The thymus extract Thymex-L potentiates the retinoic acid-induced differentiation of the human myeloid leukemia cell line HL-60.

The human promyelocytic cell line HL-60 can be differentiated with retinoic acid (RA) along the granulocytic pathway. Numerous studies have identified many synergistic combinations of RA with cytostatics, cytokines and other inducers. A combination of RA with the crude thymus extract Thymex-L increased differentiation of HL-60 cells as confirmed by two functional assays and morphology, whereas the extract itself did not show any effect. The functional markers phagocytosis-associated chemiluminescence and nitroblue tetrazolium reduction were more enhanced (up to 4-fold with 1000 micrograms/ml Thymex-L) than morphology. The effect was found over a wide RA concentration range (10(-11) - 10(-6) M) and was dependent on extract concentration. The half-maximal induction of both functional markers was reached at 400 micrograms/ml. To achieve the same effect with the combination in comparison with RA alone, an RA dose reduction of about 100-fold was estimated. The effect was also seen when the cells were pretreated with the thymus extract for two days. The enhancement of RA action by Thymex-L was not correlated with an increase of extracellular or intracellular RA concentration. The active compound in Thymex-L is heat stable and bigger than 5 kDa as confirmed by gelfiltration. The defined thymus peptides thymosin alpha 1, prothymosin alpha 1 and thymopentin were unable to synergistically enhance HL-60 differentiation. These data suggest that the treatment with a thymus extract can increase the sensitivity of HL-60 cells for RA. This may have clinical implications.

Bovine lung conditioned medium as a source of both human and murine colony-stimulating factor.

Medium conditioned by bovine lung tissue (BLCM) contains a factor (CSFBL) that can stimulate either murine or human bone marrow cells to form colonies. CSFBL was partially purified by gel filtration and DEAE-cellulose chromatography. The specific activity was determined using both target systems. In all stages of purification the specific activity on human bone marrow was about 2 times greater than on mouse bone marrow. Determination of the molecular weight of CSFBL by gel filtration and comparison with mouse lung CSF (CSFML) revealed that both are similar in size, having an apparent molecular weight of 29,000 d. CSFBL stimulates only granulocyte colonies, neither macrophage nor mixed colonies could be detected in either bone marrow assay. These results and the ease of preparation recommend BLCM as a suitable starting for large scale preparation of CSF active on human cells.

Glutathione: an in vitro granulopoiesis inhibitor at nanomolar concentration, isolated from calf spleen.

We isolated a peptide from calf spleen that specifically inhibits murine granulopoiesis in vitro. Human T-lymphocyte colony growth was not affected. Formation of erythroid colony-forming units (CFU-E) was only slightly inhibited at a concentration 1000-fold higher compared to granulocyte-macrophage colony-forming cell (GM-CFC) assay. After high performance liquid chromatography purification we isolated a monomer (reduced peptide) showing the inhibitory activity and a dimer (oxidized peptide) with stimulatory activity. The isolated peptide was identified as glutathione. The reduced form of this well-known tripeptide (GSH) specifically inhibits murine granulopoiesis in vitro at 10(-7)-10(-8) M concentration. Oxidized glutathione (GSSG) stimulates GM-CFC formation; the extent of stimulation depends on the colony-stimulating factor concentration. The striking finding that GSH and GSSG modulate granulopoiesis at nanomolar concentrations requires further studies on the molecular interaction between either peptide with GM-CSF and its receptor protein.

A micromethod to culture granulocyte colonies from human bone marrow in agar-containing glass capillaries.

A micromethod was developed to grow granulocytic colonies from human bone marrow in agar-containing glass capillary tubes. Treatment of the bone marrow samples and the culture conditions (type and quantity of serum, CSF, cell seeding density) were optimized. Up to 60 colonies were obtained from 3.5 x 10(4) nucleated cells seeded into 50 microliter of total incubation medium/capillary with horse serum (13%) and leukocyte and partially purified bovine lung conditioned medium as CSF (17 and 3%, respectively). The micromethod requires less culture materials (about 1/20), cells, CSF and less time for colony counting, but higher cell densities for seeding, resulting in an increased sensitivity for drug or factor testing. Colony morphology can be easily examined. The micromethod offers further advantages, e.g. quantitation by light scattering densitometry, and hence seems suitable for clinical investigations.

