Abstract
Penitrem A is a well-recognized tremorgenic mycotoxin produced by several Penicillium spp. However, most natural cases of penitrem A intoxication have been associated with Penicillium crustosum. Another Penicillium sp., Penicillium roqueforti, is used for the production of blue cheese and is found in silage and feeds. Penicillium roqueforti produces a mycotoxin, roquefortine C, which is also produced by P. crustosum. In contrast to a tremorgenic syndrome produced by penitrem A, roquefortine C toxicosis is characterized by a paralytic syndrome. Two cases of penitrem A intoxication in dogs are presented to investigate the use of roquefortine C as a biomarker for penitrem A exposure. The vomitus, serum, and urine were analyzed for roquefortine C and penitrem A. Results suggest that roquefortine C can be a sensitive biomarker for penitrem A intoxication. However, the detection of roquefortine C in the absence of penitrem A could merely suggest ingestion of blue cheese or spoilt silage or feed. A review of the literature did not identify any case positive for penitrem A but negative for roquefortine C. In cases in which both mycotoxins were detected, roquefortine C concentration was always higher than penitrem A concentration. In contrast, several cases have been described where the clinical history suggested penitrem A intoxication, but only roquefortine C was detected. In conclusion, roquefortine C can serve as a sensitive biomarker for penitrem A intoxication, but the clinical presentation needs to be considered for proper interpretation of its detection in the absence of penitrem A.
Penicillium spp. produce a variety of secondary fungal metabolites. These metabolites include mycotoxins, such as penitrems A through F, roquefortine C, mycophenolic acid, and thomitrems A and E. 13 Penitrem A is a well-recognized neurotoxic mycotoxin. Several cases of penitrem A intoxication, characterized by tremors and seizures, have been reported in dogs. 1,19 The sources associated with production of penitrem A include moldy cream cheese, moldy macaroni and cheese, moldy walnuts, moldy bread, rice, and compost. Recently, penitrem A intoxication associated with consumption of mold-infested canned soup was reported in humans. 9 Penicillium crustosum has been identified in all naturally occurring cases of penitrem A intoxication where a mold was isolated. This is not surprising because P. crustosum is the most common species of Penicillium isolated from food wastes of private households.
Roquefortine C, another secondary fungal metabolite, has primarily been associated with Penicillium roqueforti, which is used in the preparation of blue cheese. 14 Roquefortine C is also produced by P. crustosum. Roquefortine C has also been isolated from feed grains and silage. 7 Cases of roquefortine C intoxication, resembling penitrem A intoxication, have been reported in dogs; however, penitrem A analysis was not performed, and therefore, penitrem A exposure cannot be ruled out in those cases. In fact, all well-documented experimental cases of roquefortine C intoxication have resulted in anorexia and/or a paralytic disease, rather than a tremorgenic syndrome.
Several in vitro fungal culture studies have demonstrated that P. crustosum produce higher concentrations of roquefortine C than penitrem A. This finding is consistent with a case report involving 2 dogs in which stomach contents were analyzed for penitrem A and roquefortine C, with roquefortine C detected at higher concentrations in both animals. 19
In documented cases of penitrem A intoxication in dogs, the diagnosis was based upon the occurrence of tremorgenic clinical signs and detection of penitrem A in samples of stomach contents. To the authors' knowledge, samples other than stomach contents have not been evaluated in cases of penitrem A intoxication. In cases in which penitrem A has been detected, roquefortine C has also been found. In contrast, roquefortine C has been detected in suspect cases where penitrem A was not found. In one case series, 6 out of 7 roquefortine C–positive dogs (86%) were also positive for penitrem A. 3 One dog was positive for roquefortine C, but the sample size was insufficient to confirm penitrem A. In another survey involving 312 samples from cases involving convulsive disorders in dogs, 58 samples (19%) were positive for both penitrem A and roquefortine C, and 4 (1%) were positive for roquefortine C alone. 2 It is possible that in those 4 cases, penitrem A was present but at concentrations below the analytical test sensitivity. None of the samples were positive for penitrem A alone.
