Toxigenic aspergillus and penicillium species
by J.l. Pitt
Aspergillus
Aspergillus is a large genus, with 100 or more well on common synthetic or semi-synthetic media. The most widely used taxonomy is by Raper and Fennell (1965).,Keys and descriptions of common species may be found in Pitt and Hocking (1985a) and Klich and Pitt (1988a). Aspergillus species are of very common occurrence in the environment, principally in soils and decaying vegetation, but a number of species are also closely associated with human foods, particularly cereals and nuts (Pitt and Hocking, 1985a). Many species are xerophilic, and capable of spoiling foods only just above safe moisture limits. The most significant mycotoxigenic species are A. flavus and A. parasiticus, which make aflatoxins, A. ochraceus, which makes ochratoxin, and A. versicolor, which produces sterigmatocystin. These species are treated below: A. flavus and A. parasiticus are very closely related, and are treated together.
Aspergillus flavus and A. parasiticus
Aspergillus flavus and A. parasiticus were classified by Raper and Fennell (1965) in what was termed the "Aspergillus flavus group", an incorrect terminology now replaced by the correct term "Aspergillus Section Flavi" (Gems et al., 1985). A. flavus and A. parasiticus are closely related to A. Oryzee and A. sojae, species which are important in the manufacture of fermented foods in Asia, but which do not produce aflatoxins. For obvious reasons, accurate differentiation of these four species is important.
Identification
Differentiation of A. flavus and A. parasiticus from nearly all other species is not difficult. Both grow rapidly on standard identification media such as Czapek agar or Malt Extract agar (MEA; Raper and Fennell, 1965). or Czapek Yeast Extract agar (CYA; Pitt and Hocking, 1985a), and produce yellow green conidia on colonies which are otherwise uncoloured (Raper and Fennell, 1965; Pitt and Hocking, 1985a; Klich and Pitt, 1988a).
Distinguishing A. flavus from A. parasiticus is more difficult, with a lack of agreement among the students of these species (Klich and Pitt, 1985). A recent study has concluded that the texture of conidial walls is the most reliable differentiating feature: walls of A. flavus conidia are usually smooth to finely roughened, while those of A. parasiticus are definitely rough when viewed under a 100 X objective. A variety of other characters are also of taxonomic value (Klich and Pitt, 1988b).
Differentiating these species is of some importance: A. flavus isolates usually make only B aflatoxins (see below), and less than 50% of isolates are toxigenic, while A. parasiticus isolates make G as well as B aflatoxins, and are invariably toxigenic (Klich and Pitt, 1988b).
Toxins and toxicity: aflatoxins
Aflatoxins are produced in nature only by A. flavus and A. parasiticus. The four major naturally produced aflatoxins are known as aflatoxins B1, B2, G1 and G2. 'B' and 'G' refer to the blue and green fluorescent colours produced by these compounds under ultraviolet light illumination on thin layer chromatography plates, while the subscript numbers 1 and 2 indicate major and minor compounds respectively (Fig. 4). When aflatoxin B1 and B2 are ingested by lactating cows, a proportion (ca 1.5%; Frobish et al.,1986) is hydroxylated and excreted in the milk as aflatoxins M1 and M2, compounds of lower toxicity than the parent molecules, but significant because of the widespread consumption of cows'milk by infants.
Aflatoxins are both acutely and chronically toxic to animals, including man. They produce four distinct effects: acute liver damage; liver cirrhosis; induction of tumours; and teratogenic effects (Stoloff, 1977).
Acute toxicity of aflatoxins to humans has been observed only rarely (Shank, 1978). In 1967, 26 Taiwanese in two farming communities became ill with apparent food poisoning. Nineteen of them were children, three of whom died. Rice from affected households was blackish green and mouldy, and appeared to be of poorer quality than that from neighbouring households which were unaffected. Samples of the mouldy rice contained about 200 µg/kg of aflatoxin B1, and this was probably responsible for the outbreak. Post mortem examinations were not carried out.
