Abstract Seeds from endemic Brazilian fruits (Mangaba, Cagaita, Jatobá and Araticum), previously confirmed to be toxic using Artemia salina bioassay were exposed to different temperatures and heating times aiming to test inhibitory effects on toxicity, as a first approach to characterize and identify toxic compounds and their thermal stability. After treatment, water extracts from kernels were filtered and added to Artemia cultures for 24 hours after which the survival of cultures was calculated. The inhibitory effect of temperature on toxicity was significant, whereas toxins from Mangaba and Cagaita were the most labile, showing 90% Artemia survival after heating at 78 °C for 2 minutes and 10 minutes, respectively. The most resistant toxins were those found in Araticum seeds which remained active (0% Artemia survival), even after heating at 160 °C for 12 min. Response surfaces and statistical multivariate analysis showed a significant effect for both temperature and heating time (linear fit) on Cagaita and Mangaba toxicity reduction, while the temperature factor was more significant in Jatobá samples. The results confirm the need to adopt safety procedures during the manufacture of industrialized derivatives for these fruits, in order to minimize toxicity risks after their consumption.

Foi desenvolvido um biossensor espectrofotométrico para a determinação da concentração de glicose em amostras reais de suco de laranja, energéticos e bebidas esportivas. O biossensor consistiu de glicose oxidase (GOX) e peroxidase (HRP) imobilizadas em polianilina ativada com glutaraldeído (PANIG). Os parâmetros de imobilização foram otimizados para a GOX e o rendimento máximo de imobilização obtido foi de 16% para 5 mg de PANIG e 8,9 Unidades de Enzima (UE) preparadas em tampão fosfato 0,1 mol.L-1 (pH 7,0) com tempo de reação de 60 minutos a 4 °C, com agitação leve. Utilizando os parâmetros otimizados, encontrou-se uma faixa linear para a determinação de glicose entre 0,05 e 6,0 mg.mL-1, com boa reprodutibilidade de resposta. Os resultados obtidos foram comparados com resultados de enzimas livres (kits comerciais), sendo validados através de análise estatística pelo teste de Tukey (95% de significância).

A biosensor was developed for spectrophotometric determination of glucose concentrations in real samples of orange juice energetic drinks, and sport drinks. The biosensor consisted of glucose oxidase (GOD) and horseradish peroxidase (HRP) immobilized onto polyaniline activated with glutaraldehyde (PANIG). Immobilization parameters were optimized for GOD, and maximum immobilization yield was 16% when 5.0 mg of PANIG and 8.9 U prepared in 0.1 mol.L-1 sodium phosphate buffer (pH 7.0) reacted for 60 minutes at 4 °C with gentle stirring. The linear operational range for glucose determination using optimized operational parameters was between 0.05 and 6.0 mg.mL-1 with a very good reproducibility of response. The results obtained in the biosensor were compared with those obtained using free enzymes (commercial kits) and then validated through statistical analysis using the Tukey test (95% confidence interval).