Protection by glucose and derivatives against the lethal toxicity of mitomycin C in bacteria

Agents capable of preventing the toxicity of mitomycin C (MMC) were investigated by a cytotoxicity assay utilizing the E. coli strain WP2 uvrA, a strain sensitive to the bactericidal action of MMC. Of various compounds, mixtures, and rat tissue extracts assayed, the solution of liver extracts and yeast extracts and DULBECCO's modified EAGLE's medium (DMEM) exhibited potent activity in protecting the cells against the MMC toxicity. A further analysis of the individual components of DMEM revealed that glucose is the active principle responsible for the protection seen with DMEM. A similar protection has been observed with the use of mannose, mannitol, 2-deoxyglucose, D-glucuronic acid, glucosamine, and N-acetylglucosamine of 16 sugar derivatives tested. The protection by glucose was specific to treatment of cells with MMC but not with UV-irradiation, cis-diamminedichloroplatinum (II), 4-nitroquinoline 1-oxide, or furylfuramide. Unlike the bacterial cells, there was no protective response in the mammalian cells in culture and in mice, given a lethal dose of MMC, concurrently with glucose or each derivative. The possible mechanisms involved in this prevention of MMC toxicity by glucose are discussed.

Some observations on the localization of mitomycin C-induced aberrations in human lymphocytes

Actively cycling human lymphocytes were treated with mitomycin C for 1 h (1.4 micrograms/ml) and then grown in medium containing 10 micrograms/ml bromodeoxyuridine. Serial 5-h colcemid accumulation samples were taken up to 35 h and the air-dried methaphase spreads stained for replication banding. A complete cell-cycle subphasing analysis was made, and classified cells scored for all categories of chromatid-type aberrations and their location. In spite of the high dose which produced massive delay and cycle perturbation, there was no evidence for selective lethality of early-S cells, in fact such cells were in excess. Extreme localization of aberrations to late-replicating (mostly centromeric) regions was found at all subphases and in pre-S cells. This rules out 'localization by default' as an explanation for the observed preferential occurrence of 'break points' in these regions. The frequency of incomplete intrachanges, low in late S, rises dramatically in early S to become maximal in pre-S cells.

Spontaneous and mutagen-induced sister chromatid exchange in motor neurone disease

Spontaneous and mutagen-induced sister-chromatid exchange frequencies have been studied in the peripheral blood lymphocytes of 6 patients with motor neurone disease. Their values were compared with those obtained in age- and sex-matched healthy controls. No significant differences were observed between the 2 groups. These results do not support the hypothesis of a defect in the repair of DNA damage as the primary abnormality in the development of the disease.

Effects of smoking on spontaneous and induced sister chromatid exchanges in lymphocytes

Spontaneous and mitomycin C-induced sister chromatid exchanges (SCEs) in lymphocytes were analyzed in 24 non-smokers and 24 sex- and age-matched smokers. Mean spontaneous SCE frequency for non-smokers was 9.8 SCEs/cell, and that for smokers was 11.5 SCEs/cell. The difference was statistically significant (P less than 0.001 by t-test). These results suggest that spontaneous SCE frequency in lymphocytes is useful for evaluation of biological effects of environmental mutagens. However, we could not find any effects of smoking on the sensitivities of lymphocytes to mitomycin C in vitro. The effects of mutagens on humans may be independent of one another.

Effect of the homokaryotic or heterokaryotic state of the uvs-2 allele in Neurospora crassa on mitomycin C-induced killing and ad-3 mutation

Mitomycin C (MC) was tested for its killing and mutagenic activities in the ad-3 forward-mutation test in Neurospora crassa. The test was conducted in 4 dikaryons of N. crassa in order to determine the effect of the uvs-2 allele, which causes a defect in nucleotide excision repair, on MC-induced killing and ad-3 mutation. These dikaryons were homokaryotic for uvs-2+ (H-12), homokaryotic for uvs-2 (H-59), and heterokaryotic for uvs-2/uvs-2+ (H-70 and H-71). MC induced killing and ad-3 mutation in H-12, but the presence of uvs-2 in the homokaryotic state (H-59) resulted in a great increase in the killing and mutagenic activities of MC. This increased sensitivity to MC-induced killing and mutation conferred by uvs-2 in the homokaryotic state (H-59 vs. H-12) is a different effect than that noted by others for a defect in nucleotide excision-repair in Escherichia coli and Salmonella typhimurium or in human cells. The dikaryons heterokaryotic for uvs-2/uvs-2+ had the same sensitivity to MC as H-12, indicating that for MC-induced killing and ad-3 mutation uvs-2 is recessive to uvs-2+.

