Search

One of the most frequently studied therapeutic strategies in the field of suicide gene therapy is based on the expression by tumor cells of the Herpes simplex virus thymidine kinase gene (HSVtk) followed by a ganciclovir (GCV) treatment. In order to investigate the potential of other enzyme/prodrug strategies, we studied in vitro and in vivo the ability of the Varicella zoster virus thymidine kinase gene (VZVtk) to act as a suicide gene and to kill non-transduced bystander cells, and compared this activity to that of its HSV counterpart. Four different antiviral compounds were tested as prodrugs. Our comparative study demonstrates the superiority of the HSVtk/GCV system among the different combinations tested and underlines the importance of both the tumor cell type and the prodrug in the success of a prodrug/suicide gene strategy.

The tissue inhibitor of metalloproteinases-2 (TIMP-2) has at least two independent functions, i.e., regulation of matrix metalloproteinases and growth promoting activity. We investigated the effects of TIMP-2 overexpression, induced by retroviral mediated gene transfer, on the in vivo development of mammary tumors in syngeneic mice inoculated with EF43.fgf-4 cells. The EF43.fgf-4 cells established by stably infecting the normal mouse mammary EF43 cells with a retroviral expression vector for the fgf-4 oncogene, are highly tumorigenic and overproduce vascular endothelial growth factor (VEGF). Despite a promotion of the in vitro growth rate of EF43.fgf-4 cells overexpressing timp-2, the in vivo tumor growth was delayed. At day 17 post-cell injection, the volume of tumor derived from TIMP-2-overexpressing cells was reduced by 80% as compared with that obtained with control cells. Overexpression of TIMP-2 was associated with a down-regulation of VEGF expression in vitro and in vivo, a reduction of vessel size, density, and blood supply in the induced tumors. In addition, TIMP-2 completely inhibited the angiogenic activity of EF43.fgf-4 cell-conditioned medium in vitro using a rat aortic ring model. Our findings suggest that overexpression of TIMP-2 delays growth and angiogenesis of mammary carcinoma in vivo and that down-regulation of VEGF expression may play an important role in this TIMP-2-mediated antitumoral and antiangiogenic effects. Finally the in vivo delivery of TIMP-2, as assessed by i.v. injection of recombinant adenoviruses vectors, significantly reduced the growth of the EF43.fgf-4-induced tumors. This effect of TIMP-2 was shown to be equally comparable with that of angiostatin, a known potent inhibitor of angiogenesis.

To investigate the factors influencing the bystander effect--a key element in the efficacy of suicide gene therapy against cancer--we compared the effect triggered by four extremely efficient gene/prodrug combinations, i.e., VZVtk/BVDU, the thymidine kinase of Varicella zoster virus associated with (E)-5-(2-bromovinyl)-2'-deoxyuridine; VZVtk/BVaraU, the same enzyme associated with (E)-5-(2-bromovinyl)-1-beta-D-arabinofuranosyluracil; HSVtk/BVDU, the association of the Herpes simplex virus thymidine kinase with BVDU; and the classical HSVtk/GCV (ganciclovir) paradigm. The cells used, the human MDA-MB-435 breast cancer, and the rat 9L glioblastoma lines were equally sensitive in vitro to these four associations. In both cell types, the combinations involving pyrimidine analogues (BVDU, BVaraU) displayed a smaller bystander killing than the combination involving the purine analogue (GCV). In addition, the bystander effect induced by all the tk/prodrug systems was reduced in MDA-MB-435 cells in comparison to 9L cells; albeit, the viral kinases were produced at a higher level in the breast cancer cells. All systems induced apoptotic death in the two cell types, but the MDA-MB-435 cells, deprived of connexin 43, were noncommunicating in striking contrast with the 9L cells. That functional gap junctions have to be increased in order to improve the breast cancer cell response to suicide gene therapy was demonstrated by transducing the Cx43 gene: this modification enhanced the bystander effect associated in vitro with GCV treatment and, by itself, decreased the tumorigenicity of the untreated cells. However, the noncommunicating MDA-MB-435 cells triggered a significant bystander effect both in vitro and in vivo with the HSVtk/GCV system, showing that communication through gap junctions is not the only mechanism involved.

