The bacterial colicin active against tumor cells in vitro and in vivo is verotoxin 1.

We have identified verotoxin 1 (VT1) as the active component within an antineoplastic bacteriocin preparation from Escherichia coli HSC10 studied over two decades. Recombinant VT1 can simulate the toxicity of anticancer proteins (ACP), and the antineoplastic activity of ACP (and VT1) was abrogated by treatment with anti-VT1 antibody. Similarly, VT1 mimics the protective effect of ACP in a murine metastatic fibrosarcoma model. Prior immunization with VT1 B subunit prevents the effect of VT1 or ACP in this model. The activity of ACP against a variety of human ovarian cell lines was mimicked by VT1, and multidrug-resistant variants were significantly hypersensitive. Primary ovarian tumors and metastases contain elevated levels of globotriaosylceramide compared with normal ovaries, and overlay of frozen tumor sections showed selective VT binding to tumor tissue and the lumen of invading blood vessels. Our contention that VT1 could provide an additional approach to the management of certain human neoplasms is discussed.

Augmentation of interleukin-6 (IL-6) expression in squamous carcinoma cells and normal human keratinocytes treated with recombinant anti-neoplastic protein (ACP).

A protein purified from Eschericheri coli has previously been shown to have cytotoxic effects on neoplastic cells of several lineages both in vitro and in vivo. Accordingly, this protein has been named anti-neoplastic protein (ACP). Although ACP kills neoplastic cells by inducing apoptosis, it has negligible effects on various normal cells. In addition to the direct cytotoxic effects of ACP on tumour cells, previous studies have shown that in vivo ACP increases tumouricidal activity of cytotoxic lymphocytes. We investigated whether cytokines from host or tumour cells play a part in the enhanced cellular immunity seen in ACP-treated tumour-bearing mice. Growth of normal human keratinocytres (KC) was not significantly affected by subnanogram amounts of ACP, however ACP dose-dependently killed KHT cells, a murine fibrosarcoma cell line (LD50 = 8 x 10(4) ng/cell). as well as the human squamous carcinoma cell line COLO-16 (LD50 = 2.5 x 10(-4) ng/cell). Testing purified ACP on cultures of normal keratinocytes and squamous carcinoma cell lines revealed that ACP could induce both mRNA and protein for interleukin-6 (IL-6). Messenger RNA for IL-6 increased dose-dependently 4h after treatment of COLO-16 squamous carcinoma cells with 10(-4) to 10(-2) ng/cell ACP. Maximal increment was 50-fold. Interleukin-6 message remained elevated up to 24h later in both normal keratinocytes and squamous carcinoma cultures treated with ACP. Conditioned supernates from these cultures were analysed by ELISA and found to have 4-fold higher levels of IL-6 protein than untreated cells after 4h. After 24h, IL-6 did not increase above the 4h level. Boiling of the ACP preparation showed that the cytokine induction was not due to contaminating lipopolysaccharide. The cytoxic effect of ACP on tumour cells in vitro was not due to IL-6 protein induction since neither recombinant IL-6, nor the other proinflammatory cytokines, IL-1 alpha or Tumour necrosis factor-alpha (TNF-alpha) (0.1-10ng/ml) were able to kill malignant cells. We demonstrated IL-6 gene induction by ACP in the squamous carcinoma lines as well as in normal KC. This suggests that the in vivo effectiveness of ACP against tumours may be due to stimulatory effects of IL-6 on host immunity.

Partially purified bacteriocin kills malignant cells by apoptosis: programmed cell death.

Bacterial proteins, partially purified bacteriocin (PPB), were investigated for their selective killing of malignant cells. It is shown here that upon PPB-cell interaction DNA fragmentation starts within one hour and peaks at 6 hrs. This process requires an on-going cellular metabolism. It is prevented by both actinomycin D, a DNA dependent RNA synthesis inhibitor, and cycloheximide, a protein synthesis inhibitor. We show here that the DNA fragmentation is triggered by PPB-cell membrane-receptor interaction which signals the activation of endogenous cellular endonucleases rather than actually penetrating the cell and interacting directly with the DNA, or serving as a nuclease itself. Thus, it is suggested that the cell death initiated by the lethal bacterial proteins, PPB, is a programmed, step-wise cell death involving apoptosis.