The size of micronuclei in human lymphocytes varies according to inducing agent used

Micronuclei were induced in vitro in human lymphocytes by mitomycin C, X-rays, vincristine, and colcemid and analyzed in cells with preserved cytoplasm. The micronucleus/cell nucleus ratio was measured. It was found that micronuclei induced by mitomycin C and X-rays were significantly smaller than those formed by vincristine and colcemid. Thus, in spite of the wide size span of human chromosomes, it could be shown that it is possible to differentiate between micronuclei formed by spindle-damaging agents (vincristine and colcemid) and those induced by agents directly damaging the chromosomes (mitomycin C and X-rays). Mitomycin C-induced micronuclei were smaller than those induced by X-rays, probably because the former agent preferentially produces chromatid fragments and the latter chromosome fragments.

A loss of uvrA function decreases the induction of the SOS functions recA and umuC by mitomycin C in Escherichia coli

We have studied the levels of recA and umuC protein synthesis in Escherichia coli as a probe for regulatory and mechanistic events involved in mitomycin C mutagenesis. Both RecA and UmuC protein induction were greatly stimulated by mitomycin C in the wild-type strain, reached a peak at about 60 min for the recA gene, and at 90 min for the umuC gene, respectively, and maintained a plateau. The induction was blocked by recA and lexA(Ind-) mutations that conferred no mutagenesis on the cell. Mutation affecting uvrA protein markedly decreased induction of the recA gene as well as the umuC gene by mitomycin C. The results established that UvrA protein is involved in the induction of recA and umuC, and account, at least in part, for the mitomycin C nonmutability of uvrA mutants.

Ultraviolet-light exposure induces a heritable sensitivity to the induction of SCE by mitomycin-C

The dose-response relationship for mitomycin-C (MMC)-induced sister-chromatid exchange (SCE) has been determined in the progeny of Chinese hamster lung fibroblasts (V79) exposed to 5.0 J/m2 ultraviolet light-C (UVC, 254 nm) and in the progeny of non-UVC-irradiated controls. Progeny of UVC-irradiated cultures exhibited sensitivity to MMC-induced SCE at doses of MMC that were not detectably lethal. This sensitivity was manifest as an increase in SCE per cell in a large proportion of the cells derived from UVC-exposed cultures and thus appears not to result from the expression of a rare event such as mutation.

Effect of 5-azacytidine (5-azaC) on the induction of chromatid aberrations (CA) and sister-chromatid exchanges (SCE)

Experiments were carried out using human lymphocytes from a male donor in order to test the action of 5-azaC treatment on the induction of SCE and chromatid aberrations. The 5-azaC was found to increase the frequency of both baseline and MMC-induced SCEs. Using the same 5-azaC treatment conditions it was found that the frequency of X-ray-induced CA did not increase.

Toxic effects of chemicals on mouse post-blastocyst development--a trial to establish a testing system for embryotoxicity

We established a new testing system to assess embryotoxic chemicals using cultured preimplantation mouse embryos. Randomly bred Crj:CD-1 (ICR) mice were used. Late blastocysts collected on Day 4 of pregnancy (plug day:Day 1) were exposed for 24 hrs. to various chemicals at varying concentrations. Morphological embryonic development, structural chromosome aberrations and sister chromatid exchanges (SCEs) were examined as the parameters of embryotoxicity. Treated embryos were then transferred to pseudopregnant foster mice for the evaluation of in vivo development. Of the morphological developmental endpoints in vitro, the formation of two-layer ICMs (inner cell masses) was the most sensitive to the toxicity of chemicals, followed by ICMs and trophoblast outgrowth. Threshold limit values were confirmed regarding morphological growth. However, exposure of embryos to lower concentrations than threshold enhanced the frequency of chromosome aberrations and SCEs. Transfer of embryos treated with the lower concentration of 4-nitroquinoline 1-oxide induced not only fetal wastage but also developmental retardation. 4-nitroquinoline 1-oxide was the most embryotoxic in mouse blastocysts, followed in order by mitomycin C, bleomycin, methylmercuric chloride, diethylstilbestrol, mercuric chloride and ochratoxin-A. This new testing system appears to be a sensitive method of measuring the direct effects of toxic chemicals which yields results in a short term.

