directed against the Template Strand of DNA Knock Down the Expression of Mammalian Genes

We analyzed whether polypurine hairpins (PPRHs) had the ability to knock down gene expression. These hairpins are formed by two antiparallel purine domains linked by a loop that allows the formation of Hoogsteen bonds between both domains and Watson-Crick bonds with the target polypyrimidine sequence, forming triplex structures. To set up the experimental conditions, the human dhfr gene was used as a model. The PPRHs were designed toward the template strand of DNA. The transfection of the human breast cancer cell line SKBR3 with these template hairpins against the dhfr gene produced higher than 90% of cell mortality. Template PPRHs produced a decrease in DHFR mRNA, protein, and its corresponding enzymatic activity. In addition, the activity of DHFR PPRHs was tested against breast cancer cells resistant to methotrexate, observing high cell mortality. Given the difficulty in finding long polypyrimidine stretches, we studied how to compensate for the presence of purine interruptions in the polypyrimidine target sequence. The stability of PPRH was measured, resulting in a surprisingly long half-life of about 5 days. Finally, to test the generality of usage, template PPRHs were employed against two important genes involved in cell proliferation, telomerase and survivin, producing 80 and 95% of cell death, respectively. Taken together our results show the ability of antiparallel purine hairpins to bind the template strand of double strand DNA and to decrease gene transcription. Thus, PPRHs can be considered as a new type of molecules to modulate gene expression.

A rapid non-radioactive technique for measurement of repair synthesis in primary human fibroblasts by incorporation of ethynyl deoxyuridine (EdU)

Xeroderma pigmentosum (XP) is an autosomal recessive genetic disorder. Afflicted patients show extreme sun-sensitivity and skin cancer predisposition. XP is in most cases associated with deficient nucleotide excision repair (NER), which is the process responsible for removing photolesions from DNA. Measuring NER activity by nucleotide incorporation into repair patches, termed ‘unscheduled DNA synthesis (UDS)’, is one of the most commonly used assays for XP-diagnosis and NER research. We have established a rapid and accurate procedure for measuring UDS by replacement of thymidine with 5-ethynyl-2'-deoxyuridine (EdU). EdU incorporated into repair patches can be directly conjugated to fluorescent azide derivatives, thereby obviating the need for either radiolabeled thymidine or denaturation and antibody detection of incorporated bromodeoxyuridine (BrdU). We demonstrate that the EdU incorporation assay is compatible with conventional techniques such as immunofluorescent staining and labeling of cells with micro-latex beads. Importantly, we can complete the entire UDS assay within half a day from preparation of the assay coverslips; this technique may prove useful as a method for XP diagnosis.

Synthesis and high-throughput evaluation of triskelion uracil libraries for inhibition of human dUTPase and UNG2

Human nuclear uracil DNA glycosylase (UNG2) and deoxyuridine triphosphate nucleotidohydrolase (dUTPase) are the primary enzymes that prevent the incorporation and accumulation of deoxyuridine in genomic DNA. These enzymes are desirable targets for small molecule inhibitors given their roles in a wide range of biological processes ranging from chromosomal rearrangements that lead to cancer, viral DNA replication, and the formation of toxic DNA strand breaks during anticancer drug therapy. To accelerate the discovery of such inhibitors, we have developed a high-throughput approach for directed library synthesis and screening. In this efficient technology, a uracil-aldehyde ligand is covalently tethered to one position of a trivalent alkyloxyamine linker via an oxime linkage, and then the vacant linker positions are derivatized with a library of aldehydes. The resulting triskelion oximes were directly screened for inhibitory activity and the most potent of these showed micromolar binding affinities to UNG2 and dUTPase.

DNA synthesis and fragmentation in bacteroids during Astragalus sinicus root nodule development.