Bromelain reversibly inhibits invasive properties of glioma cells.

Bromelain is an aqueous extract from pineapple stem that contains proteinases and exhibits pleiotropic therapeutic effects, i.e., antiedematous, antiinflammatory, antimetastatic, antithrombotic, and fibrinolytic activities. In this study, we tested bromelain's effects on glioma cells to assess whether bromelain could be a potential contributor to new antiinvasive strategies for gliomas. Several complementary assays demonstrated that bromelain significantly and reversibly reduced glioma cell adhesion, migration, and invasion without affecting cell viability, even after treatment periods extending over several months. Immunohistochemistry and immunoblotting experiments demonstrated that alpha3 and beta1 integrin subunits and hyaluronan receptor CD44 protein levels were reduced within 24 hours of bromelain treatment. These effects were not reflected at the RNA level because RNA profiling did not show any significant effects on gene expression. Interestingly, metabolic labelling with 35-S methionine demonstrated that de novo protein synthesis was greatly attenuated by bromelain, in a reversible manner. By using a transactivating signaling assay, we found that CRE-mediated signaling processes were suppressed. These results indicate that bromelain exerts its antiinvasive effects by proteolysis, signaling cascades, and translational attenuation.

Expression and modulation of the Lewis x antigen (CD15) on the T cell line Molt-4.

The T cell lines Molt-4 and H9 exhibited a characteristic distribution of the cell adhesion molecule Lewis x (CD15, lacto-N-fucopentanose III) showing an unusually broad peak by flow cytometry ranging from cells without CD15 to cells with high expression. The cytokines IL-1, IL-2, IFN-beta, IFN-gamma, and TNF-alpha, known to activate T cells, did not affect CD15 expression. However, phorbol myristate acetate and the thymic peptide extract Thymex-L were able to enhance both the number of CD15-positive cells and the median fluorescence. The effects of both inducers were dose- and time-dependent. An additive or synergistic effect was not seen. These data suggest that phorbol esters and distinct thymic peptides can modulate the expression of the cell surface antigen CD15.

Bromelain protease F9 reduces the CD44 mediated adhesion of human peripheral blood lymphocytes to human umbilical vein endothelial cells.

The thiol protease bromelain has been shown to remove T-cell CD44 molecules from lymphocytes and to affect T-cell activation. We investigated the effect of a highly purified bromelain protease F9 (F9) on the adhesion of peripheral blood lymphocytes (PBL) to human umbilical vein endothelial cells (HUVEC). Preincubation of the lymphocytes with F9 reduced the adherence to about 20% of unstimulated and to about 30% of phorbol-dibutyrate (P(Bu)2) stimulated lymphocytes. Using flow cytometry, both crude bromelain and protease F9 reduced the expression of CD44, but not of LFA-1, on PBL. F9 was about 10 times more active than crude bromelain; at 2.5 micrograms/ml of F9 about 97% inhibition of CD44 expression was found. A mAb against CD44 was tested and found to block the F9-induced decrease in PBL-binding to HUVEC. The results indicate that F9 selectively decreases the CD44 mediated binding of PBL to HUVEC.

Prothymosin alpha1 effects on IL-2-induced expression of LFA-1 on lymphocytes and their adhesion to human umbilical vein endothelial cells.

Prothymosin alpha1 (Pro alpha1) is known to stimulate in vitro and in vivo natural killer (NK) and lymphokine (IL-2)-activated killer (LAK) cells against tumor cells. In this process, LAK cells first adhere to endothelial cells in vivo, raising the question whether Pro alpha1 affects this interaction as well. The binding ability of peripheral blood lymphocytes (PBL) to human umbilical vein endothelial cells (HUVEC) was increased by incubation with IL-2 in a concentration-dependent manner, reaching a maximal value at 20U/ml IL-2. Although Pro alpha1 alone was without any stimulating effect, it significantly increased PBL binding to unstimulated HUVECs in combination with suboptimal IL-2 (5 and 10 U/ml). The combination of Pro alpha1 (1 microg/ml) and 5 U/ml or 10 U/ml IL-2 is as effective as 10 U/ml or 20 U/ml IL-2 alone. This Pro alpha1 effect on IL-2-activated lymphocytes was found to be augmented on IL-1 or tumor necrosis factor (TNF)-alpha-activated endothelial cells. Analyzing the effect of Pro alpha1 on IL-2-activated lymphocytes by flow cytometry revealed an increase of CD16, CD56, and CD18 surface marker expression, whereas CD3, CD11a/b, CD49d, and CD54 were not affected. In conclusion, Pro alpha1 functions as a mediator of the adhesion of IL-2-activated lymphocytes to HUVECs.

