• antioxidant activity;
  • cell culture;
  • food analysis;
  • oxidative stress


  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References

Abstract:  Food science has progressively evolved and now there are wide evidences that foods have biological activities that are beyond their classical nutritional value. In this field, the antioxidant activity of pure compounds, food, feed, and dietary supplements has been extensively studied and numerous analytical approaches and assay models have been developed, involving various systems from simple chemical assays to animal models and human studies. This article is an overview of different cell-based models that have been used for testing the antioxidant properties of food, feed, and dietary supplements. Advantages, drawbacks, and technical problems to develop and validate suitable, robust, and high-throughput cell-based bioassays for screening food antioxidant activity will be discussed.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References

In the past years, food and feed science has progressively evolved prompted by different factors, such as the improved safety issue, the relevant changes in the European Union agriculture policy, the need to recover consumer's confidence about nutraceutical, and functional aspects of foods of animal origin, taking into account the changes in their alimentary habits. It is now well known that food may have biological activities that are beyond the classical nutritional value. This aspect has gained increasing attention in the food industry but also in animal nutrition and so-called nutraceuticals are offered both for food and feed applications. From a regulatory point of view, if food and feed are brought onto the market with ‘‘nutritional and health claims,” these claims must be objective, scientifically supported, and verifiable by the competent authorities (Regulation (EC) No 1924/2006 of the European Parliament and of the Council; Regulation (EC) 767/2009 of the European Parliament and of the Council).

In this context, it is important to develop protocols and models to evaluate the bioaccessibility, bioavailability and functionality of bioactive components. The evolution of food analysis methodology over the past years have led to improvements in chemical analysis and instrumental tools with significant enhancements in analytical accuracy, precision, detection limits, and sample throughput (Mc Gorrin, 2009). The transition from a chemical analysis approach to cell-based bioassays may support the new need for food analysis in terms of bioactivity and functional properties. The need of reliable in vitro cellular models as alternatives to animal studies has became an important issue and is considered in European legislation ( The interest of food and pharmaceutical industry to cell-based bioassays is principally for toxicological and bioavailability tests of newly developed food ingredients and drugs, inevitable for bringing products to the market. In food/feed industry, the research regarding the safety and efficacy of additives and new functional food and feed is an open issue and may take advantages from the development and validation of specific cell-based functional bioassays.

This article is an overview of different cell-based models that have been used for testing the antioxidant properties of food, feed, and dietary supplements. Advantages, drawbacks, and technical problems to develop and validate suitable, robust, and high throughput cell-based bioassays for screening food antioxidant activity will be discussed.

Oxidative stress and antioxidants

The concept of oxidative stress is becoming very important in medical and nutritional research. Oxidative stress can be defined as an imbalance between oxidants and antioxidants (Figure 1). Reactive oxygen species (ROS), continuously generated from mitochondrial respiratory chain, have powerful oxidative potential and are the major contributors to oxidative damages. Antioxidants are molecules capable of reducing the effect of free radicals, such as ROS, by stabilizing or deactivating these harmful molecules. Humans evolved a complex and efficient antioxidant defense system that involves the interaction and synergic activity of endogenous and exogenous components. This includes antioxidant enzymes, metal binding proteins, and diet-derived antioxidants. A list of the main antioxidant compounds and their dietary sources is presented in Table 1. In a normal healthy cell, equilibrium is maintained between the generation of ROS and their elimination by the antioxidant system. However, an unbalance can occur when ROS production is greater than the antioxidant defense capacities of the cell, or when the normal antioxidant defense of the cell is inhibited. ROS can react with and damage many biological molecules, such as proteins, lipids, and DNA, with dramatic consequences on cell function (Halliwell 1996; Cooke and others 2003). Oxidative stress and accumulation of ROS is implicated in the aetiology of an extremely broad spectrum of degenerative and chronic diseases, like many neurodegenerative diseases, cardiovascular diseases, diabetes, aging, as well as cancer (Adly 2010). Epidemiological studies support the important role that nutrition may play in controlling and mitigating these diseases. There is evidence that food antioxidant components and antioxidant supplementation may have a protective role against oxidative stress induced diseases, although sometimes inconsistent results have been reported (Liu and Finley 2005; Kaliora and others 2006; Adly 2010; Bisbal and others 2010).


Figure 1–. Oxidative stress and antioxidant system.

