Effects of trichothecene mycotoxins on eukaryotic cells: A review

Abstract

The major products of the trichothecene mycotoxin biosynthetic pathway produced in a species– and sometimes isolate–specific manner by cereal–pathogenic Fusarium fungi include T–2 toxin, diacetoxyscirpenol, deoxynivalenol and nivalenol. This paper briefly reviews the major effects of such trichothecenes on the gross morphology, cytology and molecular signalling within eukaryotic cells. The gross toxic effects of select trichothecenes on animals include growth retardation, reduced ovarian function and reproductive disorders, immuno–compromization, feed refusal and vomiting. The phytotoxic effects of deoxynivalenol on plants can be summarized as growth retardation, inhibition of seedling and green plant regeneration. Trichothecenes are now recognized as having multiple inhibitory effects on eukaryote cells, including inhibition of protein, DNA and RNA synthesis, inhibition of mitochondrial function, effects on cell division and membrane effects. In animal cells, they induce apoptosis, a programmed cell death response. Current knowledge about the eukaryotic signal transduction cascades and downstream gene products activated by trichothecenes is limited, especially in plants. In mammalian cells, certain trichothecenes trigger a ribotoxic stress response and activate mitogen–activated protein kinases. DON mediates the inflammatory response by modulating the binding activities of specific transcription factors and subsequently inducing cytokine gene expression. Several genes are up–regulated in wheat in response to trichothecene mycotoxins; the significance, if any, of these genes in the host response to trichothecenes has yet to be elucidated.

Keywords:

• Fusarium
• deoxynivalenol (DON)
• diacetoxyscirpenol (DAS)
• T–2 toxin
• plant
• mammalian

Introduction

Trichothecenes are secondary fungal metabolites that are harmful to human and animal health causing a range of acute and chronic symptoms (D’Mello and Macdonald 1997, D’Mello et al. 1999). This class of mycotoxins comprises a very large family of related chemicals produced by a number of taxonomically unrelated fungal genera, including Fusarium, Mycothecium, Trichoderma, Trichothecium, Stachybotrys, Verticimonosporium, and Cephalosporium (Ueno 1985). The first trichothecene type compound discovered was the antifungal trichothecin isolated from a culture of Tricothecium roseum (Ueno 1985). Fusarium species are probably the most commonly cited and amongst the most prolific of the trichothecene-producing fungi. Many trichothecene-producing Fusarium species are common causal organisms of Fusarium head blight (FHB, also known as ‘scab’ or head blight), foot rot and root rot disease of cereals (including F. culmorum, F. poae and F. graminearum). Consequently, infection and colonization of cereal heads by such fungi not only affects yield, but can also result in trichothecene contamination of grain.

Many of the known trichothecenes are toxic to cells and the gross effects of trichothecenes on eukaryotic organisms are a manifestation of their effect at the cellular level. The cellular effects of these toxins on eukaryotic cells are not yet fully understood, especially in plants. In this chapter, we will introduce the trichothecenes with respect to their diversity, chemistry and relative importance. We will briefly review the gross toxic effects of trichothecenes on eukaryotic cells, and thereafter concentrate our review on the known cytological and molecular impacts of such mycotoxins on eukaryotic cells.

Trichothecenes

Over 150 trichothecenes and trichothecene derivatives have been isolated and characterized (Gutleb et al. 2002). They are all non-volatile, low-molecular-weight sesquiterpene epoxides (Wannemacher et al. 2000). They share a tricyclic nucleus named trichothecene and usually contain an epoxide at C-12 and C-13, which are essential for toxicity (Desjardins et al. 1993). Most trichothecenes also have a C-9,10 double bond, which is important for their toxicity (Ehrlich and Daigle 1987). Their chemical structure varies in both the position and the number of hydroxylations, as well as in the position, number and complexity of esterifications (Desjardins et al. 1993).

The trichothecene skeleton is chemically stable and the 12,13-epoxide ring is stable to nucleophilic attack. Furthermore, trichothecenes are heat stable and are not degraded during normal food processing (Eriksen 2003) or autoclaving (Wannemacher et al. 2000). Trichothecenes are also stable at neutral and acidic pH (Ueno 1987) and consequently, they are not hydrolysed in the stomach after ingestion (Eriksen 2003).

