Variability of antioxidant activity among honeybee-collected pollen of different botanical origin

Norma Almaraz-Abarca, Maria da Graça Campos, J. Antonio Ávila-Reyes, Néstor Naranjo-Jiménez, Jesús Herrera-Corral and Laura S. González-Valdez

Norma Almaraz-Abarca. Biologist, Universidad Nacional Autónoma de México (UNAM). Doctor in Plant Physiology, Instituto Politécnico Nacional (IPN), Mexico. Researcher, Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional, IPN, Durango (CIIDIR-IPN-Dgo.). Address: Biotechnology Laboratory, CIIDIR-IPN. Sigma s/n Frac. 20 de Noviembre II, Durango, Dgo., Mexico. CP 34220. e-mail: nalmaraz@ipn.mx

Maria da Graça Campos. Doctor in Phrmacology, Universidade de Coimbra, Portugal. Professor, School of Pharmacy, Universidade de Coimbra, Portugal.

José Antonio Ávila-Reyes. Biologist, UNAM. M.Sc. in Science Methodology, IPN. Doctoral Candidate, UNAM. Researcher, CIIDIR-IPN-Dgo., Mexico.

Néstor Naranjo-Jiménez. Biologist, IPN, México. M.Sc. in Ruminal Nutrition Biotechnology, Universidad Juárez del Estado de Durango (UJED), Mexico. Doctoral Candidate, Universidad Autónoma de Zacatecas (UAZ), México. Researcher, CIIDIR-IPN-Dgo., Mexico.

Jesús Herrera-Corral. M.Sc. in Ruminal Nutrition Biotechnology, Universidad Juárez del Estado de Durango (UJED), México. Doctoral Candidate, Universidad Autónoma de Zacatecas (UAZ), Mexico. Researcher, CIIDIR-IPN-Dgo., Mexico.

Laura Silvia González-Valdez. Biochemical Engineer and M.Sc. in Biochemistry, Instituto Tecnológico de Durango (ITD), Mexico. Researcher, CIIDIR-IPN-Dgo., Mexico.

Resumen

Se determinaron las actividades antioxidantes de los extractos crudos de una mezcla de polen apícola y de cada uno de los seis polen constituyentes que formaban esa mezcla. Las determinaciones se hicieron por el método de sustancias reactivas al ácido tiobarbitúrico (TBARS) en preparaciones microsomales de hígado y por el método del bloqueo del radical libre 2,2-difenil-1-picrilhidracilo (DPPH*). Las actividades se compararon con la composición de flavonoles y ácidos fenólicos y con los contenidos de flavonoles en el polen. Todos los extractos mostraron actividades antioxidantes. Éstas fueron diferentes para cada especie y no estuvieron claramente asociadas al contenido de flavonoles en el polen. El polen de Amarathus hybridus mostró una alta capacidad inhibidora de la oxidación lipídica. El de la mezcla entera y el de Tagetes sp. fueron efectivos bloqueadores de radicales libres.

Summary

The antioxidant activities of total extracts of a mixture of honeybee-collected pollen and its six constituent pollens were determined by lipid peroxidation assay by the thiobarbituric acid reactive substances (TBARS) test on hepatic microsomal preparations and by free radical scavenging (2,2-diphenyl-1-picrylhydrazyl; DPPH*) method. Activities were compared to the flavonol and phenolic acid compositions and flavonol contents in pollen. All extracts showed antioxidant activities as radical scavenger substances and as inhibitors of lipid peroxidation. Antioxidant activities were different for each species and were not clearly associated to the flavonol content in pollen. Pollen of Amaranthus hybridus was a potent lipid oxidation inhibitor, and that of Tagetes sp. and the whole mixture were effective antiradical substances.

Resumo

Determinaram-se as atividades antioxidantes dos extratos crus de uma mescla de pólen apícola e de cada um dos seis polens constituintes que formavam essa mescla. As determinações se fizeram pelo método de substâncias reativas ao ácido tiobarbitúrico (TBARS) em preparações microssomais de fígado e pelo método do bloqueio do radical livre 2,2-difenil-1-picrilhidracilo (DPPH*). As atividades se compararam com a composição de flavonóis e ácidos fenólicos e com os conteúdos de flavonóis no pólen. Todos os extratos mostraram atividades antioxidantes. Estas foram diferentes para cada espécie e não estiveram claramente associadas ao conteúdo de flavonóis no pólen. O pólen de Amarathus hybridus mostrou uma alta capacidade inibidora da oxidação lipídica. O da mescla inteira e o de Tagetes sp. foram efetivos bloqueadores de radicais livres.