Enhancement of mitomycin C cytotoxicity to hypoxic tumor cells by dicoumarol in vivo and in vitro

Previous work by our laboratories demonstrated that dicoumarol can increase the enzymatic activation of mitomycin C (MC) to alkylating species by tumor cell sonicates under hypoxic conditions. To determine whether this increased generation of reactive metabolites would result in increased cytotoxicity, we examined the effect of this combination on the viability of EMT6 cells treated in vitro under hypoxic and oxygenated conditions. Dicoumarol increased the cytotoxicity of MC to these neoplastic cells under hypoxic conditions and decreased the toxicity of the antibiotic to aerobic cells. These findings suggested that dicoumarol might enhance the toxicity of MC to the hypoxic cells of solid tumors, without increasing the toxic side effects of the antibiotic to the host. Treatment of EMT6 tumor-bearing animals with both dicoumarol and MC significantly decreased the survival of the radioresistant hypoxic tumor cells from that obtained with MC alone. In contrast, the leukopenia produced by the antibiotic was not exacerbated by the addition of dicoumarol. These results suggest that a treatment regimen combining dicoumarol and MC might be a useful adjunct to radiation therapy for the eradication of the radioresistant hypoxic cells in solid tumors.

Chromosome fragments and other abnormalities induced by mitomycin C in the neuroblast of Chortophaga viridifasciata

Mitomycin C (MMC) induces acentric chromosome fragments in the neuroblast (Nb) of the grasshopper embryo (Chortophaga viridifasciata) after acute and chronic exposure to concentrations ranging from 10(-8) to 10(-4) M, the dose response being essentially linear up to 10(-5) M. Because Colcemid is not used in the Nb assay, it was possible to detect two additional effects of MMC: (1) Prolonged retardation of many cells occurs when they reach very late prophase; the chromosomes continue condensing and lose their orderly prophase orientation, and the nuclear envelope becomes increasingly fragile. Such cells, which were observed after both acute and chronic exposure, give the false impression of being c-metaphases when they are fixed and squashed. The frequency of retarded very late prophases and the duration of retardation are related to MMC concentration and time of exposure. A rationale is presented supporting the idea that the events associated with retarded very late prophase result from MMC effects on the nuclear envelope. (2) MMC significantly increases the frequency of Nb's with attenuated centromeres at the beginning of early anaphase, an effect that appears to be caused by a delay in the repulsion of sister chromatids that usually occurs immediately after centromere separation begins.

Differential response to mitomycin-C- and cis-diamminedichloroplatinum(II)-induced damage in normal human fibroblasts during confluent holding

Confluent cultures of normal human fibroblasts were treated with the chemotherapeutic agents, mitomycin C (MMC) and cis-diamminedichloroplatinum(II) (cis-Pt(II]. This treatment induced a decrease in clonogenic cell survival. Recovery of the cytotoxic effect was observed in the case of cis-Pt(II)-treated cultures maintained at confluence for 1-5 days. No such recovery was observed after treatment with MMC. These data suggest that contrary to potential lethal damage induced by cis-Pt(II) which is repaired in confluent cells, DNA damage induced by MMC is not repaired in confluent cells.

Evaluation of a new drug 7-N-(p-hydroxyphenyl)-mitomycin C [KW 2083] against carcinoma of the lung by the human tumor clonogenic assay

KW 2083, which is a new mitomycin C derivative currently under clinical investigation, was tested for its antitumor effect on the growth of human carcinoma of the lung by using an in vitro clonogenic assay system. The in vitro results in this study were as follows: We succeeded in producing clonal growth at a high rate in all histologic types of carcinoma of the lung in the assay system. That is, 61 of 81 specimens (75%) of either primary or metastatic tumors gave adequate growth for chemosensitivity testing. The in vitro responses to vindesine, adriamycin, mitomycin C and melphalan (substituted for cyclophosphamide in vitro), as standard anticancer drugs for chemotherapy of carcinoma of the lung were 18, 23, 18 and 19%, respectively, which are in good general agreement with the clinical responses to these drugs reported previously. Therefore, the clonogenic assay system might prove to be a very effective tool for an in vitro phase II study of new drugs. The rate of response to KW 2083 tested simultaneously in 51 cancer specimens was 22%, which was superior to that of mitomycin C. These results indicate that KW 2083 might be more useful than mitomycin C in clinical practice.

