Print version ISSN 1516-8913
Braz. arch. biol. technol. vol.47 no.6 Curitiba Nov. 2004
FOOD SCIENCE AND TECHNOLOGY
Rejane Maria Cirra ScaffI, II; Vildes Maria ScusselI, *
IDepartamento de Ciência e Tecnologia de Alimentos; Centro de Ciências Agrárias; Universidade Federal de Santa Catarina; C. P. 476; 88040900; Florianópolis - SC - Brazil
IIDepartamento de Ciências Morfológicas; Centro de Ciências Biológicas; Universidade Federal de Santa Catarina; 88040900; Florianópolis - SC - Brazil
Corn flour, "canjica" (corn grits), corn flakes and popcorn for human consumption, commercialized in Santa Catarina (n=82), were analyzed in order to detect the presence of fumonisins B1 (FB1) and B2 (FB2). From the samples, 92.68% showed detectable levels of Fumonisins (FBs). Corn flour showed the highest level of contamination (91.5%) with average levels of 3.811 and 5.737 mg/g for the home-processed and industrialized products, respectively. The next most contaminated product was popcorn with a average of 2.872 mg/g and an occurrence in 91.6% of the samples. All samples of corn flakes were contaminated with an average of 1.307 mg/g. The product with the lowest levels of FBs was "canjica" with a average contamination of 0.732 mg/g. These results indicated the need of monitoring corn-based products in this state.
Key words: Corn, fumonisins, Brazilian corn products
Farinha de milho, canjica, flocos de milho, milho de pipoca, destinados ao consumo humano e comercializados em Santa Catarina (n=82), foram analisados a fim de determinar a ocorrência de fumonisinas B1 (FB1) e B2 (FB2). Das amostras, 92,68 % apresentaram níveis detectáveis de FBs. A farinha de milho apresentou os maiores níveis de contaminação (91,5%) com níveis médios 3,811 e 5,737 mg/g para as de preparo artesanal e industrializadas, respectivamente. O segundo produto mais contaminado foi o milho de pipoca com uma média de contaminação de 2,872 mg/g e ocorrência em 91,6% das amostras. Todas as amostras de flocos de milho apresentaram contaminação com uma média de 1,307 mg/g. O produto com menores níveis de FBs foi a canjica com contaminação média de 0,732 mg/g. Estes resultados indicam a necessidade de monitoramento dos produtos derivados de milho em nosso Estado, ressaltando-se que os níveis mais expressivos foram encontrados em produtos comercializados no Sul e Oeste de SC, regiões agrícolas, marcadamente colonizadas por descendentes de italianos, consumidores habituais de produtos derivados de milho, particularmente a polenta, preparada a partir de farinha de milho.
Fumonisins, mycotoxins produced by Fusarium sp (F. verticillioides and F.proliferatum) are found chiefly in corn and corn-based products. In this group, the mycotoxin that most commonly contaminates corn naturally is the fumonisin B1 (FB1), which may be found in either grains used for human and animal consumption.
Since its discovery (Bezuidenhout et al.,1988) fumonisin has been associated with diseases in animals, such as LEME in equine (Kellerman et al.,1990), pulmonary edema, hydrothorax and cardiovascular disturbances in pigs (Harrison et al., 1990, Gumprecht et al., 2001), including nephrotoxic, hepatotoxic and imunosupressing effects, in various animal species (Marasas, 1996).
Although their effect on humans is still not clearly established, statistically it is associated with esophageal cancer (EC) in some regions such as Transkei/ South Africa (Sydenham et al., 1990; Rheeder et al., 1992), Linxian/China (Chu et al., 1994). FBs are considered possible carcinogenic for human, Group 2B (IARC, 2002). Francheschi et al. (1990) linked the rise in the development of EC in north-eastern Italy to increased corn consumption, in which polenta is the main basic food (Doko and Visconti, 1994) with a possible relation to a human carcinogenic In Brazil, it has been related to outbreaks of toxicosis in animals (Meirelles et al., 1994; Sydenham et al, 1992). There are records of the occurrence of fumonisins in corn and corn-based products, particularly in the Southern and Southwestern regions of the country, with levels that vary from 0.02 to 49.31m g/g for FB1 and from 0.05 to 29.16 mg/g for FB2 (Camargos et al., 2000; Hirooka et al., 1996; Ono et al., 2000, 2001, 2002; Orsi et al., 2000).
