A technique for extraction and Thin Layer Chromatography visualization of fecal bile acids applied to neotropical felid scats

Fecal bile acid patterns have been used successfully to identify scats. Neotropical felid scats are capable of this biochemical identification because they present low concentrations of plant pigments that would interfere in fecal bile acids detection. However, neotropical felid scats have poor quantities of bile acids, so we developed in this work a proper technique for their extraction, visualization and determination. Twenty eighth feces of seven different felid species, collected from Zoological and Wildlife Parks, were dried and pulverized. The procedure for analyzing feces is: Take one g of pulverized feces and shake for 3 hr at room temperature in 20 ml benzene : methanol; filter and evaporate to 5 ml. Spot on TLC plate and develop in toluene :acetic acid:water. Dry and visualize with anisaldehyde. Field collected scats could be identified by the bile acids pattern revealed by this specific technique and ,then, used as a source of information for distribution, density and food habits studies.

pampas cats (Felis colocolo).All of them were fed basically with meat, so we did not expect to find plant pigments interfering with the bile acids.
Fresh feces were dried at room temperature, pulverized with a mortar and pestle, and stored in closed bottles.One g of pulverized sample was used to prepare bile acid solution.We tested several extracting procedures (Table 1).The sample bile acid fractions were stored in closed bottles, at room temperature and without light.When solvent evaporated, solutes were redissolved in 5 ml of 1:1 benzene:methanol (V/V).
Seven known bile acids: cholic, quenodeoxicholic, deoxicholic, dehydrocholic, taurocholic, taurodeoxicholic and glicocholic (prepared in 0.1 % solution of ethanol) and cholesterol (prepared in 0.1% solution of chloroform), were used as standards during the experiment, to recognize bands from the sample extracts.Sample extracts and standards were spotted 1 cm apart, on glass plates of 20x50 cm, coated with a 250 µm thick layer of silicagel H or G60 Merk (prepared in the laboratory, activated by drying them for one hour at 100 ºC oven).We applied several amounts (15-60 µl) of each sample from each species, and of each standard, to find an acceptable amount, and cheking that the solvent dried between each spot.Plates were run in a developing bath (until the solvent front traveled 14 cm from the spotting line), air dried and sprayed with visualizing agent.We tried different developing-visualizing combinations (Table 2).In every case we used a 5% (V/V) anisaldehyde alcoholic solution to prepare the visualizing system, and this was stored at 5-15 ºC to prevent oxidation, lasting invariable for about 15 days.
Plates were placed in a 120 ºC oven for about 20 min.
We ran plates with extracts of the same sample prepared with the different procedures described above to compare them.We also compared the different developing-visualizing system combinations on the same extracts and standards, in order to find which one best separated and revealed the bands by TLC.
The Rf value (the distance traveled by a band divided by the distance traveled by the solvent front, measured with a 0.05 mm precision caliper) was calculated for each observed band.Also, color, and intensity and size of each band (related to the steroid's concentration ) were recorded.In order to quantify variation in the TLC results, we run several plates of the same sample extracts and standards, and of different samples of the same species, using the selected technique.
Plates were analyzed immediately because oxidation changes color and some bands disappear with time.When it was possible, we photographed the plates.

RESULTS
Of the procedures for bile acid extraction that were tested, Nº 3 was the most satisfactory.Technique Nº 1 was rejected because, although it could separate the same products, it was more expensive and required more work than others.Extracts obtained with absolute ethanol (in cold or heat), and with technique Nº 9, reveled fewer spots on the TLC plates, using different developing-visualizing systems.
No differences were observed in the number of bands separated from the same sample when using activated charcoal in the extraction procedure, as specified by Johnson et al. 1981.We found that 40 µl was the appropriate amount of extract to be spotted in the plates.We observed that glicocholic and dehydrocholic acid's bands were not visualized if we spotted quantities of 10 µl.
The best results were obtained when using G60 Merk silicagel on the plates.
Among developing-visualizing systems, combination Nº 10 gave the best results regarding sharpness, coloring and number of bands developed.Other combinations revealed monochromatic bands or did not separate them all.Visualizing agents containing sulfuric acid developed very dark monochromatic bands.All bands were not present in every species profile.Intensity and size of the same band differed between some species.
Means and standard deviations of Rf values measured for each standard, as well as color, are shown in Table 3. DISCUSSION TLC of fecal bile acids is a biochemical technique that has been used to identify scats of different species.Innovated procedure for analyzing felid's feces is: 1) 1 g of dry pulverized feces, 2) shake for 3 hr at room temperature in 20 ml 1:1 benzene : methanol, 3) filter and evaporate to 5 ml, 4) spot 40 µl on silica Gel G60 TLC plate , 5) develop in (5:5:1.5)toluene :acetic acid:water, 6) air dry and visualize with (0.5:50:1) anisaldehyde: glacial acetic acid:concentrated sulfuric acid.
It is the best technique regarding extraction, separation and differentiation of steroids, even by color and intensity of their respective bands.With it we detected variation among TLC plates: 1) Variations in intensity and size of a band of the same extract reflect the evenness of the visualizing spray.A glass atomizer using pressed air should be use to propel the reagent.2) Variation in the Rf values for the same band on the same plate was probably caused by non uniformity in silicagel layer's thick.We suggest the use of glass or aluminum support plates.3) All bands were not present in every sample of the same species, probably because of slight differences in the diet complement.Even so, this variation had minor effects on the bile acid profiles.
Sample extracts should always be run with standards to guarantee correct identification of the bands.Bile acids visualization may vary due to changes in the TLC procedure, so it Without using activated charcoal, in order to see if it removed some of the bile besides the pigments 10 Cazón and Sühring (unpublished) One g of sample was macerated with chloroform during 14 hs, then heated at reflux for one hour .The supernatant was concentrated in vacuo and redissolved in 5 ml of chloroform must be performed conscientiously.Many scats from each animal, and many animals from each species, must be analyzed to quantificate individual and specific variation.Several trials must be made before defining the fecal bile acids pattern for each species.Although many similarities occurred between bile acids profiles of the seven species we studied, there are no two identical profiles.When possible, visual characteristics of feces should be used to corroborate identification.With reliable identification, scats are a source of information for distribution, density and food habits studies.

ACKNOWLEDGMENTS
This project was supported by the Consejo de Investigación de la Universidad Nacional de Salta.

TABLE 1
Extraction techniques to prepare the bile acid solutions Siegfried, C. M & W. H. Elliott.1968.Separation of bile acid of rat bile by thin-layer chromatography.J. Lipid Res.9: 294 -295.

TABLE 3
Mean and standard deviations (SD) of Rf values for each steroid used as standards in the TLC plates and color for each steroid