Download figure to PowerPoint

Table 1–.  Main antioxidant compounds and dietary sources.
AntioxidantsMain dietary sources
Vitamin EVegetable oils, green vegetables, grains, nuts, eggs, milk
Vitamin CFruits (mainly citrus fruits), vegetables, tomatoes, potatoes
CarotenoidsCarrots, tomatoes, grapefruits, dark green leafy vegetables, tomatoes, papaya, watermelon
 – FlavonoidsPotatoes, tomatoes, lettuce, onions, apples, wheat bran, dark chocolate, red wine, coffee, black tea
 – Phenolic acidsBlack and green tea, coffee, red wine, berries, potatoes, broccoli
 – StilbenoidsRed grapes, berries
Lipoic acidSpinach, broccoli
CoQ10Fish, meat, vegetable oils, cereals, soy

Bioavailability and metabolism are 2 important topics to be addressed in the evaluation of the effective antioxidant activity of foods and dietary supplements. Antioxidant bioavailability and absorption are largely influenced by several factors, like structural diversity of compounds with antioxidant activity, food matrix structure, the coexistence of different nutrients, and classes of dietary antioxidants and their relative proportions in standard diets (Faulks and Southon 2005; Parada and Aguilera 2007; Reboul and others 2007; Namitha and Negi 2010; Tagliazucchi and others 2010). Antioxidants are absorbed from the gastrointestinal tract, increasing the circulating level and availability for cells. Furthermore, gut fermentation and absorption may be accompanied by antioxidant degradation and production of new molecules with antioxidant activity, extensive conjugation, and metabolism. Degradation of flavonoids to simple phenolic acids, which may be absorbed and further metabolized in the liver, has been reported in the colon by enzymes of the gut microflora (Spencer and others 2004). After ingestion, carotenoids are cleaved by an intestinal dioxygenase and monooxygenase to form retinal (Nagao and others 1996). All these processes mean that the forms of antioxidant compounds found in the blood and the targeted tissues may be different from those found in food. The precise knowledge of the processes underlying the bioavailability and metabolism of antioxidants is the basis to choose the form of antioxidant that is found in the plasma to be tested in in vitro models to obtain results of a biological significance and with relevance for the in vivo situation. The Caco-2 cell culture model was used to evaluate the bioavailability and metabolic conversion of antioxidants, such as carotenoids and flavonids, from whole foods (Murota and Terao 2003; During and others 2005; Namitha and Negi 2010; Rodriguez-Amaya 2010). At cellular level, antioxidants can at the cell membrane breaking and preventing propagation of peroxyl radical chain reactions and reducing ROS-induced peroxidation of membranes (Wang and Quinn 1999; Azzi 2007), or they can be taken up by the cell and act as free radicals scavengers (Adly 2010).

Antioxidant activity evaluation

To develop an assay to evaluate the antioxidant activity of food and additives, the precise definition of a biological antioxidant given by Laguerre and others (2010) should be carefully considered. These researchers evidenced 2 different and very important topics: an antioxidant must protect (1) a substrate from oxidation and (2) a biological system from damages coming from oxidation. The well developed chemical based assays for evaluation of antioxidants can give answers, when the 1st topic is considered, and can compare molecules for their antioxidant capacity. However chemical assays do not reflect the cellular physiological conditions. Since the mid-1990s, antioxidant research expanded and several chemical assays measuring the total antioxidant activity of compounds, food, and additives have been developed (Cao and others 1993, 1995; Frankel and Meyer 2000; Ou and others 2001; Prior and others 2003; Karadag and others 2009). The most widely used in vitro methods are: oxygen radical absorbance capacity (ORAC), total radical-trapping antioxidant parameter (TRAP), trolox equivalent antioxidant capacity (TEAC), total oxyradical scavenging capacity (TOSC), peroxyl radical scavenging capacity (PSC) methods. The chemistry at the base of these antioxidant assays and the principles of the methods have been described in several reviews (Huang and others 2005; Prior and others 2005; Magalhães and others 2008; Laguerre and others 2010). When chemical assays were compared with cell-based methods for assessing antioxidants and antioxidant activity of foods and dietary supplements, different results have been found and not always correlated each other (Wolfe and Liu 2007; Honzel and others 2008). Girard-Lalancette and others (2009) found significant differences in sensitivity for antioxidants, which activity was detected at lower concentration in a cell-based assay in comparison with the ORAC assay. All results together (1) indicate that the chemical assays may be good models for comparing molecules for their antioxidant capacity, but not a good measure of antioxidant activity in biological models; (2) highlight the importance of using more biologically relevant cell-based models for the evaluation of food protection and health effects of antioxidants.

Cell-based bioassays

A good in vitro model for food analysis has to satisfy 2 basic requirements: availability and easy handling for high-throughput testing and for getting good and reproducible results. Several important aspects must be accurately evaluated for a broader use of cell-based testing system in antioxidant activity evaluation: choice of cellular models, assessment of specific biomarkers as endpoint to measure, cell living environment. All these aspects are fundamental to ensure standardization of the model, uniformity, and sensitivity in reporting natural products antioxidant activity in biological systems. Cell-based bioassay development for food antioxidant activity analysis can include high throughput screening assays and other specific custom bioassays used to evaluate specific food properties. In vitro cell culture methods can be used in a 2-tiered approach, one by which the simple effects on cell viability and proliferation are assessed, and the second by which more complex assays are made to elucidate the mechanism of action for the compound of interest. A screening system should achieve an optimal balance between high-throughput, ease of performing experiments and analyses, adequate time, and expenses.

