Isolation, characterization and mode of antimicrobial action against Vibrio cholerae of methyl gallate isolated from Acacia farnesiana

Authors

  • E. Sánchez,

    1. Departamento de Microbiología e Inmunología, Facultad de Ciencias Biológicas, Universidad Autónoma de Nuevo León, San Nicolás de los Garza, NL, México
    Search for more papers by this author
  • N. Heredia,

    1. Departamento de Microbiología e Inmunología, Facultad de Ciencias Biológicas, Universidad Autónoma de Nuevo León, San Nicolás de los Garza, NL, México
    Search for more papers by this author
  • M. del R. Camacho-Corona,

    1. Facultad de Ciencias Químicas, Universidad Autónoma de Nuevo León, San Nicolás de los Garza, NL, México
    Search for more papers by this author
  • S. García

    Corresponding author
    1. Departamento de Microbiología e Inmunología, Facultad de Ciencias Biológicas, Universidad Autónoma de Nuevo León, San Nicolás de los Garza, NL, México
    • Correspondence

      Santos García, Universidad Autónoma de Nuevo León, Ciudad Universitaria, Apdo, Postal 124-F, San Nicolás de los Garza, NL 66451, México.E-mail: santos@microbiosymas.com

    Search for more papers by this author

Abstract

Aims

The antimicrobial activity of Acacia farnesiana against Vibrio cholerae has been demonstrated; however, no information regarding its active compound or its mechanism of action has been documented.

Methods and Results

The active compound was isolated from A. farnesiana by bioassay-guided fractionation and identified as methyl gallate by nuclear magnetic resonance (NMR) techniques (1H NMR and 13C NMR). The minimum bactericidal concentration (MBC) of methyl gallate and its effect on membrane integrity, cytoplasmic pH, membrane potential, ATP synthesis and gene expression of cholera toxin (ctx) from V. cholerae were determined. The MBC of methyl gallate ranged from 30 ± 1 to 50 ± 1 μg ml−1. Methyl gallate affected cell membrane integrity, causing a decrease in cytoplasmic pH (pHin, from 7·3 to <3·0), and membrane hyperpolarization, and ATP was no longer produced by the treated cells. However, methyl gallate did not affect ctx gene expression.

Conclusions

Methyl gallate is a major antimicrobial compound from A. farnesiana that disturbs the membrane activity of V. cholerae.

Significance and Impact of the Study

The effects of methyl gallate validate several traditional antimicrobial uses of A. farnesiana, and it is an attractive alternative to control V. cholerae.

Introduction

Vibrio cholerae, which causes cholera, is considered the most important pathogen of the last century and a threat to current global human health due to increases of up to 115% in reports of cases around the world (Kaper et al. 1995; CDC 2010; WHO 2010). There are an estimated 3–5 million cholera cases and 100 000–120 000 deaths due to cholera every year (WHO/WPRO 2012); furthermore, the numbers of cholera cases are known to be much greater than those reported (Ali et al. 2012).

Vibrio cholerae is a natural inhabitant of both freshwater and marine environments (Waturangi et al. 2011). Contaminated water with free-living V. cholerae cells is probably the main origin of epidemics (Kaysner and Hill 1994). Therefore, it is regarded primarily as a waterborne infection; although to a lesser extent, food that has been in contact with contaminated water often plays an important role in V. cholerae transmission (Sharma and Malik 2012). Antimicrobial agents such as tetracyclines and fluoroquinolones are the most common antibiotics used to treat cholera, shortening the durations of diarrhoea and V. cholerae excretion (Roychoudhury et al. 2008). However, mass administration of antibiotics is not recommended, as it has no effect on the spread of cholera and contributes to increasing antimicrobial resistance (WHO/WPRO 2012). In fact, there is a high incidence of V. cholerae strains that have been isolated from farm animals and public health sectors, which are resistant to commonly prescribed antibiotics (Mandal et al. 2012; Uddin et al. 2012).