Trichothecenes are divided into four groups (types A–D) according to both their chemical properties and their producer fungi (Ueno et al. 1973). Only a few of the known trichothecenes (all type A or B) seem to be of importance with respect to their presence in crops; for the majority of trichothecenes, there is little evidence of their occurrence under field conditions. Fusaria produce type A and B trichothecenes. Types A and B are distinguished by the presence of an oxygen or carbonyl functional group at the C-8 position, respectively. The major products of the trichothecene biosynthetic pathway are known as T-2 toxin and diacetoxyscirpenol (DAS) (type A) and deoxynivalenol (DON) and nivalenol (NIV) (type B) (Ueno 1985). Fusarium species differ in their trichothecene profiles; DON is one of the primary trichothecene metabolites found in wheat and is commonly produced by several phytopathogenic Fusarium species including F. graminearum and F. culmorum. Producers of type B trichothecenes can be grouped into two chemotypes based on whether they produce DON (or acetylated derivatives) or NIV (Lee et al. 2001, Chandler et al. 2003). Mycothecium verrucaria and other genera of fungi produce type C and D trichothecenes. The type C are characterized by a second epoxide function at C-7,8 or C-9,10 position of the pentane ring common to all trichothecenes, and include crotocin produced by Trichotheceum roseum and Cephalosporium crotocinigerum (Ueno 1985). Type D trichothecenes are those containing a macrocyclic ring between C-4 and C-5 position the pentane ring with two ester linkages, and include verrucarin produced by Myrothecium verrucaria, satratoxin produced by Stachybortys atra and roridin produced by Myrothecium roridum and Cylindrocarpon (Ueno 1985).

Gross toxic effects of trichothecenes on animals

The gross toxic effects of trichothecenes on animals have been extensively reviewed (Beasley 1989, Rotter et al. 1996). Trichothecenes are toxic to all tested animal species (SCF 2002), T-2 toxin generally being the most toxic. However, the sensitivity towards different trichothecenes varies considerably amongst organisms (as reviewed by SCF 2002).

The gross toxic effects of select trichothecenes on animals include growth retardation, reduced ovarian function and reproductive disorders, immuno-compromization, feed refusal and vomiting. These toxic effects are mediated via several mechanisms (SCF 2002). Trichothecenes such as T-2 toxin, HT-2 toxin, DON and NIV retard growth, albeit in a non-specific manner, and as a consequence of earlier toxic effects such as inhibition of protein synthesis and cell proliferation rather than as a direct mycotoxic effect. T-2 reduced body weight gain in pigs (Weaver et al. 1978) and DON and NIV caused growth retardation in mice (Ryu et al. 1988, Ohtsubo et al. 1989). Reproductive defects in mice caused by trichothecenes include embryo or foetal toxicity caused by T-2 toxin (Rousseaux et al. 1986), increased postnatal mortality caused by DON (Khera et al. 1984) and intrauterine growth retardation caused by NIV (Ito et al. 1988). DON increased the susceptibility of mice to infections (Tryphonas et al. 1986). DAS exposure caused suppression in chicken macrophage phagocitic function (Qureshi et al. 1998). Lymphocytes are more sensitive to T-2 toxin than other cell types such as kidney cells, and this corresponds well with the data from in vivo experiments that showed that trichothecenes act as immunosuppressive agents (Holladay et al. 1993). Lymphoid cells and fibroblasts were the most sensitive cell types to DON (Reubel et al. 1987, Charoenpornsook et al. 1998).

A special feature of DON toxicity is the characteristic induction of vomiting (DON is also called vomitoxin) and feed refusal, as seen in pigs, or delayed gastric emptying and feed refusal, as observed in rats and mice (SCF 1999). The emetic effect is thought to be mediated through affection of the serotonergic activity in the central nervous system or via peripheral actions on serotinin receptors (SCF 1999).