A rapid and sensitive fluorometric screening assay using YO-PRO-1 to quantify tumour cell invasion through Matrigel.

A new quantitative assay for the study of tumour cell invasion in vitro is described. Employing the novel fluorescent dye YO-PRO-1, cells that penetrate Matrigel-coated transwells are counted on the basis of dye-bound cellular nucleic acid content. Following transmigration, the cells in the lower compartments are lysed by freezing in water. After a brief incubation with YO-PRO-1, nucleic acid or DNA content is measured as fluorescence intensity in 96-well microplates and quantitated by a cell- or DNA-calibration curve. Using standard curves, a linear relationship between fluorescence intensity and cell number was found in the range tested (from 100 to 80 000 cells). The mean relative intra- and inter-assay variability of the cell quantitation in this range was 3.5 and 4.2%, respectively. When applied to Matrigel invasion studies, as few as 400 cells could be counted. The quantitation could be performed within 3 h. HCT 116, MDA MB 231 and HT 29 cells were investigated as examples of tumour cells with different invasive abilities in the 48-h Matrigel invasion assay. Using YO-PRO-1, 6.5 +/- 0.6% invasive HCT 116 cells and 52.6 +/- 4.5% MDA MB 231 cells (percentage of the inoculated cell population) were measured. HT 29 cells were practically non-invasive. These results were confirmed by visual scoring of DAPI-stained nuclei. In conclusion, the main advantages of the assay are its sensitive, reproducible and rapid quantitation of tumour cell invasion in vitro and the applicability to extended sample numbers by measuring in 96-well microplates.

Lymphokine-activated killer cells: determination of their tumor cytolytic capacity by a clonogenic microassay using agar capillaries.

A lymphokine activated killer (LAK) cell assay has been developed using a clonogenic microassay in agar-containing capillaries with KB tumor target cells. The assay avoids the problems of the commonly used 51Cr release assay and mimics physiological conditions more closely. The assay procedure has been optimized, resulting in the following conditions: LAK cells are generated by incubating nonadherent peripheral blood mononuclear cells from normal donors with 20 U/ml interleukin-2 for 3 days and cocultivated with 10(4) KB human squamous carcinoma cells/ml at 5:1, 10:1, and 20:1 effector:target ratios for 24 h. The cocultivation mixture is then seeded into agar-containing glass capillaries, allowing undamaged tumor cells to form colonies. The colony number is proportional to the number of tumor cells seeded. The present microassay requires up to 90% less cells and agent quantities compared with other clonogenic assays. It thus provides a useful tool to quantitate LAK cell activity and its alteration by immunomodulatory agents.

In vitro culture of lymphocyte colonies in agar capillary tubes after PHA-stimulation.

Human peripheral lymphocytes were stimulated with phytohemagglutinin (PHA) in liquid culture, suspended in agar and incubated in glass capillary tubes. Compact colonies of lymphocytes were found growing along the tube bottom in a buffer film, while compact clusters and rare diffuse colonies were observed inside the agar. Several parameters affecting the clonal growth were studied and optimized: PHA-dose, agar concentration, gel length (volume), quantity and density of seeded cells per capillary and gel length. Colony yield mainly depends on the seeded-cell density with a sharp optimum at 2 X 10(5) cells/ml irrespective of gel length; higher cell densities reduce the colony yield, suggesting that colony growth is the result of both stimulatory and inhibitory factors produced by cooperating cells. Following the daily clonal growth was only possible with undisturbed tubes; the number of colonies steadily increased from day 2 until day 7. Densitometric colony scanning is possible, yet problematic. Colony yield (plating efficiency is 10--50-fold higher in agar capillaries than in the usual Petri dishes. An additional advantage is that the capillaries provide a basis for a simple and reliable assay system for determining regulatory factors of lymphocyte proliferation (including chalones).

