Screening Bolivian Plants for Antioxidant Activity

Abstract

The aerial parts of 17 Bolivian plants were screened to determine antioxidant activity. A methanol extract of each plant was prepared and partitioned sequentially with hexane, chloroform, and ethyl acetate, leaving an aqueous solution. All extracts and their 5 fractions, for a total of 102 samples, were evaluated using two techniques: an adaptation of the β-carotene bleaching technique using an emulsion of linoleic acid in water as the oxidizable substrate, and the DPPH free radical trapping technique. The results with the β-carotene bleaching technique were more discriminating and better related to the rancidity process under normal conditions; with this assay, 11 species provided at least one fraction with highly promising antioxidant activity. All species gave good results under the DPPH technique, and in most cases they performed better than BHA, which was used as a reference antioxidant. We advocate the simultaneous use of these two techniques for screening purposes.

Keywords:

• Aloysia gratissima
• antioxidant activity
• Baccharis grisebachii
• Baccharis salicifolia
• β-carotene bleaching
• Croton andinus
• Croton emporiorum
• DPPH
• Erechtites valerianaefolia
• Gynoxys psilophylla
• Hedyosmum angustifolium
• Lantana camara
• Lantana magnibracteata
• Lantana trifolia
• Lippia boliviana
• Mikania buchtienii
• Ophryosporus heptanthus
• Plazia daphnoides
• Senecio herzogui
• Tagetes multiflora

Introduction

The search for new sources of natural antioxidants has been promoted by the proved toxicity of synthetic antioxidants (Ito et al., 1983) and by the demand for them to have specific functional properties that ensure their usefulness in the wide range of industrial products prone to the rancidity process, as there is not a panacea compound that can be used effectively in each of those products (Löliger, 1991). Because of the high botanical diversity of the South American flora, and because in recent years there has been an increasing interest in the screening for antioxidant activity (AA) in plants from the region (Payá et al., 1996; Aquino et al., 2001; Mensor et al., 2001; Luo et al., 2002; Schinella et al., 2002; Vanderjagt et al., 2002), a review has been published, which gives an updated view of this particular subject (Desmarchelier et al., 2000). In this paper, we present our results on the evaluation of the AA of the aerial part of 17 Bolivian plants, which had not been studied up to this moment. We chose to work with these plants because Bolivia is one of the least studied regions of the western Amazon basin, has high indexes of biological diversity, and has an important ethnobotanic culture. Our aim is to identify species with high potential as a commercial source for novel natural antioxidants.

For each plant, we evaluated the AA in the crude methanol extract, and the fractions prepared from it by partitioning with hexane, chloroform, ethyl acetate, as well as the remaining water-soluble fraction. We decided to measure antioxidant activity using two different techniques. One is based on the bleaching of β-carotene coupled with the oxidation of linoleic acid in an oxygen-saturated water emulsion. In this assay, the linoleic acid peroxide radicals produced during the propagation step of the autoxidation of the fatty acid reacts inmediately with the β-carotene, and the kinetics of the autoxidation is followed by measuring the bleaching of the β-carotene at the appropiate wavelength. So, this technique sheds light on the antioxidant capacity of an extract or fraction in the rancidity process as a whole and tends to reproduce similar conditions as found in stored foodstuffs. The other technique essentially measures the capacity of an extract or fraction to trap free radicals, generated by 2,2-di-(4-octylphenyl)-1-picrylhydrazil (DPPH) in methanol solution, a highly polar medium. A recent discussion on techniques for measuring antioxidant activity, their comparison and limitations, can be found in several papers (Arnao et al., 1999; Niki & Noguchi, 2000; Vandersluis et al., 2000; Pannala & Rice, 2001; Schwarz et al., 2001; Antolovich et al., 2002; Koleva et al., 2002).

Materials and Methods

Instrument

We used a Spectronic 3000 from Milton Roy (Rochester, NY, USA) equipped with an eight-port thermostatable multicell holder. Slit width was set at 0.35 nm.

