Immunotoxin therapy of cancer

Rationally designed anticancer agents that target cell-surface antigens or receptors represent a promising approach for treating cancer patients. However, antibodies that bind these targets are often, by themselves, non-cytotoxic. By attaching potent toxins we can dramatically improve the clinical utility of some anti-tumour antibodies. Here we describe the construction and clinical utility of several recombinant immunotoxins; each of which is composed of antibody Fv fragments fused to powerful bacterial toxins. Results from clinical trials indicate that recombinant immunotoxins and similar agents that are designed to combine antibody selectivity with toxin cell-killing potency will be useful additions to cancer therapy.     

Hemangiosarcoma and its cancer stem cell subpopulation are effectively killed by a toxin targeted through epidermal growth factor and urokinase receptors

Targeted toxins have the potential to overcome intrinsic or acquired resistance of cancer cells to conventional cytotoxic agents. Here, we hypothesized that EGFuPA‐toxin, a bispecific ligand‐targeted toxin (BLT) consisting of a deimmunized Pseudomonas exotoxin (PE) conjugated to epidermal growth factor and urokinase, would efficiently target and kill cells derived from canine hemangiosarcoma (HSA), a highly chemotherapy resistant tumor, as well as cultured hemangiospheres, used as a surrogate for cancer stem cells (CSC). EGFuPA‐toxin showed cytotoxicity in four HSA cell lines (Emma, Frog, DD‐1 and SB) at a concentration of ≤100 nM, and the cytotoxicity was dependent on specific ligand‐receptor interactions. Monospecific targeted toxins also killed these chemoresistant cells; in this case, a “threshold” level of EGFR expression appeared to be required to make cells sensitive to the monospecific EGF‐toxin, but not to the monospecific uPA‐toxin. The IC₅₀ of CSCs was higher by approximately two orders of magnitude as compared to non‐CSCs, but these cells were still sensitive to EGFuPA‐toxin at nanomolar (i.e., pharmacologically relevant) concentrations, and when targeted by EGFuPA‐toxin, resulted in death of the entire cell population. Taken together, our results support the use of these toxins to treat chemoresistant tumors such as sarcomas, including those that conform to the CSC model. Our results also support the use of companion animals with cancer for further translational development of these cytotoxic molecules.     

Diphtheria toxin-based targeted toxin therapy for brain tumors

Targeted toxins (TT) are molecules that bind cell surface antigens or receptors such as the transferrin or interleukin-13 receptor that are overexpressed in cancer. After internalization, the toxin component kills the cell. These recombinant proteins consist of an antibody or carrier ligand coupled to a modified plant or bacterial toxin such as diphtheria toxin (DT). These fusion proteins are very effective against brain cancer cells that are resistant to radiation therapy and chemotherapy. TT have shown an acceptable profile for toxicity and safety in animal studies and early clinical trials have demonstrated a therapeutic response. This review summarizes the characteristics of DT-based TT, the animal studies in malignant brain tumors and early clinical trial results. Obstacles to the successful treatment of brain tumors include poor penetration into tumor, the immune response to DT and cancer heterogeneity.     

Evaluation of monoclonal antibodies for the development of breast cancer immunotoxins

Eighty-five antibodies recognizing breast cancer-selective antigens were conjugated to ricin toxin A-chain using a disulfide linkage. The cytotoxicities of the resulting immunotoxins were determined on breast cancer cells and normal human fibroblasts. Twenty-four antibodies formed immunotoxins that were toxic to at least one breast cancer cell line at concentrations of 10 nM or less but were nontoxic to human fibroblast lines used as negative controls. Some of the breast tumor-selective immunotoxins were as toxic as a conjugate between monoclonal anti-transferrin receptor and ricin toxin A-chain (50% inhibition of cellular protein synthesis at approximately 0.1 nM). Another set of four immunotoxins were indiscriminately toxic to human breast tumor cell lines, two human fibroblast cell lines, and a human lymphoblastoid line. Several of the antibodies the toxin conjugates of which specifically killed breast cancer cell lines may be useful in cancer therapy, since they show a wide range of binding to individual breast tumors and cell lines and a limited range of binding to normal tissue types.     

Immunotoxins directed against the high-molecular-weight melanoma-associated antigen. Identification of potent antibody-toxin combinations.

