Rapid Screening of Complex Mixtures by Thin Layer Chromatography–Bioluminescence

The screening of complex mixtures such as foodstuffs, beverages, dietary supplements, and wastewater for toxins and adulterants or potential biological activity is often expensive and time consuming. To identify the active or toxic constituents in such mixtures requires the tedious isolation of single components followed by assays of their biological effects. Alternatively, standard rapid screening tests only establish the overall toxicity of a mixture disregarding the identification of the active culprit. Additionally, when these materials are tested in their mixture condition there is a high risk of artifacts or false data due to interferences and interactions with other components in the mixture. Optimally, a rapid screening assay that analyzes the activity of the individual components of a mixture in an economical and efficient fashion would be of great value. The Bioluminex™ assay (ChromaDex, Santa Ana, CA), which is based on direct bioautography, is one such technique.

The Bioluminex rapid screening assay employs direct bioautography detection by coupling the separation power of thin layer chromatography (TLC) with the biosensor properties of bioluminescent microorganisms.1–5 This technology provides a characteristic chemical and biological toxicity profile or “fingerprint” for each mixture analyzed and can be used to identify potentially bioactive compounds or adulterants. In addition, the profile can be used to help support material identity. The TLC–bioluminescence assay offers several advantages over standard toxicity screening assays. First, activity is assigned to single components of a mixture, thereby identifying the active constituent and eliminating interferences from other analytes. Second, the sample medium is evaporated from the TLC plate prior to introducing the biosensor organism, allowing for the analysis of nonaqueous samples. Third, the need for a secondary visualization agent is eliminated by employing a variety of bioluminescent organisms as the primary detection reagent. Fourth, the assay can simultaneously analyze up to 20 samples, with the organism toxicity results produced within 2 min. Last, the TLC–bioluminescence assay has been designed to be kit compatible, providing a rapid and inexpensive analysis for many complex samples.

Bioluminescent organisms of interest range from wild-type naturally bioluminescent bacteria to genetically modified (GM) bacteria or yeast, including dark variants that express the lux or GFP (green fluorescent protein) reporters under the control of a variety of promoters. Perhaps the most applicable organism for general, broad-range toxicity testing and which is used in the Bioluminex assay is the bacterium Vibrio fischeri. The bioluminescent marine bacterium V. fischeri (Beijerinck, 1889; Lehmann and Neumann, 1896) is a nonpathogenic, Gram-negative species of bacterium that thrives in the marine environment and is an ideal biosensor organism for TLC–bioluminescence applications. This bacterium has been well characterized and used in standardized and validated ecotoxicity assays for over 25 years.6 It is robust and easily cultured, and provides a consistent and reproducible response with an apparent and quantifiable endpoint. The V. fischeri bacteria are also very suitable for use in a rapid screening kit assay. The bacteria can be freeze-dried and subsequently stored at 4 °C for more than 1.5 years. Additionally, the freeze-dried bacteria can be directly inoculated and liquid cultured without an intermediate culture plate. They can be liquid cultured at room temperature using a flask, stir bar, and stir plate, eliminating the need for expensive laboratory equipment.

As V. fischeri cells reach a crucial cellular density, their lux operon expresses the reaction catalyst luciferase. In the presence of O2 and luciferase, a reduced nicotinamide adenine dinucleotide (NADH) riboflavin phosphate (FMNH2) and a long-chain fatty aldehyde are oxidized. The resulting interaction forms an excited yet highly stable intermediate, which decays slowly, resulting in the release of excess free energy in the form of a blue-green light (490 nm).7–10 The observed bioluminescence reflects the metabolic status of the cell and will decrease for cells exposed to toxic substances. Thus, a reduction in light emission is a measure of toxicity toward V. fischeri and can be selectively viewed and quantitated directly on the TLC or high-performance thin layer chromatography (HPTLC) plate. Typical limits of detection for toxic substances are in the picomole range.

In the TLC–bioluminescence assay, complex mixtures are first separated by TLC or HPTLC. These effective, inexpensive, and rapid techniques use adsorption and capillary action to separate up to 20 (20 × 10 cm plate) complex mixtures simultaneously. The matrices provide a characteristic distribution pattern of compounds that is dependent on sample composition and are routinely used to support the identity of a compound in a mixture when the retention factor (Rf) or migration distance of a compound is compared with the Rf of a known compound.

After compound separation, the mobile phase is evaporated and the matrix plate is coated with a broth of bioluminescent bacteria employing a simple dipping procedure. The broth can be used to dip multiple plates and can be buffered to use with mobile phases containing acids or bases that do not fully evaporate during the drying phase. Results occur within seconds and last until the plate dries, approximately 30 min or more with the use of the BioLuminizer (CAMAG, Muttenz, Switzerland), which optimizes the plate compartment for prolonged bacterial activity. Results identify single compounds, which inhibit luminescence, resulting in dark zones (quenched bioluminescence) on a luminescent background where the bacteria remain viable. Data can be documented by direct contact of photographic film such as X-ray and Polaroid film (Waltham, MA) or indirectly, such as with a cooled  charge-coupled device (CCD) camera, video imaging, Polaroid documentation system, or 35-mm camera.