Further studies of the action of a partially purified bacteriocin against a murine fibrosarcoma.

We have reported previously that a partially purified bacteriocin (PPB) from Escherichia coli HSC10 is toxic to KHT cells growing in vivo as micrometastases but apparently has no activity against a tumor growing i.m. We report here experiments to investigate possible reasons for this difference. The PPB was shown to become less effective against micrometastases, initiated by i.v. injection of KHT cells, as the time between cell injection and PPB treatment increased. The kinetics of the loss of efficacy did not, however, correlate exactly with the growth kinetics of the nodules as assessed by survival following radiation treatment at different times after cell injection. This suggests the possibility of a diffusion limitation although it was found that s.c. injections of PPB were nearly as effective against micrometastases as i.p. injections. We also demonstrated that the lifetime of the majority of the toxic activity of PPB in vivo was relatively short (less than 1 day) and that the majority of its effect was not caused by stimulating macrophages to act against the tumor cells. The PPB was found to be cytotoxic to KHT cells in vitro but the effect was reduced at high cell density (approximately 10(6) cells/ml). The PPB did not induce an immune reaction against itself in C3H mice nor was it toxic to either bone marrow stem cells or jejunal crypt cells at doses which were effective against KHT micrometastases. We conclude that PPB may have potential as a cytotoxic agent to act against circulating tumor cells or very small deposits of tumor cells but is limited in its efficacy against larger tumor masses probably because of diffusional and/or cell density effects.

Fluorescence assay to monitor phagocytosis by blood-clot derived polymorphonuclear leucocytes. 1. Study of patients with diabetes and phagocytosis of different staphylococcal species.

A recently developed simple and rapid fluorescence assay of phagocytosis by polymorphonuclear leucocytes (PMNs) was used to compare the phagocytic activity of PMNs from diabetic patients and healthy people. Differences in phagocytosis have been described in these systems. Different bacterial strains were used to challenge the phagocytic capabilities of bacterial ingestion and killing for both patient and control groups. Staphylococcus epidermidis derived from clinical isolates was prone to faster ingestion and more rapid killing once the bacteria were intracellular, compared with S. epidermidis derived from normal skin. The pathogen, S. aureus, though ingested rapidly, was not so easily killed. PMNs from diabetics appeared to ingest S. aureus and kill ingested pathogenic bacteria not as effectively as PMNs from healthy volunteers. These difference were not found for non-pathogenic skin isolates of S. epidermidis.

Fluorescence assay to monitor phagocytosis by bloodclot derived polymorphonuclear leucocytes. 2. Study of patients with psoriasis and the effect of exposing leucocytes to psoriatic skin scales.

Comparisons of phagocytic parameters were carried out by a recently developed fluorescence test which is reproducible, simple and fast. Phagocytosis by polymorphonuclear leucocytes (PMNs) obtained from patients with psoriasis was compared with that of healthy individuals. Psoriatic skin scales, non-sterile and sterile, were tested for stimulatory effect on PMNs and compared with the effect of normal skin scrapings. Results confirm enhanced phagocytosis of bacteria by PMNs from patients with psoriasis over that of PMNs from healthy volunteers. Furthermore, the supernatant fluid from suspensions of psoriatic skin scales, non-sterile and sterile, stimulated PMNs activity.

Acute lymphoblastic leukemia of childhood monitored by bacteriocin and flowcytometry.