Stromelysin-3 (ST3), a matrix metalloproteinase (MMP) expressed in aggressive carcinomas, has been shown to promote tumor development in different in vivo experimental models. However, the inability of its mature form to degrade extracellular matrix components casts doubt on whether ST3 functions in vivo as a protease. In this study, we evaluated whether the ST3 tumor-promoting effect could be ascribed to its proteolytic activity and whether this putative protease could be targeted with MMP inhibitors. Catalytically inactive mutant cDNA of human (h) ST3 or mouse (m) ST3 were generated and transfected into MCF7 cells. When injected into nude mice in the presence of matrigel, the mutant-bearing cells did not exhibit the enhanced tumorigenicity elicited by MCF7 cells transfected with wild-type ST3 cDNA. In a second approach, TIMP2 overproduction in MCF7 cells expressing hST3 was induced by retroviral infection. The co-expression of ST3 and TIMP2 failed to enhance the tumorigenicity of MCF7 cells. Notably, matrigel depleted of low-molecular-weight proteins and growth factors failed to promote the tumorigenicity of ST3-expressing MCF7 cells. These findings provide the first in vivo evidence that ST3 is indeed a protease that can modulate cancer progression by remodeling extracellular matrix and probably by inducing it to release the necessary microenvironmental factors. Thus, ST3 represents an interesting target for specific MMP inhibition.

The transforming properties of fibroblast growth factor 3 (FGF-3) were investigated in MCF7 breast cancer cells and compared to those of FGF-4, a known oncogenic product. The short form of fgf-3 and the fgf-4 sequences were each introduced with retroviral vectors and the proteins were only detected in the cytoplasm of the infected cells, as expected. In vitro, cells producing FGF-3 (MCF7.fgf-3) and FGF-4 (MCF7.fgf-4) displayed an amount of estrogen receptors decreased to around 45% of the control value. However, MCF7.fgf-3 cell proliferation remained responsive to estradiol supply. The sensitivity of the MCF7.fgf-4 cells, if existant, was masked by the important mitogenic action exerted by FGF-4. In vivo, the MCF7.fgf-3 and MCF7.fgf-4 cells gave rise to tumors under conditions in which the control cells were not tumorigenic. Supplementing the mice with estrogen had the paradoxical effect of totally suppressing the start of the FGF-3 as well as the FGF-4 tumors. Tumorigenicity in the presence of matrigel was similar for MCF7.fgf-3 and control cells and was increased by estrogen supplementation. Once started, the MCF7.fgf-4 tumors grew with a characteristic high rate. Remarkably, FGF-4 but not FGF-3, stimulated the secretion of vascular endothelial growth factor (VEGF165) without altering the steady-state level of its mRNA, suggesting a possible regulation of VEGF synthesis at the translational level in MCF7 cells. The increased VEGF secretion is probably involved in the more aggressive phenotype of the MCF7.fgf-4 cells while a decreased dependence upon micro-environmental factors might be part of the increased tumorigenic potential of the MCF7.fgf-3 cells.

The inhibitory effects of (E)-5-(2-bromovinyl)-2'-deoxyuridine (BVDU) and its arabinosyl derivative (E)-5-(2-bromovinyl)-1-beta-D-arabinofuranosyluracil (BVaraU) on the growth of both MDA-MB-435 human breast carcinoma and 9L rat gliosarcoma cells expressing the thymidine kinase (tk)-encoding gene of the Varicella zoster virus (VZV) or the Herpes simplex virus (HSV) were evaluated. In vitro, BVDU and BVaraU effectively killed both cell types expressing VZVtk, with 50% inhibitory concentration values ranging from 0.06 to 0.4 microM, whereas ganciclovir (GCV) lacked activity. On HSVtk+ cells, BVDU had high cytotoxic activity, with 50% inhibitory concentration values that were similar to those of GCV, whereas BVaraU was inactive. In vivo, BVDU applied intraperitoneally caused a 50% tumor growth inhibition in nude mice inoculated subcutaneously with VZVtk+ as well as HSVtk+ mammary tumor cells. In mice and at variance with the in vitro results, BVaraU had very little activity against the VZVtk+ mammary cells; GCV had the highest activity on the HSVtk+ cells, resulting in a 50% eradication of the tumors. With the 9L rat gliosarcoma model, the VZVtk/BVDU system completely failed to inhibit the development of VZVtk+ glioma tumors induced subcutaneously in syngeneic rats, although BVDU had a similar 45-minute half-life in both rats and mice. Factors other than degradation of the prodrug and related to the mode of action of these analogs are possibly involved in the observed discrepancies between the in vitro and in vivo results.