Cell-based bioassay for evaluating antioxidant activity of food: adequacy of the model

Selecting cells A specific biological question to be answered is critical to the success of a cell-based bioassay and must be very clear to choose the most adequate cells. The properties and sensitivity of the cells and their growth status are critical factors that affect antioxidant activity evaluation. In addition, cellular response to antioxidant components is quite complex, depending on the component, exposure dose, and time. Several cell types have been used to set up cell-based bioassays for assessing antioxidants activity of foods and dietary supplements at wide concentration ranges, usually selected based on the lack of cytotoxic effect and a realistic concentration range in human diet and serum.

Primary isolated rat hepatocytes have been used to evaluate the protective effects of melanoidins in adriamycin-induced oxidative stress (Valls-Bellés and others 2004). Fusi and others (2010) evaluated the protective effects of alpha-tocopherol in ochratoxin A (OTA) induced oxidative damage in primary porcine fibroblasts. These researchers found a different sensitivity of ear and embryonic fibroblasts both to the oxidative damage and the antioxidant treatment. Primary cells, isolated from human or animal tissue, could be a good cell-based model as they retain the majority of the in vivo functionality, but they survive and maintain their differentiation only few days in cell culture. If the primary cells are chosen, it must be considered that usually they derive from different individuals in each test; consequently, the reproducibility of results may significantly differ from one test to another. Therefore, when phenotypic screens with the use of primary cells are included in designing a cell-based bioassay, it should be taken into consideration that a constant supply of biologically homogenous cells to support a large scale study is necessary. Moreover, isolation procedure for primary cell culture preparation can reduce survival of the cells and cause changes in gene expression, metabolic activity, and the levels of enzymes involved in the oxidant and antioxidant systems. Halliwell (2003), discussing in details how the “culture shock” can affect cells survival and metabolic activity, concludes that the cells which do survive appear to be those that have adapted rapidly, and probably are not representative of the originally cells harvested from a tissue.

The use of immortalized cell lines for the purpose of antioxidant testing may be a desirable approach. A cell line basically comprises a phenotypically and genetically uniform population of individual cells that have been derived from a single tissue. Cell lines are robust, grow, and divide easily in culture, are simple to handle and may provide a good platform for cell-based assays. The availability of a wide number of human and animal cell lines with diverse genotypes and tissue-origins provides broad potential models for the study of various biological processes. In Table 2, examples of different cell lines that have been used for the development of cell-based bioassays for food antioxidant activity analysis are reported.

Table 2–.  Examples of different cell lines that have been used for development of cell-based bioassay for food antioxidant activity analysis.
HepG-2Liver carcinomaEpithelial-likeHumanEberhardt and others 2005; Wolfe and Liu, 2007; Goya and others 2009
AGSGastric adenocarcinomaEpithelialHumanXu and Chang, 2010
Caco-2Colon adenocarcinomaEpitheliaHumanLiu and others 2004; Boyer and others 2004, 2005; Khonkarn and others 2010
L-929Subcutaneous connective tissueFibroblast-likeMurineGirard-Lalancette and others 2009
Int 407Embrionic, intestineEpitheliaHumanElisia and others 2007
RAW264.7AscitesMonocytes/macrophagesMurineHu and others 2007
MKN-45Gastric adenocarcinomaEpitheliaHumanSerra and others 2010
MCF-7Breast cancer HumanYang and Liu, 2009
HT29Colon-rectal adenocarcinomaEpitheliaHumanBellion and others 2009
RBC and PMN cellsBloodHumanHonzel and others 2008; Jensen and others 2008; Blasa and others 2011

