In vitro cytotoxicity testing with fluorescence-based assays in cultured human lung and dermal cells

An in vitro study using human cultured cells was conducted to determine the reliability of fluorescence-based cell viability indicators with traditional in vitro cytotoxicity testing methods. Human lung epithelial carcinoma (A549) cells, and human embryonic skin (WS1) and lung (HFLI) fibroblasts were studied in culture to evaluate their potential to screen for cytotoxicity and to compare to previous protocols conducted in our laboratory. Confluent monolayers were incubated in the absence or presence of increasing concentrations of test chemicals for 24 h, and fluorescent-labeled probes were used to assess toxicity. Eight chemicals, including mercuric chloride, copper sulfate, sodium fluoride, thioridazine HCl, paraquat, amitriptyline-HCl, verapamil-HCl and chloroquine sulfate, were tested with each cell line using calcein-AM and Sytox. The data suggest that fluorescent probes are sensitive indicators of cytotoxicity and contribute to understanding the mechanisms for each chemical. In combination with previously published reports, the similarity of results among cell lines may be explained by the origin of the cell lines rather than by the diversity of the methods and indicators employed.     

Acute cytotoxicity testing with cultured human lung and dermal cells

Summary An extensive in vitro study with cultured cells was conducted to test the basal cytotoxicity theory. This theory suggests that most chemical injury, at least in vitro, is a manifestation of one or more insults to the basic cellular structures and functions common to mammalian cells. This accounts for the similarity of results in multilaboratory studies. Human fetal lung fibroblasts (HFL1), and human skin fibroblasts (WS1, Detroit551) were studied in culture to evaluate their potential to screen for cytotoxicity. Confluent monolayers were incubated in the absence or presence of increasing concentrations of test chemicals for 24 h, and the MTT assay was used to assess toxicity. Inhibitory concentrations were extrapolated from concentration-effect curves after linear regression analysis. Twenty-nine chemicals were tested with each cell line and the cytotoxicity data compared to rodent and human lethal concentrations. The data suggest that the experimental IC₅₀ values are as accurate predictors of human toxicity as equivalent toxic blood concentrations derived from rodent LD₅₀s. In addition, lung and skin fibroblasts revealed no significant differences among the three cell lines. The results support the conclusion that finite cell lines of human origin have the potential for screening chemicals for human toxicity. In combination with previously published reports, the data suggest that a basal cytotoxic phenomenon may explain the similarity of results among different human cell lines.     

Exposure to titanium dioxide and other metallic oxide nanoparticles induces cytotoxicity on human neural cells and fibroblasts

The use of titanium dioxide (TiO₂) in various industrial applications (eg, production of paper, plastics, cosmetics, and paints) has been expanding thereby increasing the occupational and other environmental exposure of these nanoparticles to humans and other species. However, the health effects of exposure to TiO₂ nanoparticles have not been systematically assessed even though recent studies suggest that such exposure induces inflammatory responses in lung tissue and cells. Because the effects of such nanoparticles on human neural cells are unknown, we have determined the putative cytotoxic effects of these nanoparticles on human astrocytes-like astrocytoma U87 cells and compared their effects on normal human fibroblasts. We found that TiO₂ micro- and nanoparticles induced cell death on both human cell types in a concentration-related manner. We further noted that zinc oxide (ZnO) nanoparticles were the most effective, TiO₂ nanoparticles the second most effective, and magnesium oxide (MgO) nanoparticles the least effective in inducing cell death in U87 cells. The cell death mechanisms underlying the effects of TiO₂ micro- and nanoparticles on U87 cells include apoptosis, necrosis, and possibly apoptosis-like and necrosis-like cell death types. Thus, our findings may have toxicological and other pathophysiological implications on exposure of humans and other mammalian species to metallic oxide nanoparticles.     