Fibroblast Growth Factors 3 (FGF-3) and 4 (FGF-4) were compared for the effects they each exert on EF43 mouse cells. This non-transformed mammary cell line appears to be myoepithelial mainly because it expresses alpha-smooth muscle actin. The EF43 cells were infected with similar vectors that carry either the short fgf-3 sequence (the product of which goes into the secretory pathway), fgf-4 or the selection gene only as control. In syngeneic animals, EF43.fgf-3 cells were tumorigenic only when orthotopically implanted whereas EF43.fgf-4 cells invariably gave rise to aggressive tumors. However, both tumor types were metastatic as evidenced by the blue micrometastases observed when the implanted cells expressed lacZ. In vitro, the FGF-3 producing cells were strongly invasive in matrigel coated chambers whereas the EF43.fgf-4 cells only were invasive in type I-collagen gels. Interestingly, FGF-3 production greatly stimulated the synthesis of pro-MMP-9 (Matrix Metalloprotease-9) and, to a lesser extent, that of pro-MMP-2. FGF-3 also up-regulated the production of plasminogen activators. In contrast, FGF-4 had no effect on these secretions and the medium conditioned by the EF43.fgf-4 cells displayed the largest plasminogen activator-inhibitor activity. These results show that FGF-3 and FGF-4 have distinct mechanisms of action on myoepithelial cells.

Co-injection of fibroblasts with human epithelial breast-tumor MCF7 cells in the presence of Matrigel enhances tumor growth in nude mice. While most of the matrix metalloproteinases (MMPs) have been shown to be produced by stromal cells, tumor cells such as MCF7 cells are unable to produce MMPs. We therefore, hypothesized that the tumor-promoting effect of fibroblasts could be related to their production of MMPs. In order to inhibit stromal proteases, over-production of TIMP-2 was induced in MCF7 cells by in vitro retroviral-mediated gene transfer. TIMP-2-producing MCF7 cells were then co-injected with fibroblasts into nude mice. Alternatively, we evaluated the effect of Batimastat, a synthetic inhibitor of MMPs, on the tumorigenicity of MCF7 cells co-inoculated with fibroblasts into nude mice. Both physiological (TIMP-2) and synthetic (Batimastat) inhibitors of MMPs were able to abolish the tumor-promoting effect of fibroblasts. On the contrary, they failed to modulate the tumorigenicity of MCF7 cells injected alone. Interestingly, Matrigel from which low-molecular-weight proteins or growth factors had been removed failed to favor the tumorigenicity of MCF7 cells inoculated with fibroblasts. These findings emphasize the importance of fibroblasts in cancer progression, and suggest that their role could be related at least in part to production of proteases which can induce the release of factors from the extracellular matrix.

The infection of normal mouse mammary EF43 cells by a retroviral vector carrying either Fgf-3 (EF43.Fgf-3) or Fgf-4 (EF43.Fgf-4) cDNA resulted in the transformation of cells displaying different tumorigenic potentials in nude mice (A. Hajitou and C-M. Calberg-Bacq, Int. J. Cancer, 63: 702-709, 1995). EF43.Fgf-4 produced rapidly developing tumors at all sites of inoculation, whereas EF43.Fgf-3 produced slowly growing tumors only in the mammary fat pad. Cells infected with the vector carrying the selection gene alone (EF43.C) were not tumorigenic. The angiogenic properties of these cells were tested in an in vitro angiogenesis model using human umbilical vein endothelial cells (HUVECs) cultured at the surface of a type I collagen gel and their capacity to form tube-like structures on invasion of the gel. Only the conditioned medium (CM) of EF43.Fgf-4 induced an angiogenic morphotype in HUVECs. In parallel, the mRNA expression of matrix metalloproteinase 1 and c-ETS-1 was increased in the HUVECs displaying a differentiated phenotype, whereas the tissue inhibitor of matrix metalloproteinase 1 mRNA level was decreased. Recombinant human fibroblast growth factor 4 (FGF-4) did not induce an angiogenic phenotype in HUVECs by itself. By Western blot analysis, a high expression of vascular endothelial growth factor (VEGF) was detected in the EF43.Fgf-4 CM. This result was confirmed by Northern blot analysis of total RNA extracted from the three cell types; the steady-state level of VEGF mRNA was low and equivalent in EF43.C and EF43.Fgf-3, whereas it was strongly increased in EF43.Fgf-4. Culturing EF43 cells carrying only the selection gene with increasing concentrations of recombinant human FGF-4 resulted in a dose-dependent stimulation of VEGF. The induction of the angiogenic morphotype and the parallel modulations of the biosynthetic phenotype in HUVECs were completely suppressed by adding a neutralizing antibody directed against VEGF to EF43.Fgf-4 CM. Furthermore, inhibition of protein kinase C by bisindoylmaleimide suppressed the angiogenic phenotype induced by the CM of EF43.Fgf-4. Our results point to an indirect angiogenic activity of FGF-4 through the autocrine induction of VEGF secretion by EF43.Fgf-4 cells, an original signaling pathway that might be significant in tumor progression and metastasis.