Materials

Linoleic acid and polyoxyethilenesorbitan monolaureate (Tween 20) were purchased from Sigma Chemical Co. (St. Louis, MO, USA); β-carotene was from BHD Chemical Ltd. (Poole, UK); 2-t.-butyl-4-methoxyphenol (BHA) was from E. Merck (Darmstadt, Germany); SEP-PAK silica cartridges were from Waters (Milford, MA, USA); 2,2′-azobis.(2-amidinopropane)dihydrochloride (ABAP) was from Wako Chemicals (Richmond, VA, USA). Chloroform, DMSO, and n.-hexane were obtained from J.T Baker Chemical Co. (Phillipsburg, NJ, USA); ethyl acetate was from BHD Chemical Ltd. (Poole, UK); 2,2-di-(4-octylphenyl)-1-picrylhydrazil (DPPH) (25,762-1) was from Aldrich Chemical Co. (Milwaukee, WI, USA); methanol was from Mallinckrodt Baker Inc. (Paris, KY, USA).

Plant collection, extraction, and fractionation

Plants were collected in different regions of Bolivia. Table 1 contains taxonomical information, common names, and the place of collection of the plants studied. We selected nine species of the family Asteraceae, one Chloranthaceae, two Euphorbiaceae, and five Verbenaceae. There are no publications in the literature on the evaluation of the antioxidant activity of any these species. Voucher specimens of the plants were identified and deposited at the Herbarium of the National Flora Reserve “Martín Cárdenas” in Cochabamba. A crude methanol extract (CE) was prepared from the aerial part of each plant. The vegetal material (30 g) was ground and extracted with 600 ml methanol. The extraction was performed at room temperature; after 24 h of maceration, the solids were filtered through filter paper and discarded. Finally, the solvent was removed under reduced pressure.

Table 1.. Identification, common name, collection place, and voucher numbers of the plants studied.

Family	Species	Common name	Collection place	Voucher numbers
Asteraceae	Baccharis grisebachii. Hieron.	Mata vacas	Independencia	MM 1824
Asteraceae	Baccharis salicifolia. (Ruiz & Pavon) Pers.	Huaycha	Carrasco	MZ 369
Asteraceae	Erechtites valerianaefolia. (Wolf Decandolle)	Encame	Chapare	MZ 319
Asteraceae	Gynoxys psilophylla. Klatt	Qupi	Ayopaya	MM 1839
Asteraceae	Mikania buchtienii. (B.L. Rob.)	Wuaca nunu	Chapare	MZ 307
Asteraceae	Ophryosporus heptanthus. (Sch.Bip. ex Wedd.) R.M. King & H. Rob.		Carrasco	MM 1821
Asteraceae	Plazia daphnoides. Wedd.	Thola	Esteban Arce	MR 66
Asteraceae	Senecio herzogui.	Falsa mora	Chapare	MZ 304
Asteraceae	Tagetes multiflora. H.G.K.	Suico enano	Independencia	MM 1823
Chloranthaceae	Hedyosmun angustifolium. (Ruiz & Pavón) Solms-Laubach	Matico menta	Chapare	MZ 316
Euphorbiaceae	Croton andinus. Mull Arg	Duraznillo	Mizque	MM 1808
Euphorbiaceae	Croton cf. emporiorum. Croizat		Independencia	MM 1829
Verbenaceae	Aloysia gratissima. (Gillies & Hook) Tron.	Cut ´ u-cut ´u	Mizque	MM 1804
Verbenaceae	Lantana camara. L.		Carrasco	MZ 385
Verbenaceae	Lantana magnibracteata.	Jank ´a-jank ´a	Mizque	MM 1803
Verbenaceae	Lantana trifolia. L.		Carrasco	MZ 387
Verbenaceae	Lippia boliviana. Rusby	Kapa-kapa	Capinota	MM 1850

The crude extract (CE) was dissolved in the minimal possible amount of methanol, and 200 ml of distilled water was added. After 24 h of refrigeration, the solution was filtered through filter paper, and the solids were discarded. This aqueous solution was thoroughly partitioned consecutively with hexane (FHX), chloroform (FC3), and ethyl acetate (FAC). Organic solvents were removed under reduced pressure. The remaining water-soluble fraction (FOH) was freeze-dried. CA is what remains after CE has been partitioned with hexane.