KEYWORDS / Antioxidant Activity / Flavonoid Profile / Free Radical Scavenging Activity / Honeybee-collected Pollen /

Received: 04/20/2004. Modified: 08/11/2004. Accepted: 08/13/2004.

Introduction

Honeybee-collected pollen is recognized as a well balanced food (González-Güerca et al., 2001). This beehive product also has several useful pharmacological properties, such as antibiotic, antineoplasic, antidiarrhoeatic and as an antioxidant agent (Campos, 1997). The antioxidant activity of honeybee-collected pollen has been recognized as a free radical scavenger and as a lipid peroxidation inhibitor (Campos et al., 1994; Campos, 1997). This activity has been associated with the phenolic pollen content (Campos, 1997). Usually, honeybee-collected pollen is a mixture of pollen pellets from different botanical origins, each one being an important source of flavonol glycosides (Wiermann and Vieth, 1983) and, in some species, of hydroxycinnamic acids (Campos, 1997). These compounds are found in a species-specific profile (Campos, 1997), which suggests that honeybee-collected pollen from different areas or seasons could have different antioxidant activities. In spite of the relevance of honeybee-collected pollen as an antioxidant substance, there is not enough systematic information about the antioxidant activity levels associated to the flavonol content and profile of honeybee-collected pollen from different botanical origins. The purpose of the present study was to evaluate the effectiveness of total extracts of a mixture of honeybee-collected pollen and its constituent pollens from Durango, Mexico, as free radical scavengers and as inhibitors of lipid peroxidation, and correlate the flavonol contents and profiles with the variability in the observed activities.

Materials and Methods

Chemicals

Standards for ascorbic acid, caffeic acid, quercetin, quercitrin and p-coumaric acid; 2,2-diphenyl-1-picrylhydrazyl (DPPH*); analytical grade absolute ethanol, n-butanol and aluminum chloride, and HPLC grade methanol, and acetonitrile were purchased from Sigma, USA.

Pollen samples

The mixture of honeybee-collected pollen was provided by local beekeepers expressly for this research. This sample represents the collection from La Parrilla, Durango, Mexico, harvested in September 1999. Pollen from the anthers of flowers of Zea mays L. (Poaceae), Tagetes sp. (Compositae), Amaranthus hybridus L. (Amaranthaceae), Solanum rostratum Dun. (Solanaceae), Bidens odorata Cav. (Compositae) and Ranunculus petiolaris HBK. (Ranunculaceae), among others, was collected from plants growing in the surroundings of the beehives. The voucher specimens were placed in the Herbarium of the Centro Interdisciplinario de Investigación para el Desarrollo Integral Regional (CIIDIR) and all the species were identified by Socorro González, herbarium botanist.

Microscopic examination

Honeybee-collected pollen was manually sorted by pellet color into different types, as previously described (Campos, 1997). Fifteen pollen pellets from each type and five samples of anther pollen were individually submitted to microscopic examination to define the botanical origin and homogeneity of the constituent pollens of the mixture of honeybee-collected pollen. Pollen samples were previously acetolyzed (Erdtman, 1966). An Olympus BX 40 microscope was used.

Preparation of extracts

Twenty grams of the mixture of honeybee-collected pollen and of each of the constituent pollens were individually extracted five times in 200ml ethanol-water solution (50% v/v) with a 60min maceration. The extracts were separated by centrifugation (15269g) for 10min. All the supernatants of each type were brought together and formed the total extracts. These total extracts were evaporated to dryness at low-pressure.

HPLC analysis

Single pellets of each type were extracted with ethanol-water (50% v/v; 1ml) and sonicated for 60min. The resultant mixtures were centrifuged (15269g) for 10min and the supernatants used for high pressure liquid chromatographic (HPLC) analysis, as previously described (Campos, 1997). Extracts (20µl) were analyzed on a Gylson 305 HPLC system, UV detector Gilson 170 and Waters Spherisorb S50D52 (5µm) column (4.6x250mm) by an acidified acetonitrile-water gradient (Campos, 1997). Standard chromatograms were plotted at 340nm. Spectral data for all peaks were accumulated in the range 220-400nm using diode-array detection. Samples of pollen collected directly from anthers were analyzed in the same manner.

