Eutypa lata produces a number of structurally related secondary metabolites, mainly acetylenic phenols and their heterocyclic analogues. Eutypine (1, Figure 2), or 4-hydroxy-3-(3-methylbut-3-ene-1-ynyl)-benzaldehyde, is the most toxic. It was isolated from the cultures of an unspecified strain of E. lata and characterised [24,25]. It has an unusual five-carbon acetylenic side chain. Some other, structurally related metabolites were isolated from organic extracts of the same fungal culture such as eutypinol, O-methyleutypin, O-methyleutypinol, eutypin carboxylic acid analogue (2-5, Figure 2), as well as the compounds produced by hydroxylation of the terminal vinyl group of eutypine, eutypinol and the carboxylic anologue (6-8, Figure 2) [25,26,27]. Under low acidic conditions, eutypine is converted into 2-iso-propenyl-5-formylbenzofuran (9, Figure 2). This bicyclic product was also detected in E. lata culture filtrates [26].

Biological assays on excised tomato plants and vine leaves have shown that of all these metabolites eutypine is the most phytotoxic [26]. A comparative study was carried out on the metabolites from three strains of E. lata: strain E120 from grapevine in California, strain E125 from grapevine in Italy, and isolate E178 from oak in California. The metabolites were grown on malt yeast broth (MYB) and on potato dextrose broth (PDB) [27]. Eutypine, occurred in only one of the strains but the main metabolite was eutypinol, which is a detoxification product of 1 (Figure 2) [27]. A novel metabolite, named eulatinol (11, Figure 2) was isolated from the Italian strain E125, together with its O-demethyl derivate, siccayene (10, Figure 2). The chromene analogue, 6-hydroxymethyl-2,2-dimethyl-2H-chromene, named eulatachromene (12, Figure 2) as well as 1 and 2 (Figure 2) were isolated from the California strain E120 [18]. As previously reported, eutypine was readily cyclised into benzofuran (9, Figure 2) in the presence of traces of acid, and eutypinol was the main metabolite under most culture conditions [27].

The phytotoxicity of E. lata metabolites was tested in a leaf disk bioassay on “Cabernet Sauvignon” leaves (Figure 3A) [28]. Activity at 50 μg/mL of metabolites 1, 2, and 9-12 (Figure 2) indicated that neither eutypinol nor siccayne were phytotoxic; while eutypine, benzofuran, eulatinol, and eulatacromene (1, 9, 11 and 12, Figure 2) had a toxic effect, producing necrotic spots on the leaves (Figure 3A). Eulatacromene and benzofuran were more toxic than the acetylenic phenols. However eulatachromene (12, Figure 2) was also examined for toxicity at various concentrations from 10 to 100 μg/mL. At its lowest concentration, 12 (Figure 2) caused necrosis on some disks, and at its highest concentration it caused necrosis on all disks (Figure 3B) [28].

Two highly oxygenated cyclohexene oxides were also isolated from E. lata: eutypoxide B [29] and the novel allenic epoxycyclohexane (13 and 14, Figure 2) [25].

To elucidate the biochemistry of E. lata, both its sterol composition and its total sterol content were investigated in solid and in liquid culture [30]. The total amount of sterols was 1.58 and 1.12 μg/mg dry weight mycelium in solid and liquid cultures respectively. The major sterol was identified as ergosterol. This accounted for 88% of the total sterols in a solid medium and 78% when the fungus was grown in a liquid medium. In addition to ergosterol, four minor sterols, ergosta-5,7,9(11),22-tetraen-3β-ol, ergosta-7,22-dien-3β-ol, fecosterol and episterol also occurred. They accounted for 1-4% and 1.8-13% in solid and liquid culture respectively. These results suggest that the sterol biosynthesis pathway of E. lata is very similar to that of other pathogenic filamentous fungi [30].

Preliminary studies on the structure-toxicity relationship, carried out on the acetylenic phenols, showed that the occurrence of the aldehydic group and a free OH group in para-position were important for phytotoxic activity [26].