Gross toxic effects of trichothecenes on plants

Shimada and Otani (1990) reported that NIV was 10–24-fold less toxic than DON to seedlings of seven Japanese wheat cultivars. Eudes et al. (2000) demonstrated the phytotoxicity of the six trichothecenes DON, 3-acetyldesoxynivalenol (3-ADON), NIV, DAS, T-2 toxin and HT2 in four wheat cultivars Norseman, Katepwa, Toropi and Nyu-Bay. They found that DON and 3-ADON were more toxic than T-2 toxin, HT-2 and DAS, and that those toxins inhibited wheat coleoptile elongation at very low concentrations. Wakulinski (1989) used seedlings of three varieties of winter wheat, Grana, Emika, and SMH-68-4 to assess the phytotoxic effects of six Fusarium metabolites. He showed that T-2 toxin was less phytotoxic than DON, but more toxic than DAS. He also found DON and 3-ADON to be the most potent inhibitors of germination, and of root and leaf mass increases. Figure 1 shows the effect of DON on germinating seeds of wheat.

Figure 1. Inhibitory effect of (A) 20 ppm DON as compared with (B) water on wheat cv. Avalon seedling growth (48 h post-treatment).

Trichothecenes are phytotoxic, and at very low level cause wilting, chlorosis, necrosis, and other symptoms in a wide variety of plants (Cutler 1988). The phytotoxic effects of DON on plants can be summarized as growth retardation, inhibition of seedling and green plant regeneration (Bruins et al. 1993, McLean 1996). Bruins et al. showed that DON inhibited growth of wheat seedlings, coleoptile segments, anther-derived callus, anther-derived embryos and green plant regeneration. DON also inhibited wheat root growth during germination (Shimada and Otani 1990). Wang and Miller (1998) observed that DON strongly inhibited coleoptile growth of wheat cultivars with different levels of resistance to FHB, however they found cultivars resistant to FHB were more tolerant to DON and 3-ADON than those that were susceptible.

Little is known about the combined (additive or antagonistic) effects of trichothecenes on plant or indeed animal tissue. A few studies reported dose additivity and antagonism among T-2 toxin, DON and NIV in vitro (yeast, lymphocyte and Vero cells); in in vivo animal studies antagonism was only observed between DON and T-2 toxin (for a synopsis of this research, see SCF 2002).

Cellular effects of trichothecenes on eukaryotes

Trichothecenes are now recognized as having multiple inhibitory effects on eukaryote cells. Table I lists some of the effects of trichothecene mycotoxins on animal and plant cell functions. These include inhibition of protein, DNA and RNA synthesis, inhibition of mitochondrial function, effects on cell division and membrane effects (McLean 1996, Miller and Ewen 1997). In animal cells, they induce apoptosis, a programmed cell death (PCD) response (Ihara et al. 1998, Shifrin and Anderson 1999, Yang, et al. 2000, Minervini et al. 2004).

Table I. Examples of the cellular effects of trichothecenes on animals and plants.

Effects	Tissue or cell type		References
	Animals	Plants
Inhibition of protein synthesis	Reticulocyte cell-free system, Vero cells, rat lymphocytes	Wheat seedling leaves, maize leaf and kernel	Carter et al. (1976), Thompson and Wannemacher (1986), Miller and Arnison (1986), Casale and Hart (1988)
Inhibition of RNA and DNA synthesis	Human (HeLa) cells, various mice and rat tissues, human erythroleukaemia cell line	–	Thompson and Wannemacher (1986, 1990), Minervini et al. (2004)
Alteration of membrane structure	L-6 myoblast, liver, spleen, kidney, thymus and bone marrow or rats	Wheat spikes and kernels	Bunner and Morris (1988), Suneja et al. (1989), Kang and Buchenauer (1999)
Mitochondria function	Rat hepatocytes, a human erythroleukaemia cell line	Inhibition of enzyme activity in maize	Pace et al. (1988), Cosette and Miller (1995), Minervini et al. (2004)
Apoptosis or programmed cell death	Jurkat human T-lymphoid cells, human leukemic cells, mice leukocytes	–	Shifrin and Anderson (1999), Yang et al. (2000) Islam and Pestka (2003)
Activation of MAPK	Jurkat human T-lymphoid cells, human leukameic cells.	–	Shifrin and Anderson (1999), Yang et al. (2000)
Activation of the cytokine	Murine CD4^+ T cells, human lumphocytes	–	Dong et al. (1994), Ouyang et al. (1995), Meky et al. (2001)

