Critical Review
The Importance of Source-Dependent Bioavailability in Determining the Transfer of Ingested Radionuclides to Ruminant-Derived Food Products N I C H O L A S A . B E R E S F O R D , * ,† R O B E R T W . M A Y E S , ‡ A N D R E W I . C O O K E , §,| C A T H E R I N E L . B A R N E T T , † BRENDA J. HOWARD,† C. STUART LAMB,‡ AND G. PAUL L. NAYLOR⊥ Centre for Ecology and Hydrology, Merlewood Research Station, Grange-over-Sands, Cumbria, LA11 6JU, U.K., Macaulay Land Use Research Institute, Craigiebuckler, Aberdeen, AB15 8QH, U.K., Department of Biological & Nutritional Science, University of Newcastle, Newcastle Upon Tyne, NE1 7RU, U.K., and Food Standards Agency, Contaminants Division, Ergon House, 17 Smith Square, London, SW1P 3WG, U.K.
The transfer of radioactivity to the milk and meat of farm animals is likely be a major exposure pathway of human populations, following an environmental release of radioactivity. The importance of source-dependent bioavailability in determining absorption from the ruminant gastrointestinal tract of the radiologically significant radionuclides (radiocesium, radiostrontium, radioiodine, and plutonium) is reviewed. The requirements for and suitability of in vitro methods of determining bioavailability for absorption of these radionuclides is also assessed. Radiocesium absorption varies over a 50-fold range, depending upon dietary source. Source-dependent bioavailability is therefore an important factor in determining the radiocesium contamination of ruminant-derived food products, and reliable in vitro techniques have been developed to rapidly determine its bioavailability. In contrast, under conditions of adequate calcium intake, the absorption of radiostrontium will not be greatly influenced by the dietary source. Results of in vitro extractions of radiostrontium could be misleading, as they indicate differences in bioavailability that are not observed in the animal. Absorption of radioiodine is complete and independent of source. There is currently a lack of data on plutonium absorption in ruminants. However, on the basis of the limited data available, in vitro incubations provide a measure of the relative bioavailabilities of plutonium from different sources. The applicability of the techniques reviewed to the study of other radionuclides and heavy metals is discussed.
Introduction Following a release of radioactivity into the environment, the potential for transfer of deposited contaminants, such as * Corresponding author e-mail:
[email protected]; tel: 44 15395 32264; fax: 44 15395 35941. † Merlewood Research Station. ‡ Macaulay Land Use Research Institute. § University of Newcastle. | Present address: Farming and Rural Conservation Agency, Woodthorne, Wolverhampton, WV6 8TQ, U.K. ⊥ Food Standards Agency. 10.1021/es0000697 CCC: $19.00 Published on Web 09/22/2000
2000 American Chemical Society
radioisotopes of cesium, strontium, iodine, and plutonium through food chains to humans may be of major concern. Predictive models used to estimate the likely contamination of food products derived from grazing animals tend to use default transfer values with no distinction being made for different ingested sources (1-3). Many of these transfer values are derived from studies using ionic tracer radioisotopes or herbage contaminated by root uptake. However, these may not always be applicable. For instance, the availability of radiocesium from the initial deposit of Chernobyl fallout for transfer to grazing animals was considerably lower than values derived for ionic radiocesium or for that incorporated into grassy herbage via root uptake (4-6). To enable robust predictions of the transfer of radionuclides to milk and meat, the potential importance of source-dependent bioavailability needs to be fully understood. In this paper, we review the results from a series of experiments designed (i) to assess the importance of sourcedependent bioavailability in the transfer of radiocesium, radiostrontium, radioiodine, and plutonium to ruminantderived food products and (ii) to develop rapid in vitro and in vivo methods of assessing the bioavailability for absorption in the gut of these radionuclides, which could be used to provide realistic parameters for predictive models in the event of a deposition of radioactive fallout.
