The impact of mycotoxins on performance and health of dairy cattle

Ruminant Processing Articles >> Biosecurity and Hygiene

Mycotoxins are a diverse group of molecules that are harmful to animals and humans. They are produced as secondary metabolites by moulds or fungi growing under propitious conditions in the field, during transport and during storage of feeds. They can be found all along the food chain.

Toxins in contaminated feeds are ingested by animals and can then be carried over to milk or edible animal tissues. Animals must not be considered as a simple ‘filter’ for mycotoxin transfer. Mycotoxins are metabolized in the liver and kidneys, and also by microorganisms in the digestive tract.

Therefore the chemical structure and the associated toxicity of mycotoxin residues excreted by animals or found in their tissues differ from the parent molecules. Due to the presence of the rumen and its dense microbial population, differences between parent mycotoxins and their metabolites are probably greater in ruminants than non-ruminants.

Part of the toxin is eliminated by animals in urine and faeces, and can be recycled in association with fungal spores in the manure and litter spread on the fields. The impact on the environment and consequences on the infection of soil and plants are unknown.

Harmful effects of mycotoxins on human and animal health have been known for about 80 years (Taubenhaus, 1920). However, studies on mycotoxins and mycotoxin-induced disease began only recently when in the 1960s a toxic molecule was extracted from Aspergillus flavus. Soon aflatoxin was identified and characterized as a very potent agent for toxicosis, even when consumed at concentrations as low as a few μg/kg (ppb) of feed.

Mycotoxins have significant economic and commercial impact. The productivity of plants is lowered by the presence of moulds, which can be considered as parasites. In addition, the nutritive value of infested cereals and forages is affected. In farm animals including lactating and growing cattle, pigs and poultry, mycotoxins have negative effects on production, reproductive rate, growth efficiency, and immunological defences; and also damage the liver and other organs such as kidneys to some extent. Export of agricultural products such as groundnuts, pistachios, cottonseed and copra originating from developing countries has been affected by the high risk of infection. Also, cereals like corn, wheat and barley produced in developed countries are now considered just as risky when produced and stored under unfavorable conditions.

In the present report an attempt has been made to identify the real impact of contaminated feeds on animal health and productivity with a focus on dairy cows and milk production. The risk due to the presence of mycotoxins in the food chain, from animal feeds to milk (and meat) supplied for human consumption will be considered, as well as possible treatments or additives for reducing such a risk.

The primary mycotoxins found in ruminant feeds and levels of contamination

It is likely that mycotoxins have been present since the beginning of organized crop production (Pittet, 1998). Ergotism was mentioned in the Old Testament of the Bible (Schoental, 1984); and fusariotoxins like T-2 and zearalenone have been considered responsible in part for the decline of Etruscans (Schoental, 1991) and for the plague of Athens during the 5th century BC (Schoental, 1994). Egyptian tombs have been suspected to harbor some mycotoxins such as ochratoxin A, which were responsible for the mysterious deaths of archaeologists (Pittet, 1998).

Forages and cereals are naturally in contact with yeast or mould spores in the field before harvest, during transport and when stored to be fed to livestock. Moulds need special conditions for growth. They require available nutrients from plants, correct temperature ranges, the presence of oxygen, unbound water and the right pH range. Forages preserved as dry hay are very dependent on the weather conditions during harvest. High humidity is the major risk for mould growth. Moulds can also grow in forages preserved in wet form as silage or wrapped big bales when anaerobiosis and acidic fermentation are incomplete. Cereals either ensiled or stored in dry form have the same potential contamination risk when harvested and stored under incorrect conditions. The level of contamination is even higher with cereals since starch or dextrins are efficiently used as substrates by moulds for growth.

Furthermore, grazing systems cannot be considered completely safe from mycotoxin contamination since green grasses can also be contaminated with moulds and fungal endophytes (Lewis and Clements, 1986) that produce toxins like ergovaline, lolitrem B and peramine. As a result, moulds and associated toxins can be present at the beginning of the silage or hay-making process. It should be pointed out that the prohibition against antifungal agents in the expanding ‘organic’ agriculture, leads to products that are probably more contaminated with mycotoxins than those coming from conventional agriculture. This indicates that elimination of moulds and their associated toxins is difficult to achieve in practical feeding systems either in developed or developing countries. However, Nelson (1993) proposed a strategy for mould control in dairy feeds. Because control of any one of the five environmental factors (nutrients, temperature, oxygen, free water, pH) prevents mould growth, management of the conditions of plant production and preservation could be a way of controlling mycotoxin hazards in the future, though this can never provide a complete solution (Park and Liang, 1993). Furthermore, because mycotoxins are secondary compounds issued from mould metabolism, environmental conditions favouring mould stress can stimulate mycotoxin production.