A number of cytotoxicity assays have been developed or adapted to study trichothecene toxicity at the cellular level. A neutral red (NR) (3-amino-7-dimethylamino-2-methyl-phenazine hydrochloride) cell-viability assay has been successfully used to determine the cytotoxic effect of trichothecenes (Babich and Borenfreund 1991, Shokri et al. 2000). This assay is based on the incorporation of neutral red, a supravital dye, into the lysosomes of viable cells. Compounds that injure the plasma or lysosomal membrane show decreased uptake and retention of the dye. Analysis of the incorporation of ³H-thymidine into DNA is a very sensitive method to study trichothecene-induced inhibition of DNA synthesis in cultured cells (Porcher et al. 1987, Charoenpornsook et al. 1998). However, this is a time consuming technique and unfortunately carries the usual risks associated with a radioactive method. 5-Bromo-2^′-deoxyuridine (BrdU) has been used as a rapid and non-radioactive immunocytochemical assay to determine DNA synthesis in proliferating cells (Dolbeare 1995). Deoxynucleotidyl transferase (TdT)-mediated fluorescein-dUTP nick end-labelling (TUNEL) method highlights DNA cleavage (Gavrieli et al. 1992) and has been used to study the effect of trichothecenes on DNA fragmentation (Miura et al. 1998, Rachid et al. 2000). 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) is a yellow salt that is reduced to dark purple formazan crystals by mitochondrial enzymes in living cells (Slater et al. 1963). MTT has been used in several studies to assess the cytotoxicity of mycotoxins (Reubel et al. 1987, Holt et al. 1988, Visconti et al. 1991). Trichothecene induced alteration of the plasma membrane can be determined by quantifying enzymes released in the cell culture medium by damaged cells. Lactate dehydrogenase (LDH) is a ubiquitous cytoplasmic enzyme and the amount released into the cell culture medium or present in intact cells has been used to evaluate mycotoxin-mediated plasma membrane damage (Bunner and Morris, 1988, Charoenpornsook et al. 1998).

Inhibition of protein synthesis

Evidence suggests that the primary toxic effects of trichothecenes may be facilitated by their action as potent inhibitors of protein synthesis in most eukaryotes (McLaughlin et al. 1977); the biochemical basis of trichothecene toxicity is non-competitive inhibition of protein synthesis (Cole and Cox 1981) (Table I). They interfere with the active site of peptidyl transferase on ribosomes, and inhibit the initiation, elongation or termination step of protein synthesis. From the data on polysomal breakdown, the trichothecenes were classified into two types: type I inhibited the initiation step of protein synthesis (T type) and type II inhibited the elongation-termination step (ET type) (McLaughlin et al. 1977). Type I includes T-2 toxin, NIV, verrucarin A and fusarenon-X, and the ET type includes DON, crotocin, verrucarol and fusarenon-X (Ueno 1985). One target organelle of trichothecenes has been identified as the 60S subunit of mammalian ribosomes, and a good correlation was observed between the inhibition of protein synthesis by trichothecenes and the affinity of trichodermin to ribosomes (Wei et al. 1974). Trichothecene-producing fungi have an altered ribosomal protein L3 (Rpl3) that protects the fungi from the primary effect of trichothecenes on protein synthesis, and a semidominant mutation in the Saccharomyces cerevisiae TCM1 gene (that encodes the Rpl3) conferred resistance to trichodermin (Fried and Warner 1981). An altered Rpl3 protein expressed in wheat and maize increased resistance to FHB and ear rot, respectively; Harris and Gleddie (2001) found that the Rpl3 gene conferred reduced sensitivity to DON in cells, tissues, and protoplasts of transgenic tobacco.