Baldi and others (2004) evaluated the protective effects of alpha-tocopherol in OTA induced oxidative damage. A panel of 5 well-characterized human and animal cell lines, SK-N-MC (human neuroblastoma), MDCK (Madin Darby canine kidney), AML-12 (mouse liver hepatocytes), LLC-PK1 (pig kidney), and BME-UV1 (bovine mammary epithelium), have been successfully used to investigate the effect of alpha-tocopherol on ROS production on OTA-treated cells. The researchers found that alpha-tocopherol, at concentrations of 10 mM, 10 μM, and 1 nM, significantly inhibited OTA-induced ROS production and that the inhibition was dose dependent. A significant difference in cell sensitivity was also found, with BME-UV1 and MDCK as the most sensitive cell lines. Vitamin E is a well known potent antioxidant and, therefore, is a good positive control to be used to set up a standardized cell-based antioxidant activity assay. Vitamin E function as a peroxyl radical scavenger that terminates chain reactions is well documented (Wang and Quinn 1999; Azzi 2007). The liver is the main target for antioxidant compounds once absorbed from the gastrointestinal tract and also the major place for xenobiotic metabolism. Therefore, studies dealing with the effect of dietary compounds at a physiological level in the liver of live animals and at a cellular level in cultured cells from hepatic origin have been widely used. HepG2 (human liver carcinoma) cells have been used to set up cell-based bioassays for assessing antioxidants and antioxidant activity of foods and dietary supplements (Eberhardt and others 2005; Wolfe and Liu 2007). Wolfe and Liu (2007) tested the antioxidant activity of pure phytochemical compounds and of selected fruit extracts. The antioxidant activity was measured as median effective dose (EC50) for a 50% inhibition of peroxyl radical-induced dichlorofluoorescin oxidation. The researchers found significant differences in the antioxidant activity of both pure phytochemicals (EC50 ranging from 5.92 to more than 250 μM) and fruit extracts, with blueberry showing the higher antioxidant activity (EC50= 3.44 μg/mL). The same researchers compared the activity of antioxidants in the cellular model with the ORAC assay, and they found no consistency in the order of antioxidant activity of fruit extracts in the different assays. Comparable results have been found in a study on the antioxidant activity of broccoli extracts at concentration of 0.5 and 2 mg/mL (Eberhardt and others 2005). Using the same HepG2 cell model, Goya and others (2009) carried out a comparative study, testing different dietary antioxidant groups, like phenolic compounds, coffee melanoidin, and selenomethyl selenocystein at concentrations of 5 to 10 μM, 0.5 μg/mL, and 1 μM, respectively. The defense against oxidative stress has been evaluated using a panel of biomarkers: cell viability, biomarkers of oxidative status, and assessment of the antioxidant non-enzymatic and enzymatic defense. The proposed cell culture model was able not only to evidence differences in the antioxidant activity of the tested compounds but also to assess differences in the mechanisms involved in the antioxidant activity. The choice of the immortalized cell lines for the purpose of antioxidant researches is critical, and it is fundamental that cell lines show no altered functional responses to oxidative stress. In the matter of this topic, the use of HepG2 received the criticism that this cell line has altered functional response (increased catalase mRNA expression) to oxidative stress and performs asymmetrical cell divisions that may cause a proportion of the cells in culture dysfunctional and in various stages of cell death (Honzel and others 2008). A human gastric adenocarcinoma (AGS) cell line has been used in a study to evaluate the antioxidant activity of several lentil cultivars (Xu and Chang 2010). The researchers chose this cell line due to the rapid proliferation properties of AGS and because of criticism received by the use of HepG2. The cell-based assay was able to detect differences in the antioxidant activity of the different cultivars of lentils, with EC50 values ranging from 0.68 to 1.44 mg/mL of lentil extracts. However, the researchers conclude that they do not know if the AGS cell line is the best choice for cellular antioxidant assay and suffers in the same manner as the HepG2 cell lines. Several other cell lines have been tested to set up sensitive cell-based bioassays. A murine fibrosarcoma cell line (L-929) has been used to test the antioxidant activity of fruit and vegetable juices (Girard-Lalancette and others 2009). These reseachers found EC50 values ranging from 14 to 1119 μg/mL for broccoli and peaches, respectively. EC50 values in the same range have been reported for the antioxidant properties of cereal extracts by Hu and others (2007), using a mouse macrophage RAW264.7 cell line. Serra and others (2010) used a human gastric cancer (MKN-45) cell line to test the antioxidant activity of apple varieties. Apple extracts from different varieties were able to inhibit the proliferation of MKN-45 cells (cell viability was measured by MTT assay), with EC50 values ranging from 9 to 26 mg/mL. The intestine, as an important internal environment where a number of processes occurs to nourish the body and protect it against the enteropathogens or harmful substances entering the gut, is another important target for antioxidant compounds. Caco-2 (human epithelial colorectal adenocarcinoma), HT29 (human colon adenocarcinoma grade II), and Int-407 (human embryonic intestinal) cells have been used with good results for food antioxidant activity evaluation (Liu and others 2004; Boyer and others 2004, 2005; Elisia and others 2007; Bellion and others 2009; Serra and others 2010).