In Santa Catarina, FBs studies reported up to 97.3 % of the corn sold is contaminated. Hermanns et al. (2000) reported levels that ranged from 1.12 to 31.88 mg/g in corn from the western region. A study involving three regions of the state (West, North and South) by van der Westhuizen et al. (2003) found levels ranging from 0.02 to 18.74 mg/g.
There is a scarcity of publications on fumonisins in corn-based products, destined to human consumption in Brazil. Machinski et al. (2000) analyzed 81 samples of corn-based products, collected in Campinas (SP), where fumonisins were found in 54 % of the samples with levels between 0.02 and 4.93 mg/g. In a study carried out in Argentina, Hennigen et al. (2000) analyzed 35 samples of corn flour and corn grits, destined to human consumption and found fumonisins in 91.4 % of the products. The levels varied from 0.038 to 4.987 mg/g and from 0.015 to 1.818 mg/g for FB1 and FB2, respectively.
These facts show that it is imperative to evaluate the risks to which the population is exposed considering, among other aspects, that the Southern Region of Brazil shows the greatest incidence of EC in the country (INCA, 1999), and that fumonisins could be one of the factors related to this problem (Scaff and Scussel, 1999; Bolger et al., 2002). In Santa Catarina, this is the first evaluation of the levels of fumonisins in corn-based products destined to human consumption, particularly corn flour.
MATERIAL AND METHOD
Samples: Home-produced and industrialized corn flour (47), pre-cooked corn flakes (11), canjica (corn grits) (12), popcorn (12), (total: 82 samples -1 kg), commercialized in different regions of SC (North, Itajai Valley, Florianópolis, Mountain Region, West and South), but processed in the Central-West, Southeast and South of Brazil were collected from January to June of 2001 according to official methodology of the Ministry of Agriculture (BRASIL, 1976). The products were collected from supermarkets, mills and cooperatives, in different regions of the state.
Each sample was ground in mill (250 mm mesh), homogenized and reduced through the technique of quartering up to 250 g (sub-sample) and from this, 50 g portions were removed for analysis.
Analysis of fumonisins
AOAC official method n° 995.15 (AOAC, 2000), based on Shephard et al. (1990), validated in a collaborative study by Sydenham et al. (1996) was used methanol:water (3:1) was utilized as extraction solvent and SPE cartridges - SAX were used for cleaning the samples. Derivatization of FBs was carried out with OPA and detection by HPLC, with a fluorescence detector (335 nm of excitation and 440 nm of emission). An isocratic mobile phase was used (methanol:sodium di-hydrogen phosphate buffer - 0.1M (77:23). Flow rate: 1 ml/min. For calibration, fumonisin standard solutions were utilized at concentrations of (a) 0.0125; 0.025 and 0.05 ml/ml for FB1 and (b) 0.005; 0.01 and 0.02 mg/ml for FB2, (ACN:H20/ v:v). The limit detection of method (LDC) was 0.04 mg /g to FB1 and 0.05 mg /g for FB2. The method recovery was of 88% and 69% for FB1 and FB2 respectively. The data were analyzed through simples and partial correlation analysis using statistical software (SAS, version 6.11)
RESULTS AND DISCUSSION
The data referring to corn-based products (corn flour, canjica, corn flakes and popcorn), commercialized in SC, but processed in different states (Goiás, São Paulo, Paraná and Santa Catarina) is shown in Tables 1 and 2. Fumonisins were found in 92.6 % of the samples (76/82) with levels that ranged from 0.060 to 21.823 mg/g. All the contaminated samples contained FB1 but FB2 was only detected in 42.6 % of the products.
According to the recommendation of the Food and Drug Administration (FDA, 2001) and Trucksess (2001), the maximum residues level (MRL) for FB1+FB2+FB3 permitted in foods for human consumption is 4 mg/g but 29.2 % of the contaminated samples in our study showed higher levels of fumonisins, referring only to the summing of FB1+FB2.
(a) Corn Flour
The product with the highest contamination level was corn flour (fubá), stressing the fact that 22 samples (46.8%) of these samples came from (a.1) home processed for their own or regional consumption, produced in conditions of hand labor, minimally processed, from the whole grain, maintaining the germ. The other 25 samples (53.2%) were (a.2) flours processed industrially, without the germ (Table 1).