The emergence of multidrug-resistant V. cholerae strains and negative perceptions of consumers towards synthetic antimicrobials have led to the development of new antimicrobial compounds (Galvao-Rodrigues et al. 2013). One area of research is natural antimicrobials from plants. Plants have long been utilized as the source of therapeutic agents worldwide and to treat many life-threatening diseases due to bacterial infections (Taran et al. 2010). The World Health Organization has estimated that 80% of the population in developing countries still relies on traditional medicines, mostly plant-based drugs, for their primary healthcare needs (Izzo 2004).

The antimicrobial effect of plant extracts against V. cholerae has been previously reported (García et al. 2006; Sánchez et al. 2010). Plants such as Costus spiralis (ethanolic extract, resuspended in water; Pérez et al. 2008), Morinda morindoïdes (ethanolic extract; Koffi et al. 2011), Terminalia arjuna (ethanolic extract; Fakruddin et al. 2011), Saraca indica, Datura stramonium (ethanolic extracts; Sharma and Patel 2009), Terminalia chebula (extracted in petroleum ether, CHCl3, dimethylformamide, EtOH and H2O) and Cassia auriculata (methanolic extract; Senthilkumar and Reetha 2011) have shown good vibriocidal activity. In addition, methanolic extracts of Ocimum basilicum, Opuntia ficus-indica, Artemisia ludoviciana and Acacia farnesiana are able to disrupt the cell membranes of V. cholerae cells, causing increased membrane permeability, a clear decrease in cytoplasmic pH, cell membrane hyperpolarization and cellular ATP (Sánchez et al. 2010).

Acacia farnesiana (L.) Willd. (sweet acacia), also known as Vachellia farnesiana (L.) Wight & Arn, is widely distributed in Mexico and Central America, although the species has a pantropical distribution incorporating Northern Australia and Southern Asia (Parrotta 2001; Sánchez et al. 2010). The plant has many uses in traditional medicine including to treat diarrhoea, dysentery, tuberculosis and as a treatment for indigestion (Parrotta 2001). In this work, A. farnesiana extract was further studied by bioassay-guided fractionation to isolate and characterize the major compound responsible for its vibriocidal activity and to determine its mode of action.

Materials and methods

Bacterial strains and culture conditions

Vibrio cholerae classical O1 strain 569-B (Inaba) and O139 strain AI-1837 (El Tor) were used in this study. Both strains were kindly provided by Dr. Elisa Elliot from the Food and Drug Administration, Washington, DC. Strains were maintained in Luria-Bertani (LB) agar (Difco) at room temperature, and periodic subculturing was performed every 3 months. Before assays, active cultures were prepared by inoculation of a loopful of each strain into 5 ml of sterile LB broth (Difco) and incubated for 18 h at 37°C.

Plant collection and extract preparations

Acacia farnesiana (sweet acacia) barks were collected from the Facultad de Ciencias Biológicas gardens of the Universidad Autónoma de Nuevo León, México. Voucher samples were authenticated by personnel of the Facultad de Ciencias Biológicas, Universidad Autónoma de Nuevo León. Barks were washed with tap water, dried at room temperature and cut into small pieces. Cut bark was dried in a forced-air oven at 50°C and ground in a grain mill. Crushed dry plant material (490 g) was placed in a Soxhlet system and heated to reflux for 12 h with 1 l of hexane, chloroform, methanol or water, in sequence. Each solvent was removed under reduced pressure using a rotary evaporator (Buchi R3000) at 45°C to dryness, while the water extract was concentrated at 55°C. Then, the dried extracts obtained were resuspended in 15 ml of the solvent used for the primary extraction, placed in amber vials and stored at 4°C until use.