Cordycepin reduces the sensitivity of BALB/Mo mouse lymphocytes to the induction of sister chromatid exchanges

Lymphocytes from the spleen of BALB/Mo mice, which carry endogenous Moloney murine leukemia virus (M-MuLV), show in vitro frequencies of sister chromatid exchanges (SCEs) significantly higher than lymphocytes from control (M-MuLV free) BALB/c mice. In vitro treatment of lymphocytes with the antiviral antibiotic cordycepin (10 micrograms/ml) lowers the level of SCEs in BALB/Mo cells to the same value of BALB/c cells. M-MuLV yield is also markedly reduced in BALB/Mo lymphocytes cultured in the presence of cordycepin. The drug also abolishes the increased sensitivity of BALB/Mo lymphocytes to the induction of SCEs by mitomycin C (MMC) either in vitro (3 X 10(-8)/10(-7)M) or in vivo (0.3/3 mg/kg). Since cordycepin is known to inhibit poly(A)synthesis thus blocking RNA maturation, it is suggested that M-MuLV proviral integration is not per se the sole factor responsible for the more pronounced susceptibility of BALB/Mo lymphocytes to SCE induction, but most likely viral gene expression and amplification are needed for this effect to occur.

Pattern of Schwann cell remyelination in a spinal cord lesion

Intraspinal injection of mitomycin C into the rat dorsal columns produced extensive demyelination, axonal degeneration and glial cell death. Five weeks post-injection Schwann cell remyelinated fibers were present along the surface of the dorsal columns and around blood vessels within the lesions. Axons near these sites either were enclosed within a Schwann cell but not myelinated or were completely devoid of any cellular ensheathment. Schwann cells were associated only with those blood vessels which no longer retained astroglial end-feet. It is concluded that Schwann cells migrate into spinal cord lesions along such vessels. The marked sub-pial and perivascular distribution of Schwann cell remyelinated fibers may reflect a failure of Schwann cells to disperse quickly elsewhere within the lesion.

Antitumor activity and toxicity in animals of RR-150 (7-cysteaminomitosane), a new mitomycin derivative

The experimental antitumor activity of a new mitomycin derivative, 7-cysteaminomitosane (RR-150), was evaluated in mice. When administered i.p. to mice bearing i.p.-implanted tumors, RR-150 was superior to mitomycin C (MMC) in increasing the life span of animals bearing P388 leukemia, B16 melanoma, and a line of L1210 leukemia partially resistant to MMC. RR-150 appeared comparable to MMC in increasing life span of mice bearing Madison 109 lung carcinoma, Colon 26 carcinoma, or parental (nonresistant) L1210 leukemia. Mice immunosuppressed with 550 rads whole-body irradiation prior to i.p. implantation of B16 still benefited (e.g., 40% cure rate) following optimal RR-150 therapy when compared to nonirradiated, B16-implanted mice given RR-150 (e.g., 70% cure rate). RR-150 had inconsistent activity in the treatment of s.c.-implanted tumors. In toxicity evaluations, RR-150 was comparable to MMC in suppression of total while blood cell counts but appeared to be less neutropenic. RR-150 also caused less cumulative leukopenia than did MMC in a weekly chronic dose experiment. Based on serum chemistries, RR-150 did not have significant nephrotoxicity, but there was evidence of possible liver toxicity at doses near the 50% lethal dose. Because of the balance of favorable antitumor and toxicity properties of RR-150, work is under way to prepare a more bioavailable form for advanced evaluation.

Mitomycin C resistance in a human colon carcinoma cell line associated with cell surface protein alterations

A human colon carcinoma cell line resistant to mitomycin C (MMC) was obtained by repeated exposure of a previously described sensitive parental line, HCT 116, to MMC in vitro. Xenografts grown from the MMC-resistant phenotype were not inhibited in MMC-treated animals, while MMC treatment produced growth inhibition in parental cell xenografts. The MMC-resistant phenotype exhibited a greater amount of a Mr 148,000 cell surface protein than did the parental line. The increase in this Mr 148,000 cell surface protein correlated positively with the degree of MMC resistance. Alkaline elution of filter-bound DNA from resistant cells exposed to MMC in vitro showed a decrease in DNA cross-link formation such that a 10-fold higher MMC concentration was required to produce similar cross-link formation in the resistant cell as compared to the parental cell. The development of MMC resistance was not associated with in vitro cross-resistance to other natural product cytotoxic drugs. This model for resistance to MMC will be useful in future studies to define the mechanisms for MMC action and resistance in human colon carcinoma cells.