β-Carotene bleaching method for antioxidant activity measurement

The antioxidant evaluation technique that we present in this paper is based on a previous work we have published (Rosas-Romero et al., 1999), which itself is based on the work of Marco (1968), later modified by Miller (1971).

Handling of the linoleic acid

The content of a 10 g sample was distributed in 0.5 mL vials. The vials were kept under dry purified argon, sealed and stored at −15°C. Marco's criteria for linoleic acid purity, solidification at −15°C in less than 16 h (Marco, 1968), was replaced by using linoleic acid with an absorbance at 234 nm, not higher than 0.1, which better indicates that the substrate is free from oxidation products.

ABAP solution

ABAP (54.2 mg) was dissolved in 500 µl argon-saturated purified water.

BHA solution

BHA (24.6 mg), recrystallized from ethanol, was dissolved in 100 ml DMSO. BHA was used as the reference antioxidant.

Preparation of the β-carotene solutions

One gram of β-carotene was recrystallized from ethanol, placed in a vial flushed with argon, sealed, and stored at −15°C. Twenty milligrams of β-carotene was dissolved in 500 µl chloroform and transferred, with petroleum ether, to a 10 ml volumetric flask. This solution, kept under dry purified argon at −15°C, may be used for several days.

Preparation of the β-carotene emulsion

Five hundred microliters of the β-carotene solution was passed through a SEP-PAK silica cartridge, previously flushed with 5 ml petroleum ether, and eluted into a 100 ml volumetric flask with 2 ml of the same solvent four times. The solution was concentrated to 10% its original volume by flushing with dry purified argon. Tween 20 (200 mg) and 13.5 mmol linoleic acid were added under argon flush. The remaining solvent was thoroughly evaporated with argon. The content of the flask was taken up to 100 ml using purified water, which had been saturated with dry purified argon. The flask was taken into an ultrasonic bath to degas and to homogenize the emulsion. The resulting emulsion should be transparent. Thirty milliters of this emulsion was saturated with oxygen for 60 s. At this point, the emulsion was used immediately.

Solutions for antioxidant activity evaluation

DMSO solutions of the material to be evaluated (CE, CA, FHX, FC3, FAC, and FOH) were prepared at 24.6 mg/100 ml, the same as the BHA solution.

Other recommendations

Water used throughout the work was purified and triple distilled, with conductivity less than 4 µS, otherwise the emulsion would not be homogeneous, nor would the kinetic data be reproducible. After each use, the SEP-PAK was washed with ethanol and with 10 ml of petroleum ether; the cartridges were discarded when channels began to appear in the silica. The preparation of the solutions, emulsions, and the oxidation reactions was carried out under a dim, diffused light. Argon was purified by flowing it through charcoal, molecular sieve, and silica gel columns.

Evaluation procedure

Cell 1 (spectroscopic reference cell, 100% T) contained 2100 µl of a solution of 100 mg Tween 20 in 50 ml purified water. All the other cells were filled with 1990 µl of the β-carotene emulsion. Six minutes were waited until thermal equilibrium was attained at 32.0 ± 0.1°C. ABAP solution (10 µl) was then added to cells 2 to 8 to initiate the linoleic acid oxidation; each cell was stirred for 20 s, and absorbance measurements were taken at 460 nm every 2 min for 10 min. After this time, 100 µl of the BHA solution was added to cell 2, and the same volumes of the solutions of the materials to be evaluated were added to cells 3 through 7, respectively. One hundred microliters of DMSO was added to cell 8 (control, no antioxidant added). Cells 2 to 8 were shaken in a vortex for 20 s. Readings were automatically taken every 2 min for 90 min.

For each extract or fraction, the control and the reference antioxidant (BHA) were evaluated simultaneously in every experiment in order to obtain good reproducibility.

Free radical capture capacity by the DPPH technique

Preparation of DPPH solutions

DPPH (2.0 mg) was dissolved in 100 ml methanol to give a 20 mg/l solution. This solution was used immediately and kept under a dim, diffused light during use.

Preparation of the extracts and fractions solutions

Methanol solutions of the extracts and fractions were prepared at 300 µg/ml and from those dilutions were made to obtain the following concentrations: 100, 50, 10, 5, and 1 mg/l.