Trichothecenes differ in their toxicity towards eukaryotic cells due to chemical features of the side chains (Ueno 1985). Thompson and Wannemacher (1986) tested nineteen 12,13-epoxytrichothecene mycotoxins for their relative capabilities to inhibit protein synthesis in Vero cells and rat spleen lymphocytes. They showed a clear chemical structure dependence of the capacity of trichothecenes and trichothecenes derivatives to inhibit protein synthesis. Their study proved that the most potent protein synthesis inhibitors were T-2, verrucarin A and roridin A; HT-2, neosolaniol or DAS showed reduced protein synthesis inhibition; T-2 triol, T-2 tetraol, 15-monoacetyl DAS, scirpentriol, fusarenon X and DON further weakened their effect; Verrucarol and deoxyverrucarol, reduced their effectiveness by over a thousand-fold compared with the most potent mycotoxins. Addition of side groups resulted in reduced effectiveness only when an acetyl group was added to the carbon 3 position of T-2 (acetyl T-2) and DON (3-acetyl DON) or on substitution of an epoxide across the 9,10 carbons of DAS. When the in vitro effects of the mycotoxins were compared with results from whole animal lethality tests, several of the trichothecenes were weak inhibitors of protein synthesis in vitro but had in vivo toxicities similar to that of T-2 toxin. Thus, the in vitro cell response of a given trichothecene is not always an accurate predictor of toxicity in whole animals. In vivo, inhibition of protein synthesis has been demonstrated in cells from bone marrow, spleen and thymus (Rosenstein and Lafarge-Frayssinet 1983, Thompson and Wannemacher 1990). In plants, low levels of DON inhibited protein synthesis in wheat (Miller and Arnison 1986). Bandurska et al. (1994) shown that the growth of Fusarium culmorum-susceptible wheat seedlings (winter, Alba, Almari, Begra, Kamila, Liwilla and Parada) was inhibited by DON. In this study DON caused an increase in the free proline levels in seedling leaves: this increase was dependent on cultivar and DON concentration, and the authors suggested that this increase in free proline levels may be attributed to decreased incorporation of the amino acid into protein. However, they added that increases in proline might reflect a plant defence mechanism in response to pathogen invasion or pathogen virulence factors.

Inhibition of RNA and DNA synthesis

In mammalian cells, trichothecenes inhibit both RNA and DNA synthesis (Ueno 1985) (Table I). However, in vitro experiments revealed no direct effects of fusarenon-X on either DNA-dependent RNA polymerases of rat liver nuclei or on nuclear ribonuclease H of rat liver and the protozoa Tetrahymena pyriformis (Tashiro et al. 1979). Substantial inhibition of ribonucleic acid (RNA) synthesis (86% inhibition) by trichothecenes mycotoxins was observed in human (HeLa) cells, while T-2 toxin showed minor effects (15% inhibition) on RNA synthesis in Vero cells (Thompson and Wannemacher 1986).

DNA synthesis is strongly inhibited in various types of cells that are exposed to trichothecenes (McLauglin et al. 1977). In mice or rat treated with trichothecene mycotoxins, DNA synthesis was suppressed in all tissues studied, although to a lesser degree than protein synthesis (Thompson and Wannemacher 1990). The pattern by which DNA synthesis is inhibited by trichothecenes is consistent with the primary effect of these toxins on protein synthesis. In appropriate cell models, for the most part trichothecenes demonstrated neither mutagenic activity nor the capacity to damage DNA (Busby and Wogan 1981). Thus, the inhibition of nucleic acid synthesis is presumed to be a secondary effect of trichothecenes (Ueno 1985).

Inhibition of mitosis

Inhibition of DNA synthesis, cell growth and mitosis are consequences of the interference of trichothecenes in the cell growth cycle (Table I). NIV and fusarenon-X affected HeLa cells in every phase of the growth cycle (Ohtsubo et al. 1968). NIV induced a G1 block before the beginning of the S phase of mitosis, and a G2 block just before mitosis (Ohtsubo et al. 1968).

