- Email: [email protected]
Cytotoxic effects of the mycotoxin beauvericin to human cell lines of myeloid origin
Pharmacological Research 49 (2004) 73–77
Cytotoxic effects of the mycotoxin beauvericin to human cell lines of myeloid origin Lucia Calò a , Francesca Fornelli a , Rossella Ramires a , Saverio Nenna a , Alfredo Tursi a , Maria Filomena Caiaffa a,b , Luigi Macchia a,∗ a
Department of Clinical Immunology and Allergology, University of Bari, Policlinico, Piazza Giulio Cesare, Bari 70124, Italy b Medical Faculty, University of Foggia, Foggia 71100, Italy Accepted 23 July 2003
Abstract Beauvericin, a cyclic hexadepsipeptide of potential importance to the health of humans and domestic animals, has been reported to exert cytotoxic effects on several mammalian cell types and to induce apoptosis. We investigated the cytotoxicity of this compound to two human cell lines of myeloid origin: the monocytic lymphoma cells U-937 and the promyelocytic leukemia cells HL-60. In some experiments HL-60 cells partially differentiated towards the eosinophilic phenotype were also used. Cultures of U-937 cells and HL-60 cells in stationary phase were exposed to beauvericin at concentrations ranging from 100 nM to 300 M for periods of time of 4 and 24 h, respectively. The effects of beauvericin on cell viability were assessed by the Trypan blue exclusion method. In another set of experiments, performed with U-937 cells, the mycotoxin was included in the culture medium at passaging, in order to assess its possible effects on cell growth. Viability of both U-937 cells and HL-60 cells was not affected by beauvericin at concentrations up to 3 M, after 4 h exposure, whereas a steady decline was seen at higher concentrations. Similarly, after an exposure time of 24 h, a decline in viability was observed in cultures exposed to beauvericin at a concentration of 10 M or higher. Thus, 50% cytotoxic concentrations at 24 h of ∼ =30 M and ∼ =15 M were estimated for U-937 cells and HL-60 cells, respectively. Similar experiments were performed with cultures of HL-60 cells partially differentiated towards the eosinophilic phenotype, revealing that, in 4 h exposure experiments (but not in 24 h experiments), the viability of these cultures underwent a significantly less pronounced decline, in comparison to undifferentiated HL-60 cultures. Interestingly, when U-937 cells were allowed to proliferate in the presence of the mycotoxin, included in the culture medium at passaging, a substantial cytotoxicity was observed at lower concentrations, compared with prevalently resting, stationary phase cultures. Accordingly, a definite inhibition of the proliferative capability of the cells was detected. The information provided by this work may be useful in selecting appropriate myeloid cell models for the development of biossays aimed at detecting beauvericin (and, possibly, other mycotoxins) in foods and other commodities. © 2003 Elsevier Ltd. All rights reserved. Keywords: Cell line models; Cytotoxicity; Food contaminants; Mycotoxins; Myeloid lineage
1. Introduction Beauvericin, a cyclic hexadepsipeptide of alternating l-N-methylphenylalanyl and d-␣-hydroxyisovaleryl residues, is synthesized by various toxigenic fungi, including the entomopathogenic Beauveria bassiana, which has been proposed as a biological control agent, active on insect pests affecting agriculturally important plants [1]. Beauvericin is also produced by several phytopathogenic Fusarium species, parasitic to maize, wheat, rice and other ∗ Corresponding author. Tel.: +39-080-5478817; fax: +39-080-5428865. E-mail address: [email protected] (L. Macchia).
1043-6618/$ – see front matter © 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.phrs.2003.07.002
important commodities [2–8]. Therefore, this mycotoxin may accumulate in the environment and enter the food chain, thus exerting possibly important and still not well investigated effects on the health of humans and domestic animals [9]. Beauvericin, at concentrations in the lower micromolar range, was found to be cytotoxic in several mammalian cell line models, including the murine P815 mastocytoma cells, Yac-1 lymphoma cells and EL-4 thymoma cells [10], the rat mast cell-like RBL-1 cells, the simian fibroblastoid CV-1 cells and the human IARC/BL 41 cells (from Burkitt’s lymphoma), HeLa cells (from cervical carcinoma), and Hep G2 cells (from hepatoma) [9,11]. Moreover, beauvericin has been shown to induce apoptosis in mammalian cells
74
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
[9,10,12,13]. However, the overall available information regarding the myeloid lineage, which is particularly vulnerable to the action of environmental toxicants, including mycotoxins [14,15], is far from being complete. In this study, we focused on the cytotoxicity of beauvericin to two human cell lines of myeloid origin: the monocytic cells U-937 [16] and the promyelocytic leukemia cells HL-60 [17], two of the best studied models of myeloid cells. In some experiments, HL-60 cells partially differentiated towards the eosinophilic phenotype [18] were also used, in order to investigate possible differences in response to the mycotoxin, related to the degree of cellular differentiation.
2. Materials and methods
The mycotoxin cytotoxic effects were assessed either in terms of cell viability by the Trypan blue exclusion method, at 4 and 24 h, respectively, after inclusion of beauvericin in the culture medium, or in terms of effects on cell growth. Individual experiments were performed in duplicate (sometimes in triplicate) and results from the duplicate (or triplicate) samples were averaged. Results are presented as arithmetic means of homogeneous experiments with standard error. The 50% cytotoxic concentration (CC50 ) was defined as the beauvericin concentration that caused a 50% decrease in cell viability. The CC50 was estimated by interpolation of the relevant cytotoxicity curve. Different curves were compared with each other by analyzing differences between homologous curve points by non-parametric statistics (Wilcoxon rank sum test for independent data). Results were adjusted according to Bonferroni inequality.