A new culture medium for human skin epithelial cells

Summary A new culture medium, NCTC 168, has been designed for human skin epithelial cells. This medium formulation was developed, by combining and testing at various concentrations, components of media NCTC 135 and 163, since a 1∶1 mixture of these two media with 10% horse serum supplement was found to promote epithelial cell outgrowth from human skin explants. The buffer system in NCTC 168 maintains the pH of the medium between 7.0 and 7.2. In contrast to other media tested, NCTC 168 with 10% horse serum is capable of initiating and sustaining larger epithelial cell outgrowths. Explants in serum-supplemented NCTC 168 in the absence of feeder cells reproducibly yield confluent epithelial cell sheets apparently free of fibroblasts after only 19 to 28 days as compared with 5 weeks or longer for the other media tested. NCTC 168 also supports passage of human epithelial cells to the sixth subculture generation without feeder cells. Electron microscopy has shown the presence of desmosomes and tonofilaments in the passaged cells indicating the epithelial nature of the cells. The addition of epithelial growth factor, hydrocortisone and insulin at 5 ng per ml, 4 μg per ml and 5 μg per ml, respectively did not appreciably enhance the growth of the epithelial cells.     

A method for the isolation and serial propagation of keratinocytes,endothelial cells,and fibroblasts from a single punch biopsy of human skin

Summary When multiple types of cells from normal and diseased human skin are required, techniques to isolate cells from small skin biopsies would facilitate experimental studies. The purpose of this investigation was to develop a method for the isolation and propagation of three major cell types (keratinocytes, microvascular endothelial cells, and fibroblasts) from a 4-mm punch biopsy of human skin. To isolate and propagate keratinocytes from a punch biopsy, the epidermis was separated from the dermis by treatment with dispase. Keratinocytes were dissociated from the epidermis by trypsin and plated on a collagen-coated tissue culture petri dish. A combination of two commercial media (Serum-Free Medium and Medium 154) provided optimal growth conditions. To isolate and propagate microvascular endothelial cells from the dermis, cells were released following dispase incubation and plated on a gelatin-coated tissue culture dish. Supplementation of a standard growth medium with a medium conditioned by mouse 3T3 cells was required for the establishment and growth of these cells. Epithelioid endothelial cells were separated from spindle-shaped endothelial cells and from dendritic cells by selective attachment toUlex europeus agglutinin I-coated paramagnetic beads. To establish fibroblasts, dermal explants depleted of keratinocytes and endothelial cells were attached to plastic by centrifugation, and fibroblasts were obtained by explant culture and grown in Dulbecco’s modified Eagle’s medium (DMEM) containing fetal bovine serum (FBS). Using these isolation methods and growth conditions, two confluent T-75 flasks of keratinocytes, one confluent T-25 flask of purified endothelial cells, and one confluent T-25 flask of fibroblasts could be routinely obtained from a 4-mm punch biopsy of human skin. This method should prove useful in studies of human skin where three cell types must be grown in sufficient quantities for molecular and biochemical analysis.     

Effect of cellular glutathione depletion on cadmium-induced cytotoxicity in human lung carcinoma cells

The effect of glutathione depletion on cellular toxicity of cadmium was investigated in a subpopulation (T27) of human lung carcinoma A549 cells with coordinately high glutathione levels and Cd⁺⁺-resistance. Cellular glutathione levels were depleted by exposing the cells to diethyl maleate or buthionine sulfoximine. Depletion was dose-dependent. Exposure of the cells to 0.5 mM diethyl maleate for 4 hours or to 10 mM buthionine sulfoximine for 8 hours eliminated the threshold for Cd⁺⁺ cytotoxic effect and deccreased the LD_50S. Cells that were pretreated with 0.5 mM diethyl maleate or 10 mM buthionine sulfoximine and then exposed to these same concentrations of diethyl maleate or buthionine sulfoximine during the subsequent assay for colony forming efficiency produced no colonies, reflecting an enhanced sensitivity to these agents at low cell density. Diethyl maleate was found to be more cytotoxic than buthionine sulfoximine. Synergistic cytotoxic effects were observed in the response of diethyl maleate pretreated cells exposed to Cd⁺⁺. Thus the results demostrated that depletion of most cellular glutathione in A549-T27 cells prior to Cd⁺⁺ exposure sensitizes them to the agent's cytotoxic effects. Glutathione thus may be involved in modulating the early cellular Cd⁺⁺ cytotoxic response. Comparison of reduced glutathione levels and of Cd⁺⁺ cytotoxic responses in buthionine sulfoximine-treated A549-T27 cells with those levels in other, untreated normal and tumor-derived cells suggests that the higher level of glutathione in A549-T27 is not the sole determinant of its higher level of Cd⁺⁺ resistance.Abbreviations BSO DL-buthionine-(R,S)-sulfoximine - DEM diethyl maleate - DMSO dimethyl sulfoxide - GSH reduced glutathione - MT metallothionein     