Bioavailability For the purposes of the following discussion, we have defined the bioavailability of a radionuclide for transfer to animal products as the degree of absorption across the gastrointestinal tract, as determined by characteristics of the ingested source. We have determined this as the true absorption coefficient (At), which has previously been defined as the fraction of ingested radionuclide that is transferred across the wall of the gut (7). In all of the studies discussed here, At has been estimated by using dual isotope techniques. By these methods, one isotope of a radionuclide is administered as the dietary source (e.g., 137Cs), and concurrent to this, a second isotope (e.g., 134Cs) is infused into the jugular vein. By collecting excreta and/or milk, the rate of endogenous fecal excretion (EF) of the dietary radioisotope or its rate of turnover in blood plasma can be determined. The true VOL. 34, NO. 21, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4455
TABLE 1. Summary of Radiocesium At Values Derived for a Range of Environmental Sources and Ionic Radiocesium (CsCl) dietary source
species (breed)
At (mean ( se)
ref
upland grasses grass hay clover hay pasture grass (Chernobyl NPP exclusion zone) CsCl CsCl CsCl CsCl CsCl Calluna vulgaris Chernobyl fallouta lowland peat soil saltmarsh siltb alluvial gley soilb peaty podzol soil peaty podzol soil
sheep (Scottish Blackface ewes) sheep (Suffolk ewes) sheep (Suffolk ewes) cattle (unknown dairy breed) sheep (Scottish Blackface ewes) sheep (ewes of mixed UK hill breeds) sheep (Texel ewes) sheep (Suffolk ewes) cattle (Friesian) sheep (Scottish Blackface ewes) sheep (crossbreeds 1-yr-old) sheep (crossbreeds 1-yr-old) sheep (Scottish Blackface ewes) sheep (crossbreeds 1-yr-old) sheep (Scottish Blackface ewes) cattle (Friesian)
0.88 ( 0.047 0.73 ( 0.070 0.76 ( 0.029 0.23 ( 0.030 0.84 ( 0.076 0.87 ( 0.008 0.78 ( 0.030 0.84 ( 0.020 0.71 ( 0.055 0.67 ( 0.034 0.65 ( 0.007 0.03 ( 0.01 0.12 ( 0.018 0.19 ( 0.03 0.02 ( 0.001 0.20 ( 0.055
13 21 21 22 13 19 21 21 23 13 7 24 13 24 23 23
a Perennial ryegrass harvested in southeast Scotland 3 d after being contaminated with fallout from the Chernobyl accident (see ref 14). of these sources were contaminated by marine discharges from the Sellafield reprocessing plant.
b
Both
absorption coefficient can then, in the example of dietary 137Cs, be estimated as either:
that estimates of absorption be standardized and determined as At (7, 13).
At )
Radiocesium
[137Cs intake (Bq d-1) - fecal 137Cs output (Bq d-1) + EF 137Cs(Bq d-1)]/137Cs intake (Bq d-1) (1)
where
EF 137Cs (Bq d-1) )
urinary 137Cs output (Bq d-1)
× urinary 134Cs output (Bq d-1) fecal 134Cs output (Bq d-1) (2)
or
At )
plasma 137Cs turnover rate (Bq d-1) 137
Cs intake (Bq d-1)
(3)
where
plasma 137Cs turnover rate (Bq d-1) ) plasma 137Cs activity (Bq L-1) 134 × Cs infusion rate plasma 134Cs activity (Bq L-1) (Bq d-1) (4) In eq 4, the activity concentration in blood plasma can be replaced by those in milk, urine, or tissues (other than those of the gastrointestinal tract) since radiocesium in each of these will be derived from the blood plasma. Similarly, measurements of urine can be replaced by measurements of milk in eq 2. The methodology relies upon the assumption that intravenously infused and gastrointestinally absorbed radioisotopes behave in the same way in plasma; it is unsuitable for radioisotopes where this is not the case. A full discussion of the relative benefits of the different methods of estimating At and suitable protocols by which such studies can be conducted are given in Mayes et al. (7). The At can be used in predictive models as the value of the parameter often described as the gastrointestinal absorption factor (f1) (8). However, previously reported f1 values have been estimated by a number of methods (8-12). Reported f1 values do not always represent the same measurement and, consequently, the actual amount of radioisotope that is truely absorbed. For this reason, we have previously recommended 4456
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 21, 2000