2.1. Cell cultures The two cell lines were maintained at 37 ◦ C in a 5% CO2 , vapour-saturated atmosphere. U-937 cells were cultured in RPMI-1640 medium, with 10% (v/v) fetal calf serum, penicillin (100 U/ml), and streptomycin (100 g/ml). Medium and fetal calf serum were from Biochrom KG, Berlin; antibiotics were from Sigma, St. Louis, MO, USA. Early stationary phase cultures, with a typical density of 1.5–2 × 106 cells/ml and a typical viability of 90–95% were passaged every 3–4 days, with a seeding density of 3 × 105 cells/ml. HL-60 cells were grown in Iscove’s medium (EuroClone Ltd, Wetherby, UK), supplemented as above, with the further addition of 1.2 mM monothioglycerol (Sigma). In certain experiments, HL-60 cells partially differentiated towards the eosinophilic phenotype were used [18]. These cells were cultivated in Iscove’s medium at pH 7.8 with 0.5 mM butyric acid (also from Sigma). Other supplements were as above. Cultures (typically with 1.5–2 × 106 cells/ml density and 90% viability) were passaged every 3–4 days; density at seeding was adjusted at 3.5 × 105 cells/ml.
3. Results 3.1. Toxicity to U-937 cells Beauvericin elicited a significant decrease in cell viability in U-937 stationary phase cultures, which occurred in a doseand time-dependent fashion (Fig. 1). Thus, in cells exposed to the mycotoxin for 4 h, the viability underwent a clear decline at concentrations of 10 M and above (the viability of controls was 95%). After 24 h
2.2. Cytotoxicity assays Beauvericin cytotoxicity to U-937 and HL-60 cells (both undifferentiated and eosinophilic-differentiated cultures) was assayed primarily by exposing micro cultures (usually 300 l, in 96-well plates; Falcon, Bedford, MA, USA) in early stationary phase to beauvericin (Sigma), at concentrations ranging from 100 M to 300 M. In another experimental setting, developed in U-937 cells only, beauvericin was included in the culture medium at passaging and the cells were allowed to grow in the presence of the mycotoxin, in order to assess its effects on cell proliferation. Density and viability of the cultures were then determined at 24, 48, 72, and 96 h. Beauvericin was dissolved in dimethyl sulfoxide (ethanol in some experiments) and the final organic solvent concentration in the cell cultures was 1% (v/v). Control cultures were exposed to the solvent only.
Fig. 1. Effects of beauvericin on cell viability in U-937 cell cultures. The cells were exposed to the mycotoxin at concentrations ranging from 100 nM to 300 M, for periods of time of 4 h (circles) and 24 h (squares), respectively. Controls were exposed to the solvent only (1%, v/v). The mycotoxin caused a dose- and time-dependent decline in cell viability, as assessed by the Trypan blue exclusion method. Single points are averages of 11 independent experiments, with standard error. The CC50 at 24 h was interpolated.
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
exposure, while no noticeable effects on viability could be seen in cultures exposed to concentrations up to 3 M, a steady decline was observed at higher concentrations (43% at 30 M beauvericin, while viability in controls was 92%). Therefore, based on data collected from 11 independent experiments, the CC50 at 24 h could be estimated as ∼ =30 M (Fig. 1). With a smaller set of experiments (n = 4), we tested the cytotoxic effects on cell growth caused by the inclusion of beauvericin in the cell medium at passaging (Fig. 2A and B). Thus, while the presence of the mycotoxin bore no apparent consequences on cell growth up to a concentration of 1 M, 10 M beauvericin caused the cultures to become extinct within 24 h (Fig. 2A). Interestingly, inclusion of beauvericin 3 M in the medium determined a decline in the growth rate around day 3, followed by a recovery by day 4. Accordingly, the cytotoxicity curve under these conditions (i.e. exposing predominantly cycling cells to the mycotoxin) underwent a leftward shift (Fig. 2B).
Fig. 2. Effects of beauvericin on U-937 cell cultures (1 ml, 24-well plates) exposed at passaging. (A) Cultures were seeded in the presence of 1 M (squares), 3 M (triangles), and 10 M (inverted triangles) beauvericin. Control cultures were originated in the presence of the solvent only (1%, v/v; circles). Averages of triplicate observations. (B) Cytotoxicity curve for cells allowed to grow in the presence of the mycotoxin for 72 h (n = 4). Controls were exposed to the solvent only (1%, v/v).
75
3.2. Toxicity to HL-60 cells The results obtained with the human promyelocytic cell line HL-60 exposed in stationary phase were similar to those obtained in the U-937 cells (Fig. 3A). After an exposure time of 4 h, cytotoxic effects were noticed at a concentration of 10–30 M, with a decrease to 29% at 100 M and 12% at 300 M (viability in controls was 86%). Accordingly, after 24 h exposure cytotoxic effects started to become apparent at concentrations of approximately 3–10 M. Cell viability
Fig. 3. Effects of beauvericin on cell viability in HL-60 cells cultures. (A) Undifferentiated cells were exposed to the mycotoxin at concentrations ranging from 100 nM to 300 M, for periods of time of 4 h (circles) and 24 h (squares), respectively. Control cultures were exposed to the solvent only (1%, v/v). The mycotoxin caused a dose- and time-dependent decline in cell viability, as assessed by the Trypan blue exclusion method. Single points are averages of 14 independent experiments, with standard error. The CC50 at 24 h was interpolated. (B) Eosinophilic-differentiated cells were exposed to beauvericin at concentrations ranging from 100 nM to 300 M, for periods of time of 4 h (circles) and 24 h (squares), respectively. Also in this case, the mycotoxin caused a dose- and time-dependent decline in cell viability, as assessed by the Trypan blue exclusion method. Single points are averages of five independent experiments, with standard error. The CC50 at 24 h was also interpolated. Statistic evaluation of the differences between the 4 h curve in (A) and the same curve in (B), carried out by Wilcoxon rank sum test adjusted according to Bonferroni inequality, indicated that the differences between concentration points 10, 100, and 300 M were statistically significant. Differences between the 24 h curves, in (A) and (B), were not significant.
