All chemicals were reagent grade. DMSO and the standard drugs were obtained from Sigma Chemical Co. (St. Louis, MO, USA). The other chemicals used in this study were purchased from J. T. Baker (Xalostoc, Edo. de Mexico, Mexico). Sterile bovine serum and PEHPS medium were prepared in our laboratory as described elsewhere [22] . Metronidazole (Sigma Chemical Co.), a nitroimidazole antibiotic and antiprotozoal agent was dissolved in DMSO to get 500 μg/mL concentration and further diluted adding 500 μL in 10 mL of dimethyl sulfoxide (DMSO), finally this solution was added to media 1:10 v:v to give 12.5 μg/mL final concentration. Ursolic acid and betulinic acid solutions (Sigma Chemical Co.), antiprotozoal pentacylclic triterpenoids, were prepared adding 2 mg of the pure compound (Sigma Chemical Co.) in 2 mL of DMSO and further diluted to 25 μg/mL. Stock solutions were stored at −20˚C until use.

Thymus vulgaris was obtained from local markets and identified at the herbarium of the Faculty of Forestry of the Autonomous University of Nuevo Leon, Mexico (Boucher number 21,344). The aerial parts of the plant were dried in electric oven at 40˚C for 72 h (J.M. Ortiz. Aparatos Eléctricos S.A. de C.V., Monterrey, N.L. México) and ground with the use of an electric mill (Molino del Rey, S.A. de C.V., San Nicolas de los Garza, NL, Mexico).

G. lamblia 0989-IMSS strain (5 × 10³ trophozoites) was cultivated in 5.5 mL TYI-S33 medium in 13 × 100 culture tubes (Pyrex, Corning life Sciences, Corning, New York, NY, USA) at 36.5˚C. The duplication time for G. lamblia trophozoites was 13.1 h (Figure 1).

E. histolytica HM1-IMSS strain (5 × 10³ trophozoites) and T. vaginalis GT-15

strain (5 × 10³ trophozoites) were cultivated in 5.5 mL PEHPS medium [22] at 36.5˚C. The duplication time for E. histolytica was 21.3 h and for T. vaginalis 11.7 h (Figure 1).

The crude extract was obtained by maceration and mixing with methanol. Briefly, 150 g of the dried, ground, aerial Thymus vulgaris parts were stirred with 500 mL methanol for 72 h on a rotary shaker (PC 622 Corning, New York, NY, USA). The supernatant of crude methanol extract was filtered using Whatman # 1 papers (Whatman International, LTD, Maidstone, England) and air-dried. This extraction was repeated two more times. The pooled filtrates were stored in the dark at 4˚C in the dark until use. Thirty mg of the extract was dissolved in 250 μL dimethylsulfoxide (DMSO), which was further diluted in distilled water to create a stock of 6 mg/mL.

Once the anti-parasitic activity of the methanol extract had been verified (Figure 2), the extract was fractionated by polarity-based partitioning: after a single

hexane:methanol (1:1, v/v) fractionation, and triple ethyl acetate:water:methanol (4:3:1, v/v) and n-butanol:water:methanol (4:3:1, v/v) fractionations, the soluble methanol fraction was recovered. All fractions were allowed to dry completely before being tested for antiprotozoal activity (see below). The best-performing fraction was extracted with chloroform, filtered, and air-dried completely. Both the chloroform-soluble and insoluble fractions were tested for anti-parasitic activity. As only the chloroform-insoluble fraction had significant anti-parasitic activity, it was re-suspended in methanol and filtered. Next, the methanol-soluble portion was three times extracted with n-hexane:methanol (3:1, v/v), whereas the methanol-insoluble part was extracted with acetone, so that four fractions were obtained: 1) a hexane-soluble fraction, 2) a methanol-soluble fraction, 3) an acetone-soluble fraction, and 4) a fraction insoluble in any of the above solvents. The purity of each fraction was determined by thin layer chromatography (TLC), (Figure 3). As the hexane and acetone-soluble fractions produced the same TLC pattern, they were combined before the various fractions were tested again for antiprotozoal activity. Only the mixed hexane/acetone-soluble fraction

had antiprotozoal activity. This fraction was characterized by high-performance liquid chromatography (HPLC) coupled to orthogonal acceleration time-of- flight mass spectrometry, mass Spectrometer and the results of data obtained from the database were compared Mascot-Blast Spectrophotometric [23] .