Role of NADPH:cytochrome c reductase and DT-diaphorase in the biotransformation of mitomycin C1

Hypoxic cells of solid tumors are difficult to eradicate by X-irradiation or chemotherapy; as an approach to this problem, our laboratories are investigating the effects of the bioreductive alkylating agent mitomycin C (MC) on hypoxic cells. This antibiotic was preferentially toxic to EMT6 mouse mammary tumor cells and V79 Chinese hamster lung fibroblasts under hypoxic conditions, but it was equitoxic to Chinese hamster ovary cells in the presence and absence of oxygen. All cell lines catalyzed the formation of reactive metabolites under hypoxic conditions and contained NADPH:cytochrome c reductase and DT-diaphorase, two enzymes which may be responsible for the cellular activation of MC. Although a correlation existed between enzymatic activities and the formation of reactive metabolites from MC, there was no correspondence between these parameters and the degree of cytotoxicity expressed by MC under hypoxic conditions. Purified NADPH:cytochrome c reductase reduced MC in the absence of oxygen, with addition of cytochrome P-450 enhancing, but not participating directly in, the reduction reaction. Addition of NADP+ to cell sonicates substantially reduced NADPH:cytochrome c reductase activity, while the formation of reactive metabolites was affected only slightly; converse results were observed using mersalyl. Exposure of cell sonicates to dicumarol inhibited DT-diaphorase activity, while the rate of formation of reactive metabolites of MC was enhanced. The findings suggest that NADPH:cytochrome c reductase and some as yet to be identified enzyme(s) are important for the reductive activation of MC. DT-diaphorase and cytochrome P-450 are not directly involved in the activation of MC, but they appear to modulate the degree of activation to reactive species, which are presumably responsible for the observed cytotoxicity.

Adoptive immune therapy in mice bearing poorly immunogenic metastases, using T lymphocytes stimulated in vitro against highly immunogenic mutant sublines

MDW3, a highly immunogenic and non-tumorigenic (tum-) mutant of the poorly immunogenic metastatic murine tumor called MDAY-D2, has been employed in an immune therapy scheme for the treatment of widespread established visceral MDAY-D2 metastases in syngeneic mice. MDW3 was selected from a mutagenized population of MDAY-D2 cells for the ability to grow in the presence of toxic concentrations of wheat-germ agglutinin (WGA) in vitro. The mutant expresses a common tumor-associated antigen (TAA) present on MDAY-D2 as well as a new antigen whose presence enhances the anti-TAA cell-mediated immune response in vivo and in mixed lymphocyte tumor cultures (MLTC) in vitro. For immune therapy, spleen cells from DBA/2 mice which had rejected an inoculum of MDW3 cells were restimulated in MLTC and injected i.v. into MDAY-D2 tumor-bearing mice. Two protocols were used. In the first, mice were given an i.v. injection of 10(3) MDAY-D2 cells ("artificial metastasis") and subsequently treated with 400 R whole-body irradiation and MDW3-stimulated T cells. Such mice had a 75% long-term survival rate, whereas 400 R alone, or no treatment, resulted in 25% and 0% long-term survivors, respectively. In the second protocol, treatment of mice bearing a 12-day-old subcutaneous MDAY-D2 tumor by surgical removal of the solid tumor, 400 R whole-body irradiation, and systemic administration of MDW3-stimulated spleen cells, resulted in a 75-100% survival rate, whereas omitting any part of the treatment resulted in 0-50% survival rates. The treatment increased splenic anti-TAA CTL activity, and the mice acquired immunity against the new antigen on MDW3, suggesting that the injected lymphocytes were proliferating in the host. The optimal combination of resection, whole-body irradiation and passive infusion of MDW3-stimulated spleen cells was ineffective when used on mice bearing a tumor-antigen-loss variant of MDAY-D2, suggesting that success of our immune therapy protocol required specific recognition of the tumor-associated antigen of MDAY-D2.