76
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
declined considerably at 30 M and was virtually equal to zero at 100 M. Thus, the CC50 at 24 h was calculated as ∼ =15 M (Fig. 3A). By using a similar approach, we tested the hypothesis that HL-60 cells partially differentiated into the eosinophilic phenotype might be affected in a different manner by the mycotoxin. In fact, the general response pattern in these cells was comparable to that observed in undifferentiated cells, with a dose- and time-dependent decline in cell viability and a CC50 at 24 h ∼ =20 M (Fig. 3B). However, statistical analysis, carried out by assessing the significance of differences between homologous points of cytotoxicity curves relevant to undifferentiated and eosinophilic-differentiated HL-60 cells, respectively, revealed that, in 4 h exposure experiments, but not in 24 h experiments, beauvericin cytotoxicity to eosinophilic-differentiated HL-60 cells was significantly less pronounced (Fig. 3A and B). The results of statistical inference of the comparisons between the nine subgroups of analytical evaluation, carried out with the Wilcoxon rank sum test, were adjusted according to Bonferroni inequality. Thus, in this analysis we used a z value of 0.05/9 = 0.006. Even under these conservative conditions, the differences in viability between the two models were statistically significant when cells were exposed to beauvericin at concentrations of 10 M and above for 4 h.
4. Discussion The hematopoietic bone marrow is particularly sensitive to a variety of cytotoxic agents, including antineoplastic drugs, cytotoxic antibiotics, radiation, benzene, and other toxicants. Evidence has been produced indicating that also mycotoxins can harm cells belonging to the myeloid lineage. Thus, gliotoxin, synthesized by several fungi, including Gliocladium virens, one of the most important commercially available fungal biological control agents, has been shown to greatly accelerate apoptosis in human granulocytes and eosinophils [14]. It was also shown that T2-toxin, another mycotoxin, at extremely low concentrations, can generate significant cytotoxic damage to bone marrow in mice, suggesting that even low level dietary contamination may be of concern [15]. Although beauvericin toxicity to mammalian cells has been known for more than a decade, a thorough appreciation of its potential myelotoxicity is still far from having been achieved, particularly in human models. Early indication of cytotoxic effects by beauvericin to mammalian cells of hematopoietic origin came from the work by Ojcius et al. [10], who in three murine cell line models, viz. P815 mastocytoma cells, Yac-1 lymphoma cells, and EL-4 thymoma cells, found a dose-dependent decrease in viability (assessed by release of 51 Cr-labeled proteins), 3 h after inclusion of beauvericin in the culture medium. Thus, a CC50 of 20 M (at 3 h) can be estimated on the basis of the data reported by these authors [10]. Another work, performed also in murine cell models, showed that macrophage-like
J774 cells and mouse peritoneal macrophages underwent a substantial decrease in viability when exposed to beauvericin for 24 h, as assessed by the Trypan blue exclusion method. A CC50 of approximately 10 M was assessed in both cell models [19]. Here, we focus on two commonly used and thoroughly studied human cell lines of myeloid origin, viz. the monocytoid U-937 cells and the promyelocytic HL-60 cells. In these two cell line models, we could show that distinct cytotoxic effects were caused by beauvericin, when the mycotoxin was included in the culture medium of prevalently resting, stationary phase cultures. Toxicity, assessed by the Trypan blue exclusion method, was dose- and time-dependent in both cell lines (Figs. 1 and 3A). A clear decline in viability was observed after 4 h incubation, which became even more pronounced when cultures of the two cell lines, respectively, were exposed to beauvericin for 24 h. Accordingly, 24 h CC50 values rather close to each other (∼ =30 for U-937 cells and ∼ =15 for undifferentiated HL-60 cells) were estimated, indicating that the two human cell lines (and, possibly, cell types) are probably affected in a similar manner by the mycotoxin. Moreover, the cytotoxic effects of beauvericin occurred in a rather restricted concentration range, since, after 24 h incubation, they were minimal at 3 M but already substantial at 30 M, as indicated by the steep profile of the cytotoxicity curves shown in Figs. 1 and 2A. These results are fully comparable with those of previous work performed by our group in other human cell lines of lymphatic origin, i.e. the human B-lymphocytic Burkitt’s lymphoma cell lines IARC/BL 28, IARC/BL 41, and IARC/BL 72 and other several non-malignant lymphoblastoid cell lines [9,11]. In the present study, we also made an attempt at assessing possible differences in beauvericin-induced cytotoxicity between promyelocytic (undifferentiated) HL-60 cells and eosinophilic-differentiated cells of the same line. In fact, although we were unable to detect any clear difference in cytotoxic response between the two models when the cells were incubated with the mycotoxin for 24 h, the decrease in viability caused by beauvericin over a shorter period of time (4 h) was significantly less pronounced in eosinophil-differentiated HL-60. These results may reflect potential mechanistic differences in short-term beauvericin toxicity, depending on the differentiation status of the cells and, therefore, be of interest in order to gain more insight into the cellular and molecular mechanisms underlying its toxicity to mammalian cells. Moreover, in the U-937 cell model, beauvericin was shown to affect in a more drastic way cells exposed at passaging (and actively cycling), compared to cells exposed in stationary phase (and, presumably, resting; Fig. 2A and B), suggesting that in naturally exposed organisms the putative toxic effects of the mycotoxin may be related to the division rate of the various tissues and may be more acute during development.