Fibroblast growth factor-3 (Fgf-3) is involved in mouse mammary tumorigenesis since the Fgf-3 gene is a main target for mouse mammary tumor virus (MMTV) insertional activation. Its action has been correlated with the appearance of pregnancy-dependent tumors. We describe here the effects on normal mouse mammary EF43 cells of the short Fgf-3 protein form which enters the secretory pathway. The genes, Fgf-3 AUG or Fgf-4 for comparison, were introduced in the mammary cells by means of retroviral vectors. Fgf-3 expression did not modify EF43 cell morphology, had no effect on growth in soft agar nor on the inhibitory action exerted on cell growth by TGF-beta 1; however, it allowed the cells to grow under low serum conditions in the absence of insulin and EGF. The Fgf-3-expressing cells were not tumorigenic in nude mice when injected s.c., but tumors developed when the cells were implanted in the mammary gland. The tumors appeared after some latency; they had a slow growth phase followed by a phase of increased growth rate. An identical tumoral growth pattern was observed in ovariectomized nude mice. These results show that the secreted Fgf-3 form can initiate tumorigenesis and that the induced tumors are hormone-independent. The mammary-gland environment, however, is required for the EF43 cells to grow and differentiate. During that process, which resembles natural cell growth during mammary-gland development at pregnancy, the cells could pass through a stage which is specifically sensitive to Fgf-3.

Tumorigenesis by mouse mammary-tumor virus (MMTV) involves proviral disruption and transcriptional activation of a number of cellular oncogenes, generically termed Int. The frequencies with which different Int genes are activated in different mouse strains can be quite variable, and previous surveys have suggested that insertions at Int-2/Fgf-3 occur primarily in strains that develop pregnancy-dependent mammary tumors. To address this issue, we have determined the relative contributions of 5 known Int genes (Wnt-1, Wnt-3, Fgf-3, Fgf-4 and Int-3) in mammary tumors from virgin BR6 and multiparous BR6, BALB/cfBR6 and RIII mice. Whereas Fgf-3 was implicated in 66%, 80% and 92% of the tumors from the respective parous animals, only 20% of the tumors from virgin mice expressed Fgf-3. This reduced involvement of Fgf-3 was compensated by proviral insertions in Fgf-4, Int-3 and Wnt-3, but the frequency of Wnt-1 activation was relatively constant. These data strengthen the link between Fgf-3 and a pregnancy-dependent phenotype and suggest that, in the strains analyzed, the frequency of Int-gene activation was influenced more by the hormonal status than by the genetic background.

The influence of various normal and malignant human cells on the level of collagen synthesis by human fibroblasts was tested in coculture. As revealed by immunoperoxidase staining, in cocultures with breast adenocarcinoma cells (MCF7, SA52, T47D) fibroblasts synthesized collagen while tumor cells did not. Fibroblasts displayed increased collagen production without change in the overall protein synthesis. Several other types of cells derived from normal human tissues (keratinocytes, normal mammary cells) or from fibrosarcoma, melanoma, cervical carcinoma, choriocarcinoma, or other breast adenocarcinoma (SW613, MDA, BT20) did not affect collagen synthesis of fibroblasts. Although to a lesser extent, this stimulating effect was reproduced by using the conditioned medium (CM) of the active cells but not with CM of the other cell types. A slight stimulation was also obtained when tumoral MCF7 cells and fibroblasts shared the same medium but were physically separated, suggesting that close contact was required for optimal stimulation of collagen synthesis. The collagen synthesis stimulating activity was not related to a modification of fibroblast proliferation rate. The production of collagen types I, III, and VI and fibronectin were increased in cocultures of fibroblasts with MCF7 cells. The increased synthesis of collagen types I and III and fibronectin was paralleled by similar changes in the steady-state level of their mRNAs. On the contrary, the increased production of collagen type VI appeared regulated at a post-transcriptional level.

Host-virus relationships were examined in mice from the two mouse mammary tumor virus (MMTV)-infected strains SWISS MB+ and RIII, which harbour the same MMTV variant, and from the derived sublines Swiss MB- and RIIIf, which were freed of milk-borne MMTV by foster-nursing. These two strains are not phylogenetically related, the SWISS strain bearing the endogenous Mtv-3 locus in its DNA. In RIII and SWISS MB+ mice, the incidence of early mammary tumors, which was of 96% and 8%, respectively, was correlated to the level of MMTV expression in milk. In the SWISS MB-line, a non-coordinate expression of the provirus associated with the Mtv-3 locus was observed in the mammary glands, the salivary glands and the spleen. This expression was not tumorigenic and was characterized by the presence of the p28 gag antigen and the absence of the gp52 env antigen, except, however, in mammary glands of elder mice where traces of gp52 were found. In the mammary glands of SWISS MB+ mice, the expression of the Mtv-3 locus was masked by large amounts of antigens resulting from exogenous virus expression. RIIIf mice were MMTV-negative. Viral antigens coexisted with anti-MMTV antibodies in the serum of infected and tumor-bearing mice, but not in the form of immune complexes as verified by a method that allowed to detect specific antigen-containing-soluble immune complexes. An anti-MMTV serum reactivity was also detected in SWISS MB- and RIIIf mice. However, the serum response was higher in the two SWISS lines than in the two RIII lines. Except in tumor-bearing mice, the anti-MMTV response was not significantly modified by the presence of exogenous virus and thus resulted essentially from exposure to endogenous MMTV expression. In experimental infection studies, RIII mice were more susceptible to MMTV infection than SWISS mice. The correlation between resistance to MMTV infection and serum response to endogenous MMTV expression, suggests that the non-tumorigenic expression of an endogenous provirus can protect at least partially, against exogenous MMTV infection.

