Antimycobacterial Activity of Constituents from Foeniculum vulgare var. Dulce Grown in Mexico

Bioassay guided fractionation of an antimycobacterial extract of Foeniculum vulgare var dulce (Apiaceae) led to the isolation and characterization of 5-hydroxyfurano-coumarin. The chemical structure of this compound was elucidated by 1H and 13C (1D and 2D) Nuclear Magnetic Resonance (NMR) spectroscopy. In addition, the active fractions were analyzed by GC-MS and seventy eight compounds were identified; the major compounds were 1,3-benzenediol, 1-methoxycyclohexene, o-cymene, sorbic acid, 2-hydroxy-3-methyl-2-cyclopenten-1-one, estragole, limonene-10-ol and 3-methyl-2-cyclopenten-1-one. Twenty compounds identified in the active fractions were tested against one sensitive and three MDR strains of Mycobacterium tuberculosis using the Alamar Blue microassay. Compounds that showed some degree of antimycobacterial activity against all strains tested were the following: linoleic acid (MIC 100 µg/mL), oleic acid (MIC 100 µg/mL), 1,3-benzenediol (MIC 100–200 µg/mL), undecanal (MIC 50–200 µg/mL), and 2,4-undecadienal (MIC 25–50 µg/mL), the last being the most active compound. To our knowledge, this is the first report of the presence of 5-hydroxy-furanocoumarin in F. vulgare.


Introduction
It is estimated that one third of the world's population is infected with Mycobacterium tuberculosis. In 2010, there were 8.8 million new cases of tuberculosis (TB), 1.1 million deaths from TB among HIV-negative people and an additional 0.35 million deaths from HIV-associated TB. M. tuberculosis has developed resistance to first and second line anti-tuberculosis drugs [1], therefore, there is a need to develop new drugs to combat multi-drug resistant (MDR) and extremely drug resistant (XDR) M. tuberculosis, and continuous efforts are under way in the search for novel bioactive compounds to develop new anti-tuberculosis drugs. To this end, bioactive compounds of natural origin, particularly from plants, are gaining significance.
Foeniculum vulgare Mill (Apiaceae), commonly known as fennel, is used as a spice and condiment in foods and cosmetics for its sweet flavor, aroma, and food preservation capability [2]. Fennel is native to Southern Europe and the Mediterranean region, but nowadays it is widely cultivated throughout the temperate and tropical regions of the World. Fennel was introduced in Mexico, where it is cultivated in home gardens and very often found free along roadsides and streams in the canyons of the Sierra Madre. It is used as a tea for stomach aches for its antispasmodic and carminative effects. In addition it promotes intestinal peristalsis and acts as an expectorant, and is associated with the diuretic action [3]. It is also recommended to cure various respiratory diseases [4].
Some chemical and biological studies have shown that fennel is an excellent source of antioxidants [5,6]. Fennel essential oil has a potent hepatoprotective action against CCl 4 -induced hepatic damage in rats, and D-limonene and β-myrcene were proposed to be the compounds potentially responsible for the activity [7]. Fennel oil was reported to exhibit estrogenic activity, promote menstruation, alleviate the symptoms of female climacterium, and increase libido [8]. Anethole, a component of fennel essential oil, inhibits TNF-induced cellular responses; TNF-mediated signaling is associated with both inflammation and carcinogenesis, thus fennel could be a chemopreventive agent [9]. Extracts and essential oils obtained from F. vulgare have shown antibacterial activity [10][11][12][13]. Dillapional and scopoletin were found to be the antimicrobial principles of fennel stems, which exhibited a potent antimicrobial activity against bacteria and fungi [14]. Our research group prepared a hexane extract from stems and leaves of F. vulgare var dulce grown in Mexico, this extract showed activity against one sensitive (MIC 200 µg/mL) and four monoresistant strains (MIC 100 µg/mL) of M. tuberculosis [15] using the Alamar Blue microassay. Thus, the aim of this study was to carry out the isolation and characterization of the antimycobacterial compounds from the hexane extract of F. vulgare var dulce grown in México.