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
Furthermore, these data, obtained in these two human myeloid cell lines, are consistent with those generated by another study carried out by our group in the lepidopteran cell line SF-9, in which a dose- and time-dependent decrease in viability induced by beauvericin was shown and a CC50 value ∼ =12 was estimated. Thus, it seems that beauvericin affects invertebrate cells and mammalian cells in virtually the same fashion [20]. Finally, this study may also provide useful indication of suitable cellular models for the development of bioassays for detection of beauvericin and other mycotoxins in potentially contaminated commodities, currently the focus of substantial research efforts by the European Union [21]. Acknowledgements This work was supported by the European Union, V Framework Programme, Project: Risk Assessment of Fungal Biological Control Agents—RAFBCA (QLK1-2001-01391). The authors are thankful to Dr. Vincenzo Solfrizzi, M.D., Ph.D., Department of Geriatrics, University of Bari, Bari, Italy, for valuable advise concerning the statistic analysis. References [1] Wagner BL, Lewis LC. Colonization of corn, Zea mays, by the entomopathogenic fungus Beauveria bassiana. Appl Environ Microb 2000;66:3468–73. [2] Logrieco A, Moretti A, Altomare C, Bottalico A, Carbonell Torres E. Occurrence and toxicity of Fusarium subglutinans from Peruvian maize. Mycopathologia 1993;122:185–90. [3] Logrieco A, Moretti A, Ritieni A, Chelkowski J, Altomare C, Bottalico A, et al. Natural occurrence of beauvericin in preharvest Fusarium subglutinans infected corn ears in Poland. J Agr Food Chem 1993;41:2149–52. [4] Moretti A, Logrieco A, Bottalico A, Ritieni A, Randazzo G, Corda P. Beauvericin production by Fusarium subglutinans from different geographical areas. Mycol Res 1995;99:282–6. [5] Logrieco A, Moretti A, Castella G, Kostecki M, Golinski P, Ritieni A. Beauvericin production by Fusarium species. Appl Environ Microb 1998;64:3084–8. [6] Munkvold G, Stahr HM, Logrieco A, Moretti A, Ritieni A. Occurrence of fusaproliferin and beauvericin in Fusarium-contaminated livestock feed in Iowa. Appl Environ Microb 1998;64:3923–6.
77
[7] Logrieco A, Rizzo A, Ferracane R, Ritieni A. Occurrence of beauvericin and enniatins in wheat affected by Fusarium avenaceum head blight. Appl Environ Microb 2002;68:82–5. [8] Desjardin AE, Manandhar HK, Plattner RD, Manandhar GG, Poling SM, Maragos CM. Fusarium species from Nepalese rice and production of mycotoxins and gibberellic acid by selected species. Appl Environ Microb 2000;66:1020–5. [9] Logrieco A, Moretti A, Ritieni A, Caiaffa MF, Macchia L. Beauvericin: chemistry biology and significance. In: Upadhyay R, editor. Advances in microbial toxin research and its biotechnological exploitation. New York: Kluwer Academic; 2002. p. 23–30. [10] Ojcius DM, Zychlinsky A, Zheng LM, Young JD-E. Ionophoreinduced apoptosis: role of DNA fragmentation and calcium fluxes. Exp Cell Res 1991;197:43–9. [11] Macchia L, Di Paola R, Fornelli F, Nenna S, Moretti A, Napoletano R, Logrieco A, Caiaffa MF, Tursi A, Bottalico A. Cytotoxicity of beauvericin to mammalian cells. In: Proceedings of the International Seminar on Fusarium Mycotoxin, Taxonomy and Pathogenicity, 1995 May 9–13; Martina Franca, Bari, Italy. Abstract book. p. 72–3. [12] Que FG, Gores GJ, LaRusso NF. Development and initial application of an in vitro model of apoptosis in rodent cholangiocytes. Am J Phys 1997;272:G106–15. [13] Harnois DM, Que FG, Celli A, LaRusso NF, Gores GJ. Bcl-2 is overexpressed and alters the threshold for apoptosis in a cholangiocarcinoma cell line. Hepatology 1997;26:884–90. [14] Ward C, Chilvers ER, Lawson MF, Pryde JG, Fujihara S, Farrow SN, et al. NF-B activation is a critical regulator of human granulocyte apoptosis in vitro. J Biol Chem 1999;274:4309–18. [15] Faifer GC, Zabal O, Godoy HM. Further studies on the hematopoietic damage produced by a single dose of T-2 toxin in mice. Toxicology 1992;75:169–74. [16] Sundström C, Nilsson K. Establishment and characterization of a human histiocytic lymphoma cell line. Int J Cancer 1976;17:565–7. [17] Collins SJ, Gallo RC, Gallagher RE. Continuous growth and differentiation of human myeloid leukaemic cells in suspension culture. Nature 1977;270:347–9. [18] Fischkoff SA, Brown GE, Pollak A. Synthesis of eosinophilassociated enzymes in HL-60 promyelocytic leukemia cells. Blood 1986;68:185–92. [19] Tomoda H, Huang X-H, Cao J, Nishida H, Nagao R, Okuda S, et al. Inhibition of acyl-CoA: cholesterol acyltransferase activity by cyclodepsipeptide antibiotics. J Antibiot 1992;45:1626– 32. [20] Calo’ L, Fornelli F, Nenna S, Marrone D, Tursi A, Caiaffa MF, Macchia L. Beauvericin cytotoxicity to the invertebrate cell line SF-9. J Appl Genet 2003, in press. [21] RAFBCA. Risk assessment of fungal biological control agents. Summary of first annual reports. RAFBCA Web Site http://www. rafbca.com/download/report en.pdf.