Recently, several efforts have been made to set up cell-based bioassays for antioxidant activity evaluation using erythrocytes (RBC) and polymorphonuclear (PMN) cells (Honzel and others 2008; Jensen and others 2008; Blasa and others 2011). According to the researchers, red blood cells may represent a good and biologically relevant model for a cell-based bioassay, as RBC play a critical role in antioxidant protection in the blood, by scavenging reactive oxygen and nitrogen species. As RBC do not have mitochondria, the use of this cell model may reduce the confounding contribution of cellular signaling. The PMN cells type could be also a useful model for assessment of overall antiinflammatory activity against an immune supportive property of a product. Honzell and others (2008) compared chemical (ORAC) and cell-based methods for evaluation of antioxidant activity of 1 animal-based, 1 microbial-based, 1 plant-based, and 1 mixed natural product. Total of 2 cellular models have been used: RBC and PMN cells, to measure antioxidant protection and reactive oxygen species formation, respectively. The researchers found different results comparing the data obtained with the 2 models and concluded that the antioxidants alone could not account for the strong antiinflammatory effect on PMN cells. The same RBC and PMN cell assays have been used by Jensen and others (2008) to evaluate the antioxidant effects of a fruit and belly juice. This juice, at concentration ranging from 0.01 to 10 g/L, was able to protect RBC from oxidative damage and reduce ROS production by PMN cells. Moreover, PMN cells showed altered migratory behavior. The possibility to evaluate the activity of antioxidants in a non-inflammatory red blood cell-based system and in inflammatory PMN cells, open new research field for the development of specific custom bioassays to evaluate specific food properties. Blasa and others (2011), testing the effects of phytochemicals and botanical extracts in a RBC assay, found significant differences in the antioxidant activity of the tested compounds. The model was useful for evaluating synergistic or antagonistic effects of combination of extracts. The phenomenon of the synergy is very important to be evaluated. Food may contain a complex mixture of bioactive compounds which are relevant in health and nutrition. The synergistic effects of phytochemicals in fruits and vegetables have been proposed to be responsible for their potent antioxidant and anticancer activity (Yang and Liu 2009). The possibility to evaluate and quantify the synergy which may occur with different compounds, could be of great interest to formulate fortified food and feed and tailor-made additives. All results taken together, underline how the choice of the cell type is critical to develop standardized cell-based assays for antioxidant research, which allow getting reliable and lab-to-lab reproducible results to be transferred into human and animal nutrition.

Selecting specific biomarkers as an endpoint to measure Antioxidants may operate via multiple mechanisms, briefly described before. Antioxidants can act at the cell membrane and break peroxyl radical chain reactions at the cell surface, or they can be taken up by the cell and react with ROS intracellularly. Therefore, the efficiency of membrane binding and/or cellular uptake, combined with the radical scavenging activity, likely dictates the efficacy of the tested compound. As the cell response to antioxidant components is quite complex, specific biomarkers as endpoints to measure must be identified. Cell-based assay format can be simple viability assays measuring the effect of compounds on cellular growth and/or viability, other can be metabolic assays. Examples of endpoint to measure and of methods and principles of cell-based antioxidant bioassays are reported in Table 3 and Figure 2. For a high throughput screening of food functional properties, biomarkers of cell viability, oxidative status, antioxidant defense, and oxidative damage are the most used.

Table 3–.  Examples of endpoints used in cell-based bioassays for food analysis.
Cell viabilityMitochondrial viabilityMTT reduction
 Release of cytoplasmic enzymesLDH release
Cell functionalityOxidative-reduction statusROS production
  Glutathione level
  Antioxidant enzyme system (SOD, CAD, GPx, GR)
 Intracellular oxidationIncorporation and oxidation of fluorescent or fluorogenic probes
 Oxidative damageMalondialdeyde

Figure 2–. AOx = antioxidants, ROS = reactive oxygen species. Examples of methods and principles of cell-based antioxidant bioassays: A) MTT assay: the yellow MTT [3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyltetrazolium bromide] is reduced to a purple formazan by mitochondrial enzymes. Antioxidants scavenge free radicals, singlet oxygen, and electrons in cellular redox reactions; B) LDH (lactate dehydrogenase) release: LDH is released into the culture medium following loss of membrane integrity resulting from oxidative damage. LDH catalyzes the reduction of NAD+ to NADH and H+ by oxidation of lactate to pyruvate; C) luminescence ATP detection assay system: ATP is a marker for cell viability because it is present in all metabolically active cells and the concentration declines very rapidly when the cells are exposed to cytotoxic stimuli; D) Exogenous probe incorporation: the DCFH-DA (2’,7’-dichlorfluorescein-diacetate) probe diffuses into the cell and it is cleaved by esterases to form DCFH (non-fluorescent 2’,7’-dichlorfluorescein). The peroxyl radicals oxidize the intracellular DCFH to DCF (fluorescent 2’,7’-dichlorofluorescein). Antioxidants bind to the cell membrane and/or enter the cell. Antioxidants prevent oxidation of DCFH and membrane lipids and reduce the formation of DCF.

Download figure to PowerPoint

When cell viability is chosen as an endpoint to measure to evaluate the antioxidant effects, the tetrazolium salts (MTT) reduction test has been widely used (Baldi and others 2004; Hu and others 2007; Heo and others 2008; Khonkarn and others 2010; Serra and others 2010). This is a reliable method and may be amenable for high-throughput formats, carried out in multi-well plates, and screening because of the cost and sensitivity. The use of a luminescent ATP cell viability assays and LDH release, as an indicator of cell viability, have been reported by several researchers in cell-based bioassays for the assessment of antioxidant properties of dietary compounds and additives (Elisia and others 2007; Goya and others 2009; Fusi and others 2010).