In 1974 an outbreak of hepatitis that affected 400 Indian people of whom 100 died, almost certainly resulted from aflatoxins (Krishnamachari et al., 1975). The outbreak was traced to corn heavily contaminated with A. flavus, and containing up to 15 mg/kg of aflatoxins. Consumption of toxin by some of the affected adults was calculated to be 2-6 mg in a single day. It can be concluded that the acute lethal dose for adult humans is of the order of 10 mg.
Kwashiorkor, a disease of children in Northern Africa and elsewhere, is usually attributed to nutritional deficiencies, but may also be related to aflatoxin intake (Hendrickse et al., 1982). Aflatoxin-induced liver damage may render children with kwashiorkor less able to assimilate the high protein diets usually recommended as the cure for this disease (Newwell, 1983).
Aflatoxin B1 has been demonstrated, in a variety of animal species, to be the most potent liver carcinogen known. Human liver cancer has a high incidence in central Africa and parts of Southeast Asia. Epidemiological data suggested a link between human liver cancer and aflatoxins. Studies in serveral African countries and Thailand showed a correlation between the logarithm of aflatoxin intake and the occurrence of primary human liver cancer (van Rensburg, 1977).
Studies in the United States produced different results. On the basis of van Rensburg's figures, Stoloff and Friedman (1976) estimated that children in rural areas of the USA consumed enough aflatoxins from corn to induce 4-10 deaths from liver cancer per 100,000 adult population. However, the actual death rate is less than 1 per 100,000. Refinements to those calculations by Stoloff (1983) reinforced his conclusion that aflatoxins are unlikely to contribute significantly to the incidence of liver cancer in the USA.
The resolution of the conflict in the conclusions drawn by van Rensburg (1977) and Stoloff (1983) lies in recent information which suggests that prior or simultaneous exposure to hepatitis B virus may be a prerequisite for the induction of this form of liver cancer in humans. It appears that aflatoxins and hepatitis B are cocarcinogens, and the probablity of cancer of the liver in humans is high only in areas where both aflatoxins and hepatitis B are prevalent (Campbell, 1983).
Recent evidence suggests that primates are able to detoxify aflatoxins by pathways which do not exist in lower animals (Masri, 1984). The toxicity of aflatoxins to humans is certainly lower than was assumed in earlier years; nevertheless the known carcinogenicity in lower animals is so high that every effort must be made to monitor aflatoxin levels in human foods and to reduce them as far as possible.
Symptoms
The symptoms of acute aflatoxin poisoning were studied following the hepatitis outbreak in India mentioned above (Krishnamachari et al., 1975). This disease outbreak was characterised by jaundice, rapidly developing ascites and portal hypertension, with the implication of a food-borne toxin involving the liver. However, the ingestion of aflatoxins at the low levels necessary to induce liver cancer is totally asymptomatic and has a very long induction period as well.
Distribution in nature and in foods
Aspergillus flavus is of ubiquitous occurrence in nature. Since the discovery of aflatoxins, it has become the most widely reported food-borne fungus, reflecting its economic and medical importance, and ease of recognition, as well as its universal occurrence. A. parasiticus is less common, but the extent of its occurrence is obscured by the tendency for A. flavus and A. parasiticus to be reported only as A. flavus.
A. flavus (and A. parasiticus) have a particular affinity for nuts and oilseeds, though the reason is not clear. Peanuts, corn and cottonseed are the three most important crops invaded by A. flavus. Earlier work assumed that invasion was primarily a function of inadequate drying or improper storage, and these factors are certainly important in the occurrence of aflatoxins in the humid tropics. However, in temperate zones, recent work has stressed the importance of A. flavus invasion of these crops before harvest. Invasion of peanuts occurs as a result of drought stress and related factors (Cole et al.., 1982).
Pre-harvest invasion in corn is partly dependent on insect damage to the developing cobs, but the fungus can also invade down the silks. cobs, but the fungus can also invade down the silks of the developing ears (Lillehoj et al., 1980). In cottonseed, invasion is now believed to occur by entry through the nectaries (Klich et al., 1984). Most other nuts are also susceptible to invasion at some time (Pitt and Hocking, 1985a).