Bacteriocin and flowcytometric analysis of 106 blood samples from children with acute lymphoblastic leukemia were correlated with clinical stages of disease. Bacteriocins interacted selectively with malignant, and not with normal, lymphocytes causing cell cycle perturbation which was rapidly and objectively recorded by the flowcytometer. The patients were grouped as: (A) newly-diagnosed (15); (B) early induction (11); (C) remission with viral infection (7); (D) remission (64); (E) bone marrow relapsed (5); (F) extramedullary relapsed (3); (G) non-malignant pediatric controls (8). Bacteriocin reacted usually with groups A, B, C, E and not with groups D, F and G. Repeated testing correlated well with the clinical status. Blood from 7 patients in remission and from 3 normal individuals, each with transient viral infection, reacted with bacteriocin. A quantitative correlation between peripheral blood blasts or surface markers for ALL and bacteriocin reactivity was not established. Unexpected results were obtained only in 13% (false-positive 11% and false-negative 3%). This test can be recommended for preliminary diagnosis and possibly prognosis of lymphoblastic leukemia and provides means of monitoring progress during chemotherapy.

Recognition of T-cell murine leukemia by bacteriocin (colicin); correlation with transplantation experiments.

The presence of potential murine leukemia and overt leukemia cells in various organs at different phases of leukemogenesis was demonstrated by transplantation experiments and sensitivity to bacteriocin (colicin HSC10). Such correlation was found in three experimental models: AKR mice developing spontaneous T-cell leukemia and BL/6 mice infected with radiation leukemia virus variants inducing a high or low overt T-cell leukemia incidence. The sensitivity to bacteriocin was evaluated by testing the cell cycle perturbation following in-vitro incubation of lymphoid cells with colicin (or Tris buffer as controls) monitored by flow-cytometry. The analysis was based on measuring relative differences in fluorescence intensity of propidium iodide stained DNA in the individual cells. The interaction with colicin of leukemic cells and lymphoid cells containing potential leukemic cells (PLC) resulted in a reduction in the cell number of the G0/G1 and SG2M phases while cells accumulated in the "pre-G1" channels. In contrast, normal lymphoid cells exposed to bacteriocin did not show such changes in the DNA histograms. The distribution pattern of PLC in the thymus and spleen (in the models tested) obtained by transplantation studies coincided with sensitivity of spleen and thymus cells to colicin. However, in most instances, the PLC in the bone marrow were not recognized by colicin, but their leukemogenic potential was reduced following interaction with colicin as shown by PLC transplantation studies. It is thus suggested that colicin might be used for identification and eradication of transformed cells.

Purified colicin as cytotoxic agent of neoplasia: comparative study with crude colicin.

Purification of a bacteriocin, colicin, from Escherichia coli HSC10, is described. A 1,800 fold purified colicin was obtained and found to be an acidic polypeptide with a molecular weight of 82,000 daltons. The effect of colicin on bacterial and mammalian cells during the transition from a crude to a pure preparation is given. A turbidimetric assay for bacterial growth inhibition and 3H-thymidine uptake inhibition for measuring the effect on mammalian cells, was used. Colicin HSC10 caused DNA loss from the bacteriocin-sensitive mammalian cells, which increased with dose and time of exposure. Therefore, flow-cytometry, which detects DNA loss from the bacteriocin-affected mammalian cells, was also used to evaluate the bacteriocin potency. Similar quantitative results were obtained to those using 3H-thymidine uptake inhibition, in terms of microgram protein of pure colicin required to affect adversely 50% of the cells. The bacteriocins were found to retain their activity following freezing and thawing, both as crude and pure preparations, against both bacterial and mammalian cells.

Bacteriocin and flow cytometry in laboratory diagnosis of leukemic peripheral blood lymphocytes and bone marrow cells.

The bacteriocin, colicin HSC10, produced by Escherichia coli HSC10, was studied as a laboratory tool for detection and differentiation of leukemic from normal lymphocytes in human peripheral blood. Flow cytometry studies detected DNA loss in bacteriocin-affected cells by computerized histograms. Differential analysis is given for the peripheral blood of 26 individuals using bacteriocin, cytochemistry and surface markers. Sensitivity to colicin was detected in 10 (83%) of the 12 patients with chronic lymphocytic leukemias and other leukemias with morphologically immature lymphocytes. Cells were lost from the G0/G1 phase and accumulated in the 'pre-G1' channels of the histogram, indicative of cells with reduced DNA content. The lymphocytes of 14 normals, however, were not or only slightly affected by the bacteriocin (P less than 0.001). Similarly, normal bone marrow cells exposed to bacteriocin remained unaffected (P greater than 0.2). Thus, immaturity per se was not recognized by bacteriocin. The bacteriocin effect was more discriminatory than other laboratory tests reported here and in most cases differentiated malignant from normal cells.