Cytotoxic effects of the mycotoxin beauvericin to human cell lines of myeloid origin Lucia Calò a , Francesca Fornelli a , Rossella Ramires a , Saverio Nenna a , Alfredo Tursi a , Maria Filomena Caiaffa a,b , Luigi Macchia a,∗ a
Department of Clinical Immunology and Allergology, University of Bari, Policlinico, Piazza Giulio Cesare, Bari 70124, Italy b Medical Faculty, University of Foggia, Foggia 71100, Italy Accepted 23 July 2003
Abstract Beauvericin, a cyclic hexadepsipeptide of potential importance to the health of humans and domestic animals, has been reported to exert cytotoxic effects on several mammalian cell types and to induce apoptosis. We investigated the cytotoxicity of this compound to two human cell lines of myeloid origin: the monocytic lymphoma cells U-937 and the promyelocytic leukemia cells HL-60. In some experiments HL-60 cells partially differentiated towards the eosinophilic phenotype were also used. Cultures of U-937 cells and HL-60 cells in stationary phase were exposed to beauvericin at concentrations ranging from 100 nM to 300 M for periods of time of 4 and 24 h, respectively. The effects of beauvericin on cell viability were assessed by the Trypan blue exclusion method. In another set of experiments, performed with U-937 cells, the mycotoxin was included in the culture medium at passaging, in order to assess its possible effects on cell growth. Viability of both U-937 cells and HL-60 cells was not affected by beauvericin at concentrations up to 3 M, after 4 h exposure, whereas a steady decline was seen at higher concentrations. Similarly, after an exposure time of 24 h, a decline in viability was observed in cultures exposed to beauvericin at a concentration of 10 M or higher. Thus, 50% cytotoxic concentrations at 24 h of ∼ =30 M and ∼ =15 M were estimated for U-937 cells and HL-60 cells, respectively. Similar experiments were performed with cultures of HL-60 cells partially differentiated towards the eosinophilic phenotype, revealing that, in 4 h exposure experiments (but not in 24 h experiments), the viability of these cultures underwent a significantly less pronounced decline, in comparison to undifferentiated HL-60 cultures. Interestingly, when U-937 cells were allowed to proliferate in the presence of the mycotoxin, included in the culture medium at passaging, a substantial cytotoxicity was observed at lower concentrations, compared with prevalently resting, stationary phase cultures. Accordingly, a definite inhibition of the proliferative capability of the cells was detected. The information provided by this work may be useful in selecting appropriate myeloid cell models for the development of biossays aimed at detecting beauvericin (and, possibly, other mycotoxins) in foods and other commodities. © 2003 Elsevier Ltd. All rights reserved. Keywords: Cell line models; Cytotoxicity; Food contaminants; Mycotoxins; Myeloid lineage
1. Introduction Beauvericin, a cyclic hexadepsipeptide of alternating l-N-methylphenylalanyl and d-␣-hydroxyisovaleryl residues, is synthesized by various toxigenic fungi, including the entomopathogenic Beauveria bassiana, which has been proposed as a biological control agent, active on insect pests affecting agriculturally important plants [1]. Beauvericin is also produced by several phytopathogenic Fusarium species, parasitic to maize, wheat, rice and other ∗ Corresponding author. Tel.: +39-080-5478817; fax: +39-080-5428865. E-mail address: [email protected] (L. Macchia).
1043-6618/$ – see front matter © 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.phrs.2003.07.002
important commodities [2–8]. Therefore, this mycotoxin may accumulate in the environment and enter the food chain, thus exerting possibly important and still not well investigated effects on the health of humans and domestic animals [9]. Beauvericin, at concentrations in the lower micromolar range, was found to be cytotoxic in several mammalian cell line models, including the murine P815 mastocytoma cells, Yac-1 lymphoma cells and EL-4 thymoma cells [10], the rat mast cell-like RBL-1 cells, the simian fibroblastoid CV-1 cells and the human IARC/BL 41 cells (from Burkitt’s lymphoma), HeLa cells (from cervical carcinoma), and Hep G2 cells (from hepatoma) [9,11]. Moreover, beauvericin has been shown to induce apoptosis in mammalian cells
74
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
[9,10,12,13]. However, the overall available information regarding the myeloid lineage, which is particularly vulnerable to the action of environmental toxicants, including mycotoxins [14,15], is far from being complete. In this study, we focused on the cytotoxicity of beauvericin to two human cell lines of myeloid origin: the monocytic cells U-937 [16] and the promyelocytic leukemia cells HL-60 [17], two of the best studied models of myeloid cells. In some experiments, HL-60 cells partially differentiated towards the eosinophilic phenotype [18] were also used, in order to investigate possible differences in response to the mycotoxin, related to the degree of cellular differentiation.
2. Materials and methods
The mycotoxin cytotoxic effects were assessed either in terms of cell viability by the Trypan blue exclusion method, at 4 and 24 h, respectively, after inclusion of beauvericin in the culture medium, or in terms of effects on cell growth. Individual experiments were performed in duplicate (sometimes in triplicate) and results from the duplicate (or triplicate) samples were averaged. Results are presented as arithmetic means of homogeneous experiments with standard error. The 50% cytotoxic concentration (CC50 ) was defined as the beauvericin concentration that caused a 50% decrease in cell viability. The CC50 was estimated by interpolation of the relevant cytotoxicity curve. Different curves were compared with each other by analyzing differences between homologous curve points by non-parametric statistics (Wilcoxon rank sum test for independent data). Results were adjusted according to Bonferroni inequality.