Cereals are also a common substrate for growth of A. flavus but, unlike the case of nuts and oilseeds, small grain cereal spoilage by A. flavus is almost always the result of poor handling. Aflatoxin levels in small grains are rarely significant (Stoloff, 1977). Spices sometimes contain A. flavus (Pitt and Hocking, 1985a) and viable counts may be very mgh. However, the quantities of spices consumed are so small that aflatoxin in spices does not appear to be a real hazard.
In developed countries, stringent sorting and clean up procedures are used to reduce aflatoxins to low levels in foods with a perceived risk. For peanuts, where fungal growth is usually accompanied by discolouration of the kernel, this includes the use of sophisticated colour sorting equipment. Statistically based sampling, the drawing of large samples, homogenising before sub-sampling and standardised aflatoxin assays are used to ensure that susceptible crops and foods meet the stringent requirements of health laws in both exporting and importing countries. Developing countries are often less fortunate. Established patterns of local consumption, where substandard nuts and corn may be consumed without any form of sorting or inspection, mean that aflatoxin ingestion remains far too high in many countries, especially in rural areas.
A. ochraceus and closely related species
A. ochraceus is the most commonly occurring species in the "Aspergillus ochraceus group" of Raper and Fennell (1965), now correctly referred to as Aspergillus Section Circumdati (Gems et al., 1985).
Identification
A. ochraceus and other closely related ochratoxin producing species - A. alliaceus, A. melleus, A. sclerotiorum and A. sulphureus - all grow moderately slowly on standard identification media such as CYA and MEA (Pitt and Hocking, 1985a). Colonies are deep but not very dense, and coloured pale brown to yellow brown; stipes are long, heads are large, vesicles are spherical and metulae and phialides are densely packed. Conidia are small, pale brown and smooth walled (Raper and Fennell, 1965; Pitt and Hocking, 1985a; Klich and Pitt, 1988a). Differentiation of these species is not always easy (Raper and Fennell, 1965; Klich and Pitt, 1988a). However, A. ochraceus is by far the most commonly occurring of these species, and differentiation of the others form A. ochraceus is rarely necessary.
Toxins and toxicity: ochratoxins
The major toxin produced by A. ochraceus and the other closely related species mentioned above is ochratoxin A. This toxin is discussed under Penicillium verrucosum, the most important species producing ochratoxins. A. ochraceus also produces penicillin acid, a mycotoxin of lower toxicity and uncertain importance in human health.
Symptoms
For a discussion of the symptoms of ochratoxin A poisoning, see under Penicillium verrucosum. Penicillic acid is a hepatocarcinogen in some animal species, and has also been reported to affect the heart (Reiss, 1988).
Distribution in nature and foods
The natural habitat for A. ochraceus and the other closely related species is drying or decaying vegetation, seeds, nuts and fruits. A. ochraceus is widely distributed in foods, especially dried foods, with records from such diverse sources as various kinds of beans, dried fruit, biltong and salt fish (Pitt and Hocking, 1985a). Nuts, including peanuts, pecans and betel nuts, are also a major source. Although A. ochraceus has been isolated from a wide range of cereals, including barley, wheat, flour and rice, records are rather infrequent (Pitt and Hocking, 1985a).
A. versicolor
The most important species in the "Aspergillus versicolor group" of Raper and Fennell (1965), A. versicolor is now classified correctly in Aspergillus Section Versicolores (Gems et a/. 1985).
Identification
Slowly growing, low, dense, green colonies on CYA or MEA are characteristic of A. versicolor, and a variety of other Aspergillus and penicillium species. Reddish drops of exudate and orange or reddish brown reverse colours on CYA are valuable additional characters. The microscopic appearance of A. versicolor is more distinctive. Stipes are relatively short, with heavy yellow walls, vesicles are usually ellipsoidal rather than spherical, and bear metulae and phialides only over the upper two thirds. Conidia are very small (22.5 µm in diameter), smooth walled, and coloured pale green (Raper and Fennell, 1965; Pitt and Hocking, 1985a; Klich and Pitt, 1988a).