Chemical Analysis
The hexane extract of F. vulgare which showed anti-tuberculosis activity [15] was submitted to fractionation by column chromatography giving 10 fractions; all fractions were further tested against M. tuberculosis H37 Rv. The only fractions that showed activity were F-7 (MIC 50 µg/mL), F-8 (MIC 12.5 µg/mL), and F-9 (MIC 50 µg/mL). The other fractions did not show activity at concentrations lower or equal to 200 µg/mL. F-8 was the most abundant fraction (4.63 g) and thus was treated with several conventional chromatographic methods, which allowed the isolation of the major compound 5-hydroxyfuranocoumarin, known as bergaptol (1, Figure 1). This furanocoumarin is reported for the first time in fennel. The structure of this compound was established unequivocally through analysis of its 1 H-and 13 C-NMR spectroscopic data and by comparison of an authentic sample. It is important to mention that furanocoumarins have been found in fennel [14,16,17], and bergaptol could have antimycobacterial activity since it has been reported that some compounds of this type had activity against M. tuberculosis [18]. In addition, GC-MS analysis was carried out on F-7, F-8 and F-9 in order to determine their volatile constituents. Retention indices for all the compounds were determined according to the Kovats retention index method using C 8 -C 23 n-alkanes as standards. The individual compounds were identified by their mass spectra fragmentation patterns using the NIST 1.7 library database, retention indices [19], and with authentic samples. From GC-MS analysis of F-7, 19 compounds were identified (Table 1), constituting 100% of the volatile constituents. These included two monoterpenes (1.53%), three sesquiterpenes (6.87%), four aromatics (33.80%), two aliphatic aldehydes (1.65%), two fatty acids (1.54%), and six aliphatic oxygenated compounds (54.61%). The major compounds identified in this fraction were 1,3-benzenediol (29.59%), 1-methoxycyclohexene (30.58%) and 2-hydroxy-3methyl-2-cyclopenten-1-one (11.81%).
From results we saw that the most active compound against the sensitive and MDR M. tuberculosis was 2,4-undecadienal, which was 12.5 times less active than the positive control ethambutol. It has been reported that aldehydes from Coriandrum sativum (Apiaceae) display antimicrobial activity, probably due their reactivity against nucleophilic groups (amino, hydroxy, sulfhydryl) in proteins associated with the cell membrane [21]. In another study, it was observed that the chain length of unsaturated aldehydes plays an important role in their antimicrobial activity, since they can behave as nonionic surfactants [22]. Thus, it is probable that the reactive ability of aldehydes groups plus the detergent-like properties make 2,4-undecadienal and undecanal active against M. tuberculosis. It is important to point out that our findings showed that the unsaturated aldehyde, undecanal showed weak antimycobacterial activity, while the α,β-unsaturated aldehyde 2,4-undecadienal was four times more active.
Apart from 2,4-undecadienal, the tested compounds that showed activity towards M. tuberculosis had MIC values higher than those found in the active fractions thus they may act in a synergistic way in the active fractions. On the other hand, it is important to mention that it is possible that one or more compounds identified in the active fractions that we did not have the chance to test could have antimycobacterial activity.
In the case of oleic acid and linoleic acid, it has been reported since 1956 that these acids affect the integrity of the cellular wall in the mycobacteria [23]. For the above effect to happen, the concentration of these fatty acids has to be considerably higher, compared to the concentration found in the OADC (0.6 µg/mL; PML Microbiological, Inc, Franklin Lakes, NJ, USA) used as enrichment medium to grow the mycobacteria. It is known that the majority of fatty acids are more effective against Gram positive than Gram negative bacteria. These facts suggest that the mechanism of their bactericidal action is due to a balance between the hydrophilic and hydrophobic portions of the molecule. The answer to why Gram positive bacteria are more sensitive than Gram negative bacteria to lipophilic agents it is due to the complexity of their cell walls [24]. In this sense it is important to mention that M. tuberculosis is considered to be a Gram positive bacterium with the additional presence of mycolic acids in its cell wall. Thus, the above could explain the activity of fatty acids on M. tuberculosis. Unsaturated long-chain fatty acids (oleic acid and linoleic acid) were found to be lethal for M. bovis and M. tuberculosis H37Ra. The inhibition of the function of enzymes associated with the cytoplasmic membrane suggested that killing was due to a detergent-like effect on this organelle [25].
On the other hand it was reported that antibacterial effect of long-chain unsaturated fatty acids such as linoleic acid, palmitoleic acid, oleic acid, and arachidonic acid was due to their inhibition of the bacterial enoyl-acyl carrier protein reductase (Fab I), an important enzyme involved in synthesis of fatty acids biosynthesis [26]. Thus, it is possible that the mechanism of inhibition could interfere in the synthesis of mycolic acids present in the cell wall of mycobacteria [27].
It is important to point out the presence of 1,3-benzenediol as one of the major compounds in the three active fractions. This compound had antimycobacterial activity against sensitive and MDR resistant strains tested; recently it has been reported that analogues of resorcinol showed activity against M. tuberculosis H37Rv with MIC values in the range of 2.5-60 µg/mL [28]. Although, the MIC values of resorcinol are higher we tested this compound against MDR clinical isolates strains obtaining positive outcomes. It is important to mention that other related phenolic compounds have demonstrated antimycobacterial activity in vitro and in vivo [29,30].