Determination of flavonol content

Flavonol contents were determined by linear regression analysis from the standard curve of quercetin (1 to 50µg/ml vs. absorbance): Abs_425nm= -0.00221 + 0.054899 [Quercetin], correlation coefficient r= 0.9973 and from calibration curves of each total extract (1 to 400µl vs. absorbance). Curves were registered after the addition of aluminum chloride. Absorbances were registered at 425nm. Flavonol contents were expressed in µg of quercetin/g of polen dry matter, according to the flavonol predominance in this reproductive structure (Campos, 1997).

Free radical scavenging activity

Modifications to DPPH* method reported elsewhere (Campos, 1997) were used to evaluate the free radical scavenging activity. Four concentrations (0 to 400µl, respective concentrations of flavonols calculated from calibration curves and standard curve of quercetin) of each sample were individually added to a DPPH* solution (3.422µg/ml in ethanol-water, 50% v/v) in such a way as to maintain a final volume of 2ml. The decrease in absorbance was determined at 523nm after 10min. The DPPH* concentrations in the reaction medium against the flavonol concentrations of samples were plotted to determine, by linear regression, the efficient concentration at 50%, defined as the amount of antioxidant needed to decrease by 50% the initial DPPH* concentration (EC₅₀). The standards for ascorbic acid, caffeic acid, quercetin, quercitrin and p-coumaric acid, were also evaluated. The following calibration curve, made with DPPH* between 1.0 and 6.6µg/ml, was used to calculate the DPPH* concentration (µg/ml) in the reaction medium: A_523nm= 0.00021 + 0.02916 [DPPH*], correlation coefficient r= 0.9999. Antiradical activities were expressed in relation to a comparable constant dry weight in terms of EC₅₀ in µg/ml.

Lipid peroxidation assay

Determination of inhibition of lipid peroxidation was made by quantification of thiobarbituric acid reactive substances (TBARS) by a modification of the methods used by Ohkawa et al. (1979) and Uchiyama and Mihara (1978) on mouse liver microsomal preparations. The amounts of thiobarbituric acid reactants were expressed in terms of the malondialdehyde (MDA) concentrations (µg/ml). The livers of decapitated CF1 mice (Instituto Nacional de Virología, Mexico) were washed with ice-cold 0.9% NaCl and homogenated in chilled 1.15% KCl in a ratio of 1g of wet tissue to 9ml of KCl solution. Microsomal fractions were prepared according to Diczfalusy et al. (1996). Protein assay was performed on microsomal fractions by the method of Lowry (García and Vázquez, 1998). The reaction mixtures composed of 100µl microsomal suspension (86µg of protein/ml), 0-100µl of assay substance (respective concentrations of flavonols calculated from calibration curves and standard curve of quercetin), 50µl of an aqueous solution of 6.95mg of FeSO4 and 17.62mg of ascorbic acid, and a variable volume of Tris-HCl buffer (10mM, pH 7.4) in such a way as to maintain a constant final volume of 500µl (Campos, 1997), were incubated at 37ºC for 30min in capped tubes. After cooling in tap water, 3ml of an aqueous solution of phosphoric acid (1%) and 1ml of an aqueous solution of thiobarbituric acid (0.6%) were added to each tube. The mixtures were heated to boiling for 60min in sealed tubes. After cooling in tap water, 4ml of n-butanol was added, and the mixture was shaken vigorously. After centrifugation at 10179g for 10min, the MDA content was determined by measuring the absorbance of the organic layer at 535nm. Reference substances (quercetin and caffeic acid) were assayed at four concentrations. Lipid inhibition activities were expressed in terms of the concentration of antioxidant required to inhibit MDA formation by 50% values ( IC₅₀  in µg/ml), calculated from absorbances against the sample flavonol concentrations curves by linear regression, using the extinction coefficient of MDA (1.56·105M^-1·cm^-1).

Statistical analysis

The analyses were carried out in triplicate. Data were separated by an analysis of variance (p£0.05) and means separated by Duncan's multiple range test. The results were processed by COSTAT computer program (1982).