Packa (1991) studied the effect of DON and DAS on actively dividing cells of rye, wheat, triticale and field bean. In all plants, treatment with DON decreased the mitotic index (MI) of germinated roots; mitosis was strongly inhibited in wheat and field bean. Germinating caryopses of rye and field beans, and to a lesser extent, triticale and wheat, were sensitive to DAS, resulting in a decrease in the MI. A decline in the MI and relative division rate was also observed in dividing root tip cells of onion treated with DON or T-2 toxin (Rahman et al. 1993). Chromosome and nuclear abnormalities caused by DON and DAS included spiralization of metaphase chromosomes, retarded equatorial chromosome arrangement, asymmetric arrangement of daughter chromosomes and numerous ana- and telophase bridges (Packa 1991). Such cytogenetic changes reflect the disturbed functioning of the karyokinetic spindle which leads to the formation of cells containing various number of chromosomes; part of the genetic material may be eliminated in the form of micronuclei or chromosomes may be duplicated. These cytological effects may be caused by a reduction in the synthesis of proteins involved in microtubule formation (Packa 1991).

Effect on cell membrane and organelles

Since trichothecene mycotoxins are amphophilic molecules, they may exert their cytotoxicity by acting on the cell membranes (MacLean 1996) (Table I). T-2 toxin alters various myoblast cell membrane functions, such as calcium efflux, rubidium uptake, or residual cellular lactate dehydrogenase activity (Bunner and Morris 1988).

Lipid peroxidation was increased in the liver, spleen, kidney, thymus and bone marrow of rats treated with a single oral dose of T-2 toxin (Suneja et al. 1989). These observations led to the suggestion that trichothecenes might induce some alterations in membrane structure, which consequently stimulates lipid peroxidation. Tsuchida et al. (1984) suggested that in animal cells (rat liver cells) T-2 toxin may be taken up by the cell membrane as an analogue of cholesterol. Toxin-stimulated alteration in mitochondria membranes contributes to the effects on cellular energetics and cellular cytotoxicity.

Once trichothecenes cross the plasma membrane barrier, they enter the cell, where they can interact with a number of targets, including ribosome (McLaughlin et al. 1977) and mitochondria (Pace et al. 1988). Trichothecenes inhibit mitocondrial function in vivo and in vitro. In a bovine kidney cell line, the mitocondrial activity was reduced after treatment with T-2 toxin (Holt et al. 1988). In vitro, low concentrations of T-2 toxin resulted in a 50% inhibition of mitocondrial protein synthesis in rat liver cells (Pace et al. 1988). T-2 toxin also inhibit electron transport activity in rat liver cells (Pace et al. 1983) as suggested by the in vitro inhibition of succinic dehydrogenase activity and mitochondrial protein synthesis in rat hepatocytes (Trusal and O’Brien 1986). T-2 toxin also affected mitocondrial electron transport system in yeast by inhibiting succinic dehydrogenease (Khachatourians 1990). Yabe et al. (1993) found that NIV increased cytochrome P-450 and glutathione S-transferase (GST) activities in rats. In mice and monkey, the hepatic cytochrome P-450 is responsible for catalysing the hydroxylation of the C-3 and C-4 positions on the isovaleryl side chain of the T-2 and HT-2 toxins (Yoshizawa et al. 1984). When oxygen is removed from the epoxide group of a trichothecene to produced the carbon–carbon bond, non-toxic de-epoxy metabolites are formed (Swanson et al. 1988). This indicates that trichothecene epoxide reduction is a single-step detoxification reaction. Cytochrome P-450 and GST of human liver cells were also found to be involved in the metabolism of other mycotoxins (e.g. aflatoxin B₁) (Guengerich et al. 1996).

In corn, DON inhibited mitocondrial enzymes and caused electrolyte loss (Cosette and Miller 1995). Kang and Buchenauer (2002) developed an immunogold-labelling method to study the subcellular localization of DON and 3-ADON in F. culmorum-infected wheat spikes and kernels. During penetration, toxins were found in the host cell wall around the infection peg and in the cytoplasm of the host cells. In host cells, these trichothecenes were localized in the cytoplasm, plasmalemma, and chloroplasts and sometimes associated with ribosomes and endoplasmatic reticula. Their association with the plasmalemma might result in the alteration of membrane permeability.