Smith et al. [31] in their study of the structure-phytotoxicity relationships compared the activity of metabolites from various strains of E. lata and from their chromene analogues, such as the corresponding aldehyde, acid and methyl esters (17-19, Figure 2) [31]. They synthesised the chromene analogue in quantities sufficient for evaluation, and carried out a rapid quantitative bioassay, involving the topical application of individual compounds to culture disks of grape leaves with subsequent measurement of chlorophyll loss to measure tissue damage (Figure 4A,B) [31]. In this assay, eulatachromene was more phytotoxic than eutypine (Figure 4A). The bicyclic product 9 (Figure 2) was also quite toxic, while the reduction product eutypinol, as well as the quinol siccayne (Figure 4A), were not toxic [31]. When eulatachromene was compared with its three synthetic analogues 17-19 (Figure 2) only the corresponding aldehyde 17 (Figure 2) had a toxicity similar to 12 (Figure 2), while 18 and 19 (Figure 2) were not toxic (Figure 4B) [31]. Consequently the active chromenes 12 and 17, along with benzofuran 9 (Figure 2), were more toxic than eutypine and eutypinol [31].

The biological action of eutypine (1, Figure 2) in grapevines has been extensively studied [32]. Eutypine is synthesised by the fungus in the trunk and is thought to be transported by the sap to the herbaceous parts of the plant, where it spreads through the leaves and inflorescences to participate in the expression of disease symptoms [33]. Eutypine has weak acid properties and a marked lipophilic character. It penetrates the vine cells by passive diffusion and accumulates in the cytoplasm due to an ion-trapping mechanism related to the ionisation state of the compound [33]. In this case the experiments were carried out on the plasmalemma of three models: V. vinifera, Beta vulgaris and Minosa pudica. These biological models were suitable for particular physiological processes. The observations obtained from these studies showed that eutypine was unspecific since it indifferently acted on all three plant cell models used [33] (Table 1).

That eutypine targets the mitochondria is suggested by the fact that it modifies the rate of respiration of the grapevine cell and its energy balance [34,35,36].

The molecular mode of action of eutypine at the mitochondrial level, and of O-methyleutypine (3, Figure 2), the non-deprotonatable derivative of 1 (Figure 2), was investigated [34]. The effects of these molecules on mitochondrial respiration and on the membrane potential were compared using isolated mitochondria from grapevine cells in suspension cultures. Eutypine caused marked stimulation of oxygen consumption and had a depolarising effect, while methyleutypine had a very slight effect on both the rate of oxygen uptake and membrane potential. High eutypine concentrations had a mixed effect, with a direct inhibition of electron transport and uncoupling. At concentrations below 200 mM, eutypine displayed a linear relationship between the oxidation rate and the membrane potential, similar to that of the traditional protonophore carbonyl cyanide m-chlorophenylhydrazone (CCCP). But unlike CCCP, eutypine induced a potential-dependent proton conductance, which was probably due to a potential-dependent migration of the dissociated form of the toxin across the membrane. It was thus concluded that eutypine uncouples mitochondrial oxidative phosphorylation and decreases the ADP/O ratio in grapevine cells by increasing proton leaks, which it accomplishes by means of a cyclic protonophore mechanism [35,36].

A further study of five mutant strains of Saccharomyces cerevisiae showed that eutypinol and eulatachromene (2 and 12, Figure 2) inhibited mitochondrial respiration in wild-type yeast, and that 2-iso-prenyl-5-formyl-benzofuran and siccayne (9 and 10, Figure 2) inhibited respiration in the strain lacking in mitochondrial respiration and vacuolar acidification. These effects of eutypinol and eulatachromene were confirmed using a strain with mitochondrial dysfunction and hypersensitivity to oxidative stress [32]. This study was not conducted on grapevine cells (Table 1).