2.1. Cell cultures The two cell lines were maintained at 37 ◦ C in a 5% CO2 , vapour-saturated atmosphere. U-937 cells were cultured in RPMI-1640 medium, with 10% (v/v) fetal calf serum, penicillin (100 U/ml), and streptomycin (100 g/ml). Medium and fetal calf serum were from Biochrom KG, Berlin; antibiotics were from Sigma, St. Louis, MO, USA. Early stationary phase cultures, with a typical density of 1.5–2 × 106 cells/ml and a typical viability of 90–95% were passaged every 3–4 days, with a seeding density of 3 × 105 cells/ml. HL-60 cells were grown in Iscove’s medium (EuroClone Ltd, Wetherby, UK), supplemented as above, with the further addition of 1.2 mM monothioglycerol (Sigma). In certain experiments, HL-60 cells partially differentiated towards the eosinophilic phenotype were used [18]. These cells were cultivated in Iscove’s medium at pH 7.8 with 0.5 mM butyric acid (also from Sigma). Other supplements were as above. Cultures (typically with 1.5–2 × 106 cells/ml density and 90% viability) were passaged every 3–4 days; density at seeding was adjusted at 3.5 × 105 cells/ml.
3. Results 3.1. Toxicity to U-937 cells Beauvericin elicited a significant decrease in cell viability in U-937 stationary phase cultures, which occurred in a doseand time-dependent fashion (Fig. 1). Thus, in cells exposed to the mycotoxin for 4 h, the viability underwent a clear decline at concentrations of 10 M and above (the viability of controls was 95%). After 24 h
2.2. Cytotoxicity assays Beauvericin cytotoxicity to U-937 and HL-60 cells (both undifferentiated and eosinophilic-differentiated cultures) was assayed primarily by exposing micro cultures (usually 300 l, in 96-well plates; Falcon, Bedford, MA, USA) in early stationary phase to beauvericin (Sigma), at concentrations ranging from 100 M to 300 M. In another experimental setting, developed in U-937 cells only, beauvericin was included in the culture medium at passaging and the cells were allowed to grow in the presence of the mycotoxin, in order to assess its effects on cell proliferation. Density and viability of the cultures were then determined at 24, 48, 72, and 96 h. Beauvericin was dissolved in dimethyl sulfoxide (ethanol in some experiments) and the final organic solvent concentration in the cell cultures was 1% (v/v). Control cultures were exposed to the solvent only.
Fig. 1. Effects of beauvericin on cell viability in U-937 cell cultures. The cells were exposed to the mycotoxin at concentrations ranging from 100 nM to 300 M, for periods of time of 4 h (circles) and 24 h (squares), respectively. Controls were exposed to the solvent only (1%, v/v). The mycotoxin caused a dose- and time-dependent decline in cell viability, as assessed by the Trypan blue exclusion method. Single points are averages of 11 independent experiments, with standard error. The CC50 at 24 h was interpolated.
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
exposure, while no noticeable effects on viability could be seen in cultures exposed to concentrations up to 3 M, a steady decline was observed at higher concentrations (43% at 30 M beauvericin, while viability in controls was 92%). Therefore, based on data collected from 11 independent experiments, the CC50 at 24 h could be estimated as ∼ =30 M (Fig. 1). With a smaller set of experiments (n = 4), we tested the cytotoxic effects on cell growth caused by the inclusion of beauvericin in the cell medium at passaging (Fig. 2A and B). Thus, while the presence of the mycotoxin bore no apparent consequences on cell growth up to a concentration of 1 M, 10 M beauvericin caused the cultures to become extinct within 24 h (Fig. 2A). Interestingly, inclusion of beauvericin 3 M in the medium determined a decline in the growth rate around day 3, followed by a recovery by day 4. Accordingly, the cytotoxicity curve under these conditions (i.e. exposing predominantly cycling cells to the mycotoxin) underwent a leftward shift (Fig. 2B).
Fig. 2. Effects of beauvericin on U-937 cell cultures (1 ml, 24-well plates) exposed at passaging. (A) Cultures were seeded in the presence of 1 M (squares), 3 M (triangles), and 10 M (inverted triangles) beauvericin. Control cultures were originated in the presence of the solvent only (1%, v/v; circles). Averages of triplicate observations. (B) Cytotoxicity curve for cells allowed to grow in the presence of the mycotoxin for 72 h (n = 4). Controls were exposed to the solvent only (1%, v/v).
75
3.2. Toxicity to HL-60 cells The results obtained with the human promyelocytic cell line HL-60 exposed in stationary phase were similar to those obtained in the U-937 cells (Fig. 3A). After an exposure time of 4 h, cytotoxic effects were noticed at a concentration of 10–30 M, with a decrease to 29% at 100 M and 12% at 300 M (viability in controls was 86%). Accordingly, after 24 h exposure cytotoxic effects started to become apparent at concentrations of approximately 3–10 M. Cell viability
Fig. 3. Effects of beauvericin on cell viability in HL-60 cells cultures. (A) Undifferentiated cells were exposed to the mycotoxin at concentrations ranging from 100 nM to 300 M, for periods of time of 4 h (circles) and 24 h (squares), respectively. Control cultures were exposed to the solvent only (1%, v/v). The mycotoxin caused a dose- and time-dependent decline in cell viability, as assessed by the Trypan blue exclusion method. Single points are averages of 14 independent experiments, with standard error. The CC50 at 24 h was interpolated. (B) Eosinophilic-differentiated cells were exposed to beauvericin at concentrations ranging from 100 nM to 300 M, for periods of time of 4 h (circles) and 24 h (squares), respectively. Also in this case, the mycotoxin caused a dose- and time-dependent decline in cell viability, as assessed by the Trypan blue exclusion method. Single points are averages of five independent experiments, with standard error. The CC50 at 24 h was also interpolated. Statistic evaluation of the differences between the 4 h curve in (A) and the same curve in (B), carried out by Wilcoxon rank sum test adjusted according to Bonferroni inequality, indicated that the differences between concentration points 10, 100, and 300 M were statistically significant. Differences between the 24 h curves, in (A) and (B), were not significant.