Preliminary evaluation of antimicrobial activity

Preliminary analysis of the antimicrobial activity was performed using the agar-disc diffusion bioassay for the evaluation of nonpolar extracts and the agar-well diffusion bioassay for polar extracts. Briefly, LB agar plates were seeded separately with 100 μl of previously activated-V. cholerae strains (approximately 6 log CFU ml−1). For the disc diffusion bioassay, sterile discs (6 mm in diameter) of Whatman filter paper No.10 were impregnated with 50 μl of each nonpolar extract. The solvents were left to evaporate at room temperature (approx. 5 min), and the discs were placed on the surface of the plates previously seeded. Paper discs impregnated with hexane and chloroform were used as controls. For the well diffusion bioassay, wells were made in the agar using an inverted sterile Durham tube (6 mm in diameter), and 100 μl of methanolic and 100 μl of aqueous extracts were deposited in the wells, respectively. Methanol and water were used as controls. Plates were incubated at 37°C for 24 h. Antimicrobial activity was detected by the presence of a growth inhibition zone surrounding the disc or well. The diameter of this zone was measured and recorded.

Isolation and purification of antimicrobial natural compounds

Bioassay-guided fractionation

Column chromatography was used as the initial purification step. The column (length, 650 mm; bore, 45 mm) was packed using 60 Å silica gel, 230–400 mesh (Sigma), previously activated at 100°C for 1 h. Chloroform was added up to 3/4 of the column length, then silica was poured on top with constant tapping to avoid air bubbles, and finally simultaneous draining was carried out to aid proper packing. Plant extract (10 ml) mixed with 20 g of activated silica gel was loaded on top of the column. The column was run with 500 ml of 100% chloroform, followed by 500 ml of chloroform/methanol solvent mixtures of increasing polarity, and finishing with 500 ml of 100% methanol (Ode et al. 2011). The flow rate was maintained at approximately 1 ml min−1, and 50 fractions were collected, concentrated at room temperature and tested by thin-layer chromatography (TLC) using Silica Gel 60 F254 (Merck, Darmstadt, Germany) (Mehrotra et al. 2010). Fractions with identical spots and retention (Rf) values were pooled together for antimicrobial evaluation using the antimicrobial assays described above.

Preparative TLC-bioautography assay

Preparative TLC plates (5 cm × 20 cm and 20 cm × 20 cm) of approximately 1 mm in thickness were prepared using Silica Gel Type G; 5–15 μm (Sigma-Aldrich, St. Louis, MO, USA). After activation at 90°C for 1 h, 500 μl of sample was applied in a band at 1 cm from the lower edge. TLC plates were developed in a presaturated solvent chamber containing chloroform-methanol (90 : 10), until the solvent front reached 1 cm from the top of the plate. Developed TLC plates were removed from the chamber and air-dried for 1 h for complete solvent removal. Compounds were visualized under UV light (254 and 366 nm), and one plate (5 cm × 20 cm) was sprayed with the universal stain cobalt chloride (CoCl2) and heated on a hot plate at 95°C to visualize and compare separated spots. Rf values of this reference plate were recorded (Motlhanka et al. 2010).

Bioautography was performed according to the method of Sharma and Patel (2009), with slight modifications. The developed TLC plates (5 cm × 20 cm) were overlaid with LB agar seeded with an overnight culture of V. cholerae. The plates were incubated aseptically for 24 h at 37°C in a moist chamber. Absence of growth (clear zones) over the separated compounds on the chromatogram indicated antibacterial activity. The Rf values of the clear zones were compared with the Rf values of the related spots on the reference TLC plate. Each compound located by the bioautography method was scraped and removed from the preparative TLC plate (20 cm × 20 cm) and then dissolved in methanol. This procedure was repeated several times until the amount required for further assays was obtained.

Isolation and purification of the active compound

The active fraction isolated from the preparative TLC bioautographical assays was analysed by analytical high-performance liquid chromatography (HPLC) using a Varian Prepstar-218 system equipped with a ProStar 335 diode array detector, a ProStar 210 binary isocratic pump, and a Varian Microsorb C18 column (250 mm × 4·6 mm, i.d.; 5 μm). Methanol–water mixtures were used as the mobile phase at a flow rate of 1 ml min−1. After standardization of purification conditions, scale-up preparative HPLC was performed using the instrument described above with a PrepStar 218 binary isocratic pump, a Varian model 701 fraction collector, and a Varian Microsorb C18 column (250 mm × 21·4 mm, i.d.; 5 μm) at a flow rate of 21 ml min−1.