Variants of an interspecies hybridoma with altered tumorigenicity and protective ability against mouse myeloma tumors.

Recent isolates of RX54-3 hybridoma cells (new cells) protect BALB/c mice against subsequent challenge with the tumorigenic myeloma parent cells used to construct this hybridoma. In contrast, hybridoma cells which have been maintained in tissue culture for long periods of time (old cells) are not protective. In the present study, we compared a number of properties of the new and old hybridoma cells and determined which line was more similar to the parent myeloma. We found that new hybridoma cells resembled myeloma cells in: (a) possessing A- and C-type viral particles on transmission electron microscopy and a relatively smooth surface on scanning electron microscopy; (b) being sensitive to a hypotonic solution containing the dye propidium iodide; (c) having similar DNA histograms on flow cytometric analysis; (d) being sensitive to the bacteriocin colicin HSC 10; and (e) being tumorigenic in nude mice. In contrast, old hybridoma cells differed in all of these characteristics from new hybridoma and myeloma cells. Therefore, in order to protect against challenge with the tumorigenic myeloma parent, hybridoma cells must retain properties of that parent.

Development and standardization of a cytotoxic micro-assay for detection of enterotoxins: survey of enterotoxins from Escherichia coli of infant origin.

The development of a practical cytotoxic micro-assay for detection of enterotoxins in crude bacterial lysates of E. coli and other Gram-negative bacteria, is described. This quantitative assay is based on growth inhibition of mouse-fibroblasts, maintained in suspension or by inhibition of uptake of DNA precursors. Guidelines for performing the assay and evaluating the results by statistical considerations, are described. The choice of a relatively cheap medium and a suitable number of target cells to achieve cell doubling in 24 h is given. The concentrations of proteins in the crude lysates from strains of different origin, are not of equal potency; a predetermined but different protein concentration for strains from infants, adults or porcine origin, are recommended for detection of the toxins and for achieving reproducible results. Production of toxic proteins is enhanced by mitomycin C in toxigenic strains and not in non-toxigenic strains. Screening of a limited number of lysates from E. coli strains originating from infants and a comparison of the cytotoxicity of several known toxigenic and non-toxigenic strains from human adults and porcine origin, are presented.

Bacteriocin typing by leakage of ultraviolet light-absorbing material.

A rapid and reproducible method of bacteriocin typing is described based on leakage of ultraviolet light-absorbing material (UVAM), detectable in supernatants of bacteriocin-sensitive cultures, by means of a spectrophotometer. The prerequisites for reproducible results, with nonsignificant fluctuations in standard error of the mean, are: a set of standardized bacteriocins, produced under defined conditions and of determined strength. These must interact with the unknown bacterial culture in suspension and at a given ratio in order to achieve an optimal multiplicity of interaction. Pyocin and colicin typing by the "scrape and streak" technique of Gillies (J. Hyg. 62:1-10, 1963) was compared with the UVAM leakage method in 275 tests; the two tests were found to be in good agreement for the strains tested.

Bacterial proteinaceous products (bacteriocins) as cytotoxic agents of neoplasia.

Several bacteriocins, bacterial proteinaceous antibiotics, are shown to markedly inhibit the division of various established (neoplastic) mammalian cell lines. The bacteriocins tested originated from Escherichia coli, Pseudomonas aeruginosa, Vibrio cholerae, and Vibrio eltor. Using exponentially growing L60T mouse fibroblasts, the inhibitory effect was concentration dependent, and a growth inhibitory unit, equivalent to cytotoxic index 50, was established. Expression of toxicity as a function of duration of exposure to pyocin required 3 to 4 hr. DNA synthesis was inhibited and reflected the effects on growth inhibition. Maximal sensitivity to the bacteriocin was observed prior to mitosis in the G2 phase of the cell cycle.