76
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
declined considerably at 30 M and was virtually equal to zero at 100 M. Thus, the CC50 at 24 h was calculated as ∼ =15 M (Fig. 3A). By using a similar approach, we tested the hypothesis that HL-60 cells partially differentiated into the eosinophilic phenotype might be affected in a different manner by the mycotoxin. In fact, the general response pattern in these cells was comparable to that observed in undifferentiated cells, with a dose- and time-dependent decline in cell viability and a CC50 at 24 h ∼ =20 M (Fig. 3B). However, statistical analysis, carried out by assessing the significance of differences between homologous points of cytotoxicity curves relevant to undifferentiated and eosinophilic-differentiated HL-60 cells, respectively, revealed that, in 4 h exposure experiments, but not in 24 h experiments, beauvericin cytotoxicity to eosinophilic-differentiated HL-60 cells was significantly less pronounced (Fig. 3A and B). The results of statistical inference of the comparisons between the nine subgroups of analytical evaluation, carried out with the Wilcoxon rank sum test, were adjusted according to Bonferroni inequality. Thus, in this analysis we used a z value of 0.05/9 = 0.006. Even under these conservative conditions, the differences in viability between the two models were statistically significant when cells were exposed to beauvericin at concentrations of 10 M and above for 4 h.
4. Discussion The hematopoietic bone marrow is particularly sensitive to a variety of cytotoxic agents, including antineoplastic drugs, cytotoxic antibiotics, radiation, benzene, and other toxicants. Evidence has been produced indicating that also mycotoxins can harm cells belonging to the myeloid lineage. Thus, gliotoxin, synthesized by several fungi, including Gliocladium virens, one of the most important commercially available fungal biological control agents, has been shown to greatly accelerate apoptosis in human granulocytes and eosinophils [14]. It was also shown that T2-toxin, another mycotoxin, at extremely low concentrations, can generate significant cytotoxic damage to bone marrow in mice, suggesting that even low level dietary contamination may be of concern [15]. Although beauvericin toxicity to mammalian cells has been known for more than a decade, a thorough appreciation of its potential myelotoxicity is still far from having been achieved, particularly in human models. Early indication of cytotoxic effects by beauvericin to mammalian cells of hematopoietic origin came from the work by Ojcius et al. [10], who in three murine cell line models, viz. P815 mastocytoma cells, Yac-1 lymphoma cells, and EL-4 thymoma cells, found a dose-dependent decrease in viability (assessed by release of 51 Cr-labeled proteins), 3 h after inclusion of beauvericin in the culture medium. Thus, a CC50 of 20 M (at 3 h) can be estimated on the basis of the data reported by these authors [10]. Another work, performed also in murine cell models, showed that macrophage-like
J774 cells and mouse peritoneal macrophages underwent a substantial decrease in viability when exposed to beauvericin for 24 h, as assessed by the Trypan blue exclusion method. A CC50 of approximately 10 M was assessed in both cell models [19]. Here, we focus on two commonly used and thoroughly studied human cell lines of myeloid origin, viz. the monocytoid U-937 cells and the promyelocytic HL-60 cells. In these two cell line models, we could show that distinct cytotoxic effects were caused by beauvericin, when the mycotoxin was included in the culture medium of prevalently resting, stationary phase cultures. Toxicity, assessed by the Trypan blue exclusion method, was dose- and time-dependent in both cell lines (Figs. 1 and 3A). A clear decline in viability was observed after 4 h incubation, which became even more pronounced when cultures of the two cell lines, respectively, were exposed to beauvericin for 24 h. Accordingly, 24 h CC50 values rather close to each other (∼ =30 for U-937 cells and ∼ =15 for undifferentiated HL-60 cells) were estimated, indicating that the two human cell lines (and, possibly, cell types) are probably affected in a similar manner by the mycotoxin. Moreover, the cytotoxic effects of beauvericin occurred in a rather restricted concentration range, since, after 24 h incubation, they were minimal at 3 M but already substantial at 30 M, as indicated by the steep profile of the cytotoxicity curves shown in Figs. 1 and 2A. These results are fully comparable with those of previous work performed by our group in other human cell lines of lymphatic origin, i.e. the human B-lymphocytic Burkitt’s lymphoma cell lines IARC/BL 28, IARC/BL 41, and IARC/BL 72 and other several non-malignant lymphoblastoid cell lines [9,11]. In the present study, we also made an attempt at assessing possible differences in beauvericin-induced cytotoxicity between promyelocytic (undifferentiated) HL-60 cells and eosinophilic-differentiated cells of the same line. In fact, although we were unable to detect any clear difference in cytotoxic response between the two models when the cells were incubated with the mycotoxin for 24 h, the decrease in viability caused by beauvericin over a shorter period of time (4 h) was significantly less pronounced in eosinophil-differentiated HL-60. These results may reflect potential mechanistic differences in short-term beauvericin toxicity, depending on the differentiation status of the cells and, therefore, be of interest in order to gain more insight into the cellular and molecular mechanisms underlying its toxicity to mammalian cells. Moreover, in the U-937 cell model, beauvericin was shown to affect in a more drastic way cells exposed at passaging (and actively cycling), compared to cells exposed in stationary phase (and, presumably, resting; Fig. 2A and B), suggesting that in naturally exposed organisms the putative toxic effects of the mycotoxin may be related to the division rate of the various tissues and may be more acute during development.
