Renal tumourigenesis in male rats in response to chronic dietary ochratoxin A

Abstract

The potency of ochratoxin A (OTA) as a renal carcinogen in the rat in response to lifetime administration by oral gavage is a basis of current concern about possible human risk from dietary exposure to the mycotoxin. In this study, dietary delivery of OTA was chosen as the mode of administration, since this mimics human intake of OTA-contaminated food more accurately than gastric intubation. Young male Fischer rats were given approximately 300 µg OTA/kg body weight (bwt) daily until they reached 333 g; thereafter their daily intake was held at about 100 µg. Renal tumours, mostly unilateral carcinomas, were first discovered at week 75 and total incidence reached 25%. Statistical comparison of total carcinoma incidence (20%) in this study with that of the classic US NTP study suggested that OTA was significantly less carcinogenic when administered in feed than when given by oral gavage. The finding may moderate perceptions of a putative risk of trace amounts of OTA in some foodstuffs to human health.

Keywords:

• Ochratoxin A
• mycotoxin
• dietary ochratoxin
• rat renal tumours
• rat renal carcinoma
• karyomegaly
• human risk assessment
• mononuclear leukaemia
• Fischer-344 rats
• Cox proportional-hazards regression

Introduction

Ochratoxin A (OTA) is one of the first fungal metabolites to be termed a mycotoxin on account of general experimental toxicity to rats (Van der Merwe et al. 1965) and was later shown to cause the disease mycotoxic porcine nephropathy (Krogh 1978). Subsequently its role as a renal carcinogen, particularly in male Fischer rats, was firmly established in the US National Toxicology Programme (NTP) (Boorman 1989).

Consequently, although there is no proven role for OTA in human nephropathy, its classification by the International Agency for Research on Cancer as a possible human carcinogen has raised questions concerning the risk to human health of trace amounts of the toxin that can be detected by sensitive analysis in some food commodities.

Since the NTP study, other lifetime experiments have been conducted also using oral gavage (Castegnaro et al. 1998; Son et al. 2003), leaving open the question of whether such administration in the rat is an ideal experimental model for assessing human responses to dietary intake. A previous acute study, for example, showed more overt toxicity when the toxin was given by gastric intubation compared to dietary delivery (Miljkovic et al. 2003). It also has to be noted that OTA occurs naturally only through biosynthesis by a small group of spoilage moulds. It is accompanied by small amounts of ochratoxin B, which is relatively inactive biologically, and sometimes by different co-metabolites as well as mycotoxins from other moulds.

The current study forms part of a research programme funded by the European Union to explore mechanisms of rat renal carcinogenesis induced by OTA. It was designed so that interim analysis of renal histology and changes in gene expression could be performed. For that purpose animals were randomly selected from the treated group, which ultimately developed a significant incidence of renal tumours.

The chosen protocol differed from previous studies in that OTA was administered in feed, which was homogenized with artificially-fermented wheat rich in ochratoxins and free of any other mycotoxins. The present paper describes the tumourigenic outcome of this lifetime study and focuses mainly on description of renal carcinoma rather than adenoma, since the occurrence of malignant tumours is the principal concern in human risk assessment.

