Review of genetic studies of susceptibility to facial eczema in sheep and dairy cattle

Abstract

Genetic responses of sheep and dairy cattle to the hepatic mycotoxin, sporidesmin, were reviewed. The mycotoxin can lead to clinical facial eczema (FE) in the most susceptible and severely challenged animals. The extent of hepatic injury is normally assessed from an enzyme secreted into the blood, gamma-glutamyltransferase (GGT). Latest heritability estimates for the natural logarithm of GGT level, 21 days or more after a sporidesmin challenge, were 0.45±0.03 in sheep and 0.34±0.02 in dairy cattle. Clinical FE follows from phylloerythrin (from chlorophyll) spilling over from the bile duct. Upon exposure to sunlight, phylloerythrin absorbs ultraviolet radiation, becoming reactive. Neither its concentration in blood nor clinical cases of FE are likely to be good indicators of liver damage from FE for ranking sires. Gaps in knowledge about genetic factors relating to FE susceptibility are highlighted.

Keywords:

• facial eczema (FE)
• genetics
• disease
• sheep
• cattle
• GGT

Description of facial eczema

Facial eczema (FE) is a metabolic disease resulting primarily from injury to the liver and bile ducts, and is caused in susceptible ruminants by the effects of a mycotoxin from the fungus Pithomyces chartarum. The main toxin is sporidesmin A, although there are other metabolites of sporidesmin that appear to be non-toxic. The term facial eczema, if not a misnomer, is at least misleading in that the most characteristic aspect of FE is hepatic injury, whereas visible damage to the skin/hide is a secondary effect (Clare 1944). Hepatic injury is also the most costly of the FE problems to the animal and its owner. The disease causes significant loss in agricultural production mainly in northern New Zealand, and occasionally in coastal Australia, South America, South Africa and the Azores.

How sporidesmin causes tissue damage has been described by Munday (1989) as involving oxidation/reduction of the sporidesmin molecule, which generates superoxide radicals and a cascade of reactive oxygen species. These lead primarily to damage of the epithelial cells lining the bile ducts and thence to bile duct damage or blockage, followed by hepatocyte damage. After exposure, sporidesmin has been found in urine, as well as in bile (Mortimer & Stanbridge 1968); clinical effects on the kidney include frequent urination, with haemolytic anaemia giving rise to red-stained urine.

Many aspects of FE have been reviewed by Morris et al. (2004) and di Menna et al. (2009), but there is no review of the latest findings on the genetics of animal resistance to FE. The purpose of this paper is to summarise findings on FE genetic resistance/susceptibility of sheep and dairy cattle, and to consider future research topics on FE in sheep and cattle.

Genetic variation in facial eczema susceptibility in sheep and dairy cattle

Measurement methods

Techniques for the phenotypic measurement of FE susceptibility are mentioned here, along with heritability estimates derived from them. Campbell et al. (1975) first reported qualitative evidence in sheep of sire-to-sire variation among sire progeny groups in susceptibility to FE; this was confirmed later with quantitative evidence, from which the heritability estimate was 0.42±0.09 (Campbell et al. 1981). These latter studies were based on post-mortem liver injury scores of each sire's progeny after an FE challenge, by either exposure to toxic pasture or dosing with sporidesmin. After finding differences among sire progeny groups, divergent selection lines were established at AgResearch Ruakura (Hamilton, New Zealand) with Romney sheep, confirming that genetic separation between selection lines could be achieved in FE sensitivity (Morris et al. 1989, 1995a). This led to a heritability estimate of 0.45±0.03, where the FE trait was defined with live animals as the natural logarithm of gamma-glutamyltransferase (GGT) level in systemic blood (log_e[GGT]) 21 or more days after a sporidesmin challenge.

Originally, for phenotypic scoring of sheep (Morris et al. 1989), the log_e[GGT] 21 days after sporidesmin challenge was expressed in relation to pre-challenge GGT level (GGT_PC). However since there was little variation in GGT_PC relative to the level 21 days after challenge, adjusting for a pre-challenge value was unnecessary. In practice, Ramguard (Morris et al. 1994b) has a protocol that requires withdrawal of any animal before FE challenge if its GGT_PC is more than ~1.4 times the group-average GGT_PC; this is done for ethical reasons and ensures that the GGT_PC of the remaining animals for FE testing has minimal variation.