The closely related species A. sydowii has many characteristics in common with A. versicolor, but colonies are coloured blue rather than grey green. A. sydowii produces no mycotoxins of significance.
Toxins and toxicity: sterigmatocystin
Sterigmatocystin is produced by several species of fungi (Cole and Cox, 1981), but A. versicolor is by far the most important. This toxin is a precursor of the aflatoxins, but acute oral toxicity is low because of very low solubility in water or gastric juices. Hence sterigmatocystin is unlikely to be responsible for acute poisoning outbreeds in man or animals (Terao, 1983). Due to this insolubility, experimental doses given to animals are only adsorbed to a small extent. As absorption is dependent on the method of administration, LD50 figures lack accuracy: literature values very from 60 to 800mg/kg body weight. Even low doses cause pathological changes to the livers of rats, however(Terao, 1983).
As a liver carcinogen, sterigmatocystin appears to be only about 1/150th as potent as aflatoxin B1, but this is still much more potent than most other known liver carcinogens. Levels as low as 15 µg/day fed continuously, or a single 10 mg dose, caused liver cancer in 30% or more of male Wistar rats (Terao, 1983). Sterigmatocystin has the potential to cause human liver cancer.
Symptoms
Again due to its very low solubility, oral ingestion of sterigmatocystin is undoubtedly asymptomatic. Detection of a disease syndrome due to this mycotoxin appears to be a very difficult epidemiological problem indeed.
Distribution in nature and in food
Distribution of A. versicotor in nature appears to be widespread though sparse. Its slow growth and exrophilic nature undoubtedly contribute to poor recovery in surveys of fungal populations.
A. versicolor has been reported from a very wide variety of foods. It occurs at harvest in some crops such as wheat, but is much more commonly isolated from stored commodities, particularly wheat, corn, flour and rice. It has also been found in a variety of nuts, fermented and cured meats and biltong, spices and cheese (Pitt and Hocking, 1985a).
Penicillium
Penicillium is a large genus, with 150 recognised species (Pitt, 1979a), of which 50 or more are of common occurrence (Pitt, 1988a). All common species grow and sporulate well on synthetic or semi-synthetic media, and usually can be readily recognised at genus level.
Classification of Penicillium is based primarily on microscopic morphology: the genus is divided into subgenera based on the number and arrangement of phialides (elements producing conidia) and metulae and rami (elements supporting phialides) on the main stalk cells (stipes). The classification of Pitt (1979a) includes four subgenera: Aspergilloides, where phialides are borne directly on the stipes without intervening supporting elements; Furcatum and Biverticillium, where phialides are supported by metulae; and Penicillium, where both metulae and rami are usually present. The majority of important toxigenic and food spoilage species are found in subgenus Penicillium.
Identification
Identification of Penicillium isolates to species level is not easy, preferably being carried out under carefully standardised conditions of media, incubation time, and temperature. As well as microscopic morphology, gross physiological features, including colony diameters, colours of conidia and colony pigments, etc. are used to distinguish species. A complete taxonomy of all species may be found in Pitt (1979a) or Raper and Thom (1949), though the latter is out of date now. Pitt and Hocking (1985a) and Pitt (1988a) provide keys and descriptions to many common species. A computer assisted key to the common species will be available shortly (Pitt, 1988b).
The most important toxigenic Penicillium species in foods are P. citreonigrum (which makes citreoviridin), P. citrinum (citrinin), P. crustosum (penitrem A), P. islandicum (cyclochlorotine, islanditoxin, luteoskyrin and erythroskyrin), and P. verrucosum (ochratoxin A). Each of these species and toxins is discussed below.