Materials and methods

Production of ochratoxin A and formulation for dietary experiment

Standardized OTA production was performed by growing Aspergillus ochraceus isolate D2306 (Harris and Mantle 2001) in shaken solid substrate fermentations at 28°C for 2 weeks to yield a product containing 5–6 mg OTA/g. More specifically, 40 g of sterilized shredded wheat (Cereal Partners UK, Welwyn Garden City, UK) in 500 ml Erlenmeyer flasks was inoculated with a concentrated spore suspension in water (16 ml) and the flasks shaken at 200 rpm and 10 cm eccentric throw. An aliquot of each fermentation product was assayed for OTA-concentration by HPLC with diode array detection (Harris and Mantle 2001). Batches of fermentation product also contained ochratoxin B (OTB) equivalent to 5–10% of the amount of OTA. No other mycotoxins (e.g. penicillic acid, citrinin) were biosynthesized in this fermentation. Each week, a weighed amount of standardized fermentation product was soaked for 4 h in warm water made slightly alkaline with sodium bicarbonate. The mixture was homogenized (Braun Espanola S.A., Barcelona, Spain) into powdered standard (RM1) commercial rat feed (Special Diets Services, UK) to a final concentration appropriate for the required mean OTA intake for the treated animals. An example of a prepared diet was analysed independently at the Central Science Laboratory, York and shown to be close to the intended concentration (e.g. 5 mg OTA/kg). Each group of five animals was given 100 g of contaminated feed daily, which was always fully consumed. Preliminary studies had shown that this amount, provided daily in an aluminium foil dish on paper, satiated five rats with negligible wastage. A group of control rats, initially 30, was given standard pelleted feed ad libitum, primarily to monitor satisfactory satiation of intake requirement of the OTA-treated group by body weight. Ultimately, verification of the consistency of OTA administration was evident from monitoring of plasma OTA concentration.

OTA was administered in the diet daily to Male Fischer-344 rats (B. & K. Universal Ltd., Hull, UK), caged in groups of 5 for up to two years. As from their initial weight of ∼175 g, daily dietary intake of OTA was 300 µg/kg bwt., but was held at 100 µg/rat after animals became full grown and reached a bwt of ∼333 g. Subsequent weight gain for male rats is largely due to accumulated lipid that is poorly vascularized. Over the period of the study, animals were housed in cages on absorbent paper (changed daily) under controlled conditions (21 ± 1°C, 55 ± 10% relative humidity, 15–20 air changes/h, 12 h light-dark cycle). Growth of animals and their welfare was monitored by daily observation and regular weighing. All handling and procedures were carried out in accordance with the UK Animals (Scientific Procedures) Act 1986.

Measurement of the concentration of OTA in plasma

At about 3-monthly intervals, blood samples (200 µl) were taken from a tail vein under mild general anaesthesia, centrifuged, and plasma separated for validated quantitative OTA analysis contracted to the Central Science Laboratory, York.

Histological preparation

Animal tissues for standard histology were fixed in buffered formalin and processed automatically for 3 µm sections stained with haematoxylin and eosin at the Breast Pathology Laboratory, Guy's Hospital, London.

Statistical analysis

Kaplan-Meier survival curves and Cox proportional-hazards regression under competing risks were used for statistical analysis of the survival in the current study in comparison to the classic NTP study. For analysis, the causes of death were classified into carcinoma, leukaemia and other. The data on individual survival from the NTP study was also classified as above. Proportional hazards assumption was satisfied. SPLUS-6 (Insightful Corporation, Seattle, WA) was used for analysis.

Results

Throughout the experiment the contaminated feed was fully consumed by all groups of rats showing that OTA delivery was complete within the group. During the latter 18 months, plasma OTA concentration stabilized around 8 µg/ml, implying that steady state toxicokinetics had been established, in which daily intake of OTA was matched by its elimination, whether as free OTA, OT-alpha or as other minor metabolites or conjugates.

The average amount of OTA ingested daily, commencing at twice the highest intake of the NTP study (see Figure 1), was apparently well tolerated, since growth rate of treated rats matched closely that of controls (Figure 2).

Figure 1. Dosing regime for dietary administration of OTA to rats over a standard two-year exposure. Until rats reached a body weight of 333 g, the mean daily dose (300 µg OTA/kg bwt) in this study (grey) was twice that of the NTP highest dose (black). Thereafter the mean daily dose was maintained at 100 µg/rat, causing a gradual convergence towards the NTP highest dose. Near the end of the study the mean daily dose was only about one third above that of the NTP. The plotted curve approximately reflects the inverse of animal weight as shown in Figure 2.

Figure 2. Mean body weight of experimental rats (grey) compared to controls (black).

Also, the general maintenance of an adult body weight characteristic of male Fischer rats was not impaired by continuous ingestion of approximately 100 µg OTA daily. Unscheduled and unexplained deaths were very rare; at the beginning of the second year of exposure to OTA, 64 rats were in the treated group and became the basis for subsequent survival analysis (Figure 3) and the occurrence of adenomas and carcinomas.