Other recommendations

Purified water with conductivity no greater than 4 µS was used. The preparation of the solutions, emulsions, and the oxidation reactions were carried out under a dim, diffused light.

Oxidation procedure

Preparation of the reaction cells: Cell 1 (reference cell, 100% T) contained 2250 µl of the solvent (2:1 methanol:water). Cell 2 (A_DPPH) contained 1500 µl of the DPPH solution and 750 µl of purified water. Cell 3 (A_blank) contained 1500 µl of solvent and 750 µl of the sample solution (an extract or one of its fractions). Cells 4 to 8 (A_sample) contained 1500 µl of the DPPH solution and 750 µl of the respective sample solution at room temperature. After 5 min, when the equilibration was completed, the absorbance was measured at 531 nm (Brand-Williams et al., 1995).

CE, CA, FHX, FC3, FAC, and FOH were evaluated following the DPPH technique.

The percentage activity for the DPPH test was calculated according to We express the DPHH results as the concentration, of the extract or fraction, necessary to inhibit 50% of the free radical oxidation promoting effect of DPPH, EC₅₀, in µg/ml. Concentrations are expressed in µg/ml.

Statistical analysis

The antioxidant activity determinations were carried out in triplicate. Student's t.-test and regression analysis were performed using Statgraphic Plus 6.0 (Manugistic Inc, Rockville, MD, USA). Differences at p. > 0.05 were considered significant.

Results and Discussion

Figure 1 shows the results of a typical β-carotene bleaching experiment. During section A thermal equilibrium is attained. At the end of this time, the initiator (ABAP) was added, and during section B, the kinetic steady-state for the reaction is established. At time C, BHA was added to cell 2 as a standard, and in cells 3 to 7 the solutions (extracts or fractions) to be evaluated were added. Cell 8 represents the oxidation of linoleic acid with no antioxidant added, as a control. The differences in slopes during section D were related to the antioxidant effect.

Figure 1 Antioxidant effect on the oxidation of linoleic acid, followed by the β-carotene bleaching technique. During A, thermal equilibrium is attained. Steady-state kinetics is established during B after the addition of the initiator (ABAP). At point C, the standard (BHA) and the solutions under evaluation are added. Section D shows the antioxidant effect. Curve 2 corresponds to BHA; curve 8 is the control (reaction mixture with no antioxidant added), and curves 3 to 7 are the samples. Absorbance was monitored at 460 nm. Cell 1 is used to set 100% transmittance.

We define the antioxidant activity, Φ, of an extract or fraction, as a percentage of that of the standard, BHA, by: where S_c = slope of the control reaction, no antioxidant added; S_f = slope of the reaction for the fraction (or extract) under evaluation; S_BHA = slope of the reaction containing the standard antioxidant, BHA; δ = (BHA concentration/fraction or extract composition, in µg/ml), equal to 1 in our case.

Slopes were calculated from the regression of absorbance over time, using the data from the first linear segment of each curve in D.

It should be noted that when we refer to the general term “antioxidant activity,” we will use AA, and use Φ when we refer to the results obtained with the β-carotene bleaching technique.

We have chosen these two tests for our screening work because they are easy to carry out. They use small amounts of sample, they are reproducible, and they measure the antioxidant activity from two different points of view. The technique based on the bleaching of β-carotene is very well related to the overall phenomenon of rancidity in fats and oils, as it works with linoleic acid in an oxygen-saturated water emulsion, whereas the other technique is more related to the capability of the material being tested to inhibit the action of the free radicals generated by DPPH on the oxidation reaction in a highly polar environment. Thus, for an extract or fraction to be considered as a good antioxidant, it is not necessary to obtain good results in both tests. Good results with the β-carotene technique means good capability for inhibiting the overall rancidity process of linoleic acid, but the same sample may give poor results under the DPPH technique, meaning that its antioxidant activity is not based on its capability to trap the DPPH free radicals, and situations where activity is observed in one system only are also common (Brand-Williams et al., 1995).