Although surveys on incidence and level of contamination are difficult to conduct owing to the need for rigorous sample collection and large laboratory facilities with reliable and sensitive analytical methods, several papers have been published on levels of feed contamination across the world (Bottalico, 1989; Galvano et al., 1996; Patel et al., 1996; Pohland and Wood, 1987; Pohland, 1993; Rustom, 1997; Tutelyan et al., 1989). Although more than 100 mycotoxins have been identified, less than 10 have been extensively studied during the last 15 years with regard to their natural occurrence and toxicity to humans and animals. Aflatoxins were extensively studied because of their carcinogenicity and widespread occurrence in tropical countries. Ochratoxin A and fumonisins have been investigated because they are typically present in temperate climates. Patulin has been detected in apples and apple juice. Deoxynivalenol is of importance as a trichothecene toxin and zearalenone is of particular concern due to its nonsteroidal estrogenic activity. Because each mycotoxin can be produced by various fungi, and the same fungus is able to produce several mycotoxins (Table 1), it is difficult to correlate mycotoxins to the presence of moulds in feeds.

Pittet (1998) summarized recent published data from around the world on the level of contamination of 27,853 samples from ingredients used for human and animal nutrition (Figure 1). Deoxynivalenol and fumonisins are the most widespread mycotoxins. Of the cereal grain samples tested, 75% and 66% were contaminated with DON and fumonisins, respectively. Also, the highest concentrations of mycotoxin were found for these two mycotoxins (62.0 and 37.6 mg/kg, respectively). Nearly 32% of cereal samples tested were contaminated with zearalenone, but the concentration in positive samples was often high (#21.4 mg/kg). Between 40 and 50% of nuts, beans or cereals contained aflatoxin B1 at a concentration lower than 6 mg/kg. The same proportion of aflatoxin M1 was found in milk and milk products; but the concentration was only 6 μg/L. Patulin in apple juice and ochratoxin A in cereals, coffee, cacao and spices were found to be present in 50% of samples at a concentration of less than 1mg/kg for the former and 0.1 mg/ kg for the latter. These concentrations are mean values across countries, however there were large differences among the 30 countries represented in this study. In addition, variations between years in the same country were significant.

A survey conducted in France on occurrence of fusariotoxins in wheat and corn harvested during 1996 and 1997 indicated that the percentage of positive samples and the level of contamination varied according to the cereal and the year (Table 2). The level of contamination (2-3 mg/kg) was the highest in corn harvested in 1996. Deoxynivalenol and fumonisin B1 had the highest level of contamination for wheat and corn, respectively, independent of the year. Zearalenone had the lowest concentration in both wheat and corn.

The FDA has enforced regulatory limits on aflatoxin in foods and feeds since 1965. Experts observed that corn is the cereal most susceptible to aflatoxin contamination in the United States, mainly in the southeastern states where the climate favors Aspergillus flavus and A. parasiticus growth (Wood, 1992). The corn belt states in the midwestern US were considered aflatoxin-free until 1983. Because of atypical weather conditions in 1988 during the latter part of growing season, corn from most areas of the country had high levels of aflatoxin contamination. This means that incidence and level of contamination is unpredictable for a particular area in any country and from year to year.

In the same survey, cottonseed and cottonseed meal were considered to be risky feed ingredients for animals, including dairy cattle as the concentration of aflatoxin varied from 20 to greater than 300 μg/kg. Milk and milk products are routinely collected for aflatoxin M1 determination in all the States. The only milk samples containing aflatoxin M1 in concentration higher than 0.5 μg/L in 1990 came from Texas and were attributed to drought conditions in that state.

Pelhate (1987) compared the microflora found in hay preserved either during favorable dry conditions or during wet conditions (Figure 2). He detected several species of Aspergillus and Penicillium in overheated wet hays that were not found in dry hays. According to Clevstroem et al. (1981), Aspergillus flavus can grow in wet hays and produce aflatoxins. In the same way, Le Bars (1982) indicated that Penicillium cyclopium colonized the wet hays and produced penicillic acid. Stored straw, if not dried effectively, probably gives rise to similar moulds and toxins.