To study factors influencing the cytotoxicity of immunotoxins (ITs), we compared the in vitro cytotoxicity of conjugates in which the plant toxin abrin and the bacterial toxin Pseudomonas exotoxin A (PE) were coupled by 2 different procedures to 2 MAbs, 9.2.27 and NR-ML-05, which bind to different epitopes on the melanoma-associated antigen p250. The individual target cell lines differed widely in sensitivity to the different ITs, as assessed by measurement of protein synthesis inhibition. The action of the ITs was highly specific, as the toxicity of abrin and PE conjugates was respectively 20-540 and 2,200-550,000 times higher in antigen-positive cell lines (FEMX, SESX, OHS) than in the antigen-negative line KPDX. The PE conjugates prepared with the 2 different MAbs differed in potency by factors of 16-126 in the target-cell lines, but were consistently more toxic than the abrin ITs. The results demonstrate that the cytotoxicity of ITs varies with the nature of both of its moieties and that optimal results require that toxins and MAbs be matched. Moreover, the 2 coupling procedures affected differentially the binding and potency of some ITs. Each of the 2 toxins was conjugated to a sheep anti-mouse antibody (SAM) and the toxicity of these 2 conjugates was tested in an indirect approach using 9.2.27 and NR-ML-05 as primary MAbs. The results showed that the indirect procedure would have correctly predicted the most potent antibody-toxin pair, indicating that the approach may be valid for selecting suitable combinations of MAbs and toxins for use as direct ITs.     

Mutant anthrax toxin B moiety (protective antigen) inhibits angiogenesis and tumor growth

Bacillus anthracis protective antigen (PA), the B subunit of the binary anthrax toxin, binds to the cellular receptors capillary morphogenesis gene 2 protein and tumor endothelial marker 8 with high affinity. Both receptors are expressed on endothelial cells during angiogenesis. We sought to determine whether one could inhibit angiogenesis by interfering with the binding of these receptors to their endogenous ligands. Here, we show that wild-type PA inhibits both vascular endothelial growth factor-induced and basic fibroblast growth factor-induced angiogenesis at moderate but statistically significant levels. Structure-activity studies identified a PA mutant that exhibited markedly enhanced inhibition of angiogenesis and also inhibited tumor growth in vivo. This mutant, PASSSR, is unable to undergo normal cellular processing and, thus, remains bound to the surface receptor. Further mutation of PASSSR so that it does not bind to these cell surface receptors abolished its ability to inhibit angiogenesis. We conclude that high-affinity anthrax toxin receptor (ATR) ligands, such as PA and PASSSR, are angiogenesis inhibitors and that ATRs are useful targets for antiangiogenic therapy. These results also suggest that endothelial cell-binding proteins from additional pathogens may inhibit angiogenesis and raise the question of the role of such inhibition in pathogenesis.     

Specific destruction of hybridoma cells by antigen-toxin conjugates demonstrate an efficient strategy for targeted drug therapy in leukemias of the B cell lineage

Many types of leukemia including multiple myeloma remain essentially incurable despite recent developments in immuno- and chemotherapy. The effectiveness of these therapies might be greatly enhanced by targeting cell surface proteins unique to the malignant clone, which for leukemias of the B cell lineage means clonotypic surface immunoglobulin (sIg). As this immunoglobulin (Ig) is necessarily epitope specific, we are developing ligand-toxin conjugates (LTCs) as a strategy for delivering toxins and other drugs to clonotypic tumor cells. Here we report in vitro studies that illustrate the effectiveness of this approach. LTC comprising the DNP hapten conjugated to ricin A toxin (DNP-RTA) were shown to specifically and effectively kill anti-DNP secreting murine hybridoma (U7.6) cells but not other hybridoma cells (1B12), a murine erythroleukemia cell line (Friend's Leukemia or) normal mouse spleen cells. In addition to direct toxicity, LTC treatment negatively affected the growth characteristics of the few surviving cells as reflected in decreased growth index and an increase in growth inhibition over 72 h post treatment. Interestingly, U7.6 cells that survived one or two LD 90 dose(s) of LTC showed no alteration in their dose response to a subsequent attack of LTC indicating that this treatment strategy may not induce drug resistance. These data suggest that LTC therapy may be a new and effective strategy for specific destruction of tumor cells such as myeloma plasma cells and could be extended to other tumors where clonotypic receptors can be identified.     