Characterization of isolated compound

The extracted and purified bioactive compound from A. farnesiana was characterized by one-dimensional nuclear magnetic resonance (NMR) techniques: 1H NMR (400 MHz, DMSO-d6) and 13C NMR (100 MHz, DMSO-d6). NMR spectra were recorded on a Bruker spectrophotometer using tetramethylsilane (TMS) as an internal standard (Gu et al. 2010). Commercial methyl gallate (Prod. # 274194; Aldrich, St. Louis, MO, USA) was used as a standard to confirm retention times of the isolated compound by HPLC.

Minimum bactericidal concentration

The minimum bactericidal concentration (MBC) was determined using the broth microdilution method. Sterile 96-well polystyrene U-shaped microtitre plates (Costar; Corning, Cambridge, MA, USA) were filled with 100 μl of 2× Mueller-Hinton Broth (Difco) plus 100 μl of varying concentrations of the isolated compound (Valtierra-Rodríguez et al. 2010). The plates were inoculated with 10 μl of a fresh culture of V. cholerae (final concentration of 6 log CFU ml−1) and incubated at 37°C for 24 h (Lambert et al. 2001). After incubation, 100 μl from each well was drop-plated on LB agar according to the method of Miles et al. (1938); then, the plates were incubated at 37°C for 24 h. Tetracycline (Sigma Aldrich, Mexico City, Mexico) was used as a control. The MBC was defined as the lowest concentration of the isolated compound at which no microbial growth was detected on the LB agar plate after the incubation period (Sánchez et al. 2010). Once the MBC was established, the effect of subinhibitory concentrations of compound (25, 50 or 75% of the MBC) on bacterial growth was also determined by colony enumeration in LB Agar.

Cell membrane integrity

The effect of the isolated compound on bacterial cell membrane integrity was determined using the LIVE/DEAD BacLight kit (Molecular Probes, Eugene, OR, USA) according to the manufacturer's instructions. Fluorogens (Syto9 and propidium iodide) were mixed in equal proportions and applied (3 μl) to 1 ml of V. cholerae cultures in LB broth, previously treated (15 min) with the antimicrobial compound and incubated in the dark for an additional 15 min. Cultures with methanol added instead of antimicrobials were used as controls. Cells were observed and counted (at least 200 from 10 fields) under an epifluorescence microscope (Axioskop 40; Carl Zeiss, Gotthingen, Germany) (Sánchez et al. 2010).

Cytoplasmic pH (pHin)

The effect of the isolated compound on pHin was determined according to the method reported by Breeuwer et al. (1996) and modified by Sánchez et al. (2010). Briefly, LB broth (5 ml) was inoculated (1%) with overnight-activated cultures of Vibrio strains and incubated at 37°C for 3 h. Cells were centrifuged (1500 g, 10 min) and washed twice with 50 mmol l−1 HEPES buffer containing 5 mmol l−1 EDTA, pH 8. The pellet was resuspended in 10 ml of HEPES buffer and loaded with the fluorescent probe carboxyfluorescein diacetate succinimidyl ester (cFDA-SE; Molecular Probes) at a concentration of 1·0 μmol l−1. Cells were incubated for 10 min at 37°C, washed once in 50 mmol l−1 potassium phosphate buffer containing 10 mmol l−1 MgCl2, pH 7·0, and resuspended in 10 ml of buffer. Finally, an additional incubation (30 min at 37°C) in the presence of glucose (10 mmol l−1, final concentration) was carried out. Cells were washed twice, resuspended in 50 mmol l−1 phosphate buffer (pH 7) and placed on ice.

Sterile 96-well polystyrene plates (Costar; Corning Incorporated, Corning, NY, USA) were filled with 100 μl of fluorescently labelled cells (1 × 107 CFU ml−1). Isolated compounds (10× the MBC) were added, and the plate was placed in the spectrofluorometer (Victor X2; Perkin-Elmer, Turku, Finland); fluorescence was measured every minute for 15 min using an excitation wavelength of 490 nm and an emission wavelength of 520 nm. The excitation and emission slit widths were 5 and 10 nm, respectively. The system was maintained at room temperature (25°C). Background fluorescence was determined on the cell-free supernatant and was deducted from the treated suspension values.