Towers & Stratton (1978) showed that blood GGT levels were proportional to the amount of sporidesmin-induced liver damage. A rise in the levels of blood GGT (measured in serum or plasma) is a general indicator of bile duct damage, although it is non-specific to the FE condition. A rise in the level of blood glutamate dehydrogenase (GDH) is another non-specific indicator, in this case associated with hepatocyte damage (Ford 1974). After a sporidesmin challenge, levels of these two liver-specific enzymes are closely correlated (e.g. a phenotypic correlation of 0.83±0.01 in dairy cattle (Cullen et al. 2011)). Thus, to score (phenotype) animals for FE susceptibility, GDH may be substituted for GGT, although the GGT enzyme is preferred to GDH under commercial conditions. This is because an elevated GGT level is more prolonged and plateaus over several days and is therefore not so prone to variations in the timing of blood sampling (Towers & Stratton 1978; N. R. Towers pers. comm. 2012).

Ante- and post-mortem rankings

Post-challenge GGT level (yielding a live-animal ranking) has been shown to be closely related to post-mortem liver injury score in at least eight groups of experimental animals—five random populations of lambs (Towers & Stratton 1978), two genetically selected FE lines of sheep (Morris et al. 2002b) and a group of randomly sampled dairy cattle (Towers & Smith 1978). In another sporidesmin-dosing experiment, Fairclough & Smith (1983) found a positive relationship between sporidesmin concentration in the bile of catheterised FE selection line sheep and their liver injury score post-mortem, indicating a positive relationship between toxin concentration at the site of injury and the extent of damage scored. A series of experiments in sheep (Ford 1974) demonstrated the time trends of serum activities/concentrations of certain enzymes and metabolites, following intravenous administration of varying dose rates of sporidesmin, with and without ligation of the bile duct. In addition to GGT, the other enzymes studied were glutamate oxaloacetate transaminase (aspartate transaminase), sorbitol dehydrogenase, GDH and arginase, and the blood metabolite urea. Although the results were difficult to interpret, most enzymes measured (i.e. sorbitol dehydrogenase, GDH and arginase), plus urea, appeared to rise earlier than GGT, especially at the lower sporidesmin dose rates.

Performance testing vs progeny testing

The first genetic FE testing in cattle was the progeny testing of Jersey bulls by scoring their two-year-old daughters in herds affected by naturally occurring FE challenge (Morris et al. 1990). The results demonstrated sire-to-sire variation in FE susceptibility, giving a heritability estimate of 0.31±0.10 for log_e[GGT]. This estimate was not significantly different from the corresponding value of 0.45±0.03 in sheep. The cattle study was followed by a second progeny-test experiment (Fig. 1) to confirm that genetic separation could be achieved in further progeny of five progeny-tested ‘FE-resistant’ and five progeny-tested ‘FE-susceptible’ Jersey sires (Morris et al. 1991a). Similar results were obtained in Holstein–Friesian and other Jersey sires (Morris et al. 1998) by performance testing them when they were younger as calves, and then later by progeny testing them. As an example, the Holstein–Friesian results are shown in Fig. 2. The current heritability estimates for log_e[GGT] and log_e[GDH] are 0.34±0.02 and 0.30±0.04 respectively, based on over 200 dairy sires not characterised in prior studies for FE resistance (Cullen et al. 2011). These latest results supersede the prior estimates, which had greater standard errors (Morris et al. 1990, 1998; Cullen et al. 2006). The current genetic correlation estimate in dairy cattle between log_e[GGT] and log_e[GDH] is close to unity at 0.93±0.03, higher than the phenotypic value of 0.83±0.01, as expected.

Figure 1.  Mean log_e[GGT] of calf groups, after dosing with the FE toxin sporidesmin and classified by the progeny-test status for FE susceptibility of their five ‘High’ (susceptible) Jersey sires or five ‘Low’ GGT (resistant) Jersey sires: results are summarised for the calves in log_e i.u./l units (with bars shown for the standard error of the difference) against days since the toxin challenge (from Morris et al. 1991a) (NB: Factor of ×1.75 between the mean log_e[GGT] values of High vs Low sire groups (74 calves)).

Figure 2.  Mean log_e[GGT] of calf groups, sired by Holstein–Friesian bulls that had been scored (performance tested) as weaned calves for response to FE susceptibility, or response to sporidesmin. Five such ‘High’ GGT and five ‘Low’ GGT bulls were selected for progeny testing for response to sporidesmin, alongside the progeny of eight untested/‘unselected’ Holstein–Friesian bulls. Results for progeny are shown in log_e i.u./l units (with bars for standard errors of means, shown in one direction) against days since the toxin challenge (from Morris et al. 1998).

Performance testing in dairy-sired cattle has been carried out with an artificial sporidesmin challenge on young calves, whilst dairy-bull progeny testing has been carried out mainly on cows (daughters of the bulls of interest), exposed unintentionally to natural sporidesmin challenge (Cullen et al. 2006). A positive relationship obtained between a performance test result for a potential young bull, using log_e[GGT], and a progeny test result for him later as a sire is evidence of a genetic contribution and a non-zero heritability for the trait (Falconer 1960).