A number of authors have shown differences ranging over at least 2 orders of magnitude in the transfer of radiocesium from environmental sources to the milk and meat of ruminants (6, 14-17). The sources have included the following: herbage contaminated by a variety of mechanisms including root uptake, marine discharges from the Sellafield reprocessing plant, and Chernobyl fallout; ewes’ milk; and a range of different soil types. These authors have expressed their results in terms of the ratio of the equilibrium activity concentration of radiocesium in milk or meat to the daily intake of radiocesium [defined as the transfer coefficient by Johnson et al. (18)]. While observed differences in this ratio may be due to variation in bioavailability, it may not be the only cause as transfer coefficients can be influenced by a number of physiological parameters such as feeding and growth rates (19) or stage of lactation (20). Consequently, comparative transfer coefficients should not be used as a measure of source-dependent bioavailability, as has previously been suggested (e.g., ref 6) unless the other factors that influence transfer coefficients are fully taken into account. True absorption coefficients of radiocesium derived for environmental sources in ruminants are summarized and compared to values for ionic radiocesium administered as CsCl in Table 1. The estimates of absorption from environmental sources in Table 1 show nearly a 50-fold range in bioavailability. Estimates of At for radiocesium from grassy vegetation contaminated by root uptake are similar to those for CsCl. The lower At value measured for pasture grass collected from close to the Chernobyl nuclear power plant (NPP) and fed to dairy cattle appears to be largely the result of an especially low herbage dry matter digestibility (22). Results from studies with sheep indicate that radiocesium absorption does not vary with age, lactation, or size/breed (13, 21, 25). Values of At for CsCl in dairy cattle are similar to those for sheep (23). However, there are discrepancies between radiocesium At values determined for sheep and cattle from other sources fed to the two species; potential reasons for this are discussed later. The data in Table 1 demonstrates that source-dependent bioavailability is one of the major parameters determining the degree of contamination in the milk and meat of ruminant livestock. An ability to be able to predict radiocesium bioavailability rapidly and without the need for animal
FIGURE 1. A comparison of the in vitro extractability of radiocesium with that of its true absorption as measured in sheep from Singleton et al. (27) and Beresford et al. (30). The relationship shown for the prediction of At is that presented by Singleton et al. (27). experimentation is desirable, especially in the event of an accidental release of radioactivity. A number of studies have considered the use of in vitro rumen liquor incubations as a method of determining radiocesium bioavailability (26-28) and also the radiocesium binding abilities of different clays (29). All of the studies to determine radiocesium bioavailability used rumen liquor obtained from sheep incubated at a temperature between 37 and 39 °C; incubation times ranged from 1 to 48 h. Different soil types were extracted using rumen liquor incubations by Cooke et al. (28); a range of conditions including the control of pH were studied. Later, the true absorption coefficient of radiocesium from two of the soils was investigated in sheep (24). The comparative availability of radiocesium from the two soils was similar in the in vitro and in vivo trials. Salbu et al. (26) considered a range of contaminated vegetation sources and forest soil litter; in vitro results were compared with measurements of radiocesium transfer to goats’ milk made by Solheim Hansen and Hove (6) for the same sources. Although the results of the (sheep) rumen liquor digests compared well with those from the animal trials, so did extractions with NH4OAc and stable cesium chloride. Consequently, the authors recommended that the inorganic extractants be used on the basis of experimental simplicity. Singleton et al. (27) considered a range of inorganic extractants and simulated digestion procedures during the development of an in vitro bioavailability protocol. The radiocesium sources used were selected with the aim of giving a range of bioavailabilities. Concurrent to the in vitro extractions, the sources were also administered to sheep to enable the determination of At. The percentage of radiocesium removed from the sources during rumen liquor incubations (without control of pH) were generally in agreement with the At estimates, although availability from one of the sources (bentonite) was markedly underpredicted. Results from 2-h incubations with 0.1 M CsCl (20 mL:1 g of sample), followed by filtration through a 0.2-µm membrane, were well correlated with At measurements for all of the sources (Figure 1). This extraction procedure was therefore recommended as a method of being able to make rapid predictions (