Calibration curves were determined using the following buffers: glycine (50 mmol l−1), citric acid (50 mmol l−1), Na2HPO4·2H2O (50 mmol l−1) and KCl (50 mmol l−1), adjusted to various pH values (4, 5, 6, 7, 8, 9 and 10). The fluorescence intensity was measured at 25°C after equilibrating the pHin and pHout by the addition of valinomycin (1 μmol l−1; Sigma-Aldrich, Mexico City, Mexico) and nigericin (1 μmol l−1; Sigma-Aldrich, Mexico City, Mexico). In this assay, a drop in relative fluorescence occurs when the pHin decreases (Breeuwer et al. 1996).

Changes in membrane potential

To measure changes in membrane potential caused by isolated compound on V. cholerae cells, the membrane potential-sensitive fluorescent probe bis-(1,3-dibutylbarbituric acid) trimethine oxonol (DiBAC4[3]; Molecular Probes) was added at a concentration of 1 μmol l−1 to 5 ml of V. cholerae cells adjusted to an optical density at 610 nm (OD610) of 0·5 (1 × 107 CFU ml−1), followed by the addition of the compound (10× the MBC). After 5 min, fluorescence was measured at excitation and emission wavelengths of 492 and 515 nm, respectively, using the spectrofluorometer described above. The results were corrected by subtracting the background fluorescence (Pag et al. 2004).

ATP determination

The effect of compound on the amount of ATP generated was determined by a bioluminescence assay according to Yuroff et al. (2003), with minor modifications. ATP generation was detected using the Enliten ATP detection kit (Promega, Madison, WI, USA). Activated cultures of Vibrio strains were dispensed into 5 ml of LB broth and incubated for 3·5 h to an OD610 of 0·5 (approximately 1 × 107 CFU ml−1). The cultures were treated with the isolated compound (10× the MBC) and incubated for 15 min. Cellular ATP was extracted with 250 μl of ice-cold 24% (v/v) perchloric acid. The mixture was centrifuged (10 000 g) for 5 min. To quantify the amount of ATP in the supernatants, 50 μl of supernatant was placed in white opaque 96-well microtitre plates (Nunc, Copenhagen, Denmark), and 100 μl of Enliten luciferase/luciferin reagent (Promega) plus 100 μl of 10 mmol l−1 Tris (pH 8·0) were added. After incubation for 10 min, light emission (bioluminescence) was determined using a multimode detector (Victor X2; Perkin-Elmer). ATP values were expressed as relative units, which were defined as the amount of light emitted per unit of cell density.

Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR)

qRT-PCR was used to determine the relative transcript amounts of cholera toxin gene (ctx) in V. cholerae with and without isolated compound (methyl gallate) treatment. The primers used for qRT-PCR were as follows: ctxA-F, 5′-ACA GAG TGA GTA CTT TGA CC-3′; ctxA-R, 5′-ATA CCA TCC ATA TAT TTG GGA G-3′; ctxAB-F, 5′-TGA AAT AAA GCA GTC AGG TG-3′; and ctxAB-R, 5′-CTG ATT TGT GTG CAG AAT ACC-3′. The housekeeping gene recA was used as an internal control to normalize the expression data of the genes of interest. RNA was extracted using the TriReagent method (Sigma-Aldrich Co., St Louis, MO, USA) according to the manufacturer's instructions. The one-step qRT-PCR was performed on an iQ5 Multicolor Real-Time PCR detection system (Bio-Rad, Hercules, CA, USA) under the following conditions: 60 min at 37°C for annealing and elongation followed by 5 min at 95°C to inactivate the reverse transcriptase enzyme, and cDNA amplification was made at 95°C for 30 s, 60°C for 1 min, and 72°C for 1 min to complete 30 PCR cycles using iScript™ Reverse Transcription Supermix for RT-qPCR (Bio-Rad). The relative change in transcripts due to the effect of methyl gallate compared with the control was calculated using the comparative CT method (Schmittgen and Livak 2008).