(a.1) Corn flour home produced: Only one sample did not show contamination by fumonisins and the levels were between 0.086 and 19.222 mg/g with an average of 3.811 mg/g. The majority of these products (18 samples) was collected in the South and West of SC, agricultural regions of the state, where were normally produced and consumed. They showed marked cultural characteristics of the Italian colonization, since they were the usual consumers of polenta. In the South, higher levels were found (up to 19.222 mg/g) for FB1 + FB2, as well as a higher frequency when compared to the West (up to 10.699 mg/g) (Table 1).
(a.2) Industrialized corn flour: Of the samples analyzed 92 % showed contamination and the levels ranged from 0.145 to 21.823 mg/g with an average value of 5.737 mg/g. The highest contamination levels of the product were also found in the South of the state where all the samples analyzed were contaminated. The samples with the highest levels of contamination were found in products from SC, although expressive levels have been detected in products from the states of Goiás and São Paulo.
Since home-processed corn flour is prepared with the whole corn and, higher contamination be expected in this product, but no significant difference was observed between the results of these two types of corn flour. Considering that in rural areas the intake of corn-based products may vary from 11 to 39g per person/day (Machinski et al., 2000), these levels represent a health risk for the population. Thus, in these regions the possible intake of fumonisins to which this population would be exposed would be much higher than the level proposed by Gelderblom et al. (1995), in which the Tolerable Daily Intake is 800ng/kg bw/day.
The results shown here reinforced previously cited data, although in this case they refered to corn-based products, destined to human consumption and not only corn, where levels ranged from 1.12 to 31.88 mg/g (Hermanns et al., 2000) and between 0.02 and 18.74 mg/g (van der Westhuizen et al., 2003) for fumonisins. In Brazil some publications, mainly in the states of the South and Southeast, pointed to the presence of fumonisins in corn (table 3) with a very high frequency of samples with positive and mean levels generally over 4 mg/g. Machinski et al. (2000) found contamination by fumonisins in all the samples of corn flour analyzed (9/9) with levels ranging from 0.56 to 4.93 mg/g, and an average of 2.32 mg/g. The average levels found in our analyses were 5.165 mg/g, higher than the above, which demonstrated the urgency of monitoring the product in our region.
(b) Other products: Popcorn, corn flakes, canjica
The average contamination levels by fumonisins B1 + B2 in canjica, corn flakes and popcorn were relatively lower than those detected in corn flour, but some samples showed levels that caused concern (Table 2).
(b.1) Popcorn: The second most contaminated product was popcorn showing FB1 ranging from 0.102 to 7.346 mg/g (frequency of 11/12) and of 0.317 to 2.427 mg/g for FB2 (6/12). The average contamination levels were 2.872 mg/g. Machinski et al. (2000) found contamination levels in popcorn in the state of SP from 0.300 to 1.72 mg/g, lower than those found in our study. According to Bolger et al. (2001) in Argentina fumonisins in popcorn were detected, ranging from 1.084 to 14.241 mg/g, in 42 samples collected between May and June of 1999.
(b.2) Corn flakes: Fumonisins were found in all of the samples, with an average contamination of 1.307 mg/g. Since this product, in addition to being ground, was submitted to thermal processing (cooking), reduction of FBs was expected, either by removal or by degradation. The grinding process, dry or moist, may lead to removal of the mycotoxin, more significant, in the latter case, (Bolger et al., 2001). The effect of the thermal processing is still being widely discussed, but no definite conclusion has been reached. There are studies that indicate reduction (Jackson et al., 1996) while others show thermal stability of fumonisins (Alberts et al., 1990; Dupuy, 1993). According to Bolger et al. (2001), temperature > 150°C is necessary for a significant loss to occur. Nevertheless, there are cases recorded in which products thermally processed generally show lower levels of FBs than those not processed or minimally processed (Stak and Eppley, 1992).
(b.3) Canjica: These products showed lowest contamination levels, with FB1 ranging from 0.297 to 2.237 mg/g and occurrence in 83 % of the samples and FB2 detected in only one sample at a level of 0.098 mg/g. The average contamination of this product was 0.732 mg/g.