Table 2 shows a summary of the results using the β-carotene technique. In our opinion, because this technique is based on the oxidation of an emulsion of linoleic acid under normal conditions, it provides an accurate description of the overall antioxidant activity as related to the actual rancidity process in fats and oils. Following the procedures explained in the experimental section, this technique gives highly reproducible results, it is easy, calls for very small amounts of sample, and is neither demanding on time nor on instrumental equipment. Actually, in our previous paper on the subject (Rosas-Romero et al., 1999), we showed how this technique could be used to obtain an absolute number to express the antioxidant activity (AA) of a compound, which solves the problem of having to express AA as a percentage (or any other ratio) of another compound defined as reference. This also allows us to estimate the AA of a compound based solely on its structural formulae, from which the data needed for the estimation can easily be calculated. For this work, we developed the shortcut of employing BHA as standard, because in this way the technique is better suited for screening purposes.

Table 2.. Antioxidant activity by the β-carotene bleaching technique.

		% Inhibition (β-carotene technique)
Family	Species	CE	CA	FHX	FC3	FAC	FOH
Asteraceae	Baccharis grisebachii.	n.a.	n.a.	n.a.	21.6	57.0	n.a.
Asteraceae	Baccharis salicifolia.	24.2	31.9	n.a.	15.7	49.0	16.2
Asteraceae	Erechtites valerianaefolia.	31.1	26.5	10.2	36.3	33.3	52.2
Asteraceae	Gynoxys psilophylla.	7.4	35.4	7.1	16.5	11.7	n.a.
Asteraceae	Mikania buchtienii.	58.0	17.8	n.a.	14.2	29.3	50.3
Asteraceae	Ophryosporus heptanthus.	20.4	n.a.	n.a.	59.1	37.2	34.1
Asteraceae	Plazia daphnoides.	51.4	4.2	n.a.	63.8	40.0	6.5
Asteraceae	Senecio herzogui.	44.3	51.7	3.8	25.6	58.6	41.9
Asteraceae	Tagetes multiflora.	18.4	21.6	13.4	13.3	n.a.	23.0
Chloranthaceae	Hedyosmum angustifolium.	44.9	20.3	45.4	16.3	59.5	5.9
Euphorbiaceae	Croton andinus.	1.4	2.5	3.2	76.1	n.a.	59.4
Euphorbiaceae	Croton cf. emporiorum.	n.a.	44.2	n.a.	12.1	21.4	49.2
Verbenaceae	Aloysia gratissima.	5.0	18.4	24.0	26.4	n.a.	22.3
Verbenaceae	Lantana camara.	42.1	12.4	n.a.	35.8	64.6	30.8
Verbenaceae	Lantana magnibracteata.	52.7	37.2	27.0	62.2	52.7	65.0
Verbenaceae	Lantana trifolia.	63.2	59.2	13.2	68.0	70.1	35.0
Verbenaceae	Lippia boliviana.	62.0	32.5	86.6	33.8	20.1	7.7

Results are expressed as % activity relative to the standard BHA solution. All samples were prepared at 24.6 mg/100 ml. Extracts or fractions with % inhibition higher than 40%, shown in boldface, are considered highly promising. n.a. means no activity detected. All results are statistically significant at p < 0.05. All experiments were carried out in triplicate.

In order to take into account the fact that we are not working with pure compounds but with crude extracts or their fractions, we established 40% of inhibition, as compared to 100% for BHA, as the minimum percentage acceptable to consider a sample as promising for our work; these values are in bold face in Table 2. The better antioxidants are those with the higher percentage of inhibition, Φ.

In all, we assayed a total of 102 samples, 6 samples for each of the 17 plants. With the exception of Gynoxys psilophylla., Tagetes multiflora., and Aloysia gratissima., the rest of our chosen vegetal material yielded at least one fraction with high Φ. A total of 31 extracts or fractions were active. We observed the same tendency as mentioned in previous work (Mensor et al., 2001), where no AA was reported in the hexane fraction (FHX), and that AA tends to concentrate in the more polar fractions; in the ethyl acetate fraction, in our case. But, in two cases we found Φ in the hexane fraction: in Hedyosmum angustifolium. and in Lippia boliviana.. Actually, the latter had the highest Φ of all the extracts and fractions studied.