Experimental

Methanol (HPLC grade), toluene (OmniSolv), and isopropyl ether (HPLC grade) were purchased from VWR (West Chester, PA). Ethyl acetate (biotech grade) and formic acid (purum 98%) were purchased from Sigma Aldrich (St. Louis, MO). The standards ochratoxin A, Capsicum annuum, capsaicin, and Bioluminex positive and negative controls were obtained from ChromaDex. 4-Androstene-3,17-dione (≥98% purity) (Sigma Aldrich), 2-chloro-N-(2-ethyl-6-methylphenyl)-N-(-2-methoxy-1-methylethyl)acetamide (Metolachlor–Pestanal®, Riedel-de-Haën, Seelze, Germany), and arsenic(III) oxide (99.995%) (Aldrich Chemical Co., Milwaukee, WI) were obtained as indicated.

Lyophilized V. fischeri cultures (ChromaDex Analytics, Inc., Boulder, CO) stored at 4 °C were inoculated and grown overnight (30 hr) in a 200-mL batch culture of complex medium (Bioluminex medium), and 6 mL/L 50% aqueous solution of glycerol (Fisher Scientific, Pittsburgh, PA), and H2O (Millipore, Billerica, MA) adjusted to a pH of 7.2 ± 0.2 at 120 rpm and 28 °C under atmospheric conditions. Directly before assay, TLC–bioluminescence buffer (Bioluminex buffer) was added to fully luminescent bacteria and dissolved at 120 rpm and 28 °C under atmospheric conditions.

Samples and standards were applied in water or methanol to prewashed (methanol elution) 10 × 10 cm HPTLC silica gel 60 F254 plates (ChromaDex) via an ATS-4 automatic TLC sampler 4 (CAMAG) at y = 8 mm using band spray application set to a methanol or water application mode as appropriate. After sample application, the plates were air dried for 15 min. Plates were then developed to 70 mm using appropriate mobile phases in a preequilibrated (30 min) 10 × 10 cm ridged-bottom TLC chamber. Postdevelopment mobile phases were evaporated from the plates in a mechanical oven at 40 °C for 2 hr.

Dried plates were coated with buffered luminescent V. fischeri (200 mL) using an automatic immersion device (CAMAG). Excess bacteria were removed from the plate using a squeegee device, and images were immediately recorded over a 10-min period using an exposure time of 120 sec with a cooled (–30 °C absolute) CCD camera and dark box (Fluorchem® 8900, Alpha Innotech, San Leandro, CA).

Discussion

Figure 1 - TLC–bioluminescence image of increasing concentrations (0.6, 1.1, and 1.6 μg of each analyte per lane) of 4-androstene-3,17-dione, arsenic oxide, and Metolachlor spiked tap water in tracks 1, 2, and 3, respectively. Corresponding application volume (8, 14, and 20 μL) of unadulterated water in tracks 4, 5, and 6, and 2 μg of each analyte, 4-androstene-3,17-dione, arsenic oxide, and Metolachlor in tracks 8, 9, and 10, respectively. Chromatogram was developed with toluene:ethyl acetate:formic acid:water (4:8:1.1:0.2, v/v/v/v) and analyzed using the Bioluminex assay.

Figure 2 - TLC–bioluminescence image of increasing concentrations (0.9, 1.8, 2.6, 3.5, 8.8, and 17.5 μg) of C. annuum (Cayenne pepper) extract (10 mL biomass, 10 mL CHCl3) in tracks 2–7, respectively, and 2 μg of capsaicin in track 1 (track 8 is a blank). Chromatogram was developed with isopropyl ether and analyzed using the Bioluminex assay.

Figure 3 - TLC–bioluminescence image of increasing concentrations (1 μL = 0.5 μg analyte, 3 μL = 1.5 μg analyte, and 3 μL = 3.0 μg analyte) of ochratoxin A spiked corn extract (1 g freeze-dried corn, 10 mL CH3OH) in tracks 1–3, respectively; equivalent amounts of unspiked corn extract in tracks 4–6; 2 μg of ochratoxin A standard in track 8; 4 μg and 8 μg of Bioluminex negative and positive controls in tracks 9 and 10, respectively (track 7 is a blank). Chromatogram was developed with ethyl acetate:methanol:formic acid:water (50:2:5:3, v/v/v) and analyzed using the Bioluminex assay.