Plant Material
Stems and leaves of F. vulgare var. dulce were collected in Pesquería, Nuevo León (Mexico) in July 2007. Plant material was provided and identified by Biologist Mauricio Gonzalez Ferrara; a sample with a voucher number 024771 was deposited at the Herbarium of the Department of Botany, Universidad Autónoma of Nuevo León, Mexico.

Equipment Utilized
1-D and 2-D 1 H-and 13 C-Nuclear Magnetic Resonance (NMR) experiments were performed on a Bruker 400 MHz spectrometer. Chemical shifts (ppm) were determined with TMS as internal standard. CDCl 3 was used as a solvent. Gas Chromatography-Mass Spectra (GC-MS) analysis of active fractions was performed using a HP Agilent Technologies 6890 gas chromatograph equipped with a MSD 5973 quadrupolar mass detector in electron impact mode at 70 eV; Volatile compounds were separated on a BPX5 MS capillary column (25 m × 0.2 mm i.d., film thickness 0.3 µm). The oven temperature was set at 60 °C for 2 min, was then programmed at 4 °C /min to 200 °C and maintained for 10 min. Mass detector conditions were as follows: interface temperature, 200 °C; mass acquisition range 10-300. The carrier gas was helium at a flow rate of 1 mL/minute. Temperature of injector and the detector were set to 250 °C and 280 °C respectively. A standard solution of C 8 -C 23 -alkanes was used to obtain the index retention times. Identification of the volatile components was performed comparing their mass spectra with those of the NIST 1.7 library. Compound identification was verified based on relative index retention time [19] and mass spectra of authentic standards from Sigma-Aldrich. Semiquantitative data were calculated from the GC peak areas without using correction factors and were expressed as relative percentage (peak area %) of the total volatile constituent identified. (1) Powdered air-dried leaves and stems (944 g) were macerated with 18 L of hexane. The hexane was evaporated in vacuo to yield 7.72 g of crude extract; this extract was tested for antimycobacterial activity. The active extract was further fractionated by column chromatography on silica gel and eluted with a hexane/ethyl acetate gradient (

Microorganisms
M. tuberculosis H37Rv ATCC 27294 sensitive to isoniazid, rifampicin, ethambutol and pyrizanamide was purchased in American type culture. The three multidrug resistant (MDR) clinical isolates were collected from sputum specimens of tuberculosis patients; these were kindly provided by Virgilio Bocanegra-García.

Antimycobacterial Assay
Mycobacteria was cultured in Middlebrook 7H9 broth supplemented with 0.2% glycerol and 10% OADC at 37 °C for two weeks in order to reach logarithmic phase growth. The testing inoculum was prepared by diluting the bacterial suspension with culture medium to adjust turbidity to McFarland's nephelometer No. 1 standard and it was then further diluted to 1:20 with same culture medium. Antimycobacterial activity was determined as previously described in the literature [15]. Briefly, in a 96-well sterile microtiter plate serial two-fold dilutions of each sample were prepared to a final volume of 100 μL. Then 100 μL M. tuberculosis testing inoculum was added to testing and control wells. The microtiter plate was incubated at 37 °C in a sealed plastic bag during five days. After incubation time, all the wells received the mixture of 20 µL Alamar blue and 12 µL 10% Tween 80. The plates were re-incubated at 37 °C for 24 h. Wells with a well-defined pink color were scored as positive growth, those wells with blue color were scored as inhibition of growth. Minimum Inhibitory Concentration (MIC) value was expressed as the lowest concentration of compound that caused 100% inhibition of mycobacterium growth. All assays were run in triplicate and ethambutol was used as positive controls. The samples were tested in the concentration range of 200 to 6.25 µg/mL.

Conclusions
The present study supports the fact that hexane extract of F. vulgare var. dulce offers an excellent opportunity to find active molecules against multidrug resistant M. tuberculosis. 2,4-Undecadienal was the most active compound found in this study. The dietary intake of fennel may lower the risk of M. tuberculosis infection, however, it is necessary to do in vivo studies to prove the above. This is also the first report of the presence of 5-hydroxyfuranocoumarin (1) in F. vulgare.

Conflicts of Interest
The authors declare no conflicts of interest.