The use of fluorescent or fluorogenic probes have been reported in several cell-based bioassays intended for the monitoring and measure of oxidation. Dichlorofluorescein, a compound that once in human cells is easily oxidized to the fluorescent compound dichlorofluorescein, has been used by several researchers in cell-based bioassays (Elisia and others 2007; Wolfe and Liu 2007; Jensen and others 2008; Girard-Lancette and others 2009; Xu and Chang 2010; Blasa and others 2011). Results indicate that the use of fluorescent or fluorogenic probes can be a good approach for a high-throughput measure of antioxidant activity of different compounds. A criticism has been raised by Laguerre and others (2010) in a review on chemical and cell-based methods for evaluating the efficacy of antioxidants. These researchers conclude that these probes may detect a general oxidative stress but cannot be used as markers of a precise oxidative pathway. Regarding this aspect, interesting biomarkers such as intracellular ROS, nonenzymatic antioxidant, and cellular antioxidant enzyme could give more insight into the specific mechanisms involved in the biological activity of antioxidant dietary compounds. Goya and others (2009) presented a HepG2 cell-based bioassay as a model for the study of different antioxidant mechanisms of dietary compounds. The evaluation of several biomarkers have been reported: LDH leakage, low density lipoprotein oxidation assay, ROS evaluation, concentration of reduced glutathione, activity of antioxidant enzymes biomarkers of damage to lipid. This model, with a choice of numerous biomarkers of the cellular antioxidant defense system is a good research model, but it could be too complex for a high-throughput cell-based screening bioassay for dietary compounds’ antioxidant activity evaluation.

In conclusion, while noting that general oxidative stress biomarkers may not be good markers of a precise oxidative pathway, when choosing an endpoint to measure for a cell-based screening antioxidant bioassay, several factors must be considered such as easiness and length of workflow, high-throughput format, specificity and sensitivity of detection, and reproducibility of data.

Selecting cell culture media and environment Cell culture has often been used to study the cellular effects of reactive species and of antioxidants, and many useful data have been obtained. To extrapolate data obtained in cell culture models to the in vivo situation, results must be cautiously discussed and some consideration must be done. The environment in culture conditions can modify the properties of cells and tested compounds. Cells are normally cultured under 95% air/5% CO2, with about 150 mm Hg of O2. In the body, most cells are exposed to O2 concentrations in the range of 1 to 10 mm Hg. Therefore, cells in culture are under an oxidative stress, which can alter their properties in multiple ways. It has been reported that rates of production of ROS by cellular enzymes (for example, xanthine oxidase) or by leakage from electron transport chains (especially in mitochondria) appear to be O2-limited at 10 mm Hg (Halliwell 2003). Another aspect to be considered, when cells are used for studies involving the oxidative stress, is the composition of the cell culture medium and the presence or absence of FCS, which may affects both the status of the cells and the properties of added compounds. Cell culture media are frequently deficient in antioxidants, especially tocopherols, ascorbate, and selenium. Therefore the beneficial effects of added antioxidants can lead to an over-interpretation of the antioxidant efficacy of tested compounds. Other aspects to be considered are the hydrophobicity, stability, and the redox properties of compounds in the culture media. Flavonoids and other polyphenols are instable in commonly-used culture media, especially Dulbecco's Modified Eagle's Medium (DMEM) (Halliwell 2003). It has been demonstrated that different chemical and biological environments can modify the redox properties of carotenoids (Palozza and others 2003; Palozza 2005). Lin and others (2007) demonstrated that the stability and uptake of lycopene in cell culture is improved by the presence of FCS. Several evidences suggest that, in cell culture, the effectiveness of retinoids, which are relatively hydrophobic and unstable, may depend upon the type of medium and the presence or absence of FCS (Klaassen and others 1999; Tsukada and others 2002; Cheli and others 2003).

As discussed previously, the disease-protective effect of polyphenols is often attributed to their powerful antioxidant activities, as established in vitro. Most of the studies have emphasized their antioxidant effects; however, polyphenols can also exert pro-oxidant activities under certain experimental conditions (Cao and others 1997; Lapidot and others 2002; Rufian-Henares and others 2006; Dai and Mumper 2010). Phenolic compounds in cell culture media undergo rapid oxidation to generate substantial amounts of H2O2 (Long and others 2000; Halliwell 2008). These researchers demonstrated that cell culture media have different “pro-oxidant” activity, with DMEM the most ‘‘pro-oxidant.” Thus, one should always be alert when adding polyphenols to cells in culture and must check for reactions taking place in the culture medium that could lead to artifacts and carefully distinguish effects of oxidation products from ‘‘real” effects of polyphenols. Addition of several components, such as catalase and pyruvate, can be used to scavenge the H2O2 (Babich and others 2009; Long and Halliwell 2009).