The TLC–bioluminescence assay has been shown to be effective in the analysis of pesticides (fungicides, insecticides, herbicides), heavy metals, organic pollutants, pharmaceuticals, and mycotoxins in a variety of complex matrices such as dietary supplements, natural products, foodstuffs, beverages, and wastewater. For example, in Figure 1, tap water was spiked with the anabolic steroid 4-androstene-3,17-dione, the heavy metal arsenic(III) oxide, and the chloroacetanilide herbicide Metolachlor. Various concentrations of these samples in conjunction with unspiked tap water and the individual analytes were examined via TLC–bioluminescence. Each separated analyte showed a concentration-dependent decrease in bioluminescence (tracks 1–3) and can be identified by comparing the Rf (compound distance relative to baseline/solvent front distance relative to baseline) of the standard analytes in tracks 8, 9, and 10 to the Rf of the compound in the mixtures in lanes 1, 2, and 3 (i.e., 4-androstene-3,17-dione Rf ≅ 0.71, arsenic oxide Rf ≅ 0 , and Metolachlor Rf ≅ 0.80).

The assay is also effective for obtaining a fingerprint profile that can be applied to the support of material identity. For example, in Figure 2 increasing concentrations of C. annuum (Cayenne pepper) extract and 2 μg of capsaicin were analyzed using TLC–bioluminescence. At lower extract concentrations, compounds in the C. annuum extract that readily inhibit V. fischeri bioluminescence can be easily identified, such as the unidentified compound detected at Rf ≅ 0.87 in track 2. At higher extract concentrations, a fingerprint identity emerges. This fingerprint is a characteristic profile that can be used to simultaneously support material identity and detect sample adulteration.

Detecting mycotoxins in complex mixtures such as agricultural commodities can also be accomplished with the method presented. The third application example shown in Figure 3 displays the TLC–bioluminescence image of increasing concentrations of corn extracts (tracks 4–6), corn extracts spiked with the mycotoxin ochratoxin A (tracks 1–3), and 2 μg of the ochratoxin A standard (track 8). Most markedly, in lane 6, a fingerprint of the analyzed corn extract is detected. In lane 3, the characteristic corn extract fingerprint is also observed, but with an extra band at Rf ≅ 0.76. This band correlates well to the Rf of the ochratoxin A standard in track 8.

Conclusion

The TLC–bioluminescence assay is an effective, rapid, and versatile tool that has the potential to become an industry standard. It can be used to support material identity, detect toxins and chemical adulterants, and control manufacturing procedures. It may also prove to be a potent research tool used to identify biologically active constituents of complex mixtures such as natural products. Furthermore, the rugged technology is kit compatible to provide a quick and inexpensive analysis of many complex samples.

References

  1. Verbitski, S.M.; McChesney, J.D.; Gourdin, G.T.; Ikenouye, L.M. The Development of Methodology Involving the Coupling of Bioluminescent Bacteria to Thin-Layer Chromatography for Detecting Active Components in a Variety of Complex Mixtures. U.S. patent pending; Jan 2006.
  2. Weisemann, C.; Kreiss, W.; Rast, H.-G.; Eberz, G. Analytical Method for Investigating Mixtures of Toxic Compounds. Eur. Patent EP 0 588 139 B1; 1998.
  3. Eberz, G.; Rast, H.G.; Burger, K.; Kreiss, W.; Weisemann, C. Bioactivity screening by chromatography bioluminescence coupling. Chromatographia  1996, 43, 5–9.
  4. Becvar, J.E.; Becvar, L.E. Luminous Bacteria and Methods for the Isolation, Identification, and Quantitiation of Toxicants. U.S. Patent 6,017,722; 2000.
  5. Becvar, J.E.; Becvar, L.E. Kit for the Isolation, Identification, and Quantitation of Toxicants. U.S. Patent 6,340,572; 2002.
  6. Use of Luminescent Bacteria for Determining Toxicity in Aquatic Environments. Aquatic Toxicity, ASTM STP 667; Markings, L.L.; Kimerle, R.A., Eds.; American Society for Testing and Materials: Philadelphia, PA, 1979; pp 98–106.
  7. Hastings, J.W.; Nealson, K.H. Bacterial bioluminescence. Ann. Rev. Microbiol. 1977, 31, 549–95.
  8. Engebrecht, J.; Nealson, K.; Silverman, M. Bacterial bioluminescence: isolation and genetic analysis of functions from Vibrio fischeri. Cell 1983, 32(3), 773–81.
  9. The Mechanism of Bacterial Bioluminescence; Muller, F., Ed.; CRC Press: Boca Raton, FL, 1990; vol II, pp 109–51.
  10. Meighen, E.A. Bacterial bioluminescence: organization, regulation, and application of the lux genes. FASEB1993, 7, 1016–22.

Dr. Verbitski is Bioluminex Manager, Mr. Gourdin is Bioluminex Project Investigator, Ms. Ikenouye is Bioluminex Research Associate, and Dr. McChesney is Consulting Chief Scientific Officer, ChromaDex Analytics, Inc., 2830 Wilderness Pl., Boulder, CO 80301, U.S.A.; tel.: 303-442-4281, ext. 226; fax: 303-442-4237; e-mail: [email protected].

Comments