To obtain a rapid and sensitive method for the analysis of both penitrem A and roquefortine C in a variety of sample matrices, including serum and urine, a liquid chromatography–tandem mass spectrometry (LC-MS/MS) method was developed by our laboratory. 16 This method was used to investigate 2 suspected cases of penitrem A intoxication in dogs.
Case 1: A 35-kg, 5-year-old, neutered, male Labrador Retriever was presented to a local emergency clinic with a 2-hr history of intermittent, whole body tremors and seizures. The dog was nonambulatory and hypersensitive to noise and touch. Routine serum chemistry and complete blood cell count parameters were within reference ranges for dogs. A gastric lavage was performed. The lavage material was visually unremarkable. Shortly after presentation, the dog developed severe aspiration pneumonia. Based on the poor prognosis, the owners elected euthanasia. No necropsy was performed. However, samples from the gastric lavage, serum, and urine were submitted to the toxicology section of the California Animal Health and Food Safety (CAHFS) Laboratory System (Davis) for analysis. The initial list of differentials included strychnine, metaldehyde, and tremorgenic mycotoxin intoxication.
Analysis of the gastric lavage sample and urine were negative for strychnine and metaldehyde with detection limits (DL) of 1.0 mg/kg and 0.1 mg/kg, respectively. However, penitrem A and roquefortine C were detected in the gastric lavage sample at 0.6 μg/ml and 1.3 μg/ml, respectively (DL 5 25 ng/ml). Additionally, the serum sample contained 4.2 ng/ml penitrem A and 13.3 ng/ml roquefortine C (DL 5 1 ng/ml for both). Although penitrem A was not detected in the urine sample, 2.1 ng/ml roquefortine C was present (DL 5 1 ng/ml).
Case 2: A 9-month-old, spayed, female dog (Yorkshire Terrier–Chihuahua cross) was presented to a veterinary clinic with generalized tremors. She was ambulatory and had a temperature of 40.5°C on presentation. The history included possible ingestion of almonds and carrots from the owner's garden. Emesis was induced using hydrogen peroxide, and the vomitus was found to contain parts of carrots, potatoes, and mushrooms. The dog was hyponatremic with a sodium concentration of 108 mEq/l (reference range: 144–160 mEq/l); all other serum chemistry and complete blood cell count parameters were within reference ranges. The dog responded positively to fluids and diazepam. The vomitus material was submitted to CAHFS for analysis of penitrem A. Roquefortine C was detected at 14.1 ng/ml. Penitrem A was present but below the limit of quantitation of 1 ng/ml. Therefore, from diagnostic purposes, penitrem A concentration, even at its DL, can be of clinical significance. In such samples, roquefortine C is likely to be a better biomarker for penitrem A exposure.
Penitrem A enhances the release of aspartate and glutamate from cerebrocortical and spinal–medullary synaptosomes. 10 In addition, there is a reduction in the stimulus-induced release of γ-aminobutyric acid (GABA) and glycine. 4 The enhanced excitatory neurotransmitter release, in combination with reduced inhibitory neuro-transmitter release, leads to the overall excitatory effects associated with penitrem A intoxication. Although it was believed that penitrem A intoxication was not associated with histopathologic lesions, 12 more recently, cytoplasmic vacuolation in the neurons of the cerebellar cortex and swelling of astrocyte and Purkinje cells have been reported in rats. These lesions have been attributed to high Ca2+ influx resulting from N-methyl-D-aspartate (NMDA) receptor activation. 4,6 No cumulative effect of penitrem A exposure has been reported. Tremors associated with acute intoxication can be effectively treated with GABA agonists, such as diazepam, or perhaps more effectively by barbituates. 6,12