In Santa Catarina the results indicate the need for control and monitoring of corn and corn-based products, destined to human consumption, in view of the regional characteristics of communities of descendents from Italian immigrants, habitual consumers of polenta, a basic food, prepared from corn flour (fubá), minimally processed.
Knowing that southern Brazil shows a relatively high incidence of esophageal cancer (EC) and that the presence of fumonisins is related to this pathology in some regions of the world, associated with smoking, consumption of alcoholic beverages and "chimarrão" (an infusion of green leaf consumed hot) (Perin et al., 1990; Scaff and Scussel, 1999) these mycotoxins may be contributing in these statistics, since cancer is now seen as a multifactorial disease.
These data, resulting from regional and partial findings and analyzed along with other results obtained point to a need for frequent studies of a broader scope, which show the overall situation in all Brazil and take into consideration climatic and seasonal differences.
Alberts, J. F.; Gelderblom, W. C. A.; Thiel, P. G.; Marasas, W. F. O.; Van Schlankwik, D. J. and Behrend, Y. (1990). Efects of temperature and incubation period on production of fumonisin B1 by Fusarium moniliforme . Applied and Environmental Microbiology, 56 : (6), 1729-1733. [ Links ]
Almeida, A. P.; Fonseca, H.; Fancelli, A. L.; Direito, G. M.; Ortega, E. M. and Corrêa, B. (2002), Mycoflora and fumonisin contamination in Brazilian corn from sowing to harvest. Journal of Agricultural and Food Chemistry, 50, 3877-3882. [ Links ]
Association of Official Analytical Chemistry -AOAC. (2000), Official Methods of Analysis of Association of Official Analytical Chemists. Edited by Cunniff (Gaithersburg: AOAC Internacional), chapter 49, pp. 1-51. [ Links ]
Bezuidenhout, S. C.; Gelderblom, W. C. A.; Gorst-Allman, C. P.; Horak, R. M.; Marasas, W. F. O.; Spiteller, G. and Vleggaar, R. (1988), Struture elucidation of the fumonisins, mycotoxins from Fusarium moniliforme. Journal of the Chemical Society, 11, 743-745. [ Links ]
Bolger, M.; Coker, R. D.; Dinovi, M.; Gaylor, D.; Geldeblom, W.; Olsen, M.; Paster, N.; Riley, R. T.; Shephard, G. and Spijers, G. J. A. (2001), Fumonisins. Safety evaluation of certain mycotoxin in food. International Programme on Chemical Safetty World Health Organization. Geneva. FAO FOOD and Nutrition Paper 74, WHO - Food Additives Series, 47, 103-279. [ Links ]
Camargos, S. M.; Soares, L. M. V.; Sawaki, E.; Bolonhezi, D.; Castro, J. L. and Bortolleto, N. (2000), Fumonisin in corn cultivars in the state of São Paulo. Brazilian Journal of Microbiology, 31, 226-229. [ Links ]
Chu, F. S. and Li, G. Y. (1994), Simultaneous occurrence of fumonisin B1 and mycotoxins in moldy corn collected from the peoples Republic of China. Applied and Environmental Microbiology, 60 : (3), 847-852. [ Links ]
Doko, M. B. and Visconti, A. (1994), Occurrence of fumonisins B1 and B2 in corn and corn based human foodstuffs in Italy. Food Addiives and Contaminants, 11, 433-439. [ Links ]
Dupuy, J.; Le Bars, P.; Boundra, H. and Le Bars, J. (1993). Thermostability of fumonisin B1, a mycotoxin from Fusarium moniliforme, in corn. Applied and Environmental Microbiology, 59 : (9), 2864-2867. [ Links ]
Francheschi, S.; Bidoli, E.,Barón, A. E. and La Vechia, C. (1990), Maize and risk of cancer of the oral cavity, pharynx and esophagus in northeastern Italy. Journal. Nat. Cancer Institute, 82, 1407-1411. [ Links ]