A colorimetric immunoassay for the detection of E-cadherin and carcinoembryonic antigen (CEA) expression on human colon carcinoma cell lines in vitro.

An immunohistochemical assay, using 96-well microtiter plates and based on the avidin-biotin peroxidase reaction, was developed for the quantitation of E-cadherin and CEA expression on HT-29 and SW620 colon carcinoma cells. The optical density of the generated dye was measured after solubilization with alkaline SDS solution. The staining procedure was evaluated with respect to reproducibility (coefficient of variation and intra-class correlation coefficient). On HT-29 cells the level of agreement for both E-cadherin and CEA were substantial, on SW620 cells almost perfect. The method allows testing compounds for their differentiation inducing capacity in screening programmes on the basis of protein marker expression.

Induction of the c-myc but not the cH-ras promoter by platinum compounds.

The recombinant plasmids p324, p330 and p323 carrying the aminoglycoside phosphotransferase (aph) gene and 5' flanking c-myc sequences linked to the reporter gene chloramphenicol acetyl-transferase (cat) were introduced into the mouse erythroleukaemic cell line F412B2TK- and stable transfectants resistant to geneticin G418 were obtained. The effects of cis-platin and two novel platinum compounds, D19466 (lobaplatin) and D17872, on c-myc promoter regions were studied using a large range of drug concentrations and correlated with cytotoxicity data. It was found that cis-platin and D19466 show a similar pattern of cytotoxicity and activation of c-myc promoter-driven cat gene expression with maximum effect (increased expression of 7.4- and 8.1-fold, respectively) at a concentration of 5 x 10(-5) M, while the less cytotoxic D17872 only slightly activates cat expression at the same concentration. However, when the F412B2TK- cell line was transfected with a plasmid carrying 5' flanking sequences of the c-Hras1 gene with promoter/enhancer function, linked to the cat reporter, no similar inductive effect was observed with any of the platinum drugs used. These data suggest that platinum compounds and possibly other DNA-damaging agents may specifically influence the expression of certain genes.

Docetaxel treatment of HT-29 colon carcinoma cells reinforces the adhesion and immunocytotoxicity of peripheral blood lymphocytes in vitro.

In vitro effects of docetaxel on the human colon carcinoma cell line HT-29 were studied with respect to the expression of different adhesion and surface marker molecules, the adhesion and immunocytotoxicity of peripheral blood lymphocytes and the secretion of IFN-gamma and TNF-alpha. Docetaxel, in a low concentration range (1-3x10-9 M), increased the expression of the adhesion molecules LFA-3, ICAM-1, CD44s, CD44v6, CD15, CD13 and VLA-4/5/6 on the tumor cells. Unstimulated and interleukin-2 (IL-2) activated killer (LAK) cells showed a better adherence to docetaxel treated HT-29 cells than to untreated cells. In neutralization experiments, anti-LFA-3, -CD44v6, -CD15, -VLA -4 and anti-CD13 mAb reduced the lymphocyte adhesion to untreated and docetaxel treated cells at different degrees, while CEA mAb increased the adhesion. Unstimulated and IL-2 activated lymphocytes exhibited significantly higher cytotoxicities against docetaxel treated cells than against untreated HT-29 cells. Unstimulated and IL-2 stimulated lymphocytes secreted more TNF-alpha and IFN-gamma when cocultured with docetaxel treated HT-29 cells than with untreated cells. These results suggest, that the increased lymphocyte mediated cytotoxicity against docetaxel treated HT-29 colon carcinoma cells may reflect an immunological process coupled with induction of differentiation, that may contribute to the clinically known cytostatic effects of the drug.

Tributyrin induces growth inhibitory and differentiating effects on HT-29 colon cancer cells in vitro.