Histo- and cytochemical techniques were used to study the DNA replication and fragmentation patterns in bacteroids formed by Mesorhizobium huakuii subsp. rengei in nodules of Astragalus sinicus. DNA replication was detected by the incorporation of 5-bromo deoxy-uridine. Signals denoting DNA synthesis were observed in plant nuclei within the nodule meristem and in bacteroids near the meristem. The TUNEL (TdT-mediated dUTP nick-end labeling) assay was used to measure DNA fragmentation. In nutrient-depleted 1-mpi (month(s) post inoculation) nodule sections, some bacteroids were in vacuoles, and DNA fragmentation signals were observed only in such bacteroids. In contrast, 1-mpi nodule sections without nutrient depletion showed neither bacteroid localization in vacuoles nor DNA fragmentation signals. The bacteroid translocation into vacuoles upon nutrient starvation might results from autophagy of the plant. In 2-mpi nodule sections, bacteroids with DNA fragmentation signals appeared within the cytoplasm of some nodule cells in the senescence zone.

A novel and rapid PCR-based method for genotyping human papillomaviruses in clinical samples.

Many human papillomavirus (HPV) genotypes are associated with cervical carcinoma. We demonstrate the utility of an innovative technique for genotyping of HPV in cervical tissue samples. This method provides an accurate means of identification of the specific HPV genotypes present in clinical specimens. By using the MY09-MY11 and the GP5(+)-GP6(+) consensus primer pairs, HPV sequences were amplified by nested PCR from DNA isolated from cervical smear samples. This led to the production of an approximately 140-bp PCR product from the L1 (major capsid) gene of any of the HPVs present in the sample. PCR was performed with a deoxynucleoside triphosphate mixture which resulted in the incorporation of deoxyuridine into the amplified DNA product at positions where deoxythymidine would normally be incorporated at a frequency of about once or twice per strand. Following the PCR, the product was treated with an enzyme mix that contains uracil N-glycosylase (UNG) and endonuclease IV. UNG removes the uracil base from the nucleotide, and endonuclease IV cleaves the phosphodiester bond at this newly formed abasic site, producing fragments of various sizes. By having end labeled one of the amplification primers, a DNA ladder which is analogous to a "T-sequencing ladder" was produced upon electrophoresis of the products. By comparing this T-sequencing ladder to the known sequences of HPVs, the genotypes of unknown HPV isolates in samples were assigned. Data showing the utility of this technique for the rapid analysis of clinical samples are presented.

In vitro antitumour activity of cis- and trans-5-fluoro-5,6-dihydro-6-alkoxy-uracils; effects on thymidylate synthesis.

A class of new 5-fluorouracil (FU) analogues, the 5-fluoro-5,6-dihydro-6- alkoxy-uracils was synthesised with a modification at the 6-position of the pyrimidine ring. At this position the analogues have a hydroxy or alkoxy group of different chain lengths either in the cis- or trans-configuration. The antiproliferative effect of these compounds was tested on five cell lines of different origin. Generally, the analogues with a cis-configuration had a higher activity than those with a trans-configuration. The growth inhibitory effect of the compounds decreased with increasing alkoxy chain length, but the compound with a hydroxy group had the lowest growth inhibitory effect. One analogue, cis-5-F-5,6-dihydro-6-methoxy-uracil had a higher antiproliferative effect than FU in one of the cell lines. Effects on thymidylate synthase (TS), the possible target of these analogues, were evaluated by thymidine rescue of growth inhibition and incorporation of tritiated deoxyuridine (3H-UdR) into DNA. In solid tumour cell lines addition of TdR reversed the antiproliferative effect. Inhibition of TS in intact cells was determined by measuring 3H-UdR incorporation in two cell lines. The effect of cis-5-F-5,6-dihydro-6-methoxy-uracil on incorporation of 3H-UdR was 2- to 5-fold stronger than that of FU in both cell lines. All other compounds produced a higher 3H-UdR incorporation than FU both at equimolar and equi-toxic concentration. Concluding from these results we regard cis-5-F-5,6-dihydro-6-methoxy-uracil as the most promising FU analogue of this series, because of its higher antiproliferative activity than FU and marked inhibition of TS in intact cells.

Biochemical responses of broiler chicks to folate deficiency.

The effects of folate sufficiency and deficiency in three pathways of folate metabolism were studied in 2-and 3-week-old broiler chicks. Erythrocyte phosphoribosylpyrophosphate concentrations and dihydrofolate reductase (EC 1.5.1.3) activity were significantly elevated in folate deficiency. Percentage incorporation of deoxyuridine into bone marrow DNA was reduced in folate deficiency. There was a trend towards reduced liver dihydrofolate reductase activity in deficient chicks. These studies identify further biochemical criteria that can be used to assess folate status of chicks.