Figure 3. Kaplan-Meier survival curve for the present experiment and for the high- and mid-doses of OTA in the NTP study.

No renal tumours were found in any control rats. In the treated group, the first renal tumour was discovered at 75 weeks (see Figure 4) in an animal that had been losing weight and was thus euthanized according to good animal practice.

Figure 4. Time points of discovery of renal tumours and weight of tumorous kidneys. Tumours occurred in the latter quarter of life. They were mostly carcinomas and unilateral, except in one animal where one microscopic adenoma was found in each kidney at the end-point of the experiment.

Notably, there was a concomitant loss of abdominal fat in this and in most other rats that had developed a large renal tumour. Subsequently, renal tumours were discovered at intervals up to and including the standard two-year end-point in a total of 16 rats (Figure 4). Occurrence of renal tumours was randomly distributed throughout the caged groups. The rate at which tumours were discovered increased from about 90 weeks. With the exception of one animal at termination of the study, tumours were unilateral and varied widely in mass and conformation (see Figures 4 and 5). In a few cases, a renal tumour was found in a rat that was obviously also developing leukaemia. No tumours were found in ureter or bladder.

Figure 5. Illustration of the range of renal carcinomas occurring in response to continuous intake of OTA. A–C: large tumours in situ, variously showing metastatic nodules and unrelated testicular seminomas; D–E: moderate-sized tumours; F–G: small tumours located at the kidney pole; H: small tumour emerging laterally, its origin indicated below in longitudinal section showing internal disposition with the main tumorous half on the right.

Generally, from external conformation and according to histology it was clear that tumours had arisen within the renal parenchyma, but in two cases (e.g. Figure 5A) it is an open question as to whether the tumour arose in the renal pelvis.

However, some larger renal tumours were associated with metastatic nodules located extensively along abdominal mesenteries, and occasionally carcinoma extended to lungs where this malignancy became probably the most life-threatening indirect feature of renal ochratoxicosis. Notable histopathological changes in kidney other than tumour were the consistent occurrence of karyomegalic nuclei in tubular epithelia, predominantly in the cortico-medullary region, and malignant leukocyte infiltration where leukaemia had already become a clinical disease. On the basis of tumour size, macroscopic conformation and/or histopathology, three of the rats with renal tumours were classified as having adenomas and in one of these the lesion was microscopic.

In some older animals from both control and treated groups, single or multiple seminomas occurred within one or both testes (Figure 5B).

The particular mononuclear leukaemia, to which ageing Fischer rats are prone (Moloney et al. 1970), was expressed overtly as jaundice and pathologically as marked splenomegaly, pale liver and dark kidneys. It presented an inevitable problem of a confounding malignant disease of rather rapid progression in 50% of the animals, thus requiring application of Cox proportional-hazard regression in order to make direct comparison with the NTP carcinoma occurrence data. A further complication was that, as in the NTP study, a few animals of both control and treated groups developed a subcutaneous sarcoma which ultimately impaired mobility and thus required euthanasia before the end of the study.

Discussion

This study illustrates, for the first time, the morphology of a range of tumorous kidneys from OTA-treated Fischer rats. Some rather small carcinomas and a few adenomas were discovered in an early stage of proliferation. Unfortunately, direct anatomical comparison of the present tumours with those in the NTP (Boorman 1989, Boorman et al. 1992) and other lifetime studies is constrained by a lack of available published data.

The present time course of incidence of rats with a renal carcinoma (20%) was compared with that in the NTP study by means of Cox proportional-hazard regression (see Table I, Figure 3). The hazard ratio of renal tubule carcinoma incidence in the present dietary study was significantly lower, 0.64 of that in the NTP high-dose (mean daily intake of 150 µg OTA/kg bwt) (p = 0.01), but not significantly higher (1.39 times; p = 0.1) than the NTP mid-dose (mean daily intake of 50 µg OTA/kg bwt).