The photosensitizing efficiency of six dyes--proflavine, 9-aminoacridine, ethidium bromide, thiopyronine, pyronine and acridine red--have been compared on the basis of the inactivation of sensitized T4 phage caused by light irradiation. This reaction was only measurable after diffusion of the dye through the phage capsid and was not observed in the presence of either chloroquine or quinacrine; it followed a single-hit kinetics as a function of the irradiation time. With each dye, a double reciprocal plot of the inactivation constant versus the dye concentration present gave rise to a linear relationship. From this relation, parameters were deduced which expressed the relative photosensitizing efficiencies. Dye-binding to the phages was measured and the proflavine-mediated inactivation appeared to be related to the amount of strongly bound molecules. Such a conclusion could not be reached in the case of 9-aminoacridine and ethidium bromide, which were much less efficient photosensitizers than proflavine, but which were also strongly bound to the phages. Thiopyronine was weakly bound to the phages; it had, however, the highest photosensitizing activity observed. These results indicate that various mechanisms are involved when the phage photosensitization is due to one dye or another.

Proflavine bound-superhelical phi XRFI DNA Molecules undergo single-strand scission upon irradiation with visible light at high fluence rate. As shown by agarose gel electrophoresis analyses, the nicking reaction is (i) oxygen-dependent, (ii) strongly inhibited by catalase and an electron scavenger such as cystamine, and (iii) totally suppressed by ceruloplasmin and radical scavengers such as t-butanol sodium benzoate. This indicates that H2O2, e-, O2 and OH, respectively, are involved in the cleavage process. NaN3, a singlet-oxygen quencher, has very little effect on strand-breakage but it prevents almost completely the alteration of guanine residues (a lesion already observed after irradiation at low fluence rate). Since, in the presence of NaN3, strand scission can occur and guanine (as the other bases) recovered intact, it follows that the radical intermediates produced during breakages are probably not involved in any permanent modification of the DNA bases.

The major antigens of a mouse mammary tumour virus (MMTV) isolated from the milk of exogenously infected MB+ Swiss mice were compared with the viral components of other MMTV strains by using the methodology developed by Teramoto et al. (1977 a). The anti-gp47 (Swiss) serum differentiated between type- and group-specific antigenic determinants on the major glycoprotein; two distinct type-reactivities were demonstrated, one of them being shared by the MMTVs (Swiss) and (RIII), and the other by the MMTVs (C3H) and (GR). The MMTVs (Swiss) and (RIII) also reacted identically in a 'type-specific' assay using an anti-p28 (Swiss) serum, but a distinct reactivity was observed with the MMTV (C3H). In the isologous test where an anti-(total RIII) serum was used to bind intact, externally labelled RIII virions, the Swiss virus was seen to possess a type-specific determinant on its surface which distinguished itself from both MMTVs (RIII) and (C3H). The Swiss/fC57BL mice (which are devoid of the milk virus and for this reason are referred to as MB-), expressed only the internal virus antigens in their mammary glands. Under the 'type-specific' assay conditions, the p28 antigen present in the mammary gland extracts of the MB+ mice was indistinguishable from the p28 antigen purified from the Swiss virion, but was clearly distinct from the p28 reactivity present in the mammary gland extracts of the MB- mice. The p28 (MB-) antigen may thus represent the expression of an endogenous virus sequence different from the milk virus genome.

Murine mammary tumor virus main glycoprotein (gp47), prepared by diethylaminoethyl cellulose and hydroxyapatite chromatography of detergent-mercaptoethanol-KCI-disrupted virion, was used as labeled antigen in a highly specific and reproducible radioimmunoassay. Seven other (glyco) proteins of the virus were antigenically distinct from gp47. Serum and organs of unifected C57BL mice did not contain gp47, but sera of infected Swiss and RIII mice did contain the antigen. Despite the high content in the mammary gland, the level of gp47 in other organs was identical in male and female mice. The titer of gp47 in serum was high in tumor-bearing females, but it varied with the mouse strain. Anti-gp47 immunoglobulins could not be detected. The investigation included 314 human sera (107 normal, 65 benign mastopathy, 89 breast vancer, and 53 digestive cancer). None contained an antigen related to gp47. One of 20 human mammary cyst fluids was positive.

