Chemical characterization of Brickellia cavanillesii (Asteraceae) using gas chromatographic methods

A methanol extract of lyophilized Brickellia cavanillesii was quantitatively analyzed using gas chromatographic (GC) techniques. The chromatographic methods employed were (i) GC-flame ionization detector (GC-FID), (ii) GC-mass spectrometry (GC-MS), and (iii) purge and trap GC-MS (P&T GC-MS). Thirteen compounds were identified with a quality match of 90% and above using GC-MS. The compounds were (1) Cyclohexene, 6-ethenyl-6-methyl-1-(1-methylethyl)-3-(1-methylethylidene)-, (S)-; (2) Bicylo (2.2.1) heptan-2-one, 1, 7, 7-trimethyl-(1S, 4S)-; (3) Phenol, 2-methoxy-4-(1-propenyl)-; (4) Benzene, 1-(1, 5-dimethyl-4-hexenyl)-4-methyl-; (5) Naphthalene, 1, 2, 3, 5, 6, 8a-hexahydro4, 7-dimethyl-1-1-(1-methylethyl)-, (1S-cis)-; (6) Phenol, 2-methoxy-; (7) Benzaldehyde, 3-hydroxy-4-methoxy-; (8) 11, 13-Eicosadienoic acid, methyl ester; (9) 2-Furancarboxaldehyde, 5-methyl-; (10) Maltol; (11) Phenol; (12) Hydroquinone; (13) 1H-Indene, 1-ethylideneoctahydro-7a-methyl-, (1E, 3a.alpha, 7a.beta.). Other compounds (14) 3-methyl butanal; (15) (D)-Limonene; (16) 1-methyl-4-(1-methyl ethyl) benzene; (17) Butanoic acid methyl ester; (18) 2-methyl propanal; (19) 2-butanone; (20) 2-pentanone; and (21) 2-methyl butane were also identified when P&T GC-MS was performed. Of the 21 compounds identified, 12 were validated using chemical standards. The identified compounds were found to be terpenes, derivatives of terpenes, esters, ketones, aldehydes, and phenol-derived aromatic compounds; these are the primary constituents of the essential oils of many plants and flowers.


Introduction
The use of complementary and alternative medicine (CAM) in disease therapy has increased tremendously in recent years. Statistical data indicate that diabetes patients are 1.6 times more likely to use CAM treatment than people without diabetes and invariably are more susceptible to the attendant consequences of improper CAM usage (Egede et al. 2002;Shane-McWhorter 2007). The dearth of valid ethno botanical information on herbal plants used in diabetes therapy remains an impediment to the utilization of CAM by clinicians. Factors such as improper identification and standardization of these herbal plants constitute a potential reason for concern. One such herbal plant used in the treatment of diabetes is Brickellia. Brickellia is a member of the Asteraceae botanical family universally known as brickellbushes. It consists of about 380 genera (~3000 species) and is among the more basal lineages of the Eupatorieae. Brickellia species are native to Mexico and the southwestern portion of the United States; they are usually perennial shrubs that have creamy florets followed by one-seeded fruits in a prominent bristly sheath (Scott 1997;Schmidt and Schilling 2000). As an intervention, Brickellia is traditionally consumed as an herbal tea for its presumed antiulcer, antimigraine, cardiotonic, and antidiabetic properties among others. The extant literature reports that the chemical components of certain Brickellia species have hypoglycemic and antioxidative properties (Marles and Farnsworth 1995;Andrade-Cetto and Heinrich 2005;Rivero-Cruz et al. 2006). Studies conducted on the chemical characterization of Brickellia veronicaefolia (Rivero-Cruz et al. 2006) have shown that 86 percent of its essential oil consists of benzoates and sesquiterpenoids. Another study indicates that one flavone, 5, 7, 3′-trihydroxy-3, 6, 4′-trimethoxyflavone, has been isolated from the leaves of Brickellia veronicaefolia Gray (Perez et al. 2002). In a study performed on Brickellia cavanillesii, a natural product, 6-acetyl-5-hydroxy-2, 2-dimethyl-2H-chromene (C 13 H 14 O 3 ), was isolated (Rodriguez-Lopez et al. 2006).
It has been acknowledged that the enormous cost of modern treatment has led many people in underdeveloped and developing countries to use traditional or alternative medicine. Herbal plants are a primary source of alternative medicine. A scientific investigation of traditional herbal plants may provide valuable leads for the development of alternative drugs and therapeutic remedies. Treatment using herbal extracts constitutes a tremendous resource for alternative remedy especially at disease onset. Presently, considerable lack of knowledge continues to exist with regards to the therapeutic and pharmacodynamic potential of the individual chemicals of herbal plants. This has made it difficult to determine the efficacy of most herbal plants used in therapy. It is therefore essential that scientific evidence be established regarding the efficacy and pharmacodynamic potential of the individual bioactive chemicals of herbal supplements. Accurate identification and determination of the chemical components of herbal plants will aid in predicting their biological actions. This will lead to the development of reliable and reproducible pharmacological and clinical assays.
Our plant of interest is B. cavanillesii known in many Latin American countries as "prodigiosa" or "hamula." It is available commercially in herbal stores and is presently and commonly used by diabetics as a cheaper alternative to insulin. For this study we chemically characterized the individual chemical components of B. cavanillesii using gas chromatographic methods. Identified compounds were validated by comparing the retention time and mass spectra of samples with standards. Gas chromatography with flame ionization detection (GC-FID) was used for method development. Subsequently GC coupled with mass spectrometry (GC-MS) analysis was used for confirmation. Purge and trap GC MS (P&T GC-MS) analysis was used for the elucidation of the more volatile compounds present in Brickellia extracts. Customarily, B. cavanillesii is prepared for consumption by boiling the plant (aerial) in water (Poss et al. 2003). Samples prepared for analysis were prepared similarly and subsequently lyophilized. The lyophilization technique was adopted because it ensures standardized composition of the herbal extracts (fixed dose), and resistance to decomposition by heat and humidity. The aim of this study was determination of the chemical composition of the tea prepared from lyophilized B. cavanillesii.