In vitro deregulation of markers characteristic of human prostate epithelial cells

We have screened primary cultures of human prostate for the expression of markers reported to be characteristic of specific cell lineages in vivo, in order to ascertain whether human prostate cells in vitro maintain and reflect their in vivo differentiated phenotypes and to evaluate the homogeneity of the populations of cells that can be derived from this tissue. Using single and dual stain immunofluorescent microscopy to analyse very early organoid and subsequently derived monolayer stage cultures, we have observed that expression of markers characteristic of human prostate epithelial cells in vivo is deregulated within 48h, indicating that dissociation of human prostate tissue and cultivation of prostate epithelial cells in culture can result in promiscuous expression of cell type specific markers of prostate epithelial cells. These observations have important implications for studies of cell lineage and differentiation of prostate cells in vitro.     

Comparative study of extracellular amino acids in culture of human liver and fibroblastic cells

Summary Amino acid concentrations are studied in the extracellular media of ten series of human fibroblast and liver cell monolayer cultures. These two cell types consume and produce ostensively the same amino acids. Among the nonessential amino acids, the most significant variations involve serine and aspartate which are decreased; α-alanine, glutamate, ornithine and proline are, on the contrary, increased. Among the essential amino acids, leucine, isoleucine and glutamine are preferentially decreased. The variations of some amino acids are correlated with the cell density. The interrelations which may exist between the variations of these different amino acids are discussed. Furthermore, the glycolytic activity of the cells studied is very high: 85% of glucose consumed is found in the form of lactate. Unité d’Hépatologie Infantile I.N.S.E.R.M. (U56) Laboratoire Central de Biochimie This work was supported by grant: INSERM CRL 73-5-057-04 and 74-5-075-04.     

Fatty acids and the selective alteration of in vitro proliferation in human fibroblast and guinea-pig smooth-muscle cells

Summary Human-foreskin fibroblast (HF) and guinea-pig aorta smooth-muscle (SM) cultures were treated with several saturated and unsaturated fatty acids. Relative plating efficiencies were used to determine the proliferative response to each treatment. At low concentrations (16 to 18 μm), proliferation in HF cultures was inhibited by 8,11,14-eicosatrienoic acid (20:3), and stimulated by both 5,8,11,14-eicosatetraenoic acid (20:4) and 9-octadecenoic acid (18:1). At these levels, proliferation in SM cultures was unchanged by 20:3, inhibited by 20:4, and enhanced by 18:1. At higher concentrations (80 to 90 μm), HF cultures were inhibited by all three unsaturated fatty acids. At these same concentrations, proliferation in SM cultures was inhibited by 20:3 and 20:4, whereas 18:1 continued to stimulate proliferation. Thus proliferative response was a specific effect of the fatty acid used, its concentration, and the cell line involved. Further treatment of SM cultures by tetradecanoic acid (14:0), hexadecanoic acid (16:0), and octadecanoic acid (18:0) showed that their relative abilities to inhibit cell proliferation increased with increasing chain length. Concentrations required for the effective inhibition of proliferation in SM cultures by 14:0, 16:0 and 18:0 were 220 μm, 95μm and 18μm, respectively. The fatty acids used in these studies are all endogenous components of sera used as growth supplements in in vitro systems. Their roles as prostaglandin and hydroperoxy fatty-acid precursors (20:3 and 20:4), inhibitors of prostaglandin biosynthesis (18:1), or as calcium ionophores (14:0, 16:0, and 18:0) may allow them to function as endogenous controls of cell proliferation. This work was supported in part by National Heart and Lung Institute Grant HL-11897.     