At present, the cheapest method of collecting FE resistance/susceptibility data on dairy sires is to progeny test them via daughters receiving natural sporidesmin challenge, and this method was used for most of the cattle data summarised above. Cullen et al. (2011) observed that, in FE-affected groups of cattle, even the binomial score for an individual animal (i.e. ‘elevated GGT value or not’) was moderately inherited. It had a heritability estimate of 0.24±0.02, which was surprisingly high, as most binomial scores have a heritability of ~0.1 or less.

Phylloerythrin

The secondary nature of clinical FE is the consequence of phylloerythrin (a photo-active breakdown product of chlorophyll), spilling over from the bile-duct-occluded liver into systemic blood, instead of being excreted directly into bile. An accumulation of phylloerythrin in the blood is believed to be the cause of clinical photosensitivity in FE-affected animals. Upon exposure to sunlight, blood phylloerythrin in lightly pigmented areas of the skin/hide absorbs ultraviolet radiation and becomes reactive, leading to local cell death. In a study of FE-affected sheep flocks and dairy herds (Morris et al. 2009), phylloerythrin concentration in systemic blood was only slightly correlated phenotypically with log_e[GGT] after sporidesmin challenge (0.32±0.04 in sheep and 0.37±0.02 in cattle). By plotting the relationship between phylloerythrin concentration and log_e[GGT], the points of inflection for GGT level were found at ~400 i.u./l in sheep and ~600 i.u./l in cattle, above which there was a greater likelihood of encountering an elevated phylloerythrin concentration. These inflection points may have been flock- or herd-specific, but they are consistent with casual field observations. For r denoting the phenotypic correlation between the two factors, the value 1−r ² (ranging from 0.86 to 0.90) indicates that most of the variation in blood phylloerythrin concentration was not associated with log_e[GGT]. Thus, high phylloerythrin and clinical cases of FE were not good indicators of liver damage from FE on an animal-to-animal basis. Likewise, measuring phylloerythrin concentration in blood (Campbell et al. 2010) will not assist in determining the extent of FE-induced liver damage. Phylloerythrin concentration was lowly inherited in cattle, with a value of 0.19±0.07 (Morris et al. 2009). Many other factors influence the levels of phylloerythrin concentration in the blood. Early studies in South Africa with sheep (Quin et al. 1935) showed between-animal variation in faecal phylloerythrin output for animals on the same diet and with similar feed intakes.

In summary, performance testing and progeny testing are about equally effective at ranking animals (sheep or dairy cattle), so relative costs and convenience become important. Ranking methods, using markers in live animals, or more invasive scoring at/after slaughter, are also equally effective. No evidence is seen of differences between the efficacy of natural versus artificial sporidesmin challenge. Measuring phylloerythrin concentrations in the blood of progeny groups or monitoring proportions of clinical FE cases in these groups is ineffective at ranking their sires for FE resistance.

Further results from sheep

Possible sources of genetic variation for FE susceptibility include

1.	animal variation in their grazing horizon on pasture
2.	animal variation in the effectiveness by which their rumen microbes destroy the toxin
3.	differences in how animals absorb sporidesmin from the digestive tract, and metabolise and excrete the toxin
4.	variation in the resistance of body tissues/organs to sporidesmin, and in their ability to repair damage caused by the toxin.

In genetic studies carried out by the authors, based on GGT trait, we were looking mainly for host genes affecting items 3 and 4.

Genetic correlations with FE susceptibility in sheep include those with production/reproduction traits (i.e. with litter size, yearling live weight and fleece weight), as estimated by Morris et al. (1999). Genetic correlations between FE susceptibility and some production traits were obtained from the AgResearch Ruakura FE selection lines of sheep. One correlation was found to differ significantly from zero, namely that with yearling fleece weight (−0.16±0.07), while those with weaning weight (0.02±0.07), 18-month live weight (−0.11±0.07) and litter size were not significant. That is not to say that FE has no effect on litter size (Morris et al. 1991b), but that the genes for FE susceptibility are unrelated or unlinked to those for litter size. The relationships of FE susceptibility with other diseases in sheep are also reviewed later in the paper.