Several reports attribute neurotoxic properties to roquefortine based on the work of Frayssinet and Frayssinet (cited by Scott et al., 1975). 15 The brief citation reported that intraperitoneal (IP) administration of 10 mg/kg of roquefortine caused muscle contractions and ataxia in mice. At higher doses (50–100 mg/kg) of roquefortine, prostration with intermittent seizures was reported. Death was observed within a few hours of such presentation, and the estimated IP median lethal dose (LD50) in male mice was estimated to be 15–20 mg/kg. However, details of their experiment were not published and remain unclear. An attempt to reproduce this observation using pure roquefortine C has been unsuccessful, raising questions about the previously reported tremorgenic effects of roquefortine C. In the current study, the estimated IP LD50 in mice ranged from 169–184 mg/kg (10 times the previously suggested LD50). Onset of clinical signs was observed within 15 min of dosing and included a clear nasal discharge. Animals that survived recovered within 4–6 hr, whereas animals that died became quiescent, anorexic, and usually died within 24–48 hr. In another study, chickens dosed with penitrem A showed tremors within 30 min after oral dosing, whereas chickens fed an undetermined concentration of roquefortine C developed ataxia, depression, and lethargy. 18 Reversible paralytic effects have also been reported in cows that ingested P. roqueforti–contaminated feed grains containing an average roquefortine C concentration of 25.3 μg/g. 7 Differences between penitrem A and roquefortine have also been reported in sheep. Oral administration of 0.75–2 mg/kg body weight of penitrem A resulted in reversible tremorgenic effects in sheep. 11 The tremors persisted for 3–24 hr depending on the dosage. Forced movement resulted in ataxia and recumbency. All animals recovered after a brief 10-min rest period. No significant pathologic observations were made in the central and peripheral nervous system of the affected animals. Roquefortine C administered orally to sheep at dietary concentrations of 5 and 25 mg/kg (0.13 and 0.62 mg/kg body weight, respectively) for 16–18 days failed to trigger any behavioral, hematologic, reproductive, or premortem or postmortem lesions. 17 In several cell lines, penitrem A was reported to have a 50% inhibitory concentration (IC50) ranging from 6.8–21.7 μg/ml, whereas roquefortine C was not toxic at the highest limit of its solubility. 5
Previously, the diagnosis of penitrem A intoxication has been based on identification of P. crustosum or detection of penitrem A in contaminated food, vomitus/lavage material, or stomach contents along with the occurrence of compatible signs in the patient. Thin-layer chromatography (TLC) and gas chromatography–tandem mass spectrometry (GC-MS/MS) methods have been used to screen for roquefortine C and penitrem A. 2 However, such methods are either nonconfirmatory (TLC), require derivatization (GC-MS/MS), or lack sensitivity. In addition, they are time consuming. In the current study, a rapid and sensitive method using LC-MS/MS for detection of penitrem A and roquefortine C in serum and urine that overcomes previous limitations [27] was developed. The kinetics of penitrem A and roquefortine C is not well established. In 1 rat study, 3.6–5.7% of roquefortine C was eliminated in urine, whereas 45–76% of roquefortine C was excreted in feces within 1 day of exposure. 8 Such data could not be found for penitrem A. Although both penitrem A and roquefortine C were detected in the lavage and serum samples from the affected dog in the case described above, it is noteworthy that only roquefortine C was found in the urine sample. In addition, roquefortine C was detected at higher concentrations than penitrem A in all tested samples where both were present. Therefore, in the absence of a sensitive analytical test, penitrem A might not be detected, but roquefortine C could still be detected.
Roquefortine C does not appear to be responsible for the tremorgenic signs associated with the ingestion of P. crustosum–contaminated food. However, it can be a valuable diagnostic marker for penitrem A exposure. Penitrem A concentration, even at its DL (1 ng/ml), is of clinical relevance in several biologic samples. In such samples, roquefortine C is likely to be present in higher concentrations and can prove to be a good biomarker. The described clinical cases suggest that penitrem A and roquefortine C can be detected in samples at low concentrations using LC-MS/MS and that roquefortine C may serve as a more sensitive way of diagnosing penitrem A intoxication. The presented cases also suggest that only roquefortine C may be found in some biologic samples in cases of confirmed penitrem A intoxication.