Eutypa lata toxins are non-specific and appear to be important virulence factors in causing the symptoms of Eutypa dieback [35]. However, a lack of toxin-deficient mutants of the fungus, as well as the long time the fungus needs to incubate in the trunk before symptoms appeared, prevented a much-needed critical study of these toxins in grapevine. Nevertheless, it was confirmed that eutypine in grapevine cells is metabolised into the corresponding alcohol, eutypinol (2, Figure 2), through the enzymatic reduction of eutypine by a NADPH-dependent enzyme [37,38]. Eutypinol was not toxic to grapevine and had no protonophoric activity. The high affinity of the eutypine reductase enzyme (ERE) for eutypine indicated that ERE may play an important role in the detoxification of eutypine. It was also found that eutypinol did not have any uncoupling activity in the mitochondria [39] unlike what was seen with S. cerevisiae [32]. When the detoxification of leaf tissue by two genotypes of V. vinifera is compared, there seems to be a relationship between the susceptibility of vine cultivars to Eutypa dieback and the degree to which these vines deactivate eutypine [40]. Novel NADPH-dependent aldehyde reductase genes conferring resistance to eutypine have been reported from Vigna radiata [40,41]. This suggests that detoxification enhances the resistance of the vines to the toxin [39].

Eutypa lata synthesises a variety of metabolites, whose role in causing dieback is quite unclear. This role needs to be elucidated before any attempt can be made to apply these metabolites in order to diagnose E. lata dieback. Some studies have been carried out to determine the best growing conditions for E. lata, and to optimise its production of toxic metabolites in an artificial environment, especially on grapevine wood and wood extracts. Molyneux et al. [27] studied in vitro variations in the metabolite production of three E. lata isolates, using two growth media [malt yeast broth (MYB) and potato dextrose broth (PDB)], and a longer term, than previously reported. These researchers examined three representative strains of E. lata, two from grapevine (E120 and E125), and one strain from oak in California, to serve as a representative non-grapevine host species (E178). Metabolite composition and yield differed significantly between strains and between growth media, but yield usually peaked after 24-30 days [27]. Metabolites were identified by gas chromatography/mass spectrometry of their trimethylsilyl ether derivatives and analysed. This method proved to be generally applicable to all the metabolites subsequently isolated, with the trimethylsilyl derivatives well resolved [27]. Eutypine, eutypinol, siccayene and eulatinol (1, 2, 10, 11, Figure 2) were quantified by high-performance liquid chromatography (HPLC) in the three strains on MYB and PDB. Eutypine (1) was detected in only one of the strains (E125) on both media. The primary metabolite was the detoxification product eutypinol (2). This metabolite was produced in large quantities by one of the strains grown on PDB, but it was detected in all strains [18]. Strain E178 from oak did not produce these metabolites [27].

Subsequently, the HPLC procedure was optimised to establish the phenolic metabolite profiles of eleven E. lata strains grown on four growth media and to determine how the growth of these strains in these media differed from their growth on “Cabernet Sauvignon” vine wood and wood extracts [28]. The same technique was used to evaluate secondary metabolite production of 30 isolates of E. lata grown on media derived from the canes of three grapevine cultivars, two of which (Merlot and Semillom) were tolerant to E. lata, and one (Shiraz) susceptible [10]. Eutypine, eutypinol, O-methyl-eutypinol, 2-iso-propenyl-5-formylbenzofuran, eulatinol and eulatachromene (1, 2, 4, 9, 11 and 12, Figure 2) were detected in all culture filtrates. The most abundant metabolites were eutypinol and O-methyl-eutypinol, which were produced by 97 and 83% of isolates respectively. No correlation was found between secondary metabolite levels in the media containing ground canes from the three grapevine cultivars, and the tolerance/susceptibility of these cultivars to Eutypa dieback. In addition, no secondary metabolites of E. lata were detected from any isolates of other fungi commonly isolated from grapevine trunks in Australia such as Cryptovalsa ampelina, Libertella sp., Pa. chlamydospora, Pm. aleophilum, and various species of Botryosphaeriaceae and Fomitiporia that were grown on ground canes. This suggests that these metabolites are specific to E. lata [10]. To detect the secondary metabolites in planta, micropropagated grapevine plantlets were treated with purified and crude culture filtrates of nine E. lata isolates grown on MYB. The secondary metabolites 1, 2, 4, 9, 11 and 12 (Figure 2) were identified in some of the treated plantlets, but no single metabolite was detected consistently in all plantlets. Eutypinol was detected in micropropagated grapevine plantlets inoculated with E. lata mycelium; however, no metabolites were detected in the sap of plotted vines that had been mechanically inoculated with the pathogen [10].