Results and Discussion

Microscopic examination

On the basis of color, six types of pollen pellets were found in the mixture of honeybee-collected pollen (Table I). Microscopic examination of all pollen samples showed that each pellet of the honeybee-collected pollen was largely homogeneous, confirming the observation (Campos et al., 1997) that pollen pellets predominantly consist of pollen grains from one species. The direct microscopic comparison between the different types of pollen pellets and the pollen collected from anthers provided evidence of the botanical origin of the six constituent pollens. The botanical identifications are showed in Table I.

These results reflect the complexity in terms of number of constituent pollens of this honeybee-collected pollen, in contrast with other reports (Serra et al., 2001) where monofloral pollen was reported as a frequent condition in honeybee-collected pollen. Regarding species dominance, in this mixture of honeybee-collected pollen co-dominance of practically five species of plants is found. These results and others as yet not reported do not agree with those of Campos et al. (1997), who reported that the major pollen types represented in any one bee pollen tends to be rather small. It has been claimed that bees are selective in their harvesting of pollen of wild over cultivated species of plants. It is of interest that in the present, as well as in other studies as yet not reported, analysis of mixtures of honeybee-collected pollen from Durango, Mexico, it is observed that pollen from Zea mays, a cultivated and anemophile plant, is found with the highest percentage. This implies that pollen collection behavior may be determined by a more complex combination of factors than it has been thought.

HPLC analysis

As found by other authors (Campos et al., 1997) the botanical origin of constituent pollens could be confirmed by direct comparison of the respective HPLC phenolic profiles with those of the pollen collected from anthers. The phenolic profile of any constituent pollen from the mixture of honeybee-collected pollen was identical to that of the species it came from. These results do not agree with those of Serra et al. (2001), who indicate that the specific plant origin of honeybee pollen can not be distinguished from its HPLC profile. Under the experimental conditions in which HPLC chromatograms were obtained (Campos, 1997), patterns comprising flavonoids and cinnamic acid derivatives were the only ones found (Table I). All pollens individually analyzed contained flavonol glycosides, specially quercetin and kaempferol derivatives, compounds with a broad spectrum of biological activity (Formica and Regelson, 1995; Campos, 1997). Pollen from Solanum rostratum was particularly rich in phenolic acid derivatives and those from Zea mays and Amaranthus hybridus were characterized by the absence of phenolic acid derivatives.

Free radical scavenging

The reduction of DPPH* concentration with increasing flavonol concentration was observed in all total extracts. The reduction was linear and dependent on the flavonol concentrations in the samples. According with the classification of kinetic behavior of Brand-Williams et al. (1995), the reaction kinetics was "rapid", reaching a steady state in less than one minute. The spectrometric recording of the DPPH* disappearance in the presence of increasing flavonol concentrations of total extract of mixture G is shown in Figure 1. These results show a clear correlation between the flavonol concentration and antiradical activity for the kinetic behavior of DPPH* disappearance, as was reported for some phenolic compounds (Brand-Williams et al., 1995), vegetable oils (Espín et al., 2000) and wine and grape fruits (Sánchez-Moreno et al., 1999). However, a clear correlation between the flavonol content in pollen and the antiradical activity seems to be more difficult to establish.

Table II contains the EC₅₀  values of the reference compounds used; caffeic acid has the highest free radical scavenging capacity (EC₅₀= 0.3µg/ml) among the tested standards and quercitrin has the lowest (EC₅₀= 1.1µg/ml). This property was intermediate for quercetin and ascorbic acid (EC₅₀= 0.4 and 0.6µg/ml, respectively). It is known that p-coumaric acid has a slow kinetic behavior, taking more than an hour to reach a steady state in the reaction with DPPH* (Brand-Williams et al., 1995). The absence of activity in this case was due to the fact that the evaluation of its antiradical activity was made after 10 minutes. Other authors have also reported a higher activity for caffeic acid than for ascorbic acid by the DPPH* method (Brand-Williams et al., 1995). Quercetin, a flavonol aglycon, is a more potent antiradical substance than quercitrin, which is a quercetin monosaccharide derivative with a C₃ rhamnosyl substituent. It is known that only C₃  disaccharide derivatives have a drastically reduced antiradical activity (Von Gadow et al., 1997).