Gelderblom, W. C. A.; Snyman, S. D.; Abel, S.; Lebepe-Mazur, S.; Smutts, C. M.; Van der Westhuizen, L.; Marasas, W. F. O.; Victor, T. C.; Knasmuller, S. and Huber, W. (1995), Hepatotoxicity and carcinogenicity of the fumonisins in rats. In: Jackson, L. S.; Vries, J. W. and Bullerman, L. B. (Eds.). Fumonisin in Food. New York : Plenum Press. pp. 279-296. [ Links ]
Gumprecht, L. A.; Smith, G. W.; Constable, P. C. and Haschek, W. M. (2001). Species and organ specificity of fumonisin-induced endothelial alterations: Potential role in porcine pulmonary edema. Toxicology, 160, 71-79. [ Links ]
Harrison, L. R.; Colvin, B. M.; Greene, J. T.; Newman, L. E. and Cole Jr., J. R. (1990), Pulmonary edema and hydrothorax in swine produced by fumonisin B1 , a toxic metabolite of Fusarium moniliforme . Journal of Veterinary Diagnostic Investigation., 2, 217-221. [ Links ]
Hirooka, E. Y.; Yamaguchi, M. M.; Aoyama, S.; Sugiura, Y. and Ueno, Y. (1996), The Natural occurrence of fumonisins in Brazilian corn kernels. Food Additives and Contaminants, 13, 2, 173-183. [ Links ]
Henningen, M. R.; Sanchez, S.; Benedetto, N. M.; Longhi, A.; Torroba, J. E. and Valente Soares, L. M. (2000), Fumonisin levels in commercial corn products in Buenos Aires, Argentina. Food Additives and Contaminants, 17 : (1), 55-58. [ Links ]
Hermanns, G.; Costa, L. L.F. and Scussel, V. M. (2000), Evaluatin of fumonisin contamination in grains of corn (Zea mays L.) produced at the western region of Santa Catarina. In: X International IUPAC Symposium On Mycotoxins And Phycotoxins, Guarujá. Anais São Paulo, Brasil. pp. 147. [ Links ]
IARC (2002), Monographs on The Evalution of Carcinogenic Risks To Humans. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Lyon, France, 82, 301. [ Links ]
INCA. Ministério da Saúde (1999), Estimativa da incidência e mortalidade por câncer no Brasil. Rio de Janeiro. [ Links ]
Jackson, L. S.; Hlywka, J. J.; Senthil, K. R. and Buççerman L. B. (1996). Effect of thermal processing on the stability of Fumonisins. In: Jackson, L. S.; Devries, J. W. and Bullerman, L. B. (Eds.). Fumonisins in Food, New York : Plenum Press. pp. 345-353. [ Links ]
Kellerman, T. S.; Marasas, W. F. O.; Thiel, P. G.; Gelderblom, W. C. A.; Cawood, M. E. and Coetzer, J. A. W. (1990), Leukoencephalomalacia in two horses induced by oral dosing of fumonisin B1. Onderstepoort Journal Veterinary Res., 57, 269-275. [ Links ]
Machinski Jr., M. and Soares, L. M. V. (2000), Fumonisins B1 and B2 in Brazilian corn-based food products. Food Additives and Contaminants, 17 : (10), 875-879. [ Links ]
Mallmann, C. A.; Santurio, J. M.; Almeida, C. A. A. and Dilkin, P. (2001). Fumonisin B levels in cereals and feeds from southern Brazil. Arquive Institute of Biology, 68 : (1), 41-45. [ Links ]
Marasas, W. F. O. (1996), Fumonisins: history, worldwide occurrence and impact. Adv. Exp. Med. Biol., 392, 1-17. [ Links ]
Meireles, M. C. A.; Correa, B.; Fischman, O.; Gambale, W.; Paula, C. R.; Chacon-Reche, N. O. and Pozzi, C. R. (1994), Mycoflora of the toxin feeds associated with equine leukoencephalo-malacia (ELEM) outbreaks in Brazil. Mycophatology, 127, 183-188. [ Links ]
Ono, E. Y. S.; Sugiura, T., Homechin, M.; Kamogae, M.; Vizzoni, E.; Ueno, Y. and Hirooka, E. Y. (2000), Effect of climatic conditions on natural mycoflora and fumonisins in freshly harvested corn of State of Paraná, Brazil, Mycopathologia, 1, 1-10. [ Links ]