Tributyrin (TB) is a prodrug of butyrate known to induce tumor cells to differentiate. We examined its effects on cell growth, viability, cellular morphology and differentiation of HT-29 colon cancer cells in vitro, as reflected by the expression of CEA, E-cadherin and the induction of alkaline phosphatase activity. TB, applied in a stable emulsion, inhibited tumor cell proliferation in a reversible and dose-dependent manner (0.5-4 mM) with significant morphological changes. The IC50 value of TB was 1 mM after 6 days. For comparison, sodium butyrate, applied in equimolar concentration, inhibited cell growth with an IC50 value of 2.2 mM. TB treatment at concentrations of 0.5 mM and 2 mM resulted in an increase of the doubling times by 18% and 160%, respectively, without any effects on cell viability. By a colorimetric immunoassay, 1.5 mM TB induced the expression of both CEA and E-cadherin by about 260% and 100%, respectively. Furthermore, the activity of the brush border enzyme alkaline phosphatase was enhanced in a dose-dependent manner, up to 60-fold at the maximum of 2 mM TB. Our results show that TB is more active than butyrate in suppressing cell growth and concomitantly promoting differentiation of HT-29 colon cancer cells. Hence it may be a promising candidate for clinical therapeutic protocols and merits further investigation.

Stimulation of clonal tumor cell growth in vitro by inhibiting the serum polyamine oxidase activity.

Several tumors are characterized by elevated levels of polyamines involved in vital cell proliferation processes. Polyamine oxidases (PAO), present in ruminant and particularily in fetal calf serum (FCS), degrade polyamines to polyaminoaldehydes and other products that inhibit cell proliferation. Since most in vitro assays for cloning tumor stem cells use FCS as an essential supplement of the nutrient media, we examined the effects of specifically inhibiting the PAO activity on the clonal growth of leukemic cells and the following normal lymphocytes: the W 25 rat chloroleukemia, the M1 mouse myeloblastic and the L 1210 rat lymphoblastic leukemia, a primary human acute myeloblastic leukemia (AML) and acute lymphocytic leukemia (ALL) as well as normal human PHA-stimulated lymphocytes. In the presence od horse serum, nontoxic doses of the PAO inhibitor 1-hydroxybenzyloxyamine did not affect colony growth of either cell type. However, in the presence of FCS, clonal growth of W 25, ALL, AML, and PHA lymphocytes was significantly stimulated by the enzyme inhibitor. Our data suggest (a) that poor cell proliferation of several tumors in vitro may result from the reaction of polyamines (from cells) and PAO (from serum), (b) that this can be easily tested by means of a specific PAO inhibitor, and (c) that the growth conditions can be optimized by adding nontoxic doses of the enzyme inhibitor or by exchanging FCS for another serum.

Comparison of cytostatic sensitivities of L 1210 cells and human stimulated lymphocytes in three cell proliferation assays.

Three methods of measuring cell proliferation, viz., cellular H-thymidine uptake, counting of cells in suspension, and counting of colonies of cells grown in agar contained in glass capillaries, were compared by studying cell growth kinetics using the L 1210 cell line. We found the agar colony culture method to be most suitable and methodologically most advantageous. Using these cytokinetic models, we investigated the differential sensitivities of exponential and stationary phase L 1210 cells and normal, human, PHA-stimulated, peripheral T-lymphocytes to methotrexate, cytosine arabinoside, azathioprine, and a partially purified lymphocyte chalone preparation. L 1210 cells in exponential growth showed a higher drug sensitivity to all the agents tested than those in stationary growth. Normal, human T-lymphocytes exhibited less sensitivity to the tested agents. We found the agar culture to be more than twice as sensitive as the suspension culture and up to 8-fold more sensitive than the 3H-thymidine uptake method.

Synergistic interactions between differentiation-inducing agents in inhibiting the proliferation of HL-60 human myeloid leukaemia cells in clonogenic micro assays.