Penicillium citreonigrum
The Oriental disease known as "beriberi" has traditionally been regarded as a nutritional disease, an avitaminosis. However, beriberi is more than a single disease, and one form of it, known in Japan as acute cardiac beriberi, has been established to be a mycotoxicosis. The work of Sakaki in the 1890s (Ueno and Ueno, 1972) implicated mouldy "yellow rice" as a probable cause of acute cardiac beriberi, and led to a ban on the sale of yellow rice in Japan in 1910. The disease subsequently disappeared from Japan.
Uraguchi (1969) and Ueno and Ueno (1972) showed that acute cardiac beriberi was due to the growth in rice of P citreonigrum (synonyms /? citreoviride, P. toxicarium), and that the mycotoxin responsible was citreoviridin.
Acute cardiac beriberi in Japan is now only of historical interest. However, P. citreonigrum and Citreoviridin may still occur in other parts of Asia. Citreoviridin is also produced by P. ochrosalmoneum (see below).
P. citreonigrum is a member of subgenus Aspergilloides. P. citreonigrum was described in 1901, but ignored in many more recent taxonomies because the description was meagre. However, on the basis of neotypification by Biourge in 1923, Pitt (1979a) revived the name, which has priority over the more commonly used name P. citreoviride.
Identification
When grown on standard identification media (Pitt, 1979a; 1988a), P. citreonigrum is a distinctive species. Colonies grow quite slowly (after 7 days: 2080 mm diameter on CYA, 22-26 mm diameter on MEA, 0-10 mm at 37Q) sparsely produce pale grey green conidia, and exhibit yellow mycelium, soluble pigment and reverse colours. Penicilli consist of small clusters of phialides only, stipes are slender and not epically enlarged, and conidia are spherical, smooth walled and tiny (1.8-2.8 µm in diameter).
Toxins and toxicity
As noted above, P. citreonigrum produces citreoviridin, the cause of acute cardiac beriberi. Citreoviridin is also produced by Penicillium ochrosalmoneum, which has an ascomycete state, Eupenicillium ochrosalmoneum. It is not closely related to P. citreonigrum.
Citreoviridin is an unusual molecule consisting of a lactone ring conjugated to a furan ring, with a molecular weight of 402 (Cole and Cox, 1981). It is a neurotoxin, acutely toxic to mice, with intraperitoneal and oral LD50s of 7.5 mg/kg and 20 mg/kg respectively (Ueno and Ueno, 1972).
Symptoms
In several animal species, Citreoviridin caused vomiting, convulsions, ascending paralysis and repiratory arrest. Less frequent signs were reported to be ataxia, enforced movements or stiffness in the extremities, and later cardiovascular disturbance, flaccid paralysis and hypothermia (Uraguchi, 1969). In higher mammals, neurological symptoms or depressed sensory responses are also evident.
In man, the disease called acute cardiac beriberi has been recognised for the past three centuries (Ueno and Ueno, 1972). The disease frequently occurred in young healthy adults, and death could occur within a few days. In many respects, the acute symptoms in man paralleled those in animals (Uraguchi, 1969).
Distribution in nature and in foods
P. citreonigrum is not a commonly isolated species, but it is widely distributed (Pitt and Hocking, 1985a). According to Miyake et al. (1940), as reported by Uraguchi (1969), P. citreonigrum grows in rice after harvest, when the moisture content reaches 14.6%. At 1% higher moisture, other fungi will overgrow it, so the moisture band for invasion is narrow. The fungus is reported to be favoured by the lower temperatures and shorter hours of daylight occurring in the more temperate rice growing areas.
P. ochrosalmoneum is also an uncommon species in most environments. However, it has been found colonising unharvested corn in the USA, where it may produce citreoviridin under natural conditions (Wicklow and Cole, 1984).
P. citrinum
Described in 1910, P. citrinum has been a well recognised species for most of this century. Its importance in the present context lies not so much in the production of a mycotoxin of particular human significance, but in its ubiquity, so that any toxins produced can be expected to be very widely distributed in food and feed supplies. P. citrinum is the major producer of citrinin, a compound discovered during the 1940s, and considered then to be a potentially valuable antibiotic. Like several other Penicillium metabolises, it proved to be too toxic for therapeutic use, and became known in time as a potentially hazardous mycotoxin.