To conclude, the effect of cell culture medium and environment seems to be often under-appreciated, but can cause artifacts in the interpretations of the cellular effects of added compounds tested for their antioxidant activity (Halliwell 2003). A numerous examples of artifacts caused by oxidation of compounds added to cell culture media have been reported by Halliwell (2008). Not all the cellular effects of antioxidant are due to artifacts, but it is necessary to consider and taking into account that the cell environment can be a potential source of error for a proper evaluation of the true effects of antioxidant compounds.

Selecting cell culture system Two-dimensional (2-D) cell culture systems are easy and convenient to set up and are very useful for routine screening of food antioxidant properties. However, normal cells experience a 3-dimensional environment, completely surrounded by other cells, extracellular matrix, fibrous layers, and adhesion proteins. Adding the 3rd dimension (3-D) to a cell's environment in “in vitro models” creates significant differences in cellular characteristics and behavior. 3-D cell culture technologies have revolutionized our understanding of cellular behavior. The principal 3-D cell culture systems currently available and employed in basic and applied research are mainly based on the use of extracellular matrix constituents, microporous membranes, bio, and nanomaterials for scaffold construction (Table 4). The current technologies of modeling tissue in 3-D have been recently reviewed by several researchers (Lee and others 2008; Justice and others 2009; Mazzoleni and others 2009). The first description of culturing cells on filters has been reported by Grobstein (1953). Since that time, mammary and intestinal epithelial cells grown on microporous membranes as 3-D cell models of the mammary gland and gut, respectively, has been widely described and used (Cheli and others 2001; Cencič and Langerholc 2010; Diesing and others 2011). Culturing cells in this 3-D model ensures the formation of a functional epithelial barrier in terms of integrity and polarity. The applications for researches in the field of food microbiology and functional food evaluation have been recently reviewed (Cencič and Chingwaru 2010; Cencič and Langerholc 2010; Chopra and others 2010). Results indicate that this cell culture model may be a powerful tool as a bioassay for the functional screening of food, food components, and supplements. Up to now, the 3-D culture technologies developed for research applications do not scale well for screening applications and therefore are still far from a real application as functional bioassays of feed. For screening applications, a significant 3-D culture expansion is still needed.

Table 4–.  Examples of most promising currently available 3-D cell culture technologies suitable for cell-based food screening.
Cell culture modelsAdvantagesDescription
Organotypic cultures: extracellular matrix (ECM)-based culturesReconstitution of tissue-like organization (polarity, function, viability)Naturally derived materials (for example, collagen, laminin, fibrin, Matrigel)
  Synthetic polymers (for example, poly[dimethylsiloxan], poly[DL-lactide-co-glycolide])
Organotypic cultures: cells cultured on insertsReconstitution of an epithelial barrier-like organization (polarized metabolic processes and function)Microporous filter membranes (hydrophilic poly[tetrafluoroethylene][PTFE], cellulose esters, polycarbonate membrane, polyethylene teraphthalate [PET])
  A variety of ECM coating can be considered.
Microcarriers-based culturesMaintenance of cell viability and tissue-like functionsSmall spheres with different coatings (gelatin, collagen, laminin)

In the continuous search for new format for cell-based assays, microfluidic technology and microarray technology represent a further and promising evolution. Microfluidic technologies, that can manipulate samples at nanoliter volumes, are now designed and developed by taking advantage of the microfabrication techniques which were originally used in patterning the electrical circuits on silicon chips. Development of new methods for implementing cell-based assays based on digital microfluidics with no adverse effects on cell viability has been reported (Barbulovic-Nad and others 2008). This method is advantageous for cell-based assays because of automated manipulation of multiple reagents associated with a reduction in reagent use and analysis time. The use of cell-based microarray technology has been investigated and a higher sensitivity than a conventional well plate assay has been demonstrated (Diaz-Mochon and others 2007). These technologies have great potential as an analytical tool for implementing cell-based assays on the microscale. They are still research models and their success, as cell-based bioassays for test screening applications, relies on the development of stable, reproducible and low cost assays.

Cell-based bioassay for evaluating antioxidant activity of food: standardization

As discussed previously, many different cell-based bioassays have been developed and used as screening tests to evaluate the effectiveness of antioxidants and for this reason the data obtained by different researchers and laboratories are extremely difficult to compare and interpret. These results highlight that the experimental models still cannot be transferred as such from the area of research to routine use. The transition from a research model to a test model needs standardization, optimization, automation and, if it is possible, miniaturization. The validation of a cell-based bioassay is a complex process and a critical point as differences exist in the cell lines, types of oxidants/antioxidants, media, concentrations of antioxidant, incubation times, and oxidative stress quantification methods. For a cell-based bioassay as an “antioxidant test protocol” important standards regarding assay procedures, choice of cellular models, and the appropriate use and interpretation of nonlinear dose-responses in cellular models must be defined to ensure more consistency in results. The problem of validation of cell culture model for toxicity testing and screening has been widely discussed by Zucco and others (2004).