Statistical analyses

All experiments were performed in duplicate at least three times. The results were subjected to analysis of variance and mean comparison, when appropriate, using SPSS software (version 10.0; SPSS, Inc., Chicago, IL, USA). Differences were considered significant at P values of ≤0·05.

Results

All extracts using various solvents were evaluated for their antimicrobial properties. The results indicated that the methanolic extract showed the greatest antimicrobial activity, obtaining inhibition zones of 2·7 ± 0·1 cm and 2·4 ± 0·1 cm against V. cholerae strains 569-B and 1837, respectively. The extracts obtained with hexane, chloroform and water showed no antimicrobial activity against either of the strains tested.

Isolation and characterization of the active compound

After column chromatography and TLC visualization, 11 pooled fractions with similar band patterns were obtained; those fractions using CHCl3:MeOH (95 : 5 to 80 : 20) as the eluent exhibited antimicrobial activity. Fractions eluted with 90 : 10 CHCl3/MeOH were selected for purification of the active compound. Other fractions showed little or no antimicrobial activity.

According to the autobiography results of the active fractions, the band responsible for antimicrobial activity had an Rf value of 0·7. The sample in the band was recovered; and after HPLC purification, three different compounds were isolated. The compound with a retention time of 4·207 min was responsible for the antimicrobial activity. After recrystallization of the compound with methanol/ethyl ether, a degree of purity >99% was achieved. The compound obtained was a white to slightly beige crystalline powder that had a melting point of 198–200°C and was soluble in methanol.

The spectroscopic data of pure active compound were as follows: 1H NMR (400 MHz, DMSO-d6) δ: 3·739 (s, 3H, OCH3), 6·937 (s, 2H, H-2, H-6), 9·213 (brs, OH); 13C NMR (100 MHz, DMSO-d6) δ: 51 (OCH3), 108·53 (C-2, C-6), 119·32 (C-1), 138·46 (C-4), 145·62 (C-3, C-5) and 166·37 (C=O). Spectral analysis of the above data allowed us to identify the compound as 3,4,5-trihydroxybenzoic acid, methyl ester (IUPAC), commonly known as methyl gallate with the molecular formula C8H8O5 (Fig. 1).

Figure 1.

Chemical structure of 3,4,5-trihydroxybenzoic acid, methyl ester (methyl gallate).

Biological activity of isolated compound

Isolated compound (methyl gallate) had MBC values of 30 ± 1 and 50 ± 1 μg ml−1 against V. cholerae strains 569-B and 1837, respectively; and the MBC of tetracycline was 0·5 and 0·2 μg ml−1, respectively. The results of the effect of methyl gallate on membrane integrity of V. cholerae indicated that after 15 min of exposure to the isolated compound, at a concentration 1× the MBC, the percentage of cells with a compromised membrane exceeded 50%; at 5× the MBC, it damaged the membrane of approximately 90% of the cells; and at 10× and 20× the MBC, it affected >99·9% of the treated cells.

Exposure of the V. cholerae strains to methyl gallate caused a significant decrease in pHin ( 0·05), when compared with the control. The pHin decreased from approximately 7·3 to ≤4 for both strains (Fig. 2).

Figure 2.

Effect of methyl gallate (isolated compound, 10× minimum bactericidal concentration) on intracellular pH of Vibrio cholerae strains. (image_n/jam12328-gra-0001) V. cholerae 1837; (image) V. cholerae 569-B; (image) V. cholerae 1837 and (image_n/jam12328-gra-0004) V. cholerae 569-B.

The membrane potential of V. cholerae strains was significantly affected by methyl gallate ( 0·05) in comparison with controls (Fig. 3). This compound caused a marked membrane hyperpolarization.

Figure 3.