Tests of a model for promoter recognition by T7 RNA polymerase: thymine methyl group contacts.

The DNA-dependent RNA polymerase from bacteriophage T7 is highly specific for a 17 base promoter sequence. Interactions between T7 RNA polymerase and its promoter DNA have been probed using modified oligonucleotides and a steady-state kinetic assay. The incorporation of deoxyuridine in place of thymidine at individual sites in the promoter sequence results in the replacement of an exocyclic methyl group by hydrogen (effectively removing the thymine methyl). This substitution has been placed individually at each of the thymines in the T7 consensus promoter. Many of these substitutions do not affect binding or catalysis; however, the thymine methyl group at position-6 is critical to recognition. The kinetic parameter Km increases approximately 10-fold while kcat is only slightly affected, suggesting that this thymine methyl is critical to binding specificity, but not to the kinetics of initiation. Two methyl groups near the start site on the template strand (at positions -1 and -3) also contribute to promoter specificity, while nearby methyl groups on the nontemplate strand do not. The implications of these results are discussed with respect to recent models for promoter binding.

In vitro antiproliferative and metabolic activity of eight novel 5-fluorinated uracil nucleosides

The in vitro growth inhibitory activity of eight novel 5-fluorinated uracil nucleosides was assessed in four human tumour cell lines, one of colon and three of head and neck squamous cell origin. These compounds are ribose or deoxyribose sugars with an acetoxy or an hydroxyl-group at the 6-position in the uracil part of the molecule, and their respective diastereoisomers. Antiproliferative effects were tested in an automated microculture assay based on the reduction of a tetrazolium dye, the MTT assay. Using a continuous drug exposure for four days, all novel nucleosides were more potent inhibitors of cell growth than 5-fluorouracil (5-FU). Most drugs were very active, having an IC 50 value at least 10 fold lower than that of 5-FU, and this was consistently found for all cell lines. The 6-acetoxy compounds were generally more active than the compounds with a hydroxyl-group at the 6-position, while diastereoisomerism did not seem to influence the antiproliferative effect. Their capacity to inhibit the incorporation of tritiated deoxyuridine into DNA, which reflects the inhibition of thymidylate synthase, was measured in a short term assay. When tested at a concentration of 10 −6 mol/l, most of the compounds were found to block this incorporation more efficiently than 5-FU.

Fluorodeoxyuridine-induced radiosensitization and inhibition of DNA double strand break repair in human colon cancer cells

The halogenated pyrimidine, fluorodeoxyuridine (FdUrd), has been used in combination with radiation for the treatment of human neoplasms. In an attempt to improve the clinical use of this combination, FdUrd-radiation interactions were studied in vitro using human HT29 colon cancer cells. It was found that FdUrd produced radio-sensitization at clinically achievable (1–100 nM) concentrations. Sensitization depended critically on the timing of exposure. When cells were irradiated after a 12-hr exposure to 100 nM FdUrd, marked sensitization was produced (mean inactivation dose (MID) = 2.01 ± 0.01, compared to control of 4.35 ± 0.16, p < .01). No radiosensitization occurred when cells were irradiated 4 hr prior to incubation (MID = 3.95 ± 0.05, p > 0.4). Radiosensitization appeared to result from an inhibition of thymidylate synthase since concentrations of FdUrd which produced radiosensitization depleted intracellular TTP pools and blocked the incorporation of deoxyuridine into DNA. Furthermore, radiosensitization was completely inhibited by co-incubation with thymidine. FdUrd also decreased the repair, but not the formation, of radiation-induced DNA double strand breaks (DSB's). These data are consistent with the hypothesis that FdUrd produces radiosensitization by depleting thymidine pools which leads to a decreased rate of DNA DSB repair. Furthermore, they suggest that in clinical trials FdUrd should be infused at least 8 hr before irradiation.