Ono, E. Y. S.; Ono, M. A.; Funo, F. Y.; Medina, A. E.; Oliveira, T. C. R. M.; Kawamura, O.; Ueno, Y. and Hirooka, E. Y. (2001), Evaluation of fumonisin-aflatoxin co-occurrence in Brazilian corn hybrids by ELISA. Food Additives and Contaminants, 18 : (8), 719-729. [ Links ]
Ono, E. Y. S.; Sasaki, E. Y.; Hashimoto, E. H.; Hara, L. N.; Corrêa, B.; Itano, E. N.; Sugiura, T.; Ueno, Y. and Hirooka, E. Y. (2002), Post-harvest storage of corn: effects of beginning moisture content on mycoflora and fumonisin contamination. Food Additives and Contaminants, 19 : (11), 1081-1090. [ Links ]
Orsi, R. B.; Corrêa, B.; Possi, C. R.; Schammass, A.; Nogueira, J. R.; Dias, S. M. C. and Malozzi, M. A. B. (2000), Mycoflora and occurrence of fumonisins in freshly harvested and stored hybrid maize. Journal of Stored Products Research, 36, 75-87. [ Links ]
Patel, S.; Hazel, C. M.; Winterton, A. G. M. and Gleadle, A. E. (1997), Surveillance of fumonisin in UK maize-based foods and other cereals. Food Additives and Contaminants, 14 : (2), 187-191. [ Links ]
Perin, I. L; Kuhnen, N. P; Miroski, L. and Ribeiro, H. G. (1990), Análise de 149 Casos de Câncer de Esôfago tratados com Acelerador Linear, entre 1980/1985, no Serviço de Radioterapia do Hospital de Caridade. Arquivos Catarinenses de Medicina, 19, 5-8. [ Links ]
Rheeder, J. P.; Marasas, W. F. O.; Thiel, P. G.; Sydenham, E. W.; Shephard, G. S. and Van Schalkwyk, D. J. (1992), Fusarium moniliforme and fumonisins in corn in relation to human oesophageal cancer in Transkey. Phytopathology, 82, 353-357. [ Links ]
Scaff, R. M. C. and Scussel, V. M. (1999), Esophageal cancer in the Southern region of Brazil - Cases from Santa Catarina State. Mycotoxin Contamination: Health Risk and Prevention Project. pp. 226-230. [ Links ]
Shephard, G. S; Sydenham, E. W; Thiel, P. G. and Gelderblom, W. C. A. (1990), Quantitative Determination of Fumonisin B1 and B by High-Performance Liquid Chromatography with Fluorescence Detection. Journal of Liquid Chromatography, 13 : (10), 2077-2087. [ Links ]
Stak, M. E. and Epley, R. M. (1992). Liquid chromato-graphic determination of fumonisin B1 and B2 in corn and corn products. Journal AOAC International, 75, 834-837. [ Links ]
Sydenham, E. W; Thiel, P. G; Marasas, W. F. O; Shephard, G. S; Van Schalkwyk, D. J. and Koch, K. R. (1990), Natural Occurrence of Some Fusarium Mycotoxins in Corn from Low and High Esophageal Cancer Prevalence Areas of the Transkei, Southern Africa. Journal of Agricultural and Food Chemistry, 38, 1900-1903. [ Links ]
Sydenham, E. W.; Marasas, W. F. O; Shephard, G. S; Thiel, P. G. and Hirooka, E. Y. (1992), Fumonisin concentration in Brazilian feeds associated with field outbreaks of confirmed and suspected animal mycotoxicoses. Journal of Agricultural and Food Chemistry, 40, 994-997. [ Links ]
Sydenham, E. W.; Shephard, G. S.; Thiel, P. G.; Stockenström,S.; Snijan, P. W. and Van Schlwyk, D. J. (1996), Liquid chromatographic determination of fumonisin B1, B2 and B3 in corn: AOAC-IUPAC collaborative study. Journal AOAC International, 79, 688-696. [ Links ]
Trucksess, M. (2001), Mycotoxins. In: General Referee Reports. Journal of the Association of Official Analytical Chemists International, 84 : (1), 202-210. [ Links ]
Van der Westhuizen, L.; Shephard, G. S.; Scussel, V. M.; Costa, L. F.; Vismer, H. F.; Rheeder, J. P. and Marasas, W. F. O. (2003), Fumonisin contamination and Fusarium incidence in corn from Santa Catarina, Brazil, Journal of Agricultural and Food Chemistry, 51 : (18), 5574-5578. [ Links ]
Received: June 25, 2003;
Revised: November 18, 2003;
Accepted: July 20, 2004.
* Author for correspondence