In this work, a method for food dyes determination in fibrous fruit pulps was developed. The analytes Tartrazine, Twilight Yellow FCF, Erythrosine, Orange G and Allura Red were analyzed by reverse phase chromatography with diode array (DAD) detection. The sample preparation consisted of a simple procedure, based on initial pH adjustment of the sample (pH = 10.0), followed by an extraction with acetonitrile (5.0 mL and agitation for 30 s), ultrasonic (5 min) and centrifugation (4000 rpm, 20 ºC, for 20 min). The supernatant had its pH adjusted to 3.0, followed by filtration. The method was validated, with the analytical curves prepared in a matrix, in the range of 50to150mg kg-1 for all the analytes studied (determination coefficients (R2) > 0.99). The method veracity was evaluated in terms of recovery, obtaining recoveries of 93-110%. The repeatability was less than 13.7%, while the intermediate precision was less than 19.0%. The limits of detection and quantification were determined, being 15 and 30 mg kg-1, respectively. The combined relative uncertainties were less than 13.8%. Therefore, it can be concluded that a simple method was developed, suitable for monitoring the presence of dyes in fruit pulps in routine analyzes.

The aim of this study was to investigate the presence of arsenic, lead, and cadmium residues in samples of liver, kidney, and muscle of cattle during the years of 2002 to 2008. A total of 1017 samples from 20 Brazilian States were used. The samples were analyzed at the National Agricultural Laboratory using the atomic absorption spectrometry technique. Arsenic residues were detected in 15.7% of liver samples and 28.7% of kidney samples although no results have exceeded the MRL. With regard to lead, 16 samples of liver and 74 samples of kidney were contaminated (5.2 and 10.9%, respectively). Among these samples, only one liver and two of kidney samples had lead levels above the MRL. Cadmium was found with levels below the MRL in 12.5% of the liver samples, and only 3 samples (1%) were quantified above the MRL. Among the kidney samples, 420 (60.8% of the total tested) had cadmium residues, and five of them exceeded the limits established by legislation. It is concluded that the Brazilian meat meets the legislation requirements without putting consumer's healthy at risk since as it satisfies the national and international food-safety conditions.

Trinta e seis (36) cultivares de milho foram avaliadas em relação à incidência de grãos ardidos, mofados e produção de fumonisina B1. Amostras de 1,2 kg de grãos foram analisadas visualmente para a quantificação de grãos ardidos (Fusarium subglutinans), mofados (Penicillium oxalicum) e para a análise de fumonisina B1. Os grãos ardidos foram submetidos à análise de sanidade (papel de filtro com congelamento) visando identificar os fungos a eles associados. A cultivar Hatã 3052 apresentou 7,6% de grãos ardidos, ultrapassando o limite de tolerância que é de 6,0%. As cultivares AG 5011, HT 7105-3, Dina 1000 e C 701 apresentaram 16,8% , 3,4%, 3,2% e 3,1% de grãos mofados, respectivamente, acima do limite de tolerância que é de 3,0%. O fungo Fusarium subglutinans (Gibberella fujikuroi var. subglutinans) foi o causador de grãos ardidos, cuja detecção variou de 50,0 a 99,0%. A análise de variância mostrou diferenças significativas entre as cultivares com relação às incidências de grãos ardidos e de grãos mofados. Com relação à produção de fumonisina B1, as cultivares Hatã 3052, NB 6077 e 983 P produziram 7,0; 6,1 e 5,9 µg.g-1 de grãos, respectivamente, diferindo significativamente da cultivar P3071 (2,2 µg.g-1 de grãos). Conclui-se que há diferenças significativas entre as cultivares de milho em relação à produção de grãos ardidos e mofados, bem como acentuada interação entre as cultivares e o fungo toxigênico Fusarium subglutinans (Gibberella fujikuroi var. subglutinans) quanto à biossíntese de fumonisina B1 em grãos de milho.