In eight of our plants, Mikania buchtienii., Plazia daphnoides., Senecio herzogui., Hedyosmun angustifolium., Lantana camara., Lantana magnibracteata., Lantana trifolia., and Lippia boliviana., the crude methanol extract gave very good results. This is interesting because it suggests the possibility of obtaining good antioxidants with a minimum of fractionation. Croton andinus., Ophryosporus heptanthus., Plazia daphnoides., Lantana magnibracteata., and Lantana trifolia. are good sources of medium polarity antioxidants. On its part, Baccharis grisebachii., Baccharis salicifolia., Plazia daphnoides., Senecio herzogui., Hedyosmum angustifolium., Lantana camara., Lantana magnibracteata., and Lantana trifolia. are good sources of polar antioxidants, while the highest polarity antioxidants are provided by Erechtites valerianaefolia., Mikania buchtienii., Senecio herzogui., Croton andinus., Croton emporioum., and Lantana magnibracteata.. The results are quite promising because in order to satisfy the industrial demand for new natural antioxidants, the candidates have to cover the whole range of polarities, given the wide variety of applications.

The results with the DPPH technique are shown in Table 3. The experimental EC₅₀ value that we obtained for BHA is 11.5 µg/ml, so for the crude extracts and fractions we defined 15 µg/ml as the highest EC₅₀ that a sample could have in order to be considered interesting for our work. These numbers are shown in bold face in Table 3. In this case, the smaller the EC₅₀, the better the antioxidant.

Table 3.. Antioxidant activity by the DPPH technique.

		EC50 DPPH
Family	Species	CE	CA	FHX	FC3	FAC	FOH
Asteraceae	Baccharis grisebachii.	12.1	6.1	n.a.	n.a.	1.7	62.9
Asteraceae	Baccharis salicifolia.	12.7	17.0	n.a.	n.a.	3.3	16.6
Asteraceae	Erechtites valerianaefolia.	6.8	10.5	n.a.	18.8	4.1	3.7
Asteraceae	Gynoxys psilophylla.	19.3	15.0	n.a.	n.a.	3.5	4.3
Asteraceae	Mikania buchtienii.	4.5	2.1	n.a.	n.a.	0.2	9.3
Asteraceae	Ophryosporus heptanthus.	26.6	15.4	n.a.	21.5	3.9	12.9
Asteraceae	Plazia daphnoides.	1.0	2.7	n.a.	2.1	5.9	13.4
Asteraceae	Senecio herzogui.	6.7	6.7	n.a.	49.4	1.4	6.9
Asteraceae	Tagetes multiflora.	18.5	6.5	n.a.	n.a.	1.9	7.3
Chloranthaceae	Hedyosmum angustifolium.	29.1	24.4	n.a.	n.a.	5.3	37.7
Euphorbiaceae	Croton andinus.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
Euphorbiaceae	Croton cf. emporiorum.	16.9	15.1	n.a.	30.7	10.5	19.3
Verbenaceae	Aloysia gratissima.	63.9	14.6	n.a.	n.a.	2.6	12.1
Verbenaceae	Lantana camara.	16.1	2.0	n.a.	n.a.	≫1	>1
Verbenaceae	Lantana magnibracteata.	n.a.	18.4	n.a.	27.8	3.6	17.3
Verbenaceae	Lantana trifolia.	23.6	3.2	n.a.	13.0	5.0	8.9
Verbenaceae	Lippia boliviana.	24.2	18.9	n.a.	45.0	10.3	29.2

Results are expressed as EC₅₀, the concentration that inhibits 50% of the DPPH oxidation promoting effect, in µg/ml. Extracts or fractions with EC₅₀ values smaller than, or close to, 15 mg/ml (shown in bold face) are considered highly promising. BHA has an EC₅₀ = 11.5 under the same conditions. All results are statistically significant at p ≤ 0.05. n.a. means no activity detected. All experiments were carried out in triplicate.