Immunosuppressive properties of chloroquinoxaline sulfonamide

Chloroquinoxaline sulfonamide (CQS), a sulfanilamide derivative with antitumor activity, was found to be toxic to lymphoid tissue during preclinical studies. The mechanism of this toxicity appears to involve profound inhibition of lymphocyte activation. Incubation of human peripheral blood mononuclear cells (PBMNCs) with CQS decreased cellular incorporation of thymidine and deoxyuridine in a dose-dependent manner. Analysis of cell cycle distribution by flow cytometry indicated that CQS blocked movement out of the G 0 G 1 phase. Drug-treated cells were smaller and expressed fewer receptors for interleukin-2 (IL-2) and transferrin than untreated mitogen-stimulated lymphocytes. These observations support the notion that CQS has cell cycle specificity in regulating lymphocyte proliferation. As little as 10 μM CQS markedly inhibited both human lymphocyte and murine CTLL cell replication in response to IL-2 containing growth factors. However, CQS did not block secretion of IL-2 into culture supernatant fractions by human PBMNCs. Finally, CQS inhibited in vitro production of immunoglobulins G and M by mitogen-stimulated lymphocytes, primarily by causing cytotoxicity. In all of these drug effects, CQS was approximately one to two logs more potent than the parent compound, sulfaquinoxaline (SQ). These studies indicate that CQS inhibits essential basic processes in human lymphocytes. This agent may find use as an immunosuppressive drug.

A new folate antimetabolite, 5,10-dideaza-5,6,7,8-tetrahydrofolate is a potent inhibitor of de novo purine synthesis

5,10-Dideazatetrahydrofolate (DDATHF) is a new antimetabolite designed as an inhibitor of folate metabolism at sites other than dihydrofolate reductase. DDATHF was found to inhibit the growth of L1210 and CCRF-CEM cells in culture at concentrations in the range of 10-30 nM. The inhibitory effect of DDATHF on the growth of L1210 and CCRF-CEM cells was reversed by either hypoxanthine or aminoimidazole carboxamide. Growth inhibition by DDATHF was prevented by addition of both thymidine and hypoxanthine, but not by thymidine alone. 5-Formyltetrahydrofolate reversed the effects of DDATHF in a dose-dependent manner. DDATHF had no appreciable inhibitory activity against either dihydrofolate reductase or thymidylate synthase in vitro, but was found to be an excellent substrate for folylpolyglutamate synthetase. DDATHF had little or no effect on incorporation of either deoxyuridine or thymidine into DNA, in distinct contrast to the effects of the classical dihydrofolate reductase inhibitor, methotrexate. DDATHF was found to deplete cellular ATP and GTP over the same concentrations as those inhibitory to leukemic cell growth, suggesting that the locus of DDATHF action was in the de novo purine biosynthesis pathway. The synthesis of formylglycinamide ribonucleotide in intact L1210 cells was inhibited by DDATHF with the same concentration dependence as inhibition of growth. This suggested that DDATHF inhibited glycinamide ribonucleotide transformylase, the first folate-dependent enzyme of de novo purine synthesis. DDATHF is a potent folate analog which suppresses purine synthesis through direct or indirect inhibition of glycinamide ribonucleotide transformylase.

Effects of Nitrous Oxide on Macromolecular Content and DNA Synthesis in Rat Embryos

Exposure of pregnant rats to the anesthetic gas, nitrous oxide (N2O), has been reported to be teratogenic. This gas decreases activity of the enzyme 5-methyltetrahydrofolate homocysteine methyltransferase (methionine synthetase) and alters the level and distribution of folic acid derivatives in embryonic tissue. Because of their role in purine, thymidine and amino acid metabolism, alterations in the levels of various folate forms could affect macromolecular synthesis. The effect of N2O exposure on the number of somites, the content of DNA, RNA and protein as well as the incorporation of thymidine, deoxyuridine or serine into DNA were determined. Pregnant rats were exposed for 24 hr to 50% N2O-50% O2 beginning on day 10 of gestation. Control animals were exposed to 50% N2-50% O2. Embryos were removed and cultured for 4 hr using a rodent whole embryo culture system in medium containing radiolabeled precursors for DNA. Treatment with N2O had no effect on somite number, RNA or protein content. However, there was a significant decrease in DNA content. There was an increase in the incorporation of thymidine into DNA, but the incorporation of deoxyuridine and serine into DNA was decreased. These data demonstrate that treatment of pregnant rats with N2O results in decreased DNA synthesis and content in exposed embryos.