Induction of apoptosis

At the cellular level, certain trichothecenes have recently been shown to induce apoptosis in a variety of cell types via mitochondrial and non-mitochondrial mechanisms (Shifrin and Anderson 1999, Yang et al. 2000) (Table I). Shifrin and Anderson (1999) classified trichothecenes according to their ability to induce apoptosis in human Jurkat T-lymphoid (Jurkat T) cells. Strong inducers included DON, scirpentriol, and T-2 triol; intermediate inducers included NIV, DAS, HT-2 toxin; and weak inducers included 3-acetyl DON, verrucarin and T-2 toxin. The authors suggested that the ability of individual trichothecenes to induce rapid apoptosis might require both translational arrest and mitogen-activated protein kinase activity (as will be discussed later). Among the many biochemical changes commonly found in cells undergoing apoptosis, is the systematic degradation and fragmentation of DNA (Bortner et al. 1995). Thus, the induction of apoptosis in mice thymocytes by fusarenon-X (Miura et al. 1998) and in the lymphoid of the bursa of Fabricus in broiler chicks by T-2 (Rachid et al. 2000) was proved by in situ labelling of DNA fragmentation (TUNEL). They also used electrophoresis of DNA in agarose gel to show that DNA cleavage due to fusarenon-X or T-2 toxin treatment resulted in fragments that were multimers of about 180 base pairs (bp), a DNA profile typical of apoptotic cells (Wyllie et al. 1984). Trichothecene-mediated apoptosis was also shown by Yang et al. (2000) using DNA fragmentation in two myeloid models, RAW 264.7 murine macrophage and U937 human leukaemic cells treated with satratoxin G, roridin A, verrucarin A, T-2 toxin, satratoxin F, H, NIV and DON.

Recently, we have found that DON negatively affects PCD morphology in Arabidiopsis thaliana cv. Landsberg erecta suspension cultures, as indicated by a reduction of the level of cells showing membrane blebbing and cytoplasmic shrinkage (Rocha et al., unpublished data).

Functional genomics of eukaryote–trichothecene interactions: Current knowledge

Current knowledge about the eukaryotic signal transduction cascades and downstream gene products activated by trichothecenes is limited, especially in plants. As stated earlier, plants may differ from animals in their trichothecene uptake or metabolism. Analysis of the functional genomics of animal and plant–mycotoxin interactions will help identify genes involved in proliferation, apoptosis (or PCD in plants) and/or plant disease resistance.

Activation of mammalian MAPKs and induction of cytokine gene expression

Figure 2 gives a basic outline of much of what is known to date regarding how trichothecenes affect signal transduction and downstream processes in mammalian cells. As mentioned above, trichothecenes inhibit protein synthesis by binding to the ribosomal peptidyltransferase. Inhibitors of the peptidyltransferase reaction can trigger a ribotoxic stress response that activates mitogen-activated protein kinase (MAPK) components of a signalling cascade that regulate cell survival in response to stress (Iordanov et al. 1997). In mammalian cells, certain trichothecenes triggered a ribotoxic stress response and activated MAPKs (Shifrin and Anderson 1999, Zhou et al. 2003b). In murine cells this induction was mediated in an unknown fashion by a dsRNA-activated protein kinase R (PKR) (Zhou et al. 2003b). Shifrin and Anderson (1999) showed that translational arrest and MAPK activation cooperated in the induction of apoptosis by trichothecenes, but postulated that different parts of the trichothecene molecule are responsible for MAPK activation and translational arrest: trichothecene derivatives that did not significantly affect protein synthesis activated MAPKs. A conformational change in the ribosome can result in MAPK activation (Iordanov et al. 1997) and therefore MAPK activation by such trichothecene derivatives may be due to their interaction with unique ribosome target sites or their displacement of the binding of ribosome-associated molecules. Moon and Pestka (2002) showed that, as a consequence of MAPK activation, DON induced expression and increased the transcript stability of cyclooxygenase-2 (COX-2) mRNA and hence COX-2 protein content in leukocytes, thus increasing the inducible synthesis of prostaglandin endoperoxides, critical components of the inflammatory response. Further evidence for DON mediation of the inflammatory response was provided by Zhou et al. (2003a) who showed that this toxin modulated the binding activities of specific transcription factors and subsequently induced cytokine gene expression.