The conditions associated with well-preserved silage, i.e. low pH and anaerobiosis, are unfavourable for the growth of moulds. The paper published by Scudamore and Livesey (1998) gave a good idea of the occurrence and significance of mycotoxins in forage crops and silage. Most Fusarium spp. associated with corn and grass in the field are unable to grow in silage. Damaglou et al. (1984) inoculated ryegrass with Fusarium roseum and F. tricinctum just before ensiling and found that only 1% of fungi survived after 15 days in the silo. Aspergillus fumigatus is found in silage and is often associated with heating. Penicillium roqueforti is the major fungus found in badly preserved corn silage (Auerback et al., 1998a) giving rise to roquefortine C (Ohmomo and Kitamoto, 1994) but, according to Nout et al. (1993), no P. roqueforti toxin is detected. Acidic conditions (pH<4) in silage could alter the production of toxins as indicated by Hacking (1979) for patulin, and make the mycotoxins less stable. Pre-formed zearalenone remained stable for a very short time after ryegrass ensiling, while it remained unchanged for a 12-week period in the study carried out by Lepom et al. (1988). In a survey conducted in France by Escoula (1974) on corn silage, 52 mould species were identified, half of them considered to be toxinogens. Most of the contaminated area in the silo had a high ammonia N/total N ratio, and a pH >8. The author indicated that more than 50% of the samples contained patulin (1.5 to 40 mg/kg) in association with the fungus Byssochlamys nivea. The concentration of patulin decreased with time in the silo after mould growth had stopped. This means that patulin is probably degraded by an unknown process in silage as indicated before for other mycotoxins. Patulin is known to have a great affinity for sulfhydryl chemical groups, so it becomes less stable when mixed with feeds rich in sulfur content (Wood, 1992).

Some fungi belonging to the Aspergillus, Eurotium and Rhizopus genera are able to convert aflatoxin B1 to aflatoxicol, and also to reconvert aflatoxicol to aflatoxin B1 (Nakazato et al., 1990). These facts suggest the possibility of finding aflatoxicol in foods and feeds contaminated with aflatoxin B1.

Effect of mycotoxins on feed intake

The biological effects of mycotoxins observed are often related to dosage levels and duration of exposure. This point must be kept in mind when considering all of the data presented in this paper.
Feed refusal from grain or forage infected by mycotoxins has been a recurring problem in animal feeding. Ruminants seem less sensitive than monogastrics where loss of appetite is concerned (Bersjo et al., 1992; 1993; Friend et al., 1992; Prelusky et al., 1994; Rotter et al., 1995). Similarly, adult ruminants are less sensitive compared with unweaned calves.

According to Charmley et al. (1993), concentrations of DON up to 12 mg/kg in the diet failed to influence intake of concentrate or forage in lactating cows. Trenholm et al. (1985) observed a slight decrease in feed intake of lactating cows receiving DON at 6 mg/kg. Such differences between experiments could be due to the presence of other toxins that were not identified.

Aflatoxin B1 given to lactating cows in amounts up to 13 mg/day (around 8 mg/kg feed) did not affect feed intake (Applebaum et al., 1982). In the same experiment, aflatoxin B1 was given either pure or mixed with other aflatoxins and fungal metabolites. The authors showed that feed intake fluctuated when cows received 13 mg/day impure aflatoxins. In comparative slaughter feeding trials using steers, Helferich (1984) and Helferich et al. (1986a) observed that aflatoxin concentrations from 60 to 300 μg/kg had no effect on feed intake. On the other hand, 600 μg/kg aflatoxin significantly decreased ingestion. The dose level and additive effects of aflatoxins in multicontaminated feeds are certainly determining factors for feed intake.

Osweiler et al. (1993) observed that daily feed intake was not changed after addition of 148 mg/kg fumonisin to the solid diet of calves. Only the rate of ingestion was altered since, occasionally, calves failed to consume the portion before the next feeding. Thus, there was no treatment-related effect on feed intake of calves on a daily basis, although contaminated feeds seemed less palatable than the control.

The negative effect of mycotoxins sometimes observed on feed intake could be due to a repellent taste or odour of contaminated feeds. Dust in dry forages associated with the presence of moulds and spores can be involved in decreased feed palatability. In addition, digestive and metabolic disorders caused by mycotoxins are probably involved in reducing voluntary intake, as administration of 50 to 150 mg day (3 to 9 mg/kg feed) aflatoxin by bolus in gelatin capsules induced a persistent drop in feed consumption of cows (Mertens and Watt, 1977).

Behaviour of mycotoxins in the digestive tract and liver of ruminants

Metabolism of mycotoxins by animals may alter the manifestation of adverse effects. Bioconversion occurring in the digestive tract or within the tissues may have additional implications for quality of carcass and milk aimed at human consumption. In general, bioconversion leads to hydrosoluble and less toxic forms of the parent molecule, particularly in the case of hydrolyzed, reduced or conjugated metabolites like deepoxytricothecenes, ochratoxin ", aflatoxicol or glutathione conjugates of aflatoxin B1 (Figure 3). Hydrosolubilization is one important step in the excretion of foreign metabolites from the animal body (Brattsten, 1991). These detoxifying processes take place in the rumen, and some occur in the liver. By contrast, oxidative reactions involving cytochrome P450 are generally considered to produce more highly toxic metabolites like hydroxylated or epoxide derivatives. These aspects are of importance for their possible consequences on the safety of edible animal products.