Effect of methyl gallate (isolated compound, 10× minimum bactericidal concentration) on the membrane potential of two Vibrio cholerae strains. (image_n/jam12328-gra-0005) V. cholerae 569-B and (image_n/jam12328-gra-0006) V. cholerae 1837.

Methyl gallate also significantly affected ATP levels of V. cholerae ( 0·05). The amount of ATP detected in the supernatants was minimal, indicating that ATP was no longer produced by the treated strains (Fig. 4).

Figure 4.

Effect of methyl gallate (isolated compound, 10× minimum bactericidal concentration) on ATP synthesis of Vibrio cholerae strains. (image_n/jam12328-gra-0005) V. cholerae 1837 cells; (image_n/jam12328-gra-0007) V. cholerae 1837 supernatant; (image_n/jam12328-gra-0006) V. cholerae 569-B cells and (image_n/jam12328-gra-0008) V. cholerae 569-B supernatant.

Using a qRT-PCR assay, the relative transcription level of the ctx gene was compared among V. cholerae strains treated with different concentrations of methyl gallate using the recA gene as an internal control. Different concentrations of isolated methyl gallate (75, 50 and 25% of the MBC) did not significantly affect (> 0·05) ctx gene expression levels (data not shown).

Discussion

Natural products such as plant extracts are a source of endless opportunities for the discovery of new drugs due to the wide chemical diversity present in plants (Cos et al. 2006). In this work, the phenolic compound methyl gallate was isolated and purified from A. farnesiana, and its antimicrobial activity against V. cholerae was demonstrated. This compound has been found in plants of the genus Acacia such as A. nilotica and A. leucophloea (Chaubal et al. 2005), as well as in Rhus javanica, Toona surei, Acer ginnala and Diospyrous kaki (Ahna et al. 2005; Abou-Zaid et al. 2009; Choi et al. 2009; Ekaprasada et al. 2009; Kaur et al. 2011). This compound has been also isolated from Entada abyssinica Stend ex A. Satabie (Teke et al. 2011) and Paeonia lactiflora Pallas (Ngan et al. 2012). The isolated compound exhibited antimicrobial activity with MBC ranging from 19 to 312 μg ml−1 against Gram-positive and Gram-negative bacteria (Teke et al. 2011). In addition, the antimicrobial properties of methyl gallate have been reported against micro-organisms such as Shigella dysenteriae, Streptococcus mutans, Candida albicans, Bacillus cereus, Staphylococcus epidermidis and Escherichia coli (Kim et al. 2005; Choi et al. 2009) as well as Salmonella and Enterobacter (Choi et al. 2008). However, no studies have been conducted on its efficacy against V. cholerae and its mechanism of action.

As expected, the MBC of the isolated methyl gallate was much less than the previously reported MBC of A. farnesiana whole extract (500 and 900 μg ml−1, Sánchez et al. (2010); against 30 and 50 μg ml−1 for strains 1837 and 569-B, respectively). The antimicrobial mode of action of methyl gallate was determined by the evaluation of several parameters in V. cholerae cells, such as cytoplasmic pH, membrane potential and proton motive force, which are indicators of cellular homeostasis; all these parameters help determine the functionality of the bacterial cell membrane (Lambert et al. 2001; Ultee et al. 2002).

pHin is an important parameter of bacterial cell physiology, which involves the control of ion transport systems and is closely related to the cellular membrane permeability (Booth 1985). Changes in pHin disrupt the transmembrane proton motive force, which is a very important process because it drives the bioenergetics network (Krulwich et al. 2007). Living cells are critically dependent upon pH homeostasis because most proteins have distinct pH ranges for function (Krulwich et al. 2011). In this regard, a reduction in the specific activity of the enzymes hexokinase and acetate kinase, as well as the overall glycolytic activity, occur at pH <5 (Warth 1991). Likewise, many other proteins, including ribosomal-subunit proteins and transcription factors, are repressed due to reduced intracellular pH (Lambert et al. 1997). Furthermore, magnesium efflux at pH <4 and malfunction of membrane structures and solute leakage occur at low pHin (Bender et al. 1986).