Effects of feeder layers made of human,mouse, hamster,and rat cells on the cloning efficiency of transformed human cells

Summary The effect of feeder layers on cloning efficiency of transformed human cells was investigated. Embryonic human skin or lung fibroblasts; adult human skin fibroblasts; early passage cells from embryos of mouse, rat, and hamster; established mouse cell lines; 3T3 and 10T1/2 were used as feeder layers after they were lethally exposed to Co-60 gamma-rays at 3,000 rad. As test cells to study the effect of feeder layers on cloning efficiency, WI-38 CT-1 cells transformed in vitro by Co-60 gamma-rays and HGC cells cultured from a human gastric cancer were used. The effect of feeder layers on the cloning efficiency of the test cells was dependent on cell density of feeder layer cells, sources of the feeder layer cells, and kinds of test cells. An optimal density of feeder cels produced cloning efficiencies 3 to 15 times higher than in cultures without a feeder layer. Generally, high density of cells in feeder layers decreased the cloning efficiency of the test cells, presumably owing to contact inhibition of growth and depletion of essential nutrients by the feeder layer cells. Regarding the effect of the feeder layers made of human fibroblasts, there were no significant differences in population doubling levels; tissue origins of fibroblasts; or fibroblasts derived from normal individuals, patients with cancer, or with a genetically high familial incidence of cancer, hereditary adenomatosis of the colon and rectum. This study was supported by a Grant-in-Aid for Cancer Research from the Ministry of Education, Science and Culture, Japan.     

Anin vitro study on the cytotoxicity of chlorhexidine digluconate to human gingival cells

Chlorhexidine digluconate is the active ingredient in mouthrinses used to prevent dental plaque and gingivitis. Thein vitro cytotoxicity of chlorhexidine was evaluated with the Smulow-Glickman (S-G) gingival epithelial cell line. The potency of chlorhexidine was dependent on the length of exposure and composition of the exposure medium. The midpoint cytotoxicity values for 1-, 24-, and 72-h exposures were 0.106, 0.011, and 0.0045 mmol/L, respectively. S-G cells exposed for 2 h to chlorhexidine and then maintained for 48 h in chlorhexidine-free medium were unable to recover from the initial insult. The adverse effects of chlorhexidine on the plasma membrane were suggested by the leakage of lactic acid dehydrogenase from chlorhexidine-treated S-G cells and by the increased permeability of chlorhexidine-treated liposomes to Ca²⁺. The toxicity of a 24-h exposure to chlorhexidine to the S-G cells was progressively lessened as the content of fetal bovine serum (FBS) in the exposure medium was increased from 2% to 8%. The potency of a 1-h exposure to chlorhexidine was reduced in medium amended with albumin, lecithin, and heat-killedEscherichia coli. These reductions in toxicity were presumably due to the binding of the cat onic chlorhexidine to the negatively charged chemical moieties of the components of FBS and of albumin and lecithin and of sites on the surfaces of bacteria. Combinations of chlorhexidine and carbamide peroxide were additive in their cytotoxicities.Abbreviations ANOVA analysis of variance - [Ca²⁺]i calcium concentration in internal medium of liposomes - DMEM Dulbecco's modified Eagle medium - EDTA ethylenediamine tetraacetic acid - FBS fetal bovine serum - Hepes N-(2-hydroxyethyl)piperazine-N'-2-ethanesulfonic acid - HGF-1 human gingival fibroblast cell line - HSD honestly significant differences - KB cell line derived from a human epidermoid carcinoma in the mouth - LDH lactic acid dehydrogenase - NADH nicotinamide adenine dinucleotide, reduced form - NR neutral red - NR₅₀ concentration inhibiting neutral red uptake by 50% - PBS phosphate-buffered saline - SEM standard error of the mean - S-G Smulow-Glickman human gingival epithelial cell line     

Pleiotropic effects of corticotropin releasing hormone on normal human skin keratinocytes

The hypothalamic-pituitary-adrenal (HPA) axis is the major stress response system. Several components of the HPA axis, such as corticotropin-releasing hormone (CRH) and POMC peptides and their receptors are also present in the skin. In earlier studies, we showed that CRH inhibits cellular proliferation of immortalized human keratinocytes. We now examine further the functional activity of the HPA axis in the skin, by characterizing the actions of CRH on normal foreskin keratinocytes. The CRH receptor was detected as CRH-R1 antigen at 47 kDa in the cultured keratinocytes by Western blotting, and immunohistochemistry demonstrated its presence in the epidermal and follicular keratinocytes. CRH is also biologically active in cultured keratinocytes, where it inhibits proliferation and enhances the interferon-gamma-stimulated expression of the hCAM and ICAM-1 adhesion molecules and of the HLA-DR antigen. These effects were concentration-dependent, with maximal activity at CRH 10(-7) M. Thus, in the keratinocyte, the most important cellular component of the epidermis, CRH appears to induce a shift in energy metabolism away from proliferation activity, and toward the enhancement of immunoactivity. Therefore, similar to its central actions, cutaneous CRH may also he involved in the stress response, but at a highly localized level.     