As pointed out previously, ruminants are more resistant to some mycotoxins than monogastrics. This could indicate that biotransformation of toxins can occur in the rumen. King et al. (1984), Côté et al. (1986) and Swanson et al. (1987) reported that rumen microbes degraded the DON molecule by opening the epoxy cycle to form a diene called de-epoxydeoxynivalenol (DON-1) (Côté et al., 1986). It is likely that a specific epoxide reductase may be involved in such a transformation. Because the epoxide group on the trichothecene nucleus explains the cytotoxicity and vomiting activity (Sato and Ueno, 1977), its bioconversion into a diene function is the main reason for the lower toxicity of deoxynivalenol in ruminants. Such a biotransformation occurs also in the large intestine of monogastrics (Galtier and Alvinerie, 1976). However, the degradation is less and has a less protective effect on the animal since the parent toxin has been absorbed in the digestive tract before reaching the large intestine.

Conflicting results have been reported regarding rumen biodegradation of aflatoxin B1. Some authors found a marked decline in initial aflatoxin B1 concentration when incubated with rumen juice (Engel and Hagemeister, 1978), while others did not (Kiessling et al., 1984). Because metabolites from aflatoxin B1 such as aflatoxicol and aflatoxin M1 have been detected in the rumen (Trucksess et al., 1983), Auerback et al. (1998b) looked for these molecules when aflatoxin B1 was incubated in an in vitro system inoculated with rumen juice. They showed that aflatoxicol is produced from labeled aflatoxin B1 by rumen microbes, but they could not find aflatoxin M1. They hypothesized that aflatoxin M1 found in the rumen may be ascribed to hepatic biotransformation and subsequent recycling to the rumen via the rumino-hepatic pathway. Because the biodegrading capacity of rumen microbes towards aflatoxin B1 is poor and the toxicity of aflatoxicol is close to the toxicity of the native toxin, there is no evidence of a true detoxification process in the rumen for this toxin.

Ochratoxin A contains a methyl-isocoumarin molecule (ochratoxin ") linked to L-ß-phenylalanine by a peptide bond. Hydrolysis of the peptide bond by rumen microorganisms leads to the excretion of ochratoxin ", which is not toxic (Hult et al., 1976), and phenylalanine which is used as free amino acid by microbes. This explains why ruminants are much less sensitive to the acutely toxic effect of ochratoxin A (Ribelin et al., 1978). However, ochratoxin A hydrolysis in the gastrointestinal tract of the ruminant is perhaps less than previously described based on in vitro experiments, especially when the toxin is ingested in combination with concentrate diets (Höhler et al., 1999). Kiessling et al. (1984) showed that protozoa are more active than bacteria in detoxifying ochratoxin A, as well as for the degradation of zearalenone, T-2, diacetoxyscirpenol and DON. Westlake et al. (1989) made the same observation. However, these authors indicated that the strongest activity of protozoa against trichothecenes is counterbalanced by the higher sensitivity of protozoa to the toxic compounds, mainly to the T-2. Some strains of Butyrivibrio fibrisolvens and Selenomonas ruminantium are able to use T-2 toxin as an energy source. T-2 needs at least two enzyme systems to be hydrolyzed, one of which is associated with the membrane of B. fibrisolvens (Westlake et al., 1987a; 1987b). T-2 and diacetoxyscirpenol are rapidly reduced in a similar manner to DON by rumen microbes into several products (Swanson et al., 1987). Contrary to DON, there is no direct deepoxidation of T-2 and diacetoxyscirpenol, but intermediate compounds are produced before the diene structure appears.

In liver, the cytochrome P450 oxidase generally metabolizes xenobiotics into more polar conjugated metabolites more easily eliminated via bile and urine, and theoretically less toxic compounds (Galtier, 1998). However, oxidation increases the electrophilic nature of lipophilic compounds rendering them more reactive with cellular nucleophiles like glucuronides, sulfates, amines, mercapturates and amino acids. Therefore, more highly toxic metabolites such as hydroxylated or epoxide derivatives of aflatoxins, fusarin C, trichothecenes and penicillic acid are produced. Aflatoxins are oxidized in the liver into very reactive molecules (Figure 3) capable of binding critical biomolecules such as nucleic acids or functional proteins in cells thus originating the initial steps of cancer formation. This hepatic bioactivation is of considerable importance to health because the active metabolites are formed in situ, within the animal tissues. Aflatoxin M1 is produced in the liver by hydroxylation of aflatoxin B1. Other aflatoxins of the M series are found in milk and probably come from the liver: M2, GM1, GM2, M2a, GM2a. Although many scientists have focused their attention on aflatoxin M1 because of public health concerns, its production represents less than 3% of ingested aflatoxin B1.