Antibody conjugate therapeutics: challenges and potential

Antibody conjugates are a diverse class of therapeutics consisting of a cytotoxic agent linked covalently to an antibody or antibody fragment directed toward a specific cell surface target expressed by tumor cells. The notion that antibodies directed toward targets on the surface of malignant cells could be used for drug delivery is not new. The history of antibody conjugates is marked by hurdles that have been identified and overcome. Early conjugates used mouse antibodies; cytotoxic agents that were immunogenic (proteins), too toxic, or not sufficiently potent; and linkers that were not sufficiently stable in circulation. Investigators have explored 4 main avenues using antibodies to target cytotoxic agents to malignant cells: antibody-protein toxin (or antibody fragment-protein toxin fusion) conjugates, antibody-chelated radionuclide conjugates, antibody-small-molecule drug conjugates, and antibody-enzyme conjugates administered along with small-molecule prodrugs that require metabolism by the conjugated enzyme to release the activated species. Only antibody-radionuclide conjugates and antibody-drug conjugates have reached the regulatory approval stage, and nearly 20 antibody conjugates are currently in clinical trials. The time may have come for this technology to become a major contributor to improving treatment for cancer patients.     

Identification of c-Yes expression in the nuclei of hepatocellular carcinoma cells: involvement in the early stages of hepatocarcinogenesis

It is thought that the subcellular distribution of Src-family tyrosine kinases, including c-Yes binding to the cellular membrane, is membranous and/or cytoplasmic. c-Yes protein tyrosine kinase is known to be related to malignant transformation. However, the expression patterns of c-Yes in hepatocellular carcinoma (HCC) remains unknown. In the present study, we report that c-Yes is expressed not only in the membrane and cytoplasm, but also in the nuclei of cancer cells in some human HCC tissues and in a human HCC cell line. We examined the expression and localization of c-Yes in human HCC cell lines (HLE, HLF, PLC/PRF/5 and Hep 3B) by Western blotting and immunohistochemical analyses; we also examined the expression of c-Yes by immunohistochemistry and Western blotting in the tissues of various liver diseases, including 39 samples from HCC patients. We used an antibody array to detect proteins that bind to nuclear c-Yes in PLC/PRF/5 cell line. c-Yes was found to be expressed in the membranes and cytoplasm of HLE, HLF and Hep 3B HCC cells; it was also detected in the nuclei in addition to the membranes and cytoplasm of PLC/PRF/5 HCC cells. HCC with nuclear c-Yes was detected in 5 of 39 cases (13.0%), and nuclear c-Yes expression was not detected in normal, chronic hepatitis or cirrhotic livers. All HCCs with nuclear c-Yes expression were well-differentiated, small tumors at the early stages. In the PLC/PRF/5 cell line, the nuclear localization of c-Yes with cyclin-dependent kinase 1 was confirmed by a protein antibody array. In conclusion, nuclear c-Yes expression was found in cancer cells at the early stages of hepatocarcinogenesis, suggesting that nucleus-located c-Yes may be a useful marker to detect early-stage HCC.     

Human melanoma cell lines showing striking inherent differences in sensitivity to immunotoxins containing holotoxins

The in vitro sensitivity of human melanoma cell lines to conjugates of whole abrin or ricin linked through disulfide bonds to the monoclonal antimelanoma antibody 9.2.27 was studied. After passage of the conjugates through a Sepharose 4B column to remove molecular species with exposed binding sites on their B-chains, toxicity of the conjugates to different melanoma cell lines and nonmelanoma tumor lines was assessed by measuring their ability to inhibit cellular protein synthesis. The abrin conjugate was far more toxic to the target cells than the corresponding ricin conjugate. The 8 melanoma cell lines studied differed widely in their sensitivities to the abrin conjugate. The differences were associated with concomitant large differences in the sensitivities of the cells to the native toxins, and the significance of the level of the antigen expression became apparent only when the sensitivities of the different cell lines were normalized with respect to their sensitivity to native abrin. The observed relationship could not be accounted for by unspecific binding via the B-chain binding site of the immunotoxin. The differential sensitivity of the melanoma cell lines to the immunotoxin seems to be related to inherent differences between the cells in their ability to internalize and to process immunotoxins and toxins. The findings may have considerable practical implications.     

In vitro toxicity of A-431 carcinoma cells with antibodies to epidermal growth factor receptor and epithelial glycoprotein-1 conjugated to radionuclides emitting low-energy electrons.