The first topic is the choice of the cell type for the purpose of antioxidant testing. It is fundamental to check whether the cells used in a cell-based bioassay maintain the integrity of cell structure, and the signaling pathways of a nonaltered functional response to oxidative stress. Most of the articles simply indicate the cells used and did not specify if the cells meet this requirement. As already discussed, altered functional response to oxidative stress have been described for HepG2 cells and therefore criticism have been raised for their use. The reduction of ROS production is one of the mechanisms of antioxidant-rich natural products. ROS production is recognized as a tightly controlled process in the cell and therefore the regulatory mechanisms must be conserved in the cells. The process of apoptosis leads to some production of ROS by the cell's mitochondria (Jou and others 2002). Honzel and others (2008) conclude that if an antioxidant-rich natural product reduces ROS production in such a cell model (tumoral cells), the interpretations are far from simple and that the product possibly protects tumor cells, which is not a desirable conclusion. Therefore, several researchers proposed the use of erythrocytes as a cell model (Honzel and others 2008; Jensen and others 2008; Blasa and others 2011).

Another important topic is testing the activity of pure antioxidant compounds as positive controls to standardize and validate a cell-based method for a routine use as test for dietary compounds’ antioxidant activity. The activity of quercetin, caffeic acid, gallic acid, and alpha-tocopherol have been tested by Wolfe and Liu (2007) and Girard-Lalancette and others (2009). These researchers recommend quercetin, that it is found in most fruits, vegetables, and other plants, as a positive control because of its high activity and because the pure compound is relatively stable and can be easily and economically obtained.

For a cell-based assay standardization, a correct choice of the cellular biomarker as an endpoint to measure must be done. As previously discussed, several endpoint measures have been used. If cell viability assays are chosen, the detection sensitivity varies with the cell type, plate format, and number of cells/well and the parameter being measured. Cell-based bioassays that are designed to detect a change in viability in a population of 10000 cells, such as a MTT test, may not require the most sensitive assay technology. On the other hand, assay model systems that use low cell numbers in a high-density multiwell plate format may require maximum sensitivity of detection, such as that achieved with the luminescent ATP assay technology. The use of fluorescent or fluorogenic probes to monitor cellular oxidation may represent a sensitive and easy to use endpoint to measure for a high-throughput testing cell-based bioassay. In fact, the use of these probes allow to detect a general oxidative stress, although it cannot be used as markers of a precise oxidative pathway (Laguerre and others 2010).

Finally, the need of an accurate evaluation and selection of the cell culture medium and environment for a cell-based assays to be used in studies involving antioxidant activity evaluation have been previously discussed. This aspect cannot be under-appreciated, as the cell environment can be a potential for a non correct determination of the true effects of antioxidant compounds.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References

Antioxidant research expanded widely during the last years and several assays measuring the total antioxidant activity of compounds, food, feed, and additives have been developed. The chemical assays may be good models for comparing molecules for their antioxidant capacity, but more biologically relevant cell-based models for the determination of antioxidant activity are needed to evaluate the biological relevance of multifunctional food, antioxidant activity and functional properties. Several cell-based assays have been developed for testing dietary compounds for their antioxidant activity. To represent more accurately what occurs in vivo in the human body, the antioxidant activity of numerous molecules have been tested in vitro at wide concentration ranges, usually selected based on the lack of cytotoxic effect and a realistic concentration range in human diet and serum. Up to now, a wide divergence of results has been reported and for this reason, the data obtained by different researchers and laboratories are extremely difficult to compare and interpret. The transition of cell-based bioassays from research models to test models still needs optimization, standardization, and validation of analytical protocols. Moreover, despite the antioxidant activity has been demonstrated in vitro, caution must be taken when transferring these results to an in vivo situation, suggesting benefits from antioxidant in the diet or dietary supplementation. Sometimes inconsistent results have been reported regarding the association between dietary antioxidants and low risks of diseases. This reflects the limitation and complexity of human clinical studies, in which several confounding factors must be considered in the evaluation of the results, like diet differences, total antioxidant status, sex, age, type of antioxidant formulation, duration of the supplementation, stage of disease, interaction between dietary modulation, and genetic composition of the individuals.

In conclusion, cell-based bioassays can improve our understanding of the effects of antioxidant compounds. They represent a promising analytical tool for the initial antioxidant screening of food compounds and dietary supplements, and therefore an important support in antioxidant research prior to animal studies and human clinical studies.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References