Not much information is available on the effect of mycotoxins on ruminant digestion and supply of nutrients to the animal. According to Edrington et al. (1994) and Auerbach et al. (1998b), aflatoxin B1 had no effect on the digestion of cellulose nor on VFA production measured either in vivo or in vitro. However, the presence of large amounts of mycotoxin and a possible synergistic effect among diverse toxins could have a negative impact on ruminant digestion. Also, because the composition of the rumen microflora and microfauna depends on the diet, feeding conditions can affect the biological activity of microorganisms involved in the detoxification process. This aspect has not previously been studied.

Effect of mycotoxins on dairy cow health

Gastrointestinal absorption governs toxin entrance into the systemic circulation and subsequently to the rest of the body. As a general rule, absorption through the digestive tract is accomplished by one of three major ways: simple diffusion in the liquid phase for polar compounds, diffusion in the lipid phase for non-ionized compounds, and active carrier transport. Lipophilic molecules of low molecular weight like aflatoxins and zearalenone are candidates for passive diffusion transport. Ochratoxin A has both phenolic and carboxylic groups that give it weakly acidic properties like citrinin, cyclopiazonic and penicillic acids. Diffusion of the non-ionized form through the lipid membrane is the primary means of mycotoxin absorption in the digestive tract.

Bioavailability of mycotoxins depends on the type of molecule. Fifty to 65% of DON and ochratoxin A are bioavailable in monogastrics (Galtier et al., 1981) while less than 6% of fumonisin B1 is absorbed in cows (Prelusky et al., 1995). Cyclopiazonic acid, diacetoxyscirpenol, fusarin, patulin and penicillic acid are rapidly and intensively absorbed in the digestive tract of monogastrics (Galtier, 1998).

As expected, aflatoxin induces severe hepatic dysfunction confirmed by modifications in serum or liver enzymes (Table 3). Because of its high metabolic activity, the liver of high producing lactating dairy cows is often in poor condition, which makes it very sensitive to toxins. It is now clear that aflatoxin B1 and aflatoxicol both induce hepatocarcinogenesis (Table 4) in animals and humans. Aflatoxin M1 was classified in group 2B according to the International Agency for Research on Cancer (IARC), as an agent possibly carcinogenic to humans. In addition to the liver, the thymus is also a target organ for aflatoxins.

Because of the hydrolysis of ochratoxin A in the rumen into the non-toxic ochratoxin ", ruminants appear to be naturally protected against the risk of ochratoxicosis. However, symptoms can appear when the toxin is supplied in too large an amount in the diet and the capacity of microbial detoxification is exceeded. The fact that the kidney is the main target for ochratoxin A is explained by selective uptake of the toxin by the proximal tubules of the kidney cortex. The main clinical patterns associated with acute ochratoxicosis in rats, pigs and poultry are anorexia, weight loss, vomiting, increased body temperature, conjunctivitis, polyuria, bloody mucus from the rectum, dehydration, prostration and death within two weeks after toxin has been administered (Marquardt and Frohlich, 1992). Ochratoxin A and C have similar toxicity, but ochratoxin B is approximately 10 times less toxic than the former for these animal species. In chronic toxicosis, ochratoxin A elicits both nephrotoxicity and hepatotoxicity (Prelusky, 1994). It has teratogenic effects on embryos in the mouse and rat, but not in pigs and chicks (Juszkievicz et al., 1982). Its immunotoxicity has been described in poultry (Dwivedi and Burns, 1985) and pigs (Szczech et al., 1973) as well as incidence of renal adenomas and carcinomas in rats (Bendele et al., 1985).

Toxins from Fusarium such as zearalenone and trichothecenes are common mycotoxins found in cereal grains and they play a great part in mycotoxicoses in cattle in North America (Cheeke and Shull, 1985). They can be also found in significant amounts in pastures grazed by ruminants in New Zealand (DiMenna et al., 1991). Trichothecenes produce a wide variety of disorders at the gastrointestinal level such as vomiting (DON), diarrhoea and inflammation. Dermal irritation, abortion and haemorrhages also appear. Immune response is altered by trichothecenes through a complex mechanism (Sharma and Kim, 1990; Sharma, 1993) which makes the animals more sensitive to pathogens. In dairy cows, T-2 toxin given at 0.6 mg/kg resulted in animal death and blood in faeces, enteritis and digestive ulcers (Pier et al., 1980). Soon after ingestion, a burning sensation in the mouth and digestive tract appears. A severe leukopenia follows, and necrotic lesions extend. This could be due to a protein synthesis inhibition resulting from binding of toxin to ribosomes. This mechanism was proved for T-2 toxin (Midelebrook and Letherman, 1989).