L. Cal`o et al. / Pharmacological Research 49 (2004) 73–77
Furthermore, these data, obtained in these two human myeloid cell lines, are consistent with those generated by another study carried out by our group in the lepidopteran cell line SF-9, in which a dose- and time-dependent decrease in viability induced by beauvericin was shown and a CC50 value ∼ =12 was estimated. Thus, it seems that beauvericin affects invertebrate cells and mammalian cells in virtually the same fashion [20]. Finally, this study may also provide useful indication of suitable cellular models for the development of bioassays for detection of beauvericin and other mycotoxins in potentially contaminated commodities, currently the focus of substantial research efforts by the European Union [21]. Acknowledgements This work was supported by the European Union, V Framework Programme, Project: Risk Assessment of Fungal Biological Control Agents—RAFBCA (QLK1-2001-01391). The authors are thankful to Dr. Vincenzo Solfrizzi, M.D., Ph.D., Department of Geriatrics, University of Bari, Bari, Italy, for valuable advise concerning the statistic analysis. References [1] Wagner BL, Lewis LC. Colonization of corn, Zea mays, by the entomopathogenic fungus Beauveria bassiana. Appl Environ Microb 2000;66:3468–73. [2] Logrieco A, Moretti A, Altomare C, Bottalico A, Carbonell Torres E. Occurrence and toxicity of Fusarium subglutinans from Peruvian maize. Mycopathologia 1993;122:185–90. [3] Logrieco A, Moretti A, Ritieni A, Chelkowski J, Altomare C, Bottalico A, et al. Natural occurrence of beauvericin in preharvest Fusarium subglutinans infected corn ears in Poland. J Agr Food Chem 1993;41:2149–52. [4] Moretti A, Logrieco A, Bottalico A, Ritieni A, Randazzo G, Corda P. Beauvericin production by Fusarium subglutinans from different geographical areas. Mycol Res 1995;99:282–6. [5] Logrieco A, Moretti A, Castella G, Kostecki M, Golinski P, Ritieni A. Beauvericin production by Fusarium species. Appl Environ Microb 1998;64:3084–8. [6] Munkvold G, Stahr HM, Logrieco A, Moretti A, Ritieni A. Occurrence of fusaproliferin and beauvericin in Fusarium-contaminated livestock feed in Iowa. Appl Environ Microb 1998;64:3923–6.
77
[7] Logrieco A, Rizzo A, Ferracane R, Ritieni A. Occurrence of beauvericin and enniatins in wheat affected by Fusarium avenaceum head blight. Appl Environ Microb 2002;68:82–5. [8] Desjardin AE, Manandhar HK, Plattner RD, Manandhar GG, Poling SM, Maragos CM. Fusarium species from Nepalese rice and production of mycotoxins and gibberellic acid by selected species. Appl Environ Microb 2000;66:1020–5. [9] Logrieco A, Moretti A, Ritieni A, Caiaffa MF, Macchia L. Beauvericin: chemistry biology and significance. In: Upadhyay R, editor. Advances in microbial toxin research and its biotechnological exploitation. New York: Kluwer Academic; 2002. p. 23–30. [10] Ojcius DM, Zychlinsky A, Zheng LM, Young JD-E. Ionophoreinduced apoptosis: role of DNA fragmentation and calcium fluxes. Exp Cell Res 1991;197:43–9. [11] Macchia L, Di Paola R, Fornelli F, Nenna S, Moretti A, Napoletano R, Logrieco A, Caiaffa MF, Tursi A, Bottalico A. Cytotoxicity of beauvericin to mammalian cells. In: Proceedings of the International Seminar on Fusarium Mycotoxin, Taxonomy and Pathogenicity, 1995 May 9–13; Martina Franca, Bari, Italy. Abstract book. p. 72–3. [12] Que FG, Gores GJ, LaRusso NF. Development and initial application of an in vitro model of apoptosis in rodent cholangiocytes. Am J Phys 1997;272:G106–15. [13] Harnois DM, Que FG, Celli A, LaRusso NF, Gores GJ. Bcl-2 is overexpressed and alters the threshold for apoptosis in a cholangiocarcinoma cell line. Hepatology 1997;26:884–90. [14] Ward C, Chilvers ER, Lawson MF, Pryde JG, Fujihara S, Farrow SN, et al. NF-B activation is a critical regulator of human granulocyte apoptosis in vitro. J Biol Chem 1999;274:4309–18. [15] Faifer GC, Zabal O, Godoy HM. Further studies on the hematopoietic damage produced by a single dose of T-2 toxin in mice. Toxicology 1992;75:169–74. [16] Sundström C, Nilsson K. Establishment and characterization of a human histiocytic lymphoma cell line. Int J Cancer 1976;17:565–7. [17] Collins SJ, Gallo RC, Gallagher RE. Continuous growth and differentiation of human myeloid leukaemic cells in suspension culture. Nature 1977;270:347–9. [18] Fischkoff SA, Brown GE, Pollak A. Synthesis of eosinophilassociated enzymes in HL-60 promyelocytic leukemia cells. Blood 1986;68:185–92. [19] Tomoda H, Huang X-H, Cao J, Nishida H, Nagao R, Okuda S, et al. Inhibition of acyl-CoA: cholesterol acyltransferase activity by cyclodepsipeptide antibiotics. J Antibiot 1992;45:1626– 32. [20] Calo’ L, Fornelli F, Nenna S, Marrone D, Tursi A, Caiaffa MF, Macchia L. Beauvericin cytotoxicity to the invertebrate cell line SF-9. J Appl Genet 2003, in press. [21] RAFBCA. Risk assessment of fungal biological control agents. Summary of first annual reports. RAFBCA Web Site http://www. rafbca.com/download/report en.pdf.