Thirty six corn cultivars were evaluated in relation to incidence of burned (Fusarium subglutinans) and moldy grains (Penicillium oxalicum), and fumonisin B1 production. The burned grains were submitted to the grain health (deep freezing), seeking to identify the fungi associated to the grains. The cultivar 'Hatã 3052' presented 7.6% of burned grains, above the tolerance limit that is of 6.0%. The cultivars 'AG 5011', 'HT 7105-3', 'Dina 1000' and 'C701' presented 16.8, 3.4, 3.2 and 3.1% of moldy grains, respectively; above the tolerance limit that is of 3.0%. Fusarium subglutinans (Gibberella fujikuroi var. subglutinans) was the promoter of burned grains, whose detection varied from 50 to 99%. The variance analysis showed significant differences among the cultivars regarding the incidences of burned grains and moldy grains. Regarding the fumonisina B1 production, the cultivars 'Hatã 3052', 'NB 6077' and '983 P' produced 7; 6.1 and 5.9 µg.g-1 of grains, respectively, with significative difference from the cultivar 'P3071' (2.2 µg.g-1 of grains). There were significative differences among the corn cultivars with respectm to the burned and moldy grains production; as well as interaction among cultivars and the toxigenic fungous Fusarium subglutinans (Gibberella fujikuroi var. subglutinans) regarding fumonisin B1 production in corn grains.

Um método para determinação de aflatoxina M1 em leite empregando cromatografia em camada delgada foi otimizado e validado por procedimentos intralaboratoriais. Foram realizados testes para otimização das etapas de extração, purificação, detecção e quantificação por análise visual e densitométrica. Para validação do método foram realizados ensaios de recuperação com soluções padrões e amostras artificialmente contaminadas em níveis entre 0,027µg/L e 0,970µg/L. Foram avaliados linearidade, especificidade, exatidão, precisão, limites de detecção e quantificação. Os valores de porcentagem de recuperação na faixa de quantificação do método variaram de 84,2% a 99,0% na análise visual e de 85,2% a 105,2% na análise densitométrica, com coeficientes de variação de 2,3% a 9,8% e de 3,6% a 13,9%, respectivamente. O limite de detecção foi de 0,036µg/L e o de quantificação de 0,045µg/L, contemplando os níveis de tolerância estabelecidos na legislação nacional e internacional para aflatoxina M1 em leite.

A method for aflatoxin M1 determination in milk by thin layer chromatography was otimized and in house validated. Steps of extraction, clean-up, detection and quantification by visual and densitometric analysis were standardized. In method validation were employed standard solutions and spiked samples in levels varying from 0.027µg/L to 0.970µg/L. Linearity, specificity, trueness, precision, detection and quantification limits were the parameters evaluated. Average recovery values in quantification range were from 84.2% to 99.0% for visual analysis and from 85.2% to 105.2% for densitometric analysis, with CVs values from 2.3% to 9.8% and from 3.6% to 13.9%, respectively. The detection limit was 0.036µg/L and the quantification 0.045µg/L, according to national and international regulations for aflatoxin M1 in milk.

Procedimentos para validação intralaboratorial de kits de ELISA foram adotados na verificação da eficiência de um kit para detecção e quantificação de aflatoxina M1 em leite. Foram realizados ensaios com soluções padrões fornecidas pelo kit, amostras artificialmente contaminadas e amostras naturalmente contaminadas. Os valores médios de recuperação obtidos para os padrões do kit apresentaram-se entre 85,6 e 114,8%, com valores de coeficiente de variação entre 11,6 e 23,1%. Nos ensaios com amostras artificialmente contaminadas, foi definida uma faixa ótima de trabalho para análises quantitativas entre 0,019 e 0,090µg/L. A presença de aflatoxina M1 em 110 amostras de leite do estado de Minas Gerais foi investigada. Cinco das 27 amostras positivas na triagem por ELISA foram confirmadas por cromatografia em camada delgada.

Procedures for intralaboratory validation of an ELISA kit were adopted to verify its efficiency in aflatoxin M1 detection and quantification in milk. Assays were accomplished with standard solutions provided with the kit, spiked samples and naturally contaminated samples. The mean recovery values obtained for the kit standard solutions were between 85.6 and 114.8%, with coefficient of variation ranging from 11.6 to 23.1%. Working with spiked samples, a concentration range for quantitative analysis (0.019 to 0.090µg/L) was defined. The occurrence of aflatoxin M1 in 110 milk samples from the State of Minas Gerais was investigated. Five of the 27 positive samples in the screening for ELISA were confirmed by thin layer chromatography.