Later on Wayner Burton Ingold Barclay

Later on, Wayner, Burton, Ingold, Barclay, and Locke (1987) modified the method by using lipid to examine the ability of an antioxidant Deferoxamine protect against lipid peroxidation generated by azo-compound. It was later improved using β-phycoerythrin (β-PE) as a fluorescent probe, and the ability of the plasma to protect β-PE from peroxidation was monitored by spectrofluorometer (Table 3) (DeLange & Glazer, 1989). This assay was applied in vitro for the evaluation of antioxidant activity of beverages and foods (Sánchez-Moreno, 2002). It was reported that the measurement of lag time was the main shortcoming of this assay since not only each antioxidant had different lag phases, but also the antioxidant potential after lag phase was ignored in this assay. Based on a similar principle, biologically occurring low-density lipoproteins (LDL) isolated from blood sample was used as an oxidizable probe and it was oxidized with an azo-initiator or a transition element followed by monitoring the conjugated dienes spectrophotometrically (Table 3) (Frankel, Kanner, German, Parks, & Kinsella, 1993). However, it was emphasized that problem arising from the difficulty in lag phase determination of each antioxidant and limitation in the LDL isolation were the disadvantages of this method. Later on, Cao, Alessio, and Cutler (1993) developed another physiological relevance method that measured the total oxygen-radical absorbing capacity (ORAC) of serum samples by monitoring the fluorescence of phycoerythrin kinetically followed by calculating the area under the decay curve of phycoerythrin (Table 3). This approach has overcome to disadvantages of TRAP assay since not only it takes into account both the initial and progression rate of oxidation but also it can be applied to antioxidants have distinct lag phases or even those that have no lag phases. It has gained recognition both in biological and food systems but it was realized that phycoerythrin as a fluorescent probe had some shortcomings: (1) the assay results were inconsistence, as the reactivity of phycoerythrin toward to peroxy radicals could vary with time, (2) phycoerythrin bleached after exposure to excitation light, (3) polyphenols could bind to phycoerythrin via protein-phenol interaction (Prior, Wu, & Schaich, 2005). Therefore, fluorescein was approved as an alternative fluorescent compound in ORAC assay with more stable and less reactive properties. In very recently, the scavenging ability against peroxyl and hydroxyl radicals has been also determined by monitoring the absorbance of indigo carmine as a redox indicator. It was demonstrated that this method was able to make more accurate predictions of antioxidant capacity of food samples (Pérez-Burillo, Rufián-Henares, & Pastoriza, 2018). However, it was emphasized that all types of ORAC assay measured only the antioxidant capacity against peroxyl and hydroxyl radicals, not all other oxygen species such as superoxides and singlet oxygen (Apak, Güçlü, Özyürek, & Karademir, 2004). By the way, several limitations of ORAC assay need to be acknowledged. This method is extremely sensitive to temperature and oxygen concentration during generating the peroxyl radicals and these are critical parameters to rigid control. Therefore, automated systems, which are available in just a few laboratories, have been developed to improve efficiency and accuracy in addition to decrease the personal mistakes (Huang, Ou, Hampsch-Woodill, Flanagan, & Prior, 2002). Some researchers has applied ORAC procedure with manual handling, and using a conventional fluorescence plate reader; however the parameters in procedures (such as required reaction time, concentration of target probe and radical initiator) differ from paper to paper and the results are not reproducible (Davalos et al., 2004, Ranamukhaarachchi et al., 2017, Schaich et al., 2015, Wang et al., 2017, Zhou et al., 2017). It is obvious that non-automated systems cause confusion in reaction parameters and inconsistency in the results. When required temperatures for thermal decomposition of radical initiator (azo-compound) are not reached, reactions will be slow and incomplete, that cause poorly reproducible results. In addition, if dissolved oxygen, which is critical parameter for the generation of peroxyl radicals efficiently from azo-compound, are not sufficient, the reaction takes longer time leading to irreproducible results. Ambiguity both in reaction end-point and in concentration of reagents, and inconsistency of results from different laboratories make interlaboratory comparison difficult. Therefore, it may not be a standardized and robust method that can be efficiently utilized by researchers in every laboratory to measure the antioxidant capacity of food or biological samples.