As mentioned above, eutypinol is not toxic when applied to grapevine leaf disks [19,23]. However, this lack of toxicity does not preclude the use of this metabolite as a chemical marker, since the pathogenicity of E. lata is not due solely to the phenolic metabolites it produces, but also to the fact that it colonises grapevine wood and degrades the xylem. Consequently, whenever a non-toxic metabolite of E. lata such as eutypinol is detected in a vine, it is a positive indication of the occurrence of E. lata itself [28].

DNA-based markers to identify E. lata in infected vine wood have been developed [42,43]. Although identification techniques based on DNA analysis are sensitive enough, they are destructive and the particular vine portion sampled must actually contain E. lata; if healthy tissues from an infected vine are sampled, a false negative reading may result. However, the metabolites of E. lata are likely to be distributed throughout the vascular tissue and foliage of infected vines, especially in spring when the foliar symptoms are most evident. A diagnosis based on identifying such specific metabolites of E. lata in planta can therefore be carried out early in the season, even before the pathogen itself has spread throughout the vine. False negatives are in any case possible since foliar symptoms, and hence also toxin translocation, fluctuate from year to year on infected spurs or branches, as is common in leaf stripe disease [44].

The toxic metabolites produced by F. mediterranea, the basidiomycetous fungus which, in Europe, is most frequently associated with grapevine white rot, were also investigated. Tabacchi et al. [55] reported that the culture filtrate of F. punctata (current nomenclature: F. mediterranea) produced 4-hydroxy-benzaldehyde (28, Figure 9), dihydroactinolide and a novel chromanone, called 6-formyl-2,2-dimethyl-4-chromanone, and biogenetically related to eutypine (34-35, Figure 9).

White [80] reported that ten basidiomycetes isolated from grapevine in South Africa (8 novel species in the genera Fomitiporella, Fomitiporia, Inonotus, Inocutis, and Phellinus, and two species in an undetermined genus) produced in vitro 4-hydroxybenzaldehyde at fairly low concentrations ranging from 0.005 to 0.06 mg/L. The level of 4-hydroxybenzaldehyde was not related to a specific taxon and varied within each taxon [80].

The mode of action of the toxins, essentially naphthoquinones, produced by the fungi involved in esca may be related to their oxidant property. The quinones are extensively used in studies of oxidative stress because of the important role they play in plant defence [81]. Scytalone levels in particular may be related to the intensity of the brown-black colour of infected vine wood and to the occurrence of dark material in the xylem vessels.

Scytalone and the naphthoquinones produced by Pa. chlamydospora and Pm. aleophilum are intermediate metabolites resulting from the biosyntesis of dihydroxynaphthale (DHN) melanins. Some studies examined the pentaketide pathway of melanin synthesis in some black fungi pathogenic to humans and plants [82,83]. As shown in Figure 10, scytalone may be produced by the oxidation of 1,3,8-THN.

Melanins are dark pigments of a phenolic nature with a high molecular weight, which occur in animals, plants and micro-organisms. The structure of these pigments is unknown, and they probably are not necessary for the life of the producing organism, but they increase the capacity of that organism to survive under certain conditions. For example the occurrence of melanins in the cell wall of some fungi [84,85,86] increases the resistance of these fungi to physical agents such as UV radiation [86,87].