Table I. Hazard ratios for renal tubular carcinoma in NTP highdose (mean daily intake of 150 mg OTA/kg bwt) and mid-dose (mean daily intake of 50 mg OTA/kg bwt) compared to present study (results from Cox proportional-hazards regression).

NTP-dose	Hazard ratio in present study vs. NTP study	95% confidence interval	p-value
High-dose	0.637	[0.453, 0.898]	0.0099
Mid-dose	1.385	[0.935, 2.053]	0.1000

Direct comparison with the NTP is made more difficult by the different dosage regimens (feed vs. gavage). Nevertheless, it is estimated that OTA has significantly less potency as a renal carcinogen when acquired as a dietary component, e.g. if ingested naturally, than when given once daily by oral gavage in corn oil during the ‘night’ phase of the animal's diurnal cycle. This finding is consistent with that of the acute study of Miljkovic et al. (2003), in which the effects of the toxin differed not only according to the mode but also the time of intake in the diurnal cycle. Approximate translation of the experimental protocol used in the NTP study to adult human circumstances would require that people wake in the night and drink about 300 ml of OTA-contaminated cooking oil. This does not mimic normal human food intake very well.

As in the NTP study, some of the male Fischer rats developed a spontaneous mononuclear leukaemia during the second year of life, presenting as jaundice, though without notable weight loss. Inevitably this was an undesirable experimental feature by causing premature death of some individuals and necessitating complex statistical analysis of the results concerning renal tumour incidence to take into account the competing risks of two malignancies. However, there was an advantage in that the random occurrence of leukaemia provided some interim samples of kidneys before most of the renal tumours were discovered. It was notable that no obvious adenoma was found in these tissues, indicating that the development of renal tumours had probably occurred over a period of rather few months during the second year of life.

Some renal tumours found in leukaemic rats were too small alone to have justified euthanasia at that stage. Since the confounding factor of leukaemia caused the animal's termination, predictions as to the potential tumour morphology, if the animal had lived on, cannot be made. Nevertheless, a unilateral adenoma may not be life-threatening, whereas a carcinoma will have serious consequences.

In the present study, leukaemia occurred in 50% of the animals, but in only about a quarter of the Fischer rats used in the mid-1980s for the NTP's OTA experiments. Notably, this apparent adverse drift in leukaemia incidence in ageing males of the Fischer strain had occurred also in the NTP study programme, since a 50% incidence was recorded during a recent study on fumonisin (NTP 2001).

The current study formed a context also for exploring changes in gene expression at intervals during the first year of continuous dietary administration of the toxin to investigate mechanisms of OTA-induced nephrotoxicity and carcinogenicity (Marin-Kuan et al. 2005, Cavin et al. 2005).

This study has emphasized some disadvantages in using the male Fischer as a model for OTA carcinogenicity. Dark Agouti males, for example, are not afflicted by leukaemia, do not become as obese as Fischer rats, and with about half the mature adult weight of the Fischer strain would require significantly less OTA for toxicological studies. Male Dark Agouti rats have therefore much to commend them for any subsequent lifetime studies on OTA-induced cancer.

The new finding concerning less potent carcinogenicity of OTA when given in diet, together with a small natural amount of the co-metabolite ochratoxin B, should be included in human risk assessment calculation. However, it is important also to bear in mind that inducing renal carcinoma experimentally in the laboratory rat may not be the ideal model for extrapolating to the human. Renal carcinomas do not normally develop spontaneously in the laboratory rat whereas in the human, a range of idiopathic renal cell and transitional cell carcinomas occur naturally.

Acknowledgements

We gratefully acknowledge skilled animal husbandry from staff of Imperial College Central Biomedical Services; contracted measurement of OTA in plasma by validated methodology at the Central Science Laboratory, York; histological preparation at the Breast Pathology Laboratory, Guy's Hospital, London; Dr G. Kirkwood and Dr M. Dobrota for helpful discussions, and the European Union for funding (Contract No. QLK1-CT-2001-011614).