Vibrio cholerae is a neutralophilic micro-organism with a pHin of approximately 7·5. Methyl gallate produced reductions in pH values to ≤4, and these changes may be related to damage of the aforementioned membrane. Generally, bacteria can regulate pHin when the pH of the medium (pHex) changes (Hackstadt and Williams 1983). However, antimicrobial compounds cause a breakdown in the pH gradient leading to acidification of the cytoplasm, which in our experiments was detected by a decrease in fluorescence (Breeuwer et al. 1996). A reduction in the internal pH of micro-organisms such as Staphylococcus aureus has been previously reported as a result of treatment with oregano essential oil, thymol and carvacrol (Lambert et al. 2001).

The resting membrane potential is one of the most important parameters of living cells (Wright 2004). Methyl gallate caused membrane hyperpolarization, which has been recognized as an important type of membrane damage (Yuroff et al. 2003). The hyperpolarization could be explained as the result of an efflux of positively charged ions from the cytoplasm into the cell exterior, specifically K+, through a specific K+ channel in the plasma membrane affecting cell homeostasis (Bot and Prodan 2009) or by the movement of negative charge from the extracellular space into the cytoplasm (Gresík et al. 1991). This redistribution of K+ would result in an increase in membrane potential, that is, a positive outside (Gresík et al. 1991).

The determination of intra- and extracellular ATP in V. cholerae cultures treated with methyl gallate revealed that the levels of cellular ATP decreased, without detecting a proportionate increase in ATP outside the cell. These results suggest that the rate of ATP synthesis was diminished or that the hydrolysis rate of ATP was increased (Chandler and Segel 1978).

Several natural products such as essential oils produce changes in the bacterial membrane which collapse the integrity of this structure (Ultee et al. 1999). An increased excretion of intracellular metabolites to external medium is usually observed (Helander et al. 1998). This accelerated excretion could provoke a decrease in the intracellular pH (Ultee et al. 1999) and, as consequence, increased outward pumping of H+ ions is provoked, in a manner quantitatively consistent with their role as a substrate for the proton pump (Sanders et al. 1981). This increased pumping is often manifested as a transient hyperpolarization of cell membrane (Sanders et al. 1981). This situation could induce an accelerated activity of ATPase; however, ATP hydrolysis is also accelerate because a substantial portion of the ATP produced was used to counteract effects of acidification of the cytoplasm (O'Sullivan and Condon 1999).

The use of natural compounds as inhibitors of virulence factors is a novel approach to control infection (Chatterjee et al. 2010). The main virulence factors of V. cholerae are cholera toxin (CT) and toxin coregulated pili (TCP), which are encoded by the genes ctxAB and tcpA, respectively (Nair et al. 2006). In this work, no effect of methyl gallate on the expression of the enterotoxin gene (ctxAB) was detected, which undoubtedly indicates that the inhibitory activity reported is mainly due to its effect on the membrane of V. cholerae.

We concluded that methyl gallate exhibited antimicrobial activity by damaging the membrane integrity, causing physiological changes leading to a significant reduction in the pHin, affecting the membrane potential, and decreasing cellular ATP levels. Acacia farnesiana has been used as an antimicrobial for the treatment for tuberculosis, dysentery and other infections (Parrotta 2001). The antimicrobial activity of methyl gallate validates the antimicrobial activity of A. farnesiana extracts and partially supports its various traditional medicine uses. More studies are necessary to demonstrate the efficacy of these medical treatments and the safety of its use. Additional experiments are under way to provide information regarding other physiological sites of relevance that may be affected in this micro-organism, especially in regard to the search for inhibitors of different virulence factors other than the enterotoxin of V. cholerae.

Acknowledgements

This research was supported by the Consejo Nacional de Ciencia y Tecnología de México (CONACYT, Grant # 105389) and PAICYT-UANL. Eduardo Sánchez was supported by a scholarship from CONACYT.

Conflict of interest

No conflict of interest declared.

Ancillary