The role of melanins in virulence in may be due to various factors. Bürki et al. [88] reported that the naphthoquinones are cytotoxic, and they are thought to act by forming a covalent adduct through a Michael 1,4 addition between the quinone and a protein thiol or amino group [89]. Of these compounds, the most widely studied is juglone, which is highly toxic to plants and animals [61]. In fungal cells, juglone generates superoxide radicals, diminishing the cellular pool of reduced pyridine nucleotides which are particularly necessary for cells exposed to oxidative stress. It is well known that one of the main ways of plants to respond to phytopathogenic invasion is by generating O₂⁻ and H₂O₂ [90,91]. Many plants also produce auto-oxidisable quinone defence compounds, which oxidise NAD(P)H in plant and pathogenic cells, leading to the formation of superoxide radicals and hydrogen peroxide [92]. In response to these attempts of the plant to defend itself, the pathogenic fungi synthesise the naphthoquinones, which are toxic and suppress the defence reaction [93].

The naphthoquinones lower the resistance of plants to pathogenic fungi and hence enhance the virulence of those fungi. The production of naphthoquinone pigments appears to be an important component of the disease process as these pigments act as non-specific virulence factors inhibiting the plant defence reaction, resulting in plant hypersensitivity [94].

It has been reported that actively melanising cell-like appressoria of P. oryzae produce higher levels of melanin intermediates or their derivatives (phenols and quinones) in the presence of agents blocking melanin biosynthesis such as Tricyclazole^® or Carpropamide^® [95]. As phenols and quinones are highly toxic to living cells, an abnormal accumulation of metabolites such as 2-HJ probably prevents the appressoria from functioning [96].

Not much is known about how these metabolites act in the vine cells or tissues. The physiological changes that occur in the chlorotic portion of leaves with tiger-stripes, in the surrounding green portion of these leaves, and in the asymptomatic leaves of vine arms that also bear tiger-striped leaves, indirectly suggest that these changes are caused by toxic metabolites. Physiological changes include alterations in the rate of gas exchange, chlorophyll concentration, net photosynthesis, stomatic conductance, intercellular concentration of CO₂, and transpiration rate [97,98]. Tiger-striped leaves also show higher levels of fructose and glucose and lower levels of sucrose. Starch goes down while abscissic acid goes up, as do the free amino-acids, especially proline [98,99]. The mechanism whereby these disorders are produced has not yet been elucidated, but the changes may be an indirect effect of the toxins produced by the fungi colonising the vine wood, and perhaps even by some non-pathogenic fungi which also abundantly colonise the vine wood. In that case these changes would be a response of the vine to the disease. For example, the increase of abcissic acid in the leaves, as also the increase in soluble sugars and amino-acids, appears to be a stress-response to an impairment of the conducting vessels [100].

Scytalone and isosclerone also cause peroxidation of the membrane lipids, and this is known to be related to leaf senescence [56]. Toxic polypeptides could also be involved in this early senescence process, since they cause an increase in the anthocyanin levels of vine leaves [78].

As a rule, esca is first detected by visual inspection of the foliar symptoms. Visual inspection is completed by isolating on a growth medium the fungal colonies growing out of a piece of infected wood tissue and by identifying the growing fungus using microscopic inspection of the mycelium and the conidia, and/or by molecular methods.

A new means of detection is to employ HPLC or immunological techniques to identify the secondary metabolites secreted by the pathogen that is to be detected. Studies based on these techniques have established that secondary metabolites have a role as vivotoxins [101]

To study the metabolic changes in grapevine plants affected with esca, leaves were analysed from both symptomatic and non-symptomatic cordons of V. vinifera cv. Alvarinho, collected in the Vinho Verde region, Portugal. The metabolite composition of leaves from infected cordons with visible symptoms [diseased leaves (DL)] and from asymptomatic cordons [healthy leaves (HL)] was evaluated by 1D and 2D ¹H-nuclear magnetic resonance (NMR) spectroscopy [106].

Principal component analysis (PCA) of the NMR spectra showed a clear separation between DL and HL leaves, indicating that the compounds produced by these two types of leaves differed. NMR/PCA analysis identified the compounds belonging to each of the groups.