Breed comparisons

A standard test of whether genetics are involved in a trait is to compare breeds. It can thus be deduced that FE resistance/susceptibility is an inherited trait by comparing the Romney with the Finnish Landrace breed and their crosses (Morris et al. 1994a): Finns were the most FE-resistant breed, the Romneys least resistant and the crosses intermediate. Similarly, Romney control line animals were compared with East Friesians and FE-resistant line Romneys (Morris et al. 2001), where the East Friesians were the most FE-susceptible breed. Smith et al. (1980) measured the responses of four sheep breeds to sporidesmin challenge (Merino, Romney, Border Leicester and Romney×Border cross) and the detoxification ability of these breeds with pentobarbitone (sleeping times being used as a measure of how quickly the liver could metabolise the anaesthetic). They found that Merinos were significantly more resistant to both sporidesmin and pentobarbitone than the other breeds and crosses.

A test of the liver's ability to dispose of toxic materials was described by Phua et al. (2009b), in sheep from AgResearch's FE-resistant and FE-susceptible lines. Using acetaminophen (‘paracetamol’), the authors showed that the resistant line was more tolerant of, and recovered faster from, the drug.

Using the candidate gene approach, Phua et al. (1998) tested genes encoding antioxidant enzymes for linkage to the FE traits in outcross families with resistant×susceptible cross-line sires. They found no significant co-segregation of the superoxide dismutase-1, superoxide dismutase-2, glutathione peroxidase or glutathione reductase genes with the disease trait. However, in studies designed to compare FE-resistant and FE-susceptible lines of sheep for antioxidant enzymes in the blood, the resistant line showed less superoxide dismutase activity, more catalase activity and, in two out of three experiments, higher glutathione peroxidase activity than the susceptible line (Hohenboken et al. 2004).

Susceptibilities to facial eczema and other mycotoxins

In addition to sporidesmin, other mycotoxins are relatively common (Towers 2006) on summer and autumn pastures in New Zealand. Studies on the resistance of sheep to these mycotoxins and drugs are informative for FE, including: lolitrem (which causes ryegrass staggers), ergovaline (which causes heat or cold stress in New Zealand and tall fescue toxicosis in the USA) and zearalenone (a cause of infertility).

Lolitrem (ryegrass staggers)

Lines of sheep were selected at AgResearch Ruakura for increased resistance or susceptibility to ryegrass staggers (RGS). By testing the resulting RGS selection lines for FE susceptibility and the FE selection lines for RGS susceptibility, a genetic correlation estimate of 0.31 was obtained between the two diseases (Morris et al. 1995b). This indicated that at least some genes were in common, possibly those coding for enzymes that were part of the phase I detoxification process (Smith et al. 1980; Parkinson 2001). Lynch & Price (2007) identified six such enzymes in humans (CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP3A4 and CYP3A5), and the properties of these enzymes may be worth testing in ruminants, with sporidesmin as a substrate.

Ergovaline (heat or cold stress and tall fescue toxicosis)

Ruminants show varying degrees of susceptibility to ergovaline. Clinically, this leads to vasoconstriction of the skin surface and, depending on diet and season, affected animals may show heat or cold stress. In New Zealand, animals may experience heat stress in summer/autumn when grazing endophyte-infected perennial ryegrass (Lolium perenne, L.), and cold stress in winter when eating conserved endophyte-infected L. perenne. Tall fescue toxicosis is the result of ergovaline toxicity of livestock in the USA.

Gooneratne et al. (2011) reported on ergovaline susceptibility in AgResearch RGS-resistant and RGS-susceptible lines of sheep. The lines were exposed to heat stress following an artificial challenge with perennial ryegrass seed containing ergovaline at 30 ppm. Rectal temperatures of the resistant line animals were significantly lower (by an average of 0.53°C) than those of the susceptible line animals.

In the southern USA, tall fescue (Festuca arundinacea) is the forage of concern as it carries ergot toxins that lead to tall fescue toxicosis when susceptible cattle are exposed to serious challenge (Paterson et al. 1995). Genetic differences in response to the challenge among sire groups of cattle have been reported (Lipsey et al. 1992) and breed differences in weight gain have been recorded in affected cattle in Arkansas (Morrison et al. 1988). Related studies of ergovaline susceptibility in mice have been carried out, where heritable differences in tolerance to ergovaline in the diet were found among selection lines (Hohenboken and Blodgett, 1997); following an artificial challenge with sporidesmin in the selected lines, some of the genetic variation among lines for ergovaline tolerance was found to be in common with that for resistance to FE (Hohenboken et al. 2000). This suggests that some genes for the detoxification of ergovaline and sporidesmin may be in common, as was the case of lolitrem and sporidesmin, described earlier.