Classified in subgenus Furcatum section Furcatum (Pitt, 1979a), P. citrinum is a very well circumscribed species, accepted without controversy for many years.
Identification
The most distinctive feature of P. citrinum is its benicillus, which consists of a cluster of three to five divergent metulae, usually epically swollen. Under the stereomicroscope, the phialides from each metula usually bear conidia as a long column, producing a distinctive pattern which can be of diagnostic value. Colonies of this species on CYA and MEA are of moderate size (25-30 mm and 14-18 mm, respectively), with the smaller size on MEA also a distinctive feature. Growth normally occurs at 37°C but colonies seldom exceed 10 mm after 7 days.
Toxins and toxicity
Citrinin is the only mycotoxin produced by P. citrinum. P. citrinum is the major producer of this toxin, but production by P. expansum and P. verrucosum has also been reliably reported. Literature citations indicate that at least 22 Penicillium species have been reported to produce citrinin, but the great majority of these are either regarded as synonyms, or require confirmation (Pitt and Leistner, 1988).
Citrinin is a significant renal toxin to monogastric domestic animals, including pigs (Frlis et al., 1969) and dogs (Cariton et al., 1974). Domestic birds are also susceptible: citrinin causes watery diarrhoea, increased food consumption and reduced weight gain due to kidney degeneration in chickens (Mehdi et al., 1981), ducklings and turkeys (Mehdi et al., 1984). LD50 figures have been reported to be imprecise due to delayed deaths. The oral LD50 in mice is 110 mg/kg (Scott, 1977).
Chronic kidney degeneration in Danish pigs was at first considered to be due to citrinin (Frlis et al., 1969), but it was later shown that ochratoxin A was more significant (Krogh et al., 1973).
The effect of citrinin on humans remains undocumented. However, kidney damage appears to be a likely result of prolonged ingestion.
Symptoms
Like many other mycotoxins, citrinin is essentially asymptomatic in animals (as distinct from birds), causing a nonspecific deterioration in kidney function.
Distribution in nature and foods
P. citrinum is a ubiquitous fungus, and has been isolated from nearly every kind of food surveyed for fungi. The most common sources are milled grains and flour, and whole cereals, expecially rice, wheat and corn (Pitt and Hocking, 1985a). Instances of spoilage are rare, but growth, and toxin production, are likely to be a common occurrence.
P. crustosum
Described in 1930, P. crustosum was regarded by Raper and Thom (1949) as uncommon, with its major source has been shown to be P. crustosum in consequence, this species remained more or less unrecognised until recently. The occurrence of serious outbreaks of tremorgenic and other neurotoxicity in domestic animals was linked to other species, notably P. cyclopium (now P. aurantiogriseum; Pitt, 1979a), P. palitans (now P. commune; Pitt, 1988a) and P. viridicatum. The toxin responsible is now known as penitrem A, and its major source has been shown to be P. crustosum (Pitt, 1979b). P. crustosum is now recognised as a very common species in foods and feeds (Pitt and Hocking, 1985a).
Identification
A member of Penicillium subgenus Penicillium, P. crustosum produces the large penicilli with rami, metulae and phialides characteristic of this subgenus. It is one of the faster growing species in section Penicillium, within that subgenus, producing dull green colonies with a granular texture on both CYA and MEA. Microscopically, P. crustosum is characterised by large rough walled stipes and smooth walled, usually spherical conidia. However, the most distinctive feature of typical isolates is the production of enormous numbers of conidia on MEA, which become detached from the colony when the Petri dish is jarred.