General experimental procedure
Chromatographic measurements of tea sample: GC-FID measurements were conducted on a Hewlett-Packard (Palo Alto, CA) 6890 GC with a FID. The GC column (75 m 9 530 lm diameter) was a DB-624 capillary column (J&W Scientific, Folsom, CA). The temperature program was linear with a 5 min hold at 35°C, ramping to 260°C at a rate of 5°C/min and holding for 4 min, for a total run time of 54 min. The carrier gas used was He (47 mL/min). Injection volume was 3 lL using the split less mode. For the GC-MS analyses we employed identical chemistry as outlined for GC-FID. The GC-MS measurements were conducted on a Hewlett Packard 6890 series GC system with a HP 5973 Mass Selective Detector. The mass spectrometer was operated in the full scan mode with electron impact (EI+) ionization of 70 eV. The temperature of electron ionization source and MS quadrupole were kept at 230 and 150°C, respectively. For qualitative analysis, the MS detector was operated in full-scan mode from m/z 50 to 550 at 2.92 scans/sec. Control of the equipment, data acquisition, processing, and management of chromatographic information were performed by the Enhanced MSD D.02.00.275 software program. Constituents of B. cavanillesii were identified by matching their 70 eV mass spectra with those of the reference library. For P&T GC-MS analysis, an aliquot of 0.5 mL of tea extract in methanol was dissolved in 2.5 mL of Milli-Q (Billerica, MA) water. Measurements were conducted on an Agilent 6890 series GC system with an Agilent 5973 Mass Selective Detector (Santa Clara, CA), OI Analytical 4560 Sample Concentrator (College Station, TX) (purge pressure -25 psi [170 kpa]) and OI Analytical DPM-16 Discrete Purging Multisampler. The GC column was a capillary column with model number HP 1909IV-402 (HP-624) (25 m 9 200 lm diameter). The linear temperature range was from 35 to 240°C, at a rate of 12°C/min. The carrier gas used was He (17 mL/min).

Plant material
Leaves and twigs of a specimen of B. cavanillesii were purchased from an herbal store "Hieberia Claria" in Piedras Negras, Coahuila, Mexico.

Extraction and isolation
The leaves and twigs of B. cavanillesii were pulverized to a fine powder using the Black and Decker Coffee Bean Grinder (Towson, MA) "Smart Grind TM" C3G100W model. Five grams of B. cavanillesii was dissolved in 237 mL (+237 mg [0.1% ascorbic acid]) of hot (95°C) Milli-Q water (18.3 MΩ, pH 6.8). The ascorbic acid was added to prevent oxidation. The solution was thoroughly mixed and allowed to steep for 3 min. The filtrate was immediately frozen by immersion in liquid nitrogen and subsequently lyophilized using LABCONCO Freezone 1 lyophilizer (Kansas City, MO). Five grams of B. cavanillesii yielded~1 g of lyophilized tea sample on average. Extraction was done by introducing 2 g of the lyophilized B. cavanillesii into 100% high-performance liquid chromatography (HPLC) grade methanol using Dionex ASE 200 Accelerated Solvent Extractor (Salt Lake, UT). Each extraction cycle included a 5 min preheating (100°C) step, immediately followed by a 5 min static extraction process with 100% HPLC grade methanol at constant temperature (100°C) and pressure (1500 psi). The resulting extract (15-20 mL/sample) was transferred and concentrated to 1 mL using nitrogen evaporation. The samples were subsequently filtered (0.45 lm; Acrodisc, Port Washington, NY) into autosampler vials.

Preparation of chemicals used as standards for P&T GC-MS study
Eight chemicals were used as standards for P&T GC-MS study. Standards were purchased from VWR (Radnor, PA), TCI America (Portland, OR), and Alfa Aesar (Ward Hill, MA). The compounds were (14)

Results and Discussion
GC-FID was used to develop a method that ensured high separation efficiency. The chromatogram obtained using this method exhibited good separation as illustrated in Figure 1.
Consequently this method was used for GC-MS. A total of 109 peaks were obtained using GC-MS 70 eV mass spectra.
Thirteen of the compounds identified using GC-MS (see Fig. 2) had a quality match of 90% and above ( Table 1). The compounds identified were 1-13 (Fig. 3).
We validated 12 of the 21 identified compounds using reference standards purchased commercially from VWR, TCI America, and Alfa Aesar. Nine of the 21 compounds that were identified were not available for chemical confirmation. Validation of these compounds was performed by comparing the retention times of the isolated compounds against that of their reference standards. The retention times of the identified compounds were found to match that of the reference standards thus providing a reasonable confirmation of their identity. The validated compounds were 2, 6, 9, 12, 14-21. An example using one of the validated compounds, 15 (D-Limonene), is illustrated using Figures 6, 7A, and 7B.