Zearalenone often occurs in combination with trichothecenes in contaminated cereal grains. Symptoms of zearalenone toxicosis are uterine enlargement, swollen vulva and mammary glands, decline in ovulation rate and cycle length (Smith et al., 1990). Conception rate is decreased in dairy heifers, but transfer of zearalenone or its metabolites into milk is very low (Prelusky et al., 1990). The oestrogenic effect of zearalenone is mediated by binding of mycotoxin to the cytoplasmic oestrogen receptor (Klang et al., 1978).

Fumonisin B1 causes leucoencephalomalacia in horses and has been shown recently to be a cancer promoter (Dragacci, 1999). The mechanism of fumonisin carcinogenicity does not appear to be due to an interaction with DNA as described for aflatoxins but probably involves an epigenetic procedure (Coulombe, 1993).

While some of the mycotoxins have been discussed individually in this report, little is known about synergy among toxins present simultaneously in animal feeds.

Effect of mycotoxins on dairy cow performance and toxic residues in milk

Acute aflatoxicosis following oral doses of 50 or 150 mg aflatoxin daily for five days caused a quick reduction in feed intake and milk production and an increase in milk fat content (Mertens and Watt, 1977). A 13 mg daily dose of impure aflatoxin B given to cows in mid-lactation for seven days induced a decrease in milk production, but no signs of ill health were observed. A dairy herd receiving corn contaminated with 120 μg/kg feed had severe difficulties in milk production and reproduction associated with health problems during the year 1977 in the US (Guthrie and Bedell, 1979). The authors noted a 28% increase in milk production within three weeks after removing the contaminated corn from the ration. In the UK, Jones and Ewart (1979) followed 140 cows receiving a diet containing some decorticated extracted groundnut meal contaminated with aflatoxin B1, B2, G1 and G2.

Concentration of B1 was about 0.02 mg/kg complete diet. Diarrhoea stopped and cows started to eat more during the 3-4 days after the groundnut meal was removed, and production of milk increased within the next 5-8 days.

The body condition of cows also greatly improved. However, when compared to a control group of cows that never received any aflatoxin, the group fed the contaminated diet did not recover fully.

In an experiment conducted by Charmley et al. (1993), 18 primiparous cows received a diet formulated to reach a daily intake of 0.59, 42 and 104 mg DON from contaminated wheat and corn. Increasing DON in the diet did not affect intake of concentrate or forage nor milk production. However, milk fat responded quadratically with cows given the toxin at the 42 mg daily dose having the lowest milk fat content and fat output. The authors observed no transfer of DON or deepoxydeoxynivalenol to milk.

Lactating dairy cows fed corn heavily contaminated with fusariotoxins from the Gibberella species and containing 12-13 mg DON and 500 μg zearalenone per kg DM, had slightly reduced feed intake but no effect on milk or fat production was noted (Noller et al., 1979). The authors concluded that cereals infested with Gibberella could be fed to ruminants without any apparent detrimental effect. Although ruminants are less sensitive to zearalenone than monogastrics, few experiments have been performed to determine whether the toxin affects performance of cows. Reduced milk production and fertility associated with hyperestrogenism in cows have been described when zearalenone is present in cereals (Mirocha et al., 1974) or in hay (Mirocha et al., 1968). Dairy cattle fed a ration with 385 to 1,982 μg/ kg of zearalenone for seven weeks had normal milk production and no toxin residue was found in the milk, urine, serum or tissues (Shreeve et al., 1979).

Excretion of xenobiotics in milk can be achieved through the same three basic processes described for the absorption of toxins by the digestive tract, i.e. intercellular filtration, passive diffusion through membrane cells or active transport by means of secretion vesicles (Figure 4). The intracellular filtration concerns only small sized hydrophilic molecules, and is important in the formation of the milk liquid phase (water and electrolyte entrance) but is minor in the excretion of mycotoxins. Passive diffusion through cell membranes concerns lipophilic mycotoxins in non-ionized form, or in ionized hydrophilic form. On the contrary, polar and acidic compounds such as fumonisins have a low excretion rate for milk. Prediction of mycotoxin excretion in milk is difficult mainly because of changes in molecular structure during their metabolism in the animal body.