Irradiation in the presence of O2, with near-UV light of five promazine (PZ) derivatives added to erythrocyte ghost membranes, causes covalent cross-linking between proteins as revealed by a progressive decrease in the amounts of proteins separable by electrophoresis after denaturation. The induction of cross-links in the two spectrin subunits is a single-hit process as a function of the irradiation time; relatively the rate constants (in min-1) of the photoreactions were 0.060 with chlorpromazine (CPZ), 0.039 with methoxypromazine (MTPZ), 0.031 with PZ, 0.029 with triflupromazine (TFPZ) and 0.006 with acepromazine (ACPZ). A main photochemical intermediate implicated in the spectrin aggregation seems to be the cation radical of the PZ derivatives. Indeed, (i) the chemically generated cation radicals can induce the reaction in the dark; (ii) the photoaggregation is regularly reduced upon addition of increasing concentrations of NaN3; (iii) NaN3 similarly affects the amount of cross-links induced by the isolated cation radicals. Hydroxyl radicals are also involved in the photocross-linking when the reaction is initiated only by MTPZ and not by the other sensitizers. In the absence of oxygen during irradiation, PZ, MTPZ and ACPZ completely loose their cross-linking activities whereas CPZ and TFPZ remain as efficient as in the presence of oxygen.

The main protein of the core of murine mammary tumor virus, with a molecular weight of 28,000 (p28), was solubilized by deoxycholate treatment of the virus and purified by Ultrogel ACA-54 filtration and hydroxyapatite chromatography. This protein was used as labeled antigen in a highly specific and reproducible radioimmunoassay. Organ extracts of uninfected C57BL mice did not contain p28, but organ extracts of infected RIII mice did contain the antigen. Despite the high content in the mammary gland, the level of p28 in the other organs was identical in male and female mice. Sera of uninfected mice and the majority of the sera of infected mice did not contain the antigen. The investigation included 338 human sera (50, normal; 157, breast cancer; 77, polycystic disease; 32, benign mastopathy; 12, fibroadenoma; 10, at risk of developing breast cancer). None contained an antigen related to p28. Eight of 24 extracts of human breast cancer gave results that appeared weakly positive, possibly as a result of proteolysis. Extracts of healthy breast tissue and the serum from the breast arterial and venous blood of corresponding patients were negative.

Persistence of mouse mammary tumour virus (MMTV) components in the digestive tract of suckling mice was investigated by immunoperoxidase staining of the main viral antigens and micro-immunoenzyme assays of gp52 and p28; these latter assays were also performed after ingestion of milk enriched in viral antigens using Cr2O3 as a marker for the alimentary bolus migration. When compared to the ingested antigens, the amounts of gp52 and p28 decreased during transit, p28 being more rapidly digested than gp52. The antigens were, however, destroyed to a much larger extent in the gut of the adult than in that of the newborn mouse. A fraction of the marker remained for a long time in the stomach; a prolonged retention was also observed with gp52 and especially with p28. Significant amounts of viral antigens were detected in the intestinal walls: both p28 and gp52 were found in the duodenum and small intestine. Moreover, the four viral antigens gp52, gp36, p28 and p8 were clearly observed in very large supranuclear vacuoles inside the epithelial cells of the distal part of the gut. Total particles can reach the intestine; the viral material could then be either destroyed or taken up in the epithelial cells by endocytosis, so that the intestinal epithelium might serve as a portal of entry for MMTV in the suckling mouse.

The coliphage phi gamma, though capable of genetic recombination with phi 80, is morphologically distinct from the phi 80 virion. It has a prolate head (58.4 X 46.7 nm) bearing a tail (143 nm) which is strikingly flexible. On the basis of their buoyant density in CsC1, both infectious and transducing phi gamma particles form a single population. This density value is slightly higher than that of the phi 80 virion. The phi gamma chromosome is a double-stranded linear DNA molecule of 13.4 mum in length (corresponding mol. wt.: 27.6 X 10(6)). From its melting temperature and its buoyant density in CsC1, phi gamma DNA appears to have a base composition very close to that of Escherichia coli DNA.