PURPOSE: The ability of antibodies (Abs) conjugated to radionuclides emitting low-energy electrons to specifically kill nonadherent lymphoma target cells in vitro was demonstrated previously. This study extends this work to adherent carcinoma cells. The fact that these cells are spread out on plastic can potentially make it more difficult to deliver radiation to the nucleus from decays in the cytoplasm or on the cell surface. EXPERIMENTAL DESIGN: The Abs tested were anti-epidermal growth factor receptor and anti-epithelial glycoprotein-1, conjugated to indium-111 or iodine-125, which emit low-energy Auger and conversion electrons. Conjugates of the beta-particle emitter, iodine-131, also were tested, for comparison. Abs were incubated with the cells for 2 days, and then the treated cells were assayed for colony-forming units. The radiation dose delivered to the nucleus was calculated from the cumulative decays per cell. RESULTS: With conjugates of (111)In, very potent killing was obtained with both of the Abs, with 100% kill (approximately 4-5 logs) even at subsaturating Ab concentrations. Lower levels of kill were obtained with (125)I or (131)I conjugates. Conjugates with (131)I, a beta-particle emitter, produced greater nonspecific toxicity. The greater potency of (111)In could be attributed to the higher specific activity that was obtained routinely with this radiolabel, up to 70 mCi/mg. Uptake of radioactivity peaked at approximately 200 cpm per cell. Dosimetry calculations, using subcellular S values, demonstrated that the toxicity observed was consistent with the amount of radiation delivered to the nucleus. CONCLUSIONS: These results are similar to previous results obtained with B lymphoma cells and indicate that this approach is applicable to a wide range of tumor types. Radionuclides emitting low-energy electrons are effective at killing target cells with relatively little nonspecific toxicity, if sufficient activity is delivered to the cell. Most Abs to high-density cell surface antigens would probably be effective.     

Specific thermal ablation of tumor cells using single‐walled carbon nanotubes targeted by covalently‐coupled monoclonal antibodies

CD22 is broadly expressed on human B cell lymphomas. Monoclonal anti‐CD22 antibodies alone, or coupled to toxins, have been used to selectively target these tumors both in SCID mice with xenografted human lymphoma cell lines and in patients with B cell lymphomas. Single‐walled carbon nanotubes (CNTs) attached to antibodies or peptides represent another approach to targeting cancer cells. CNTs convert absorbed near‐infrared (NIR) light to heat, which can thermally ablate cells that have bound the CNTs. We have previously demonstrated that monoclonal antibodies (MAbs) noncovalently coupled to CNTs can specifically target and kill cells in vitro. Here, we describe the preparation of conjugates in which the MAbs are covalently conjugated to the CNTs. The specificity of both the binding and NIR‐mediated killing of the tumor cells by the MAb‐CNTs is demonstrated by using CD22⁺CD25^? Daudi cells, CD22^?CD25⁺ phytohemagglutinin‐activated normal human peripheral blood mononuclear cells, and CNTs covalently modified with either anti‐CD22 or anti‐CD25. We further demonstrate that the stability and specificity of the MAb‐CNT conjugates are preserved following incubation in either sodium dodecyl sulfate or mouse serum, indicating that they should be stable for in vivo use. © 2009 UICC     

Immunotoxins: a clinical review of their use in the treatment of malignancies

Immunotoxins are a new class of antitumor agents consisting of tumor-selective ligands (generally monoclonal antibodies [MoAbs]) linked to highly toxic protein molecules that have been modified to remove their normal tissue-binding domains. These immuno-conjugates combine the potency of the parent toxin with the specificity of the attached ligand. Toxins used in the construction of immunotoxins belong to a group of peptides that catalytically inhibit the elongation step of protein synthesis, and include ricin, abrin, pokeweed antiviral protein, gelonin, Pseudomonas exotoxin A, diptheria toxin, and alpha-sarcin. To synthesize immunotoxins, the normal cell-binding function must be removed by chemical cleavage or modification, or in the case of toxins that have been cloned, genetic engineering used to delete amino acids critical to cell binding. Covalent linkage of toxin to ligand generally involves a disulfide or thioether bond, though recently, recombinant toxin molecules with ligands that are genetically engineered into the protein have been made. The most successful clinical application of immunotoxins has been in the depletion of T cells from allogeneic bone marrow grafts to prevent graft-versus-host disease (GVHD). Clinical trials have been conducted using immunotoxins for the systemic treatment of chronic lymphocytic leukemia (CLL), GVHD, and selected solid tumors. With the possible exception of GVHD, responses have been limited. Obstacles have included rapid systemic clearance, poor delivery to extravascular tumor deposits, and humoral immune responses to the immunotoxin. Research to overcome these problems is in progress and should lead to a better definition of the role of immunotoxins in the therapy of malignancies.     