Toxins and toxicity
Naturally occurring compounds which can cause sustained trembling are rare, and most of those known are produced by fungi. One of the most potent is penitrem A, which is produced by P. canescens, P. crustosum, P. glandicola ( = P. granulatum; Seifert and Samson, 1985) and P. janozewskii ( = P. nigricans; Pitt, 1979a). P. crustosum is by far the most important source, as virtually all known isolates are producers, and it is by far the most commonly occurring of these species in foods. P. crustosum has also been reliably reported to produce cyclopiazonic acid and roquefortine, though rarely (Pitt and Leistner, 1988).
Penitrem A is a potent neurotoxin, with an intraperitoneal LD50 of 1 mg/kg in mice. Oral LD50 data do not appear to be available, but death or severe brain damage have been reported in field outbreaks involving sheep, cows, horses and dogs (Wilson et al., 1968; Hocking et al., 1988).
Symptoms
In laboratory animals, the main symptom of poisoning by penitrem A is the onset of sustained trembling, which may continue for long periods without appearing to interfere with the normal functions of the animal. Trembling has been sustained in experiments without apparent ill effects or residual effects for as long as 18 days (Jortner et al., 1986). However, relatively small increases in does (5 to 20 fold) can be rapidly lethal (Hou et al., 1971 a). Post mortem diagnosis of tremorgenic toxins such as penitrem A is virtually impossible, as no pathological effects are evident.
The symptoms of penitrem A are essentially the same as those of a range of other fungal tremorgens, including those from Claviceps paspali growing in Paspalum grass. Acremonium lolii growing in Lolium perenne (ryegrass), or any of several Penicillium and Aspergillus species growing in foods or feeds (Pitt and Leistner, 1988).
The potential hazard of penitrem A to man remains unknown, and puzzling. Its known toxicity to large domestic animals and dogs is such that it is unlikely to be non-toxic to humans. However, the only symptoms in man which can be attributed to P. crustosum have been unlikely, though quite well documented, instances of dizziness and vomiting after consuming beverages which contained mould growth . Recovery of patients was complete in all cases. Available evidence is fragmentary, and direct experimentation impossible. The role of penitrem A and perhaps other fungal neurotoxins in human illness or neurological disorders still awaits elucidation.
Distribution in nature and foods
P. crustosum is an ubiquitous spoilage fungus. Pitt and Hocking (1985a) reported isolating it from the majority of cereal and animal feed samples examined by them over more than a decade. P crustosum can cause spoilage of corn, processed meats, nuts, cheese and fruit, as well as being a weak pathogen on pomaceous fruits and cucurbits (Pitt and Hocking, 1985a). The occurrence of penitrem A in animal feeds is well documented (Wilson et al., 1968; Hou et al., 1971b). Its occurrence in human foods appears equally certain.
P. islandicum
Described by Sopp in 1912, P. islandicum attracted little attention from taxonomists until Raper and Thom (1949) recognised it. It has been accepted unchanged in concept since that time.
When tested against experimental animals, P. islandicum has been shown to produce several highly toxic compounds. However, the significance of P. islandicum as a toxigenic fungus and of the toxins themselves remains in doubt. The species is included here as a potential problem, rather than because of known outbreaks of disease.
Neither the name of this species nor its circumscription has materially altered during the 80 years since it was described. P. islandicum is a representative of subgenus Biverticillium, which is characterised by the production of penicilli with more than five appressed metulae, of metulae and phialides of approximately equal length, and of phialides which are acerose (shaped like a pine needle) (Pitt, 1979a; 1988).
Among species which produce the penicillus type characteristic of subgenus Biverticillium, P. islandicum can be readily distinguished by its slowly growing, dense colonies with brilliant orange to brown colours in both mycelium and reverse. Colonies at 37°C are usually similar to those at 25°C Conidia are usually blue.
Toxins and toxicity
P. islandicum produces at least four mycotoxins, unique to the species. Cyclochlorotine and islanditoxin are chlorine-containing cyclic peptides which have the same toxic moiety, a pyrrolidine ring with two attached chlorine atoms, and share a number of other physical and chemical properties (Scott, 1977). Both compounds are very toxic: cyclochlorotine has an oral LD50 in mice of 6.5 mg/kg, while that of islanditoxin by subcutaneous injection was 3 mg/kg. Fed to mice at the rate of 40 µg per day, cyclochlorotine caused liver cirrhosis, fibrosis and tumors (Uraguchi et al., 1972).