Induction of 6-thioguanine-resistant mutants and SCEs by 3 chemical mutagens (EMS, ENU and MMC) in cultured human blood lymphocytes

Human peripheral blood lymphocytes were exposed in vitro to a series of graded concentrations of ethyl methanesulphonate (EMS) or N-ethyl-N-nitrosourea (ENU) or mitomycin C (MMC) to: (a) estimate the frequency of thioguanine-resistant (TGr) cells using the T-cell cloning technique, (b) examine the induction of sister-chromatid exchanges (SCEs) by these chemicals in the lymphocytes of the same blood sample used to study the TGr cells, and (c) assess the nature of correlations between these two biological end-points. The frequencies of TGr cells as well as those of SCEs increased with increasing concentration of the chemicals studied. For EMS and ENU, the increases were consistent with linear dose-effect relationships. There was a linear relationship between SCEs and mutation induction for all 3 chemicals; but the ratio of induced SCEs to induced mutants was different for the different chemicals, being highest for ENU, followed by EMS and MMC, in that order. The basis for these differences is discussed in the light of what is known about the relationships between chemical reactivity patterns and the resultant biological effects of these chemicals.

Two-step cell-death kinetics in vitro during cis-platinum, hydroxyurea and mitomycin incubation

A human leukaemic cell line (REH) growing in suspension was incubated with cis-platinum, hydroxyurea and mitomycin C at various concentrations causing complete cell-cycle arrest. At different times the cell suspensions were harvested, diluted 1:1 with a buffer solution, stained without further treatment with a mixture of acridine orange (AO) and ethidium bromide (EB) and analysed with a biparametrical flow cytometer. Fluorescent plastic beads were introduced into the suspensions to provide an internal numerical reference for the control of cell loss. The fluorescence distributions showed three groups of cells: vital cells (V) which were only stained with AO; dead cells in which EB stained cytoplasmic components but not the nuclear DNA (D1), and dead cells which allowed EB to stain both cytoplasm and nuclear DNA (D2). The kinetics of cells entering D1 depended on drug concentration and showed equal characteristics for cis-platinum and mitomycin, but were different for hydroxyurea. The subsequent entry into D2 occurred about 15 hr later and showed no pronounced dependence on drug concentration. Parallel trypan-blue (TB) exclusion tests revealed that TB only stained D2 cells and therefore is not useful for investigating cell-death kinetics during exposure to cell-killing agents.

Induction of sister-chromatid exchanges and chromosome aberrations in human peripheral lymphocytes by 2,4,7-trinitro-9-fluorenone

The genotoxicity of 2,4,7-trinitro-9-fluorenone (TNF) was investigated using the sister-chromatid exchange (SCE) and chromosome aberration assay systems with human peripheral lymphocytes. The SCE frequency increased from the control value of 9.45 to 27.80 (mean of 50 cells scored from each of two donors) after the lymphocytes were treated with the highest dose of TNF (3.0 microgram/ml). The results of the chromosome aberration experiment show that TNF induced more than a 4-fold increase in aberrations in each donor. The results, which indicate that TNF is clastogenic, add further evidence that this compound is potentially genotoxic to exposed populations.

The DNA repair host-mediated assay as a rapid and sensitive in vivo procedure for the determination of genotoxic factors present in various organs of mice. Some preliminary results with mitomycin C

The DNA repair host-mediated assay, in which repairable DNA damage is determined in E. coli cells present in various organs of mice exposed to genotoxic agents, was further developed to broaden the range of organs under study and to simplify the procedure of assessing differential bacterial cell survival. A pair of derivatives of E. coli K-12 strain 343/113 was constructed which differed vastly in DNA repair capacity (uvr+/rec+ vs uvrB/recA), as a means of assessing DNA damaging effects; furthermore, the strains differed in their ability to ferment lactose (delta Lac vs Lac+), so that the individual survival of both strains could be determined on a single agar medium (containing neutral red as pH indicator), on which the strains had different colony colour morphology (red, Lac+ vs white, Lac- colonies). Finally, the strains were made streptomycin-dependent, to prevent uncontrolled growth of the bacterial cells within the various organs and also to inhibit contamination of the survival agar medium by representatives of the normal intestinal microflora. The experimental procedure consisted of injecting mixtures of stationary cells of the two strains (ca. 3-5 X 10(8) viable cells per mouse) both intravenously and orally into mice, either pretreated or subsequently treated with test chemicals. Ninety minutes after injection of the bacteria, the liver, spleen, lungs, kidneys, stomach, intestine, colon, and ca. 50 microliter blood, were removed, suspended in buffer, homogenized, and the survival of the two strains determined on neutral red agar supplemented with streptomycin.(ABSTRACT TRUNCATED AT 250 WORDS)