Gadolinium in human glioblastoma cells for gadolinium neutron capture therapy

157Gd is a potential agent for neutron capture cancer therapy (GdNCT). We directly observed the microdistribution of Gd in cultured human glioblastoma cells exposed to Gd-diethylenetriaminepentaacetic acid (Gd-DTPA). We demonstrated, with three independent techniques, that Gd-DTPA penetrates the plasma membrane, and we observed no deleterious effect on cell survival. A systematic microchemical analysis revealed a higher Gd accumulation in cell nuclei compared with cytoplasm. This is significant for prospective GdNCT because the proximity of Gd to DNA increases the cell-killing potential of the short-range, high-energy electrons emitted during the neutron capture reaction. We also exposed Gd-containing cells to thermal neutrons and demonstrated the GdNC reaction effectiveness in inducing cell death. These results in vitro stimulated in vivo Gd-DTPA uptake studies, currently underway, in human glioblastoma patients.     

Preparation and biological characterization of conjugates consisting of ricin and a tumor-specific non-internalizing MAb

MOv18, a non-internalizing monoclonal antibody (MAb) with restricted tumor specificity, was conjugated to ricin toxin (RT). According to their ability to bind to galactose residues of Sepharose 6B, the immunoconjugates were fractionated into Bound and Unbound MOv18-RT. The two conjugates could be distinguished by SDS-PAGE, in vivo toxicity and agglutination capability. When the binding activity of both fractions was compared by solid-phase RIA to that of native MAb, it proved to be similar on the relevant target cells but significantly increased on the non relevant cells. On the latter, galactose totally cancelled the binding of the Unbound immunoconjugate, whereas it could only partially reverse that of the Bound MOv18-RT. By in vitro cytotoxic activity, either in the presence or absence of galactose, only a slight selectivity for relevant versus non-relevant target cells was observed for both conjugates. It seems that in the presence of a MAb which is incapable of internalization, the conjugate cytotoxicity could only be attributed to RT, with a loss of the MAb's specificity.     

Towards a ligand targeted enzyme prodrug therapy: single round panning of a beta-lactamase scaffold library on human cancer cells

A novel beta-lactamase scaffold library in which the target-binding moiety is built into the enzyme was generated using phage display technology. The binding element is composed of a fully randomized 8 amino acid loop inserted at position between Y34 and K37 on the outer surface of Enterobacter cloacae P99 cephalosporinase (beta-lactamase, E.C. 3.5.2.6) with all library members retaining catalytic activity. The frequency and diversity of amino acids distributions in peptide inserts from library clones were analyzed. The complexity of the randomized loop appears consistent with standards of other types of phage display library systems. The library was panned against SKBR3 human breast cancer cells in 1 round using rolling circle amplification of phage DNA to recover bound phage. Individual beta-lactamase clones, independent of phage, were rapidly assessed for their binding to SKBR3 cells using a simple high throughput screen based on cell-bound beta-lactamase activity. SKBR3 cell-binding beta-lactamase enzymes were also shown to bind specifically using an immunochemical method. Selected beta-lactamase clones were further studied for their protein expression, enzyme activity and binding to nontumor cell-lines. Overall, the approach outlined here offers the opportunity of rapidly selecting targeted beta-lactamase ligands that may have a potential for their use in enzyme prodrug therapy with cephalosporin-based prodrugs. It is expected that a similar approach will be useful in developing tumor-targeting molecules of several other enzyme candidates of cancer prodrug therapy.     

Cell killing by antibody-drug conjugates

Antibody-drug conjugates (ADCs) are designed to specifically bind to and kill cells expressing their target antigens. In addition to the obvious requirement of the presence of the target antigen on the cell surface, several other factors contribute to the sensitivity of target cells to the action of ADCs. These include (i) the rate of internalization of the ADC, (ii) its proteolytic degradation in late endosomes and lysosomes and the subsequent release of cytotoxic drug, and (iii) the intracellular concentration of the released drug. In addition to killing antigen-expressing cells, some ADCs were found to kill bystander cells irrespective of their antigen expression. This review summarizes the current knowledge of the mechanisms of killing of antigen-expressing and bystander cells by antibody-drug conjugates.     