Table 5 summarizes the results from experiments on the carryover of mycotoxins. Transmission of aflatoxin from feed to milk is variable, ranging from 0.3 to 2.2% (Stoloff, 1979; Patterson et al., 1980). Using [14C]-labeled aflatoxin, Helferich et al. (1986b) observed that in goats the proportion of the oral dose recovered in milk, urine and faeces was 1.05, 30.9 and 52.3% respectively. With the exception of liver, tracer recovery from the tissues was negligible. The cumulative recovery of aflatoxin metabolites in milk after 120 hrs was approximately 1%. Considering such a value as likely, the upper limit for aflatoxin B1 content in feeds for dairy cows producing 20 kg milk and ingesting 6 kg contaminated concentrate per day was evaluated to 5 μg/kg by the European Community in July 1998 (Dragacci, 1999).

No fumonisin B1 and only traces of ochratoxin A were found in milk (Table 5). This is explained by filtration by the mammary gland as previously discussed for the former, and through degradation by rumen microbes for the latter.

High levels of trichothecenes need to be ingested to produce detectable residues in milk. Prelusky et al. (1984) found only traces (<4 μg/L) of DON and metabolites in milk after administration of a single oral dose of 920 mg.

Longer periods of feeding probably increase the amount of residue. Feeding 66 mg DON per kg feed, Côté et al. (1986) found 30 μg/L of the metabolite DOM-1 in milk, but the parent molecule was not detected. Following an administration of 880 mg/kg of DON to lactating ewes for three days, Prelusky et al. (1987) found 220 μg/L in milk with most of the residue being in the conjugated form of DOM-1. In a survey carried out in North Carolina,

Whitlow et al. (1986) and Whitlow and Hagler (1999) noted a significant decline in milk production as content of DON in the concentrate increased to 0.8 mg/kg feed DM. Such a result could be explained by a synergistic effect of mycotoxins associated with DON, although these other mycotoxins were not identified. The presence of DON residue in animal tissues has not been studied.

Due to its lipophilic nature, more T-2 toxin than DON is transferred into milk (Table 5). No accumulation of T-2 residues occurs in animal tissues according to Beasley et al. (1986) and Chatterjee et al. (1986). Daily intake of 50 to 165 mg zearalenone by cows for 21 days gave no residues in milk (Prelusky et al., 1990). Ingestion of 544.5 mg for 21 days induced the presence of zearalenone associated with "-zearalenol. Cumulative excretion in milk represented 1.4% of mycotoxin intake. The same rate of excretion was found when a single dose of 1.8 g was given, but "-zearalenol was found in addition to the two other residues. A dose as large as 6 g gave the same molecules in milk, but the rate of excretion in milk was only 0.01%. It is likely that the amount of residue increases with the length of mycotoxin exposure.

Methods of reducing the negative impact of mycotoxins in animals and edible animal products

Because mycotoxins are secondary metabolites of moulds and fungi growing on plants, the first idea is to limit the level of contamination in the field. Several crop management strategies have been proposed to reduce mycotoxin contamination (Nelson, 1993). Treating the susceptible crop in the field with fungicide is actually the main procedure used by plant producers. An alternative method is to use genetic modification or classic plant-breeding strategies to produce plants that eliminate or reduce mould contamination (Brown et al., 1999). Harvested plants need then to be preserved in controlled conditions so as to avoid the growth of moulds.

Feeds can be also treated before being given to animals to partially eliminate mycotoxins. Treatment of aflatoxin-contaminated cottonseed with 1.5% ammonia and 10% water for 21 days reduced the amount of aflatoxin M1 in milk by approximately 85-90% (Price et al., 1982; Ewaidah, 1984) or in animal tissues (Allam et al., 1999). A similar treatment was applied to infested groundnut meal with the same efficiency (Thiesen, 1977). Treating feeds with 61% hydrogen peroxide solution and then heating at 80°C for 0.5 hr was nearly as efficient as ammonia (Allam et al., 1999). Oxidative degradation and detoxification by ozone was tested by McKenzie et al. (1997). Addition of 4% calcium hydroxide and 0.5% paraformaldehyde at 2 atmospheres for 20 minutes in an autoclave decreased aflatoxin from 401 to 29.5 μg/kg (Piva et al., 1985). Several heat treatments were tested by Purchase et al. (1972).