Zearalenone and infertility in sheep

Zearalenone is a mycotoxin with a steroid-like chemical structure similar to that of oestrogens. It may be abundant in summer/autumn and may interfere with reproductive function in sheep during that time. Resistance to zearalenone had an estimated heritability of 0.32±0.10 in experimental sheep (Morris et al. 2005) and 0.19±0.07 in sheep under commercial conditions (Amyes & Morris 2008). In a study at AgResearch Ruakura with FE selection line lambs grazing zearalenone-affected pastures, the concentration of zearalenone breakdown products in urine was 38% higher in FE-resistant than in FE-susceptible lines (Smith & Morris 2006), although the authors concluded that there was no significant association between susceptibilities to the two toxins. A possible advantage to FE-resistant animals of faster toxin breakdown could be negated by the fact that one of the zearalenone breakdown products, α-zearalenol, is known to be more oestrogenic than zearalenone itself, at least in monogastrics (Galtier 1999); this is a principle described by Lynch & Price (2007) for ‘prodrugs’, which require metabolic conversion into their active compounds.

DNA studies of facial eczema selection lines of Romney sheep

The Romney FE selection lines bred by AgResearch have provided the resources to search for causative disease genes. Three approaches have been taken—a candidate gene method, a whole-genome screen experiment and selection sweep studies. The ultimate aim of this work was to develop commercial DNA tests for marker-assisted selection of FE-resistant sheep.

Candidate gene method

Loong et al. (1986) used polyacrylamide gel electrophoresis to search for variation in plasma proteins between FE-resistant and FE-susceptible line sheep. They detected significant differences in transferrin, a major iron-transporting protein in blood: resistant animals carried predominantly the acidic transferrin A, whereas susceptible ones carried mainly the basic transferrin D. Further study in an independent breeding trial was unable to confirm this initial finding (Morris et al. 1988).

The candidate gene approach relies on the biology of FE and known functions of genes. A reported mechanism of sporidesmin toxicity is the generation of superoxide radicals through redox cycling of the disulphide group in the epidithiopiperazinedione moiety (Munday 1989). The superoxide marks the start of a cascade of increasingly toxic free radicals, including hydrogen peroxide. Within a cell, the main anti-oxidant enzymes for removal of superoxide are superoxide dismutase-1 and superoxide dismutase-2, while for hydrogen peroxide they are selenium-glutathione peroxidase and catalase. Since selenium deficiency in sheep showed no effect on the level of sporidesmin-induced liver damage (Sissons et al. 1982), it appeared that the removal of hydrogen peroxide relied heavily on catalase activity. Two flanking microsatellite markers and one single nucleotide polymorphism (SNP) marker within the catalase gene were analysed in the FE-resistant and FE-susceptible line animals. The markers showed significant allele-frequency differences between the two lines, indicating the involvement of the catalase gene in the FE trait (Phua et al. 1999) (Table 1).

Table 1 . Findings from different approaches used to identify causative FE genes and loci in the FE selection lines of Romney sheep. Locations are given in Ovis aries (OAR) chromosomes.

Experimental approach	Gene/locus	Chromosome location
Candidate genes	Catalase ABCG2	OAR15 OAR6
R×S QTL experiment^1	QTL (significant) QTL (suggestive) QTL (suggestive) QTL (suggestive) QTL (suggestive)	OAR3 OAR1 OAR8 OAR13 OAR15
Selection sweep studies^2	SNP (Peddrift) SNP (Peddrift) SNP (Peddrift)	OAR1 OAR11 OAR12
^1 FE-resistant×FE-susceptible (designated R×S) cross-bred rams were used to generate four half-sib families in this genome screen experiment to detect quantitative trait loci (QTL).
^2 Resistant and susceptible selection lines were used in selection sweep studies (Phua & Dodds 2011), with the ‘Peddrift’ analysis method of Dodds & McEwan (1997).

Bissinger & Kuchler (1994) isolated a Saccharomyces cerevisiae gene, PDR5 (or STS1) that, when present in multiple copies in yeast, conferred on it resistance to sporidesmin and other structurally unrelated drugs. Similarly an over-expression of this gene gave yeast increased resistance to the same drugs and, conversely, PDR5-deleted cells showed hyper-sensitivity to sporidesmin. The PDR5 gene codes for a membrane protein belonging to the mammalian super-family of ATP-binding cassette (ABC) transporters. The closest human homologues of the yeast PDR5 gene belong to the ABCG subfamily (Sheps et al. 2004). Mammals in general have five ABCG members, of which the ABCG2 gene is known to function as a xenobiotic transporter in both humans (Staud & Pavek 2005) and cattle (Jonker et al. 2005), while the other members are involved in cholesterol and lipid homeostasis (Schmitz et al. 2001). Duncan et al. (2007) hence tested the ovine ABCG2 gene for its involvement in the FE trait. An intronic SNP marker showed significant allele-frequency differences between resistant and susceptible line animals, confirming the gene's involvement (Table 1). Recently it was found that FE resistance is positively associated with ABCG2 expression and is inversely correlated with DNA methylation state of CpG sites within the regulatory region of the gene (Babu et al. 2012).