Adenylate cyclase toxin from Bordetella pertussis enhances cisplatin-induced apoptosis to lung cancer cells in vitro

The present study examined the possibility to enhance lung cancer cell cytotoxicity and apoptosis of the anticancer drug cisplatin by exposure with adenylate cyclase (AC) toxin from Bordetella pertussis. A malignant mesothelioma cell line (P31) and a small-cell lung cancer cell line (U1690) were exposed to increasing concentrations of cisplatin and AC toxin, alone or in combination. Cytotoxicity was determined by a fluorescein-based assay and apoptosis by flow cytometry quantification of annexin V binding. Caspase-3, -8, and -9 activities were measured by enzyme activity assays. The cytotoxicity of AC toxin was time and dose dependent with an LD50 value at 72 h of 3 and 7 mg/L for P31 cells and U1690 cells, respectively. Cisplatin showed a similar time- and dose-dependent cytotoxicity, which was increased in the presence of a low toxic concentration (1 mg/L) of AC toxin. Furthermore, cisplatin caused a dose-dependent increase of annexin V binding cells of both cell lines after 24-h incubation, which was also enhanced in combination with AC toxin. AC toxin (1 mg/L) increased cisplatin-induced caspase-3, -8, and -9 activities in U1690 cells. Only minor increases of caspase-8 and -9 were noted for P31 cells. The present results, together with the knowledge that bacterial toxins decrease side effects of traditional cancer treatment, suggest a possibility to use them to enhance the therapeutic effect of cancer chemotherapy with reduced clinical adverse effects.     

Characterisation of the Binding Properties of Bacillus Thuringiensis 18 Toxin on Leukaemic Cells

Background

Various strains of Bacillus thuringiensis (Bt) have been found to produce parasporal proteins that are cytotoxic to human cancer cells. This study aims to establish the binding affinity of purified Bt 18 toxin for CEM-SS (T lymphoblastic leukaemia cell line), to determine if competition exists between the toxin and commercial anticancer drugs for the binding site on CEM-SS and to localise the binding site of the toxin on CEM-SS.

Methods

In homologous competitive binding study, the purified toxin was labelled with biotin and allowed to compete with unlabelled toxin for binding sites on CEM-SS and its dissociation constant (Kd) was determined. Comparisons were made with CCRF-SB, CCRF-HSB-2 and MCF-7. In heterologous competitive binding study, biotinylated toxin competition was determined with two other Bt toxins (crude Btj and crude Bt 22) and anticancer drugs (cisplatin, doxorubicin, etoposide, navelbine and methotrexate). To localise the binding site under the confocal microscope, the biotinylated toxin was tagged with FITC-conjugated streptavidin.

Results

Homologous competitive binding assays revealed decreasing binding affinity of Bt 18 toxin for CEM-SS, CCRF-SB, and CCRF-HSB-2 with Kd of 8.44 nM, 14.98 nM and 17.71 nM respectively. Kd for MCF-7 was not determined as the inhibitory concentration (IC₅₀) was not reached. Heterologous competitive study showed little competition (< 30%) between biotinylated Bt 18 toxin and all test compounds used. Confocal microscopy revealed binding of toxin at the periphery of the cell.

Conclusions

It was postulated that purified Bt 18 toxin binds on the cell surface of CEM-SS and the mechanism of cell death may differ from that of Btj toxin, Bt 22 toxin and all five anticancer drugs used in this study, since it did not significantly compete with these compounds for the same binding site.     

Kinetic analysis of 5-fluorouracil action against various cancer cells

Based on our recent kinetic analysis, which made it possible to distinguish between the cell-killing actions of cell cycle phase-specific and non-specific agents, we attempted to elucidate the actions of 5-fluorouracil (FUra) on three different cancer cell lines. By colony-forming assay, the concentrations of fluorouridine (FUrd), fluorodeoxyuridine (FdUrd) or FUra giving 90% cell kill (IC90) at various exposure times (texps) were obtained. With P388 cells, the curve of texps-IC90 for FUrd on a log-log scale was linear with a slope of -1, which is typical for cell cycle phase-nonspecific agents. In contrast, the curve for FdUrd showed a much steeper slope than -1, which is characteristic for cell cycle phase-specific agents. We found that the curve for FUra was exactly the same as that for FUrd, indicating that the mode of FUra action on P388 leukemia is analogous to that of FUrd. Similar results were observed with human colon and renal cancer cell lines, HT-29 and KU-2, although when the cells were exposed to relatively low concentrations of FUra for a long time, a cell cycle phase-specific action became evident.