They observed that spray-drying was the most efficient treatment since it lowered by 86% the initial content of aflatoxin M1 in milk. The decrease was 81% for sterilization, 75% for roller-drying at 4.9 kg/cm2, 64% for roller drying under reduced pressure, 61% for evaporation, 61% for pasteurization at 80°C for 45 seconds, 45% for pasteurization at 72°C for 45 seconds, and 32% for pasteurization at 62°C for 30 minutes. Irradiation of raw whole milk artificially contaminated with aflatoxin M1 with ultraviolet energy for 20 minutes at 25°C decreased the amount of toxin by 60.7% (Yousef, 1986). Addition of 0.05% hydrogen peroxide in milk increased the efficiency of ultraviolet treatment as evidenced by 89.1% toxin disappearance. The effects of sodium bisulphite and extrusion cooking under high temperature on DON in wheat grain and mill fractions were investigated, but none of them significantly altered the toxin content (Accerbi et al., 1999). According to Xie Mao Chang and Wang Ming Zu (2000), dehulling reduced the DON content of naturally contaminated wheat grain by 23.6 to 34.7%. The decontamination efficiency of 0.1 mol/L sodium carbonate treatment reached 83.9%, while addition of 1% sodium hydrogen bisulphite decreased the DON content by 69.9%. The average DON reduction rate of 4% hydrogen peroxide and 5% limewash treatments were 45.1 and 21.8%, respectively. Washing contaminated corn for 48 hrs with daytime rinses every 2 hrs probably eliminated part of the fusariotoxin since it greatly improved both feed intake and daily gain in pigs (Forsyth et al., 1976).

Adsorbents can be added to contaminated feeds to selectively bind mycotoxins in the digestive tract, allowing the toxins to pass through the animal without any negative effects. Phillips et al. (1988) demonstrated that a special hydrated sodium calcium aluminosilicate (HSCAS) was able to bind to aflatoxin and limit its bioavailability in animals. The chemisorption of aflatoxin to HSCAS by the ß-ketolactone or bilactone groups of aflatoxin is efficient either in vitro or in vivo (Ramos and Hernandez, 1997), but the binding capacity of HSCAS to the other toxins is very low when measured in vivo (Patterson and Young, 1993; Ledoux and Rottinghaus, 1999). Other products like bentonite, anion exchange resin, cation exchange resin, and vermiculite-hydrobiotite were tested. They had no significant effect on T-2 toxin excretion in rats (Carson and Smith, 1983). Attapulgite and Novasil were shown to be efficient against aflatoxin B1, while kaolin was less active (Kane et al., 1998). Synthetic zeolites (NaX, NaY, NaA, CaA) were evaluated in vitro for their ability to bind aflatoxin B1. From an in vivo test, Miazzo et al. (2000) observed that zeolite NaA counteracted some of the toxic effects of aflatoxin B1 in growing broiler chicks.

Patulin found in mouldy fruit and vegetables and in badly preserved silages is known for its high reactivity towards nucleophiles such as compounds with thiol function. Biological or chemical compounds harboring such a function can be used to limit the bioavailability of patulin (Fliege and Metzler, 1998).

Facing the relative inefficacy of the clay binders towards mycotoxins other than aflatoxins, a natural product made of modified yeast cell wall has been proposed (Dewegowda et al., 1998; Evans and Dawson, 2000). According to the authors, the modified (esterified) cell wall glucomannan fraction is able to bind 58-75% aflatoxin over a large pH range (Table 6). Zearalenone and fumonisins are also bound in vitro by esterified cell wall material, but a larger amount of adsorbent is needed as the binding capacity is lower than for aflatoxin. Only 10% of DON can be bound in vitro.

Lactobacilli, bifidobacteria and propionibacteria were also tested and showed a large binding effect against aflatoxins either in vitro or in vivo (Table 6). Because non-viable bacteria are as efficient as live bacteria, it is suggested that covalent interactions and metabolic activation are not involved in the binding process. Haskard et al. (1998) proposed that a cation exchange mechanism involving the cell wall peptidoglycan of bacteria is concerned. The same explanation was suggested to explain the mode of action of yeast cell walls (Evans and Dawson, 2000).

Understanding the chemical mechanisms involved in the binding processes between modified yeast cell wall glucomannans and mycotoxins has the potential to provide improved biological material for reducing the bioavailability of toxins and limit the risks for animals and humans that consume animal products.

Conclusion
In relation to concerns about food safety and animal welfare, feed manufacturers and farmers should consider that mycotoxins associated with animal feeds and ultimately food animal products could be a very significant consumer issue affecting the dairy industry. Contamination in the field and during preservation of feeds can be reduced by appropriate cultural and harvesting techniques. However, we must consider that total elimination of moulds and associated toxins is impossible.

For this reason addition of mycotoxin-ligands in the diet is a way to reduce the bioavailability of the toxins. Because of the presence of a dense and efficient microbial ecosystem in the first digestive compartment of ruminants, any manipulation of the rumen microbes aimed to increase the degradation of toxins is another way to reduce their impact on animal health and safety of edible animal products. These are fascinating topics for research in the coming years.

by J.P. Jouany - Alltech Inc.

This article hasn't been commented yet.