All-trans-retinoic acid, hexamethylene bisacetamide and 5-azacytidine are inducers of granulocytic differentiation of HL-60 human myeloid leukaemic cells, which eventually leads to inhibition of cell proliferation. The effect of graded concentrations of all-trans-retinoic acid (RA) (1 nM-1 microM), hexamethylene bisacetamide (HMBA) (0.5-4 mM) and/or 5-azacytidine (5azaC) (1 nM-1 mM), alone and in combination with each other on colony formation and growth of HL-60 cells was studied in agar capillary clonogenic micro assays in order to identify new potential therapeutic regimens for elderly patients with acute myeloid leukaemia. ED90 concentrations, inducing 90% inhibition of colony formation for RA, HMBA and 5azaC, were 128 nM, 2.7 mM and 40 microM, respectively. The drug interactions between these differentiating agents were analysed by Berenbaum's general algebraic solution. The combinations: RA + HMBA, 5azaC + HMBA and RA + 5azaC were significantly synergistic in inhibiting HL-60 colony formation. Their interaction indices were 0.62, 0.83, and 0.97, respectively, at a specific effect level of 15%. The addition of 1 mM HMBA to 100 nM 5azaC- and 1 nM RA-treated cultures significantly increased the colony-formation inhibition from only 2.6% and 7.0% to 46.4%, and 43.1%, respectively. Also, HMBA showed marked synergism with RA and 5azaC in inhibiting colony growth. The interaction indices (I) of HMBA + RA and HMBA + 5azaC were 0.013 and 0.009, respectively, at the same specific level of 15%. Moreover, the triple combination of RA + HMBA + 5azaC showed synergism in inhibiting both the colony formation (I = 0.7) and colony growth (I = 0.4) at the same specific level of 15%. Since RA, HMBA and 5azaC were effective when administered alone in phase I clinical trials of myeloid leukaemic patients, their synergistic combinations could provide shorter and less toxic courses of treatment in elderly myeloid leukaemic patients. I is less than 1, = 1 or greater than 1 in synergistic, additive or antagonistic interactions, respectively.

Synergistic interactions between interferon beta and carboplatin on SK-MEL 28 human melanoma cell growth inhibition in vitro.

Carboplatin and interferon beta (IFN beta) were tested alone and in combination for their antiproliferative activity on the human melanoma cell line SK-MEL 28 in vitro. Cells were incubated for 4 days in the presence of carboplatin (0.1 mM and 0.1 microM) and interferon beta (5 pM and 5 nM) and cell growth inhibition was determined by the sulphorhodamin B assay. The antiproliferative effects of the drug combinations were analysed using Berenbaum's hyperplane theorem to determine additive, synergistic and antagonistic effects. IFN beta was found to be 10,000 times more active in inhibiting cell growth of SK-MEL 28 cells than carboplatin on the basis of IC50 values (IFN beta: IC50 = 1.24 nM, carboplatin: IC50 = 18.2 microM). The addition of IFN beta at 0.5 nM reduced the IC50 value of carboplatin 18.0-fold; with IFN beta at 0.05 nM a dose reduction of 1.84 was measured. At the carboplatin: IFN beta molar concentration ratios of 2000:1 and 6000:1, interaction indices (I) of 0.66 and 0.83 were determined respectively, indicating synergistic interactions between the two drugs. At higher carboplatin: IFN beta molar ratios (20,000:1 and 60,000:1) an additive interaction was observed (I = 1.07 and 1.20). However, further in vitro studies with several melanoma cell lines are necessary to evaluate the potential effectiveness of the drug combination of carboplatin and IFN beta for eventual clinical utilisation.

Recombinant human interferons alpha, beta and gamma reduce the antiproliferative action of cytarabine in K562 human myeloid leukaemia clonogenic cells.

The effects of recombinant human interferons alpha, beta and gamma (IFN) on the antiproliferative activity of cytarabine in K562 human myeloid leukaemia clonogenic cells were studied in an agar capillary microassay. The addition of IFN-alpha did not affect the antiproliferative activity of cytarabine in K562 cultures treated with low concentrations of cytarabine (10-50 nM), whereas in those treated with high concentrations (100-150 nM) IFN alpha increased the IC50 of cytarabine on day 5 from 102 nM to 214 nM, i.e., cytarabine combined with IFN alpha was about two-fold less potent than cytarabine alone. Similarly, low concentrations of IFN beta and IFN gamma did not affect the antiproliferative activity of cytarabine on K562 colonies, but high concentrations of these two interferons: 4 x 10(3) U/ml and 10(4) U/ml respectively, increased the IC50 of cytarabine on day 5 to 304 nM and to 316 nM respectively, i.e. cytarabine combined with IFN beta or IFN gamma was about threefold less potent than cytarabine alone. The evaluation of the present negative interactions of interferons with cytarabine is warranted in fresh cells from myeloid leukaemia patients in primary culture.