The total extracts of honeybee-collected pollen mixture and the six different constituent pollens were effective antioxidants as free radical scavengers, although with lower levels of antiradical activities than all the standards tested. Antiradical activity (EC₅₀) and flavonol content in pollen were calculated in all the cases, looking for a correlation between the two measurements. The antiradical activities and the flavonol contents in the mixture and in the individual pollens appear in Table III, where samples are listed in decreasing flavonol content. Contrary to reports for the extracts from reproductive organs of Crataegus monogyna (Bahorun et al., 1994), a correlation between the flavonol content in pollen and antiradical activity is not apparent for the total extracts.

Significant differences were found among the EC₅₀ values of pollen of species of plants analyzed. The mean separation by Duncan's multiple range test is shown in Table III. Pollen from Amaranthus hybridus (G3) had the lowest antiradical activity (EC₅₀= 14.0µg/ml, labelled a). Those from Zea mays (G1), Ranunculus petiolaris (G6) and Bidens odorata (G5) had intermediate levels of activity (EC₅₀= 10.3, 9.9 and 9.3µg/ml, respectively) without significant differences among them despite large differences in flavonol content. Pollen from Solanum rostratum (G4) was grouped individually and showed a high level of antiradical activity (EC₅₀= 8.4µg/ml), while that from Tagetes sp. (G2) and the whole mixture had the highest antiradical activity (EC₅₀= 6.8 and 6.4µg/ml, respectively), without significant differences between them.

Comparing the antiradical scavenging activity of the total extract of the mixture of honeybee-collected pollen and those of its six constituent pollens individually, no correlation seems to exist between the flavonol content and the antiradical activity for pollen from different botanical origins. The results suggest that the flavonol and phenolic acid composition, rather than the concentration, could be the determinant factor. The particular combination of flavonol glycosides and phenolic acids could define the level of antioxidant capability of pollen of different origin.

Lipid peroxidation

Extracts were evaluated for their capability of inhibition of lipid peroxidation. In all cases, the reduction of MDA concentration with increasing flavonol concentration was linear. An example, obtained with the total extract of pollen from Amaranthus hybridus (G2), is shown in Figure 2.

The corresponding IC₅₀ values are shown in Table III, while those of the standards quercetin and caffeic acid are shown in Table II. Under the present experimental conditions, all the total extracts were effective inhibitors of lipid peroxidation although, as in the case of antiradical activity, a correlation between flavonol content in pollen and lipid inhibition activity is not clear. Significant differences were found among the  IC₅₀  values from pollen of the plant species analyzed, and the mean separation by Duncan’s multiple range test is included in Table III. Pollen from Amaranthus hybridus (G3) and the mixture of honeybee-collected pollen had the highest activities as lipid oxidation inhibitors ( IC₅₀ = 0.7´10^-1µg/ml in both), even higher than those of the caffeic acid ( IC₅₀ = 1.7´10^-1µg/ml) and quercetin ( IC₅₀ = 2.2´10^-1µg/ml) standards, this last one considered as a powerful antioxidant against lipid peroxidation (Terao, 1999). As with antioxidant activity, the results suggest that the diverse levels of lipid peroxidation activity could be mainly associated to the species-specific phenolic profile.

Conclusion

Honeybee-collected pollen can be a complex mixture of pollen from different botanical origins. Pollen from a cultivated and anemophile plant like Zea mays is found in a high proportion in mixtures of honeybee-collected pollen from Durango, Mexico.

Honeybee-collected pollen can be an effective antioxidant substance. Its antioxidant capacity is based on free radical scavenging and on lipid peroxidation inhibition activities. Pollen from different botanical origin had different antioxidant capacity. Individually, pollen from Amaranthus hybridus is among the ones with the highest lipid oxidation inhibition capacity, and that from Tagetes sp. is among the ones with the highest level of antiradical capacity. The species-specific flavonol and phenolic acid profiles may be more important than flavonol content to determine the particular antioxidant capacity of pollen of different botanical origin. This beehive product can be considered as an important source of natural flavonol antioxidants.

To reach a better understanding and to take advantage of the antioxidant properties of this apicultural product it is necessary to carry out in vivo assays and systematic analyses with standard methods so as to create databases that will make it possible to compare the antioxidant activities of honeybee-collected pollen from different botanical origins.