The above examples illustrate how a candidate gene is chosen for in-depth investigation of its possible involvement in FE. The method is very limited in its scope as it depends heavily on existing knowledge. Furthermore, though the genes are associated biologically with FE resistance, they may be minor genes of small effect. A more comprehensive approach to capture all genes of detectable effect size would be a whole-genome screen experiment.

Whole-genome screen experiment

This experiment used reciprocal crosses between FE-resistant and FE-susceptible line animals to generate four line-cross rams (designated R×S) (Phua et al. 2009a). Each ram was then mated to ~150 unselected Romney ewes to generate four half-sib families, consisting of 130–180 progeny per family. All the progeny were challenged with a fixed dose rate of sporidesmin (at 0.13 mg sporidesmin per kg live weight) and their FE status was measured in terms of GGT and GDH levels in the blood 21 days after dosing. About 240 microsatellite markers, evenly spaced throughout the sheep autosomes, were tested across the resource families. The genotype data were analysed with the GGT and GDH responses using the method of Knott et al. (1996), on the conditional probability of inheriting the sires' alleles. One significant and four suggestive quantitative trait loci (QTL) were detected in this experiment (Table 1).

Selection sweep studies in the facial eczema selection lines of sheep

With the availability of the Illumina OvineSNP50 BeadChip, 66 resistant and 66 susceptible FE selection line animals were genotyped across these chips, which contained about 50,000 SNP markers. The data obtained were analysed using the ‘Peddrift’ method of Dodds & McEwan (1997). The analysis incorporated the pedigree data and compensated for genetic drift, and evaluated the allele-frequency differences of markers between the two lines. These studies detected three highly significant SNP markers, which denote the selection signature sites in the resistant and susceptible line animals (Phua & Dodds 2011) (Table 1).

In summary, studies on how or why FE-resistant and FE-susceptible animals differ can provide information on which host traits are important for FE. Analogies with sheep showing resistance to other mycotoxins, such as to RGS, ergovaline and zearalenone, also provide hints as to FE-resistance mechanisms. Three types of DNA study are described in sheep in our search for FE markers—candidate gene studies, whole-genome screen methods and selection sweep methods. A complete set of markers of FE resistance in sheep has still to be assembled.

Further results from dairy cattle

DNA test

A DNA-based test would provide a cheaper, less invasive and more convenient FE-ranking than a sporidesmin challenge test. However, to develop such a test in cattle, more DNA markers need to be tested to match genotype with phenotype in order to identify a suitable marker.

Ranking animals

When estimating a sire's FE breeding values from his mature daughters' GGT data in FE-affected herds, the daughters can be classified into treatment groups per herd according to whether there was an (unsuccessful) attempt to protect them from FE with a zinc treatment (Towers & Smith 1978). Cullen et al. (2011) found that the heritability of log_e[GGT] tended to be slightly higher in unprotected herds (0.37±0.03) than in herds attempting FE protection using zinc or other methods (0.33±0.03), as expected, but the two estimates were not significantly different from each other. Thus, pooling these two sources of data served as an approximate starting point for sire breeding value estimation. In the future, if the ranking of candidate rams and bulls by FE breeding value were to rely increasingly on field records, the collection of phenotypes by industry would require greater uptake of testing for blood GGT levels to detect FE so that accurate sire selection could be achieved before the widespread use of any particular sire for mating.

Other indicators of liver damage

Many physiological and biochemical markers have been investigated as indicators of FE damage in dairy cattle. Although improved predictors have been found, in comparison with GGT level (Table 2), no improvement could be made by substituting another single predictor for GGT. The potential indicators tested included levels of GDH, aspartate transaminase, 5′-nucleotidase, ferroxidase, bilirubin and bile acids. With the exception of ferroxidase, all showed a positive genetic correlation with log_e[GGT] level.

Beef cattle

The commonly held belief that beef cattle are less susceptible to FE than dairy cattle can probably be explained by differences in their daily sporidesmin intakes. Beef cows eat considerably less pasture, especially in autumn (the FE season), than lactating dairy cows of similar live weight and season of calving, and therefore their daily toxin intake is less than that of dairy cows. Pasture types being grazed may also differ between beef and dairy cows. Though the animals may be of different coat colour, damage to the liver and bile ducts by sporidesmin is likely to be independent of coat colour. No formal cattle breed comparisons have been carried out for FE susceptibility. However, where similar levels of toxic challenge have been encountered, no evidence—circumstantial or otherwise—has been found that beef cattle breeds in New Zealand are less susceptible to FE than dairy breeds.