As we expected, because this technique measures the least specific property (as related to the rancidity problem) of capturing free radical generated by DPPH, we found a larger number of samples with acceptable EC₅₀ (46) than with the β-carotene technique (31). Also, as mentioned above, there is not an univocal correspondence between these two techniques, in part because of the different polarity of their reaction media and, mainly, because those assays work through different mechanisms. Ophryosporus heptanthus. (FC3), Hedyosmum angustifolium. (CE and FHX), Croton andinus. (FC3 and FOH), Croton cf. emporiorum. (FOH), Lantana camara. (CE), Lantana magnibracteata. (CE and FOH)., Lantana trifolia. (CE) and Lippia boliviana. (CE and FHX) are the cases where although high Φ were encountered, their corresponding EC ₅₀ values were below the accepted level. On the other hand, Baccharis grisebachii. (CE and CA), Baccharis salicifolia. (CE), Erechtites valerianaefolia. (CE, CA, FAC, and FOH), Gynoxys psilophylla. (CA, FAC, and FOH), Mikania buchtienii. (CA, FAC, andFOH), Ophryosporus heptanthus. (CA, FAC, and FOH), Plazia daphnoides. (CA and FOH), Tagetes multiflora. (CA, FAC, and FOH), Croton cf. emporiorum. (CA and FAC), Aloysia gratissima. (CA, FAC, and FOH), Lantana camara. (CA and FOH), Lantana trifolia. (FOH), and Lippia boliviana. (FAC) gave good results with DPPH but were below the accepted value in the β-carotene technique.

When using the DPPH technique, our results agree with those reported in Mensor et al. (2001) in that we did not find activity in any of the hexane fractions. The majority of our higher polarity fractions (FAC and FOH) are very active, better than BHA in most cases, probably because in those fractions tannins and the flavonoid family of compounds are concentrated, which confirms previous reports (Pietta, 2000).

So, when trapping free radicals is the desired property, six crude extracts (CE) gave promising results: Baccharis grisebachii., Baccharis salicifolia., Erechtites valerianaefolia., Mikania buchtienii., Plazia daphnoides., and Senecio herzogui..

In general, CA, the remaining fraction after the crude extract (CE) is partitioned with hexane, improved free radical trapping activity. In this case, most of the fractions were active: Baccharis grisebachii., Erechtites valerianaefolia, Gynoxys psilophylla, Mikania buchtienii, Ophryosporus heptanthus, Plazia daphnoides, Senecio herzogui, Tagetes multiflora, Croton cf. emporiorum, Aloysia gratissima., Lantana camara., and Lantana trifolia.. Although only six CE were active, 12 became active after the hexane soluble compounds were separated, and in most of the cases, the EC₅₀ of those CA is lower than that of their respective CE.

Only the chloroform fractions of two species showed good activity: Plazia daphnoides. and Lantana trifolia..

All species but Croton andinus. gave very good results as source of polar free radical trappers (FAC); in particular, Lantana camara. EC₅₀ is well below 1 µg/ml.

As for highly polar antioxidant, Lantana camara. is by far the best candidate, although good results were also found in the FOH fraction of Erechtites valerianaefolia., Gynoxys psilophylla., Mikania buchtienii., Ophryosporus heptanthus., Plazia daphnoides., Senecio herzogui., Tagetes multiflora., Aloysia gratissima., Lantana camara., and Lantana trifolia..

We believe it is worthwhile to do the screening work with these two techniques simultaneously, because a better picture is obtained of the potentiality of the extracts or fractions under study and on their eventual field of application. In this kind of screening work, comparison of the results obtained by different research teams is made difficult by the large variety of techniques used to evaluate AA (Emmons et al., 1999; Arnao et al., 2000; Niki & Noguchi, 2000; Antolovich et al., 2002; Koleva et al., 2002). We believe that the simultaneous use of the β-carotene bleaching technique and the DPPH assay will help to face this problem.

We have frequently noticed how the AA disappears as one progress through the separation routine to isolate the active compounds. This is explained because, in these cases, the AA is not due to one single compound but is a consequence of the combined effect of several compounds working through synergistic mechanisms. This points to the possibility of commercially using, without further purification, the extracts or fractions that have produced high activity, once it is ascertained that they comply with the regulations for their incorporation into industrial products.

Acknowledgments

This work was partially funded by the IberoAmerican body, CYTED (Programa de Ciencia y Tecnología para el Desarrollo), Project IV.11; by the Decanato de Investigación y Desarrollo of Universidad Simón Bolívar, Venezuela; and by Project QF-13 (CONICIT-BID, Venezuela).