Identification of the Antimicrobial Substances Produced by Solanum Palinacanthum (solanaceae)

To find out natural antimicrobial agents as alternative in therapeutics and to preserve food, the methanol extract of Solanum palinacanthum aerial parts was submitted to purification steps guided by antibacterial and antifungal assays. As a consequence, the flavonoid rutin and 3,5-dicaffeoylquinic acid were isolated by column chromatography and high performance liquid chromatography, and identified by mass and nuclear magnetic resonance spectrometry. Minimal inhibitory concentrations (MIC) of the quinic acid derivative against Aeromonas hydrophila, Bacillus subtilis, Staphylococcus aureus and the fungus Aspergillus ochraceus were 250, 1000, 1000 and > 568μg/mL, respectively. Against the same microorganisms, MIC for rutin were 1000, > 1000, > 1000 and 35μg/mL, respectively. Rutin was very promising for A. ochraceus control, since its MIC against such fungus was close to the one observed for benzalkonium chloride, which is used as a fungicide in Brazil.


INTRODUCTION
Food-borne diseases correspond to a world problem that can be caused by microorganisms or their toxic metabolites. An example is ochratoxin A, a nephrotoxic, hepatotoxic and carcinogenic substance produced by some fungi of the Aspergillus genus that have been found in stored food such as coffee beans. It is also possible to mention enterotoxins produced by Staphylococcus aureus, which cause headache, diarrhea and vomit (Jay 2000). As a consequence, the use of additives to protect food against microorganisms is of great interest to the food industry. However, consumers are increasingly avoiding products prepared with preservatives of chemical origin due to their undesirable effects on human health. A promising alternative to achieve quality improvement in food commodities and a high degree of safety with respect to pathogenic microorganisms resides on the use of natural products (Rauha 2000). As plants have been important sources of new antimicrobial agents (Rios and Recio 2005), a preliminary evaluation of local plant extracts to identify those with antibacterial and antifungal properties was carried out. One of the best results was observed for the aerial parts of Solanum palinacanthum Dunal (Solanaceae), which presented ac-tivity against Aeromonas hydrophila, Bacillus subtilis, Staphylococcus aureus and Aspergillus ochraceus.
S. palinacanthum is a perennial herb or subshrub which spreads by means of slender, horizontal rhizomes. It is common in pastures, roadsides and similarly disturbed areas in Brazil (Coleman and Coleman 1982), where it has been used to treat skin diseases (Alves et al. 2006). Although its in vitro antimicrobial activity has already been briefly described in the literature by another research group (Alves et al. 2006), to the best of the authors' knowledge no work has ever been done to identify the antimicrobial metabolites produced by S. palinacanthum. Thus, the methanolic extract of such plant leaves was submitted to purification steps guided by antimicrobial assays in order to isolate and identify those substances.

GENERAL EXPERIMENTAL PROCEDURES
All reagents used were of recognized analytical grade. Acetic acid and methanol were HPLC-grade. During purification steps, solvent concentrations were carried out in a rotary evaporator at 35 • C followed by 24 h in a freeze-drier. Except when mentioned otherwise, all fractions were submitted to antibacterial diffusion assays and antifungal assays to direct fractionation. Column chromatography (CC) was carried out on silica gel 60 (230-400 mesh, Merck) or Amberlite XAD-16 (Sigma). Mass spectrometry (MS) analyses were performed on an Agilent 1100 LC/MS Trap equipped with an electrospray interface. Samples (1.0 mg) were dissolved in water:methanol (1:1, 1.0 mL) and 20μL were directly injected into the interface. Deuterated dimethyl sulphoxide (DMSO-d6) was used as solvent for nuclear magnetic resonance (NMR) analyses, performed on a Varian instrument ( 1 H NMR: 500 MHz and 13 C NMR: 125 MHz) using solvent peak as reference. Two dimensional NMR techniques (COSY, HMQC, HMBC and NOESY) were performed using standard Varian programs.

EXTRACTION AND ISOLATION PROCEDURE
Fresh leaves of Solanum palinacanthum Dunal (Solanaceae), collected in Lavras city, Minas Gerais State (Brazil), and identified at Herbarium ESAL (ESAL 06644), at Universidade Federal de Lavras, were sub-mitted to exhaustive methanol extraction at room temperature. Part of the crude extract was submitted to the antibacterial (10 mg/mL) and antifungal (4 mg/mL) assays. Then, 10 g of the S. palinacanthum crude extract were extracted exhaustively with hexane, ethyl acetate and methanol. A 3.5 g aliquot of the fraction soluble in methanol was submitted to CC on silica gel. As eluents were used methanol, water and 0.1% HCl. Part of the active material eluted with methanol (400 mg) was submitted to CC on Amberlite, using water and methanol as mobile phases. Part of the fraction eluted through the resin with methanol (176 mg) was purified on a HPLC system (Varian equipped with a 9050 UV detector at 254 nm, 9012 ternary pump and 9300 automated injector) using a Phenomenex Gemini silica C18 column (5μm, 250×10 mm). Gradient of water:methanol (80:20 to 32:68 in 16 min, 32:68 to 0:100 during 10 min) at 4.5 mL/min was used to elute substances, yielding seven fractions. Fraction six (F6, 27 mg, eluted between 14.5-15.0 min) and fraction seven (F7, 68.2 mg, eluted between 15.0-25.0 min) presented antimicrobial activity. F6 (20 mg) was purified on the same HPLC column using a 0.1% acetic acid:methanol (50:50) solution at 4.5 mL/min as mobile phase, which resulted in only one active fraction (rutin, 5.0 mg, elution at