In summary, searching for additional FE markers in dairy cattle has been unsuccessful so far, including investigating other indicators of liver injury. Studying FE in beef cattle is unhelpful because their levels of sporidesmin challenge are generally lower than in dairy cows.

Possible follow-up experiments on the underlying biology of facial eczema susceptibility

Experiments have been carried out in New Zealand to measure responses by sheep and cattle to artificial or natural FE challenges. These experiments could be followed up by some of the trials suggested below in order to provide a more in-depth understanding of the underlying FE damage and cellular detoxification processes involved.

Time series for metabolite and enzyme changes associated with facial eczema damage

A sporidesmin challenge was carried out in 528 weaned calves over three birth-year cohorts (Morris et al. 1998). The metabolites and enzymes reported alongside GGT and GDH are listed in Table 2. Other future targets could include serum concentrations of minerals, glucose and haematocrit; additional targets in dairy cows should include daily milk yield, somatic cell count and other milk components. A further time series was reported for nine-month-old lambs (Smith 2000), including monitoring changes in liver, kidneys and bladder. These time series need to be repeated in other stock classes, namely ewes and in-milk and dry cows. It is suggested that investigating sporidesmin effects on the mammary gland, kidneys and bladder may reveal further aspects of tissue damage by the toxin.

Table 2 . Genetic correlation (r _g) estimates (±standard errors) in dairy cattle between GGT and other potential indicators of liver injury (log transformations of enzyme levels or concentrations of metabolites, in blood) and heritability (h ²) estimates for these indicators (from Morris et al. 1998; Cullen et al. 2011).

Biochemical indicator	r g with GGT	h ^2
Glutamate dehydrogenase	0.93±0.03	0.33±0.05
Aspartate transaminase^1	1.00±0.04	0.28±0.09
5′-nucleotidase	1.00±0.01	0.23±0.12
Ferroxidase	Not significant	0.34±0.15
Bilirubin	0.92±0.14	0.12±0.06
Bile acids^2	0.94±0.07	0.50±0.26
^1 Also known as glutamate oxaloacetate transaminase.
^2 Jersey-sired animals only.

In commercial dairy herds, earlier indicators of response than GGT would make it possible to achieve better FE control at the herd level. It is important that a ‘day 1–3 post-dosing’ response indicator be found (apart from a drop in daily milk yield), but the results to date appear to replicate delayed secondary responses rather than identifying immediate metabolites or enzymes responding. Ideally, a day 1–3 indicator is needed that is correlated with the later 21-day GGT and GDH responses of sporidesmin-treated animals.

Does zinc act at the gut level or in the liver?

Zinc (Zn), among other elements, is known to provide prophylaxis against the effects of sporidesmin toxicity (Towers and Smith 1978). Ruminants have no storage organ for Zn in the body to provide daily protection through Zn mobilisation (Grace 2010). On FE-prone farms, owners can protect their livestock from FE by treating them with commercially available slow-release capsules. These are rumen devices that provide a continuous release of Zn to the treated animal, as ZnO in the case of the Time Capsule® (Munday et al. 2001) and with 88% elemental Zn in the case of Faceguard® (Bennison et al. 2010a). Zinc from each type of device is readily absorbed from the gut. In the case of the Time Capsule®, protection against FE appears to be associated with increased Zn concentration in the blood. The mode and site of action of the Faceguard® device have not been confirmed. Administration of the Faceguard® boluses yields smaller increases in serum Zn concentration than does the Time Capsule®. Bennison et al. (2010a, 2010b) suggest that a minimal concentration of Zn in the gastro-intestinal tract is required for protection against FE and question whether the liver is the main organ controlling serum Zn concentration in ruminants. There are other suggestions that active control of Zn supply is via the abomasum and lower small intestine, and is affected by animal age (Miller & Cragle 1965; Grace 2010).

Using an in vitro HepG2 human hepatoma cell line, Duncan et al. (2005) treated cells with various concentrations of sporidesmin and observed a sigmoidal dose–response curve with an LC50 (concentration lethal to 50% of cells) of 5 µg/ml toxin. The authors showed that cells could be protected from the toxin by pre-treatment for two or more hours with ZnSO₄, in a concentration-dependent manner; significant protection was obtained at 50 µM Zn and with a maximum at 200 µM Zn. These in vitro findings were consistent qualitatively with the in vivo results obtained by Munday et al. (2001) on administering ZnO boluses to calves before sporidesmin challenge. Interestingly, the Zn protection did not require concomitant de novo gene transcription (Duncan et al. 2005) and it may be inferred from the above studies that Zn protection might act possibly at the gut level as well as in the liver.