ACKNOWLEDGMENTS

The authors wish to thank Antonio Rivas Orozco for providing the honeybee-collected pollen samples.

REFERENCES

1. Bahorun TF, Trotin JP, Vasseur J, Pinkas M (1994) Antioxidant activities of Crataegus monogyna extracts. Planta Medica 60: 323-328.        [ Links ]

2. Brand-Williams W, Cuvelier ME, Berset C (1995) Use of free radical method to evaluate antioxidant activity. Lebensmit. Wiss. Technol. 28: 25-30.        [ Links ]

3. Campos RMG (1997) Caracterização do polen apícola pelo seu perfil em compostos fenólicos e pesquisa de algumas actividades biológicas. Thesis. School of Pharmacy. Universidade de Coimbra. Portugal. 318 pp.        [ Links ]

4. Campos MG, Cunha PA, Navarro MC, Utrilla MP (1994) Free radical scavenger activity of bee pollen. Polyphenols 94: 415-426.        [ Links ]

5. Campos MG, Markham KR, Proença da Cunha A (1997) Quality assessment of bee-pollens using flavonoid/phenolic profiles. Polyphen. Comm. 96: 53-54.        [ Links ]

6. Diczfalusy MA, Björkhem I, Einarsson K, Alexson SEH (1996) Acyl-coenzyme A: cholesterol O-acyltransferase is not identical to liver microsomal carboxylesterase. Arteroscl. Thromb. Vasc. Biol. 16: 606-610.        [ Links ]

7. Erdtman G (1966) Pollen morphology and plant taxonomy. Angiosperm. An introduction to palinology. Afner. New York, NY, USA. 145 pp.        [ Links ]

8. Espín JC, Soler-Rivas C, Wichers HJ (2000) Characterizations of the total free radical scavenger capacity of vegetable Oils and oil fractions using 2,2-diphenyl-1-picrylhydrazyl radical. J. Agric. Food Chem. 48: 648-656.        [ Links ]

9. Formica JV, Regelson W (1995) Review of the biology of quercetin and related bioflavonoids. Food Chem. Toxicol. 33:1061-1080.        [ Links ]

10. García AH, Vázquez DR (1998) Cuantificación de proteínas: una revisión. BioTecnología 3: 77-88.        [ Links ]

11. González-Güerca MC, Almaraz-Abarca N, Ávila-Reyes JA, Herrera-Corral J, Naranjo-Jiménez N (2001) Polen apícola: una alternativa alimenticia y terapéutica. Apitec 28: 19-23.        [ Links ]

12. Ohkawa H, Ohishi N, Yagi K (1979) Assay for lipid peroxides in animal tissues by thiobarbituric acid reaction. Anal. Biochem. 95: 351-358.        [ Links ]

13. Sánchez-Moreno C, Larrauri JA, Saura-Calixto F (1999) Free radical scavenging capacity and inhibition of lipid oxidation of wines, grape juices and related polyphenolic constituents. Food Res. Int. 32: 407-412.        [ Links ]

14. Serra BJ, Soliva TM, Centelles LE (2001) Evaluation of polyphenolic and flavonoid compounds in honeybee-collected pollen produced in Spain. J. Agric. Food Chem. 49: 1848-1853.        [ Links ]

15. Terao J (1999) Dietary flavonoids as plasma antioxidants on lipid peroxidation: significance of metabolic conversion. In Packer L, Hiramatsu M, Yoshikawa T (Eds.) Antioxidant food supplements in human health. Academic Press. San Diego, CA, USA. pp. 255-267.        [ Links ]

16. Uchiyama M, Mihara M (1978) Determination of malonaldehyde precursor in tissues by thiobarbituric acid test. Anal. Biochem. 86: 271-278.        [ Links ]

17. Von Gadow A, Joubert E, Hansmann CF (1997) Comparison of the antioxidant activity of aspalathin with that of other plant phenols of Rooibos tea (Aspalathus linearis), a-tocopherol, BHT, and BHA. J. Agric. Food Chem. 45: 632-638.        [ Links ]

18. Wiermann R, Vieth K (1983) Outer pollen wall, an important accumulation site for flavonoids. Protoplasma 118: 230-233.        [ Links ]