Figure 2. Trichothecene-mediated signal transduction and downstream processes in mammalian cells. Protein synthesis inhibition: trichothecenes interfere with the active site of peptidyl transferase on ribosomes. Mitogen-activated protein kinase (MAPK) activation: certain trichothecenes induce ERK1/2, JNK and p38 MAPKs (Laskin et al. 2002). A double-stranded RNA-activated protein kinase R (PKR) mediates DON-induced phosphorylation of ERK1/2, p38 and JNK MAPKs (Zhou et al. 2003b). DON activation of the proinflammatory response: it induces cytokine transcription and, as a consequence of MAPK activation, induces cyclooxygenase-2 (COX-2) and increases synthesis of prostaglandin endoperoxides. DON-induced apoptosis (Shifrin and Anderson 1999, Yang et al. 2000, Poapolathep et al. 2002): increases linearly with inhibition of protein synthesis and the peptidyl transferase site is postulated to be a regulator of both MAPK activation and apoptosis in Jurkat T cells (Shifrin and Anderson 1999).

Induction of wheat EF-1α, adenosine kinase and genes of unknown function

Recently, as part of a differential gene expression study, we have found translation elongation factor 1α (EF-1α), adenosine kinase (ADK), and several genes of unknown function were overexpressed in wheat roots in response to DON treatment (Ansari et al. unpublished data). We are now studying the expression profiles of these genes in plants in response to DON and DON producers. Retardation of growth by DON may not be accompanied by cell senescence since, in various eukaryotes, decreased EF-1α activity correlated with senescence and increased activity with longevity (Cavallius et al. 1986, Silar and Picard 1994).

Activation of wheat retrotransposon-like transcripts

Retrotransposons are mobile genetic elements found throughout the plant kingdom (Fedoroff 2000) and they move to new chromosomal locations via an RNA intermediate (Boeke and Corces 1989). In previous research, we found three mRNA transcript fragments, overexpressed in wheat roots treated with DON, with full or partial homology to retrotransposons (Ansari et al, unpublished data). The level of transcript activation was genotype-specific. This and other research (Ivashuta et al. 2002) implies that there may be genetic or epigenetic transposon activation in plant cells in response to stress and the degree of activation may be genotype-specific (perhaps due to genotype-specific sites of genomic integration). Even if they are transpositionally inactive, the transcription of redundant RNA could affect the expression of other genes through RNA–RNA interaction or RNA-directed DNA methylation (Grant 1999, Wolffe and Matzke 1999).

Conclusions

All trichothecenes and trichothecene derivatives isolated to date differ in their toxicity towards eukaryotic cells due to their chemical structure. The gross and cellular toxic effects of trichothecenes have been more extensively studied in animals than in plants. Regarding the relative toxicity of trichothecenes, it may not always be possible to extrapolate from animal to plant systems that may differ in their trichothecene uptake or metabolism. The common effects of trichothecenes on both animal and plant cells include inhibition of protein, DNA and RNA synthesis, inhibition of mitochondrial functions, cell division and membrane effects. In animal cells, trichothecenes induce apoptosis, a programmed cell death (PCD) response via mitochondrial and non-mitochondrial mechanisms. The effects of trichothecenes on plant PCD have not yet been extensively investigated. The amphophilic nature of trichothecenes facilitates their cytotoxic effect on cell membranes (McLean 1996). Once trichothecenes cross the plasma membrane barrier, they enter the cell, where they can interact with a number of targets, including ribosome (McLaughlin et al. 1977) and mitochondria (Pace et al. 1988). Protein synthesis inhibition is the primary toxic effect of trichothecenes. The dependence of all cell metabolic process on protein synthesis suggests that many of the other effects of trichothecenes may be secondary effects of protein synthesis inhibition.

Acknowledgements

The authors thank the European Union Framework Programme 5 Project QLRT-2000-02044, the Irish Department of Agriculture Research Stimulus Fund (National Development Plan), and Enterprise Ireland Basic Research awards (National Development Plan) for associated research funding.