The lower threshold for serum Zn concentration to provide adequate protection against FE is thought to range in dairy cows from 18 to 34 µmol/l (Anon 2012, as derived from Smith 1987). Studies in Holstein–Friesian, Jersey and cross-bred cows following a sporidesmin challenge (Cullen et al. 2011) suggest that proportions of cows with elevated GGT levels in summer/autumn may have risen genetically in the last 25 years, so that Smith's (1987) earlier recommended range for a protective lower circulating Zn concentration may need revising.

What is potentiation?

Potentiation in FE occurs when, following previous exposure to sporidesmin, animals show greater sensitivity to the toxin on further exposure, often after small doses within a short time period (Fairclough & Smith 1983; Smith 2000). The reason for this phenomenon is not known. A possible explanation comes from the observation that recovery from the first toxin challenge may sometimes not be complete prior to re-exposure. Phua et al. (2009b) dosed both FE-resistant and FE-susceptible line sheep with the drug acetaminophen and found faster recovery of the former animals from the challenge. It is unknown whether this translates into less potentiation in genetically resistant animals compared with susceptible ones. In general, informal comments from dairy farmers suggest that some cows, clinically affected by FE in one autumn, may have recovered fully from the past FE effects by the following spring season. However, formal analysis (Morris et al. 2002a) of two-year-old cows with high blood GGT after exposure to FE showed reduced longevity compared with unaffected herd mates.

Similarities between responses in dairy cows after a sporidesmin challenge and a mastitic infection

Akers & Nickerson (2011) summarised the consequences in lactating cows of a mastitic infection as a temporary increase in somatic cell count and a temporary fall in daily milk yield, with increases in α-lactalbumin and casein concentrations in milk. Some of these changes are similar to the effects in autumn of natural sporidesmin challenge in spring-calving cows still in lactation, including increases in somatic cell count, changes in electrolyte concentrations (N. R. Towers, pers. comm.), short-term reductions in milk yield, delayed increases (as before) in the levels of GGT and ornithine carbamoyl transferase (OCT) and temporary or permanent drying off in more serious cases (Towers & Smith 1978). The experiment of Towers & Smith (1978) was designed to test the protective effect of ZnSO₄ on experimental spring-calving dairy cows' performance after a sporidesmin challenge. At the herd level, the timing would have coincided with sporidesmin presence in commercial New Zealand herds at pasture (January to April), and the yield reduction observed in the experiment could possibly have been explained in part by mammary gland reactions to sporidesmin. These reductions could go undetected in commercial herds as symptoms of FE because blood samples are seldom taken routinely to monitor GGT levels in a herd.

Brown et al. (1981) investigated the consequences in Staphylococcus-infected lactating cows of another mycotoxin (aflatoxin) that has immunosuppressive properties. The toxin was administered orally, every day for 12–14 days. The research identified increased counts of bacteria, ‘variable’ reductions in daily milk yield, inappetance, loss of body weight and increased serum activities of OCT and sorbitol dehydrogenase. These parameters should also be monitored in sporidesmin-challenged cows.

The evidence reviewed briefly above is consistent with sporidesmin generating a reaction similar to mastitis, producing an inflammatory (acute phase) response (Stuart & Lewis, 1988). Clinical mastitis is also characterised by increases in plasma protein concentrations and interleukin-6, as mediated by cytokines. More experimental studies are required to characterise the mode of action of each common mycotoxin in sheep and dairy cattle.

Conclusions

Much is known about the biology of FE outbreaks. Large animal-to-animal differences in response to sporidesmin are apparent, with relatively high estimates of heritability, using GGT level in blood at least 21 days after a challenge as an index of liver damage. Traditional quantitative genetic techniques can now be applied to potential seedstock to achieve genetic change in FE susceptibility, if required. Opportunities with DNA testing may be applied when available.

Much less is known, however, about what early changes in the liver and other organs may result from a sporidesmin challenge. Preliminary results suggest that the mammary gland may also be a sporidesmin target and that, under FE conditions in autumn, unobserved losses in milk yield may occur in commercial dairy herds as a result of natural challenge. Experimental studies need to be set up to answer the questions posed in the previous section.

Acknowledgements

SHP was in receipt of Ovita funds for part of this work, and CAM and NGC gratefully acknowledge partial funding from Meat & Wool NZ (now Beef +Lamb NZ), CRV-AmBreed, Dairy NZ and AgResearch.