Altogether, levels of phenolic compounds were significantly higher in DL leaves than in HL leaves, and carbohydrate levels were significantly less, suggesting that the diseased leaves were rerouting carbon and energy from primary to secondary metabolism. The diseased leaves also had higher levels of methanol, alanine, and γ-aminobutyric acid, which may have been caused by the activation of other defence mechanisms [106]

Five species isolated from declining grapevines in Spain, Botryosphaeria dothidea, Diplodia seriata, Dothiorella viticola, Neofusicoccum luteum and N. parvum were examined for toxin production in liquid cultures of Czapek-Dox broth for different lengths of time [74]. All fungi produced high-molecular weight hydrophilic compounds with toxic properties (Figure 14).

Toxin production of D. seriata and N. parvum peaked after 14 days in culture, while that of other species peaked after 21 days. The effect of 14 day-old N. parvum culture filtrate on grapevine leaves cv. Tempranillo is shown in Figure 15.

The high-molecular weight hydrophilic toxic compounds produced by N. parvum, which were later identified as EPSs, were further tested. GC-MS analysis of the acetylated O-methyl glycosides of these EPSs revealed that these compounds consisted mainly of glucose, mannose and galactose, and that they differed from the botryosphaerans [70]. The botryosphaerans are branched β-(1→3; 1→6)-D-glucans produced by B. rhodina [117] and by an unidentified ligninolytic Botryosphaeria sp. [70]. The role of EPSs in bacterial and fungal diseases has been revised in depth [118]. In addition, N. luteum and N. parvum produce low molecular weight lipophilic phytotoxins (Figure 14), that do not invariably occur in all the remaining species [74].

The sections following deal with the toxins produced by the most familiar Botryosphaeriaceae species associated with grapevine, N. parvum and D. seriata. However, further investigation, such as that into the toxins produced by other botryosphaeriaceous fungi and that into the toxic compounds of infected grapevines, is needed to better understand the role that these and other toxic compounds play in symptom expression.

In a recent study has isolated the metabolites produced by N. parvum (strain CBS 121486) in optimised culture condition and characterised them chemically and biologically [119]. Four toxic metabolites were isolated from the organic extract and identified by spectroscopic and physical examination as (3R,4R)-(-)-4-hydroxy- and (3R,4S)-(-)-4-hydroxy-mellein, isosclerone, and tyrosol (36, 37, 21, 29, Figure 16), which were reported as being produced by N. parvum [119] for the first time.

When assayed on tomato cuttings all four metabolites (21, 29, 36, 37, Figure 16) were toxic, producing symptoms ranging from slight to severe leaf wilting (Table 3). (3R,4R)-(-)-4-hydroxymellein and isosclerone were the most toxic. All these toxins have already been reported as being produced by many phytopathogenic fungi [54,120,121,122,123,124,125,126,127].

Moreover, isosclerone (21, Figure 16) was reported for the first time as produced by a Botryosphaeriaceae species [119]. (3R,4R)-(-)-4-hydroxymellein (36, Figure 16) is produced, together with other melleins and tyrosol (29, Figure 16), by a strain of B. obtusa (syn. D. seriata) that causes frogeye leaf spot and black rot of apple [125,128]. The same 4-hydroxymellein as well as its steroisomer (3R,4S)-(-)-4-hydroxymellein and mellein are also produced by Diplodia pinea (Desm.) Kickx, another Botryosphaeriaceae species, which causes decline of Pinus radiata D. Don in Sardinia, Italy [127].