Selective cytotoxicity of marine algae extracts to several human leukemic cell lines

Extracts from 8 species of marine algae which showed selective cytotoxicity in our previous screening program, were further examined for cytotoxic spectra to five human leukemic cell lines. The extract from a red alga, Amphiroa zonata exhibited strong cytotoxicity to all human leukemic cell lines tested and murine leukemic cells L1210 at the final concentrations from 15 to 375 μg ml−¹. Then the cytotoxicity was not found in normal human fibroblast HDF and murine normal cells NIH-3T3. The active extract fraction from this alga was soluble in higher polar organic solvents and water and heat-stable. The extract from a brown alga Dilophus okamurae with weak selective cytotoxic activity to L1210 cells exhibited not only strong cytotoxicity to L1210, but also to human leukemic cells, HL60 and MOLT-4 at 50 μg ml-¹. While, the extract from a green alga, Cladophoropsis vaucheriaeformis with most selective cytotoxic activity, did not show cytotoxicity to any human leukemic cell lines tested at 50 μg ml-¹. However, this extract showed strong cytotoxicity to two human leukemic cell lines and NIH-3T3 at 100 μg ml−¹. Thus, it was considered that a red alga, Amphiroa zonata might be suitable natural source for development of anti-cancer agents without side-effect. This revised version was published online in June 2006 with corrections to the Cover Date.     

Effects of doxorubicin on human dental pulp cells in vitro

There is substantial information concerning the effects of continuous exposure to supratherapeutic or therapeutic concentrations of doxorubicin on human molar pulpal cells; the effects of continuous exposure to subtherapeutic concentrations of this agent are undetermined. To this end, we studied the proliferation of human fibroblasts and pulpal cells and their pattern of mineralized nodule deposition in vitro. Cell proliferation was assessed at 1, 3, 5, and 7 days from populations with either no exposure (control) or exposure to 10⁻⁶–10⁻⁹ mol/L doxorubicin. Mineralized nodule deposition and calcium-45 incorporation were assessed at 7 and 21 days of culture. Data were compared by factorial ANOVA and a post-hoc Tukey test. 10⁻⁶ and 10⁻⁷ mol/L doxorubicin significantly reduced the total number of viable pulpal cells in cultures from days 1 to 3 (p < 0.05); doxorubicin 10⁻⁶–10⁻⁹ mol/L significantly inhibited cell proliferation (p < 0.05) and DNA synthesis 5 days after plating (p < 0.001). After 21 days, doxorubicin 10⁻⁶–10⁻⁸ mol/L significantly decreased calcium-45 incorporation into pulpal cultures (p < 0.001); all dilutions significantly reduced the number of mineralized nodules within the 21-day pulpal cultures (p < 0.05). In addition, all dilutions of doxorubicin significantly inhibited fibroblast cell proliferation and incorporation of [³H]thymidine. In contrast, the fibroblast cultures did not produce mineralized nodules, suggesting that the mineralized nodules within the pulpal cell cultures did not result from dystrophic calcification. Thus, exposure to subtheraputic doxorubicin concentrations has potential adverse effects on mineralized tissue formation within the pulp, which could affect the rates of reparative dentin deposition within the tooth pulps of patients receiving this chemotherapeutic agent.     

In vitro cytotoxicity of heavy metals,acrylamide, and organotin salts to neural cells and fibroblasts

The cytotoxicity of neurotoxic agents was determined for a series of brain-derived cell types and compared with their toxic effects on BALB/c 3T3 fibroblasts, using the neutral red assay. Ranking of toxicants according to their potencies was the same for all cells tested and was in the order of methylmercury>cadmium> mercury>zinc>acrylamide. For a series of di- and triorganotins the ranking order was dibutyl>diphenyl>dibenzyl>diproyl> diethyl>dimethyltin and triphenyl>tribenzyl>trimethyltin, respectively. The test was sensitive enough to detect structure activity relationships between the degree of toxicity and the hydrophobic characteristics of the agents tested.Abbreviations Be₂Sn dibenzyltin dichloride - Be₃Sn triphenyltin dichloride - Bu₂Sn dibutyltin dichloride - CdCl₂ cadmium chloride - CH₃HgCl methylmercuric chloride - DMEM Dulbecco's Modified Eagle's Medium - Et₂Sn diethyltin dichloride - FBS fetal bovine serum - HgCl₂ mercuric chloride - Me₂Sn dimethyltin dichloride - Me₃Sn trimethyltin hydroxide - Phe₂Sn diphenyltin dichloride - Phe₃Sn triphenyltin hydroxide - Pr₂Sn dipropyltin dichloride - ZnCl₂ zinc chloride     