This paper presents a study of the interactions between Escherichia coli cells and nifurzide, a nitrofuran derivative which is used as an intestinal antiinfectious agent. At low concentrations of nifurzide, the growth rate of the cultures decreased, and elongated, nonseptate cells appeared. At high concentrations, complete growth inhibition occurred, accompanied by a rather strong bactericidal effect, but the appearance of the cells was normal; in particular, no bacteriolytic effect was observed. A very large number of antibiotic molecules were bound per bacterial cell. After cell disruption, similar amounts of nifurzide were found in the cytoplasm, cytoplasmic membranes, and cell wall, respectively. Most of the bound nifurzide was rapidly degraded or became protein bound. The structure of the outer membrane lipopolysaccharide appeared to have little influence on the activity of nifurzide.

The indirect immunofluorescence technique was used to investigate the expression of virus antigens in a Swiss albino mouse strain infected by mouse mammary tumour virus (MMTV). The antisera used were monospecific for: the glycopolypeptides gp160 and gp47, and the internal polypeptides p28 and p8. In mice infected with milk-borne MMTV, all four antigens were detected in the mammary gland during the first pregnancy and then throughout life, and in mammary tumours. Antigen gp160, located at the secretory border of all glandular cells, is shared by the plasma membrane of the mammary cell, whether virus-producing or not, and by the virus envelope, as shown by the use of absorbed antisera. The other three antigens are virus-specific. Anti-p47 serum stained the secretory border of cells whereas the fluorescence related to p28 and p8 consisted of a few large intracytoplasmic spots. The specific antibodies for p28 and p8 were only absorbed by disrupted virions, confirming that these polypeptides are internal antigens. With serial sections of glands from pregnant and old female mice, the alveoli stained with anti-p28 or anti-p8 serum were found to react also with anti-gp47 serum. However, during lactation, the presence of p28 and p8 could not be detected in the alveoli stained with anti-gp47 serum. Thus, infected alveoli express virus antigens differently. In the mammary glands of all mice studied, some alveoli did not display any staining when the latter three antisera were used. The specific antigenic pattern, maintained in differentiated mammary tumours, became a diffuse and unlocalized staining after transformation into the undifferentiated state. With the exception of the kidney, where gp47 was found in the glomeruli, all other organs of the mouse did not stain with any of the antisera. The presence of gp47 in the glomeruli, probably related to immune-complex deposits, increases with the age of the mouse and is most noticeable in tumour-bearing females. In mice free of milk-borne MMTV, examined during the first week of lactation, antigen gp47 could not be detected. Internal antigens p28 and p8 were detected in nearly all mammary cells as numerous small cytoplasmic dots. This suggests a limited expression of an endogenous virus genome; this expression would be modified when milk-borne virus is produced.

We have investigated the intracellular proteins synthesized in rat XC and feline kidney cells transfected with endogenous mouse mammary tumor virus (MMTV) proviral DNA. The endogenous provirus GR40, associated with the Mtv-8 locus, directs the synthesis of gag proteins indistinguishable from those found in MMTV-infected cells. The env precursor Pr73env and the mature gp52 proteins could not be detected in these cells. Instead an env-related protein of 68K is synthesized. In contrast to this endogenous provirus, a cloned exogenous proviral variant directs the synthesis of apparently normal env proteins upon transfection into the same cell lines. These results suggest that the env gene of the endogenous MMTV provirus GR40 is defective. The exogenous proviral variant is not expected to synthesize virus particles since it carries a rearrangement in the gag gene. In order to obtain an MMTV provirus capable of correctly expressing both gag and env functions, we have constructed a hybrid endogenous-exogenous provirus containing the 5' long terminal repeat (LTR)-gag of GR40 and the pol-env-3' LTR of the exogenous provirus. Upon transfection into feline kidney cells, this hybrid provirus directed the synthesis of apparently authentic gag and env proteins. Further, virus particles can be detected in the culture medium of the transfected cells by electron microscopy. Viral proteins obtained from viral particles banded in a sucrose gradient were detected by immunoprecipitation.

Quantitative determinations of gp52, the main envelope glycoprotein, and p28, the main core protein, of MMTV, have been performed in about 1000 individual samples of milk of breeding females from our colony of MMTV-infected Swiss mice, a line characterized by a moderate incidence of mammary tumors. A computer analysis of the results showed: 1-- an important individual variation, ranging from 0 to 120 micrograms per ml of milk for p28, and from 0 to 320 micrograms per ml of milk for gp52; 2-- a variation of the release of both antigens during a single lactation, with a maximum on the 7--8th day of nursing; 3-- an increase of the release of both antigens with parity up to the 6th lactation, followed by a marked decrease during later lactations; 4-- a higher degree of infection in the offspring of 2nd and 3rd litters. The possible dependence of viral expression and transmission of infection upon factors such as cyclic activity of the mammary gland and progressive immunization of mice against MMTV is analyzed. The status of our laboratory line of MMTV infected Swiss mice is discussed in comparison with high and low tumor incidence strains.