Folate- and deoxyuridine-sensitive chromatid breakage may result from DNA repair during G2

A substantial increase in chromosome breakage was seen in proliferating human lymphocytes treated with 1-β-D-arabinofuranosyl cytosine (cytosine arabinoside, cytarabine, araC) during the last 2 h of culture. The increase was larger in low folate media than in high folate media and larger still in low folate media supplemented by deoxyuridine. Similar results were found when cells were exposed to aphidicolin for the last 2 h of culture. The results provide additional evidence that folate-sensitive chromosome breakage is a consequence of abnormal DNA synthesis (in particular incorporation of deoxyuridine) and subsequent attempts during G2 to repair the abnormal DNA.

Effects of ciprofloxacin on eucaryotic pyrimidine nucleotide biosynthesis and cell growth.

Several of the new 4-quinolones significantly increase the incorporation of [3H]thymidine into the DNA of mitogen-stimulated human lymphocytes. This study suggests that ciprofloxacin inhibits de novo pyrimidine biosynthesis, thereby resulting in a compensatory increase in the uptake of pyrimidine precursors through salvage pathways, and that additional effects may affect eucaryotic cell growth. Incorporation of deoxyuridine, uridine, and orotic acid as well as thymidine was increased in the presence of ciprofloxacin, one of the antibacterially most active of the new 4-quinolones. In contrast, the uptake was decreased in very high concentrations of the drug. Culture in HAT (hypoxanthine, aminopterine, thymidine) medium, which blocks de novo thymidylate synthesis, abrogated the increase in [3H]thymidine incorporation induced by ciprofloxacin. Ciprofloxacin also failed to increase the uptake of [14C]hypoxanthine or leucine, indicating a selective effect on pyrimidine and not on purine nucleotide biosynthesis. N-(Phosphonacetyl)-L-aspartate, an inhibitor of pyrimidine nucleotide biosynthesis, also increased [3H]thymidine incorporation in phytohemagglutinin-stimulated lymphocytes in a fashion similar to ciprofloxacin. The growth of several cell lines was partially inhibited by ciprofloxacin at 20 micrograms/ml and completely inhibited at 80 to 160 micrograms/ml. Growth inhibition by ciprofloxacin could not be restored by the addition of uridine to the medium. Chromosome breaks, gene amplification, or other genetic alterations could not be detected in human lymphocytes incubated with up to 25 micrograms of ciprofloxacin per ml.

Effect of liposomes sensitized with methotrexate-γ-dimyristoylphosphatidylethanolamine on cells that are resistant to methotrexate

This study compares the ability of methotrexate and liposomes, in which the drug is anchored to the lipid bilayers via methotrexate-γ-dimyristoylphosphatidylethanolamine, to inhibit proliferation of human leukemic cells (CEM/O) and cells derived from this line that are resistant to methotrexate because of either a defective transport system (CEM/MTX cells) or elevated levels of dihydrofolate reductase (CEM/R 1 cells). Whereas CEM/O and CEM/MTX cells show a 120-fold difference in their susceptibility to methotrexate (as measured by the incorporation of tritiated deoxyuridine into DNA), both lines are equally sensitive to the liposomes. In contrast, proliferation of CEM/MTX cells is not inhibited significantly by methotrexate-γ-glycerophosphorylethanolamine (MTX-γ-glyceroPE), the water-soluble analog of MTX-γ-DMPE. Both the ability of the liposomes to circumvent the transport defect, and the inability of MTX-γ-glyceroPE to do so, were anticipated on the basis of previous experiments which show that thiamine pyrophosphate could antagonize inhibition of mouse 3T3 and L1210 cell proliferation by methotrexate and MTX-γ-glyceroPE, but not inhibition by liposomes. Human cells (CEM/O) behave similarly. The present experiments also suggest that liposomes prepared with MTX-γ-DMPE can partially reverse the methotrexate resistance of CEM/R 1 cells that is due to overproduction of the target enzyme.