Luteoskyrin is a dimeric anthroquinone and erythroskyrin a heterocyclic red pigment. Both are liver and kidney toxins, though less acutely toxic than cyclochlorotine. Luteoskyrin is also carcinogenic.
Because of the toxic "yellow rice" syndrome, described above under P. citreonigrum, Japanese scientists have taken a particular interest in P. islandicum, which also can cause yellowing of rice (Saito et al., 1971). However, the significance of the toxins produced by P. islandicum remains unclear.
Symptoms
Little information exists about the symptoms caused by the P. islandicum toxins.
Distribution in nature and foods
Reports of P. islandicum in nature have been infrequent (Pitt and Hocking, 1985a). Considering the striking appearance of colonies and the ease with which this species can be identified, the indications are that it is uncommon, at least in the temperate zones where most studies of Penicillia in foods have been undertaken.
P. varrucosum
For half a century or more, nephropathy has been an important disease in Danish pigs. Etiological studies first showed it to be associated with mouldy grain, and then with a fungus identified as P. viridicatum (Krogh and Hasselager, 1968). A representative isolate was first shown to produce oxalic acid and citrinin, and then ochraoxin A (Krogh et al., 1973). The major source of fungus and toxin was barley (Krogh, 1978).
Pitt (1987) showed that the major Penicillium species producing ochratoxin A was P. verrucosum, not P. viridicatum. P. verrucosum is the main source of ochratoxin A in temperate regions. In the tropics, Aspergillus ochraceus may be more significant.
Ochratoxin A is fat-soluble, and not readily excreted, so it accumulates in fatty tissues. In consequence, it poses a serious health risk to humans, especially in rural areas where pigs are not subject to rigorous inspection, and pork and bacon may contain high levels of ochratoxin.
P. verrucosum is classified in subgenus Penicillium, a large subgenus, and in section Penicillium, which includes many mycotoxigenic species of common occurrence in foods.
Identification
P. verrucosum is distinguished by slow growth on CYA and MEA at 25°C (17-24 mm and 10-20 mm after 7 days, respectively; Pitt, 1987), bright green conidia, clear to pale yellow exudate, and rough stipes (Pitt, 1988a). It is similar in general appearance to P. viridicatum, differing most obviously by slower growth, and to P. solitum, from which it differs by having green rather than blue conidia.
Toxins and toxicity
P. verrucosum is the principal producer of the nephrotoxin ochratoxin A. Pitt (1987) reported that 47 of 84 P. verrucosum isolates (56%) from West German and Australian sources produced this toxin. A few isolates also produce citrinin (6 of 84, 7%; Pitt, 1987).
Ochratoxin A is also produced by some isolates of Aspergillus ochraceus (5 of 17, 30%; Ciegler, 1972). Ochratoxin A is an acute nephrotoxin, with oral LD50 values of 20 mg/kh in young rats and 3.6 mg/kg in day old chicks. It is also lethal to mice, trout, dogs and pigs (Scott, 1977). Necroses of the renal tubules and periportal liver cells were the main pathological changes observed after fatal doses.
In humans, ochratoxin A appears to be responsible for kidney degeneration, which in extreme cases can lead to death. Kidney failure rates in rural Scandinavian populations are high, and a possible cause is the ingestion of pig tissues containing excessive levels of ochratoxin A (Krogh et al., 1974).
Distribution in nature and foods
P. verrucosum has been reported almost exclusively from temperate zones. It is associated with Scandinavian barley: in one survey of farms where pigs were suffering from nephritis, 67 of 70 barley samples contained high levels of P. verrucosum, and 66 contained ochratoxin A (Frisvad and Viuf, 1986). This species has also been isolated quite frequently from meat products in Germany and other European countries. It appears to be uncommon elsewhere (Pitt and Hocking, 1985a).