The antiprotozoal activity of the methanol extract and its fractions were evaluated as previously described [24] [25] [26] . Trophozoite suspensions (2 × 10⁴/mL for E. histolytica, 1 × 10⁵/mL for T. vaginalis, and 2 × 10⁵/mL for G. lamblia) in their respective media supplemented with 10% bovine serum were incubated with the various extracts and control solutions at 36.5˚C for 72 h for E. histolytica and 24 h G. lamblia or T. vaginalis. These incubation periods were chosen because of their differential growth kinetic characteristics under standard conditions (Figure 1). Test samples included the crude methanol extract and the pooled hexane/acetone-soluble fraction at 0 - 300 μg/mL. Positive antiprotozoal solutions included two-fold serial dilution series (0 - 25 μg/mL) in appropriate media of ursolic acid and betulinic acid, and metronidazole at 1.25 μg/mL, whereas 5% DMSO served as a negative control. Incubations were stopped by placing the suspensions in ice-cold water for no less than 20 min. The number of trophozoites/mL was determined using a hemocytometer.

The percentage of growth inhibition was calculated with respect to untreated controls. The 50% inhibitory concentration (IC₅₀) of each drug was calculated by PROBIT analysis [27] . Each drug was assayed in triplicate in three independent experiments for each protozoan species, and the mean and 95% confidence limits were calculated.

Vero cells (African green monkey epithelial kidney cells, ATCC CCL-81) were maintained in complete RPMI media/1 mM pyruvate/10% fetal bovine serum at 37˚C and 5% CO₂. To evaluate cytotoxicity, 1 × 10⁵ cells/well were seeded in 96-well plates and incubated for 24 h. Next, the supernatant was replaced with 100 μL of test or control solutions and incubated for another 24 h. Test samples included crude methanol extracts and the hexane/acetone-soluble fractions (9.37 - 300 μg/mL); the positive control was 10% DMSO and the negative control was fresh medium. After the incubation time, the viable cells/well were counted with the use of the trypan blue exclusion method and a hemocytometer. The IC₅₀ was determined by PROBIT [27] [28] . All assays were performed in triplicate [29] .

The plant extracts with the most promising antiprotozoal activity were analyzed by TLC on pre-coated silica gel plates with benzene:acetone 9:1 v/v as the mobile phase. The air-dried TLC plates were observed at both 254 nm and 366 nm, and developed by iodine vapors and also exposed to different spraying reagents, in order to determine the Rf values of the observed spots.

A drop of the extract was mixed with 1 drop of chloroform and three drops Liebermann-Burchard reagent (sulfuric acid/anhydrous acetic acid/chloroform, 20:1:1 v/v) on a porcelain plate. A change of color to orange, red, blue or green color is indicative of the presence of cholesterol-like compounds.

For each experiment the mean and standard deviation of three independent experiments performed in triplicate were determined. The IC₅₀ for each protozoa- antiprotozoal extract combination was calculated with the PROBIT trial [27] . The t student test was performed to compare the means of IC₅₀ of each antiprotozoal extract combination using the statistical software SPSS version 15.0. A p-value < 0.05 was considered as statistically significant.

The 300 µg/mL crude methanol extract of Thymus vulgaris had antiprotozoal activity against G. lamblia (IC₅₀ 86.41 ± 2.9 µg/mL) and T. vaginalis (IC₅₀ 115.41 ± 2.29 µg/mL), but not against E. histolytica trophozoites (Table 1 and Table 2).

The subsequent fractionation steps were bio-guided by the antiprotozoal assay. Only fractions with antiprotozoal activity underwent further work-up as in-

dicated in Figure 2. After the first step, the fraction with highest antiprotozoal activity was methanol extract from the ethyl acetate/water/methanol fractionation, causing a ≥ 97% growth inhibition of T. vaginalis and G. lamblia cultures. Next, it was the chloroform-insoluble fraction that caused growth inhibition of about 97% in aforementioned cultures. After the third step, both the hexane and acetone fractions from the chloroform-insoluble, methanol-recovered extract yielded antiprotozoal activity against both species. The mixed hexane-acetone fraction induced had excellent antiprotozoal activity, inducing 96% - 99% growth inhibition (Table 1).

The color change to red in the Liebermann-Burchard test indicated the presence of triterpenes in the mixed hexane-acetone fraction. Further characterization by mass spectrometry chromatography allowed the identification of oleanolic acid, betulinic acid and ursolic acid. These compounds were commercially obtained and their antiprotozoal activity was compared with the mixed hexane-acetone fraction. Only ursolic acid had antiprotozoal activity against both T. vaginalis and G. lamblia. The antiprotozoal activity of ursolic acid was 5 - 14 times stronger than that of the mixed hexane-acetone fraction, which was 1.5 - 2 times stronger than that of the crude methanol extract (Table 2).

Although the crude methanol extract, the hexane-acetone fraction, and commercial ursolic acid were all toxic for Vero cells, the concentration required to achieve an IC₅₀ on Vero cells was 2.2 - 57.8 times higher than the IC₅₀ for the protozoa (Table 2).