Botryosphaeriaceae species were reported by Larignon et al. [95] to cause foliar symptoms (i.e., tiger-striped chlorosis and necrosis) but these cannot be clearly distinguished from the symptoms of grapevine leaf stripe (previously young esca). Those symptoms were described as belonging to a disease called black dead arm (BDA) which was attributed to several species of Botryosphaeriaceae. The difficulty encountered in distinguishing between different diseases having basically the same symptoms has been commented on by several authors [5,53,129]. Since isosclerone (21, Figure 16) and EPS are both produced by esca-associated fungi such as Pa. chlamydospora, Pm. aleophilum [54,55,58], N. parvum [74,119], and possibly other Botryosphaeriaceae species as well, it is necessary to clarify whether these toxins are related to the foliar symptoms of leaf stripe disease (young esca) or to BDA, if this disease is confirmed. If the symptoms cannot be distinguished, it would explain the confusion about how to diagnose these diseases if the diagnosis is based solely on the foliar symptoms. Although the cause of the chloro-necrotic foliar symptoms of esca still needs to be fully elucidated [5], it is possible to hypothesise that the foliar symptoms are the result of the synergistic action of the toxic metabolites produced by these pathogens, all of which colonise grapevine.

While isosclerone is produced by both the pathogenic botryosphaeriaceous fungi and the fungi causing esca, the melleins and their derivatives appear to be produced only by the botryosphaeriaceous fungi [125,127,130]. Tyrosol (29, Figure 16) is another toxic metabolite produced by plants and fungi, including B. obtusa [123,125,131].

The fact that N. parvum produce isosclerone and exopolysaccharides is very interesting not only because these compounds are also produced by Pa. chlamydospora but because they are also involved in causing foliar symptoms of young esca and grapevine leaf-stripe disease [58,103]. Although the exact cause of these chloro-necrotic foliar symptoms is still unclear [5], they could be the result, wholly or in part, of the synergistic action of the toxic metabolites produced by species of Botryosphaeriaceae and Pa. chlamydospora [132].

Diplodia seriata (teleomorph: Botryosphaeria obtusa) is frequently associated with wood cankers of grapevine, sometimes named black dead arm or BDA [112,133], although this disease was initially reported by Lehoczky [134] as being caused not by D. seriata but by D. mutila. D. seriata also causes black rot of apple fruit, and produces several phenolic dihydroisocoumarins, such as mellein, cis-(3R,4R)-4-hydroxymellin and 5-hydroxymellein [125,128]. However, Djoukeng et al. [130] identified four phytotoxic compounds from a culture filtrate of D. seriata strain F-99-2, isolated from the vine cv. Cabernet Sauvignon. From an organic extract of the culture filtrates they isolated three known melleins, which on the basis of their spectroscopic data were identified as mellein, (3R,4R)-4-hydroxymellein, (3R)-7-hydroxymellein and (3R,4R)-4,7-dihydroxymellein (38, 36 and 39, Figure 17). Using the same technique they further isolated a unknown mellein, characterised as (3R,4R)-4,7-dihydroxymellein (40, Figure 17). A bioassay of vine cv. Gamay leaves found that (3R,4R)-4,7-dihydroxy-mellein was the most active metabolite, causing full leaf necrosis with a minimum inhibitory concentration (MIC) of 2 μg mL⁻¹, although compounds 36, 38, and 39 (Figure 17) had a similar degree of toxicity with a MIC of 3 μg mL⁻¹[130]. These authors using a HPLC DAD-MS analysis found that mellein (38, Figure 17) also occurred in infected vine wood inoculated with D. seriata. In that case, mellein would be a good diagnostic marker for D. seriata in diseased vines, and could be used to differentiate between esca and BDA at an early stage, since the melleins and their derivatives are not produced by esca-associated pathogens such as Pa. chlamydospora, Pm. aleophilum or F. mediterranea [130]. Mellein and its derivatives are also produced by many other non-botryosphaeriaceous fungi, including the genera Aspergillus, Cercospora, Cryptosporiopsis, Hypoxylon, Phoma, Pezicula, Plectophomella, Septoria and Xylaria, and they have phytotoxic, zootoxic and moderate antifungal activities [124,127].

As regards other toxins of D. seriata, it has been reported that tyrosol and p-hydroxybenzaldehyde were also produced by this pathogen, but only from an isolate obtained from apple [125,128]. Recently tyrosol has further been detected from N. parvum [119], and it cannot be excluded that some day D. seriata isolates will also be found to produce this toxin.