Growth of normal human mammary cells in culture

Summary Reduction mammoplasty tissue was used to obtain short-term cultures of human epithelial cell populations. Digestion of tissue with collagenase and hyaluronidase resulted in cell clusters (organoids) resembling ductal and alveolar structures; these could be separated by filtration from the stromal components. Epithelial outgrowth from these organoids was greatly enhanced by the addition of conditioned medium from other human epithelial and myoepithelial cell lines. Additionally, the mammary epithelial growth was stimulated by insulin, hydrocortisone, epidermal growth factor, and steroid hormones. With this enriched nutritional environment, active cell division could be maintained for 1 to 3 months and cells could be serially subcultured 1 to 4 times. This research was supported by Grant PDT-72 from the American Cancer Society and Grant CP-70510 from the National Institutes of Health.     

Preservation of microplate-attached human hepatoma cells and their use in cytotoxicity tests

We investigated the feasibility of hypothermic- orcryogenically-preserved human hepatoma Hep G2 cell preculturedin 96-well plates in cytotoxicity testings. First, we observedthat microplates precoated with both collagen (CN) and pronectin (PN) showed significantly improved living cell adhesion (71.0 ± 5.5%) after 48 hr of cryopreservation with 10%-DMSO containing culture medium, whereas non-coated surfaces gave very low living cell adhesion (33.5 ± 2.1%). Hypothermic preservation was most suitable for short-term storage, and cryogenic preservation at –20 °C allowed cells to be used within a week of the storage period. Only cryopreservation in a deep freezer (–85 °C) gave satisfactory results in much longer period of storage. Second, we evaluated the cytotoxicity of ten chemicals during 48 hr of exposure using hypothermically – (4 °C for 2 days) or cryogenically – (–85 °C for 7 days) preserved cells cultured inCN/PN-precoated microplates in comparison with results fromfreshly inoculated cells. Although almost the same LD₅₀values were obtained, LD₁₀ values of relatively hydrophilic chemicals obtained with cryopreserved cell were significantly lowered. These results shown that CN/PN-precoating is effective in keeping cells attached even in recultivation of preserved cells and that the toxicities of relatively hydrophilic chemicals tend to be overestimated when we use preserved cells in that manner.     

Apoptotic primary normal human dermal fibroblasts for in vitro models of fibrosis

Recent studies show that apoptosis affects surrounding tissue, playing a role in diseases such as fibrosis, a significant global disease burden. Elucidating the mechanisms by which the different apoptotic cells present during fibrotic wound healing affect their environment would enable development of new therapies. We describe here a simple, rapid, and cost-effective method for inducing apoptosis of primary normal human dermal fibroblasts without affecting the overall cell viability of the population. Such population could be used for in vitro models of fibrotic wound healing in co-culture with other cells involved in this process to study events such as apoptosis-induced proliferation.     

Relative resistance to methothexate by proliferating normal rabbit epidermal cells in vitro

Summary Primary cell cultures of normal rabbit epidermal cells (keratinocytes) were established without the use of enzymatic techniques. Six experiments were carried out on cells from six different rabbits. When these cells were exposed to methotrexate (MTX) for 24 h at 1 μg/ml, proliferation, as measured by cells entering mitosis, was significantly inhibited (P<0.05) in only one experiment. When the dose of MTX was elevated to 100 μg/ml, only two experiments showed significant inhibition of mitosis. This minimal inhibition of mitosis by MTX was contrasted by the dramatic inhibitory effect of this antimetabolite on DNA synthesis. At 1 μg/ml MTX for 24 h, DNA synthesis, as measured by [³H]deoxyuridine uptake, was inhibited >95%. We can conclude that under certain conditions, the rabbit keratinocyte may represent a normal cell type that is inherently resistant to the antiproliferative effects of methotrexate. The research was supported by National Cancer Institute Grant CA 11536.