When injected into the peritoneal cavity, ethidium bromide can strongly inhibit the multiplication of mouse Ehrlich ascites tumour cells. This antitumour effect is increased when ethidium bromide is linked to DNA and also injected into the peritoneal cavity. The cellular alterations are identical after a treatment with E.B. either free or bound to DNA. However, when the cells are treated with E.B.-DNA they contain E.B. for a longer period than after a treatment with E.B.

Ethidium bromide, either free (EB) or bound to DNA (EB-DNA), is injected into the peritoneal cavity of adult rats or mice. EB is then detected by fluorescence microscopy in peritoneal cells and by spectrophotometry in the peritoneal fluid. EB-DNA persists for a longer period of time in the peritoneal cavity than free EB does.

The photosensitizing efficiencies of eight dyes have been compared; two acridines, two xanthene derivatives, one sulphur-containing dye and three chemotherapeutic agents. The analysed reaction was the photosensitized induction of free radicals in calf-thymus DNA at low temperature. The binding of these dyes to DNA was first measured. Both strong (process I) and weak (process II) binding, with different intensities, either alone or together, were observed as mode of fixation. Whatever the nature of their binding, all the dyes used revealed a photosensitizing power as inducers of peroxide radicals in DNA. Their relative efficiencies, expressed as a function of the amount of dye molecules bound to DNA, were found to be very different. Intercalation, however, appeared to favour the free-radical induction as the first strongly bound molecules were more efficient.

Induction of peroxide free radicals (detected by Electron Paramagnetic Resonance at 77 K) due to the photodynamic activity of proflavine was measured on bacteriophage phi X174 DNA either single-stranded (ss) as isolated from the virion, or double-stranded supercoiled (RFI) as isolated from the infected bacteria. Comparison was made with calf thymus DNA photosensitization. In order to use equivalent DNA-proflavine complexes, binding of the dye to the three DNA's was first determined under those conditions of high ionic strength favourable to the photodynamic reaction. Free radical induction was maximal for definite amounts of bound proflavine (which varied depending upon the DNA substrate) and at an ionic strength value of 0.5. The level of the maximal reaction increased in the following order: from phi Xss DNA to calf thymus DNA and finally to phi XRFI DNA. The conformation of the proflavine-DNA complex was thus a determinant for the efficiency of the photodynamic process. The ionic strength effect could not be explained by the evolution of the proflavine triplet state in irradiated proflavine-calf thymus DNA complexes.

Inactivation of øX174 bacteriophages as a function of the irradiation time in the near-UV and in the presence of triflupromazine (TFPZ), promazine (PZ), chlorpromazine (CPZ) or methoxypromazine (MTPZ) proceeds according to single hit kinetics. Acepromazine (ACPZ) has no significant activity. At low concentrations (0.1 mM) TFPZ and PZ are the most active compounds. Higher concentrations (up to 5 mM) result in a protective effect by these two compounds but cause increased inactivation rates in the case of MTPZ or CPZ. Photoinactivation mediated by TFPZ or CPZ increases the reversion frequency of a øXamber mutant. Neither MTPZ nor PZ sensitization induces mutagenesis. The effect of NaN3 on the phage inactivation rate varies depending upon both the sensitizer and the concentration of the quencher. Phage inactivation in an N2 atmosphere is measurable only in the presence of high concentrations of CPZ and MTPZ. The drugs do not show any selectivity for calf thymus DNA or bovine serum albumin, at least as measured by dialysis equilibrium experiments.

Bacteriophage phi X174 single-stranded DNA molecules were primed with five different restriction fragments and irradiated with visible light in the presence of proflavine. This photodamaged DNA was used as template for the in vitro complementary chain synthesis by E. coli DNA polymerase I (Klenow fragment). Chain terminations were observed by polyacrylamide gel electrophoresis of the synthesized products and localized by comparison with standard sequencing performed simultaneously on the untreated template. 90% of the chain terminations occurred one nucleotide before a guanine residue in the template strand. More than 80% of the sequenced guanine residues were blocking lesions demonstrating the absence of 'hot-spots' for the photodamaging effect of proflavine. At a defined position, the chain termination frequency increased linearly with the irradiation time and was directly influenced by the proflavine concentration present. An important part of lesions resulted from the action of singlet oxygen produced by excited proflavine as shown by the effect that both NaN3 and 2H2O exerted on the reaction. The induced blocking lesions must be important in vivo since no complete replicative forms could be extracted from cell infected with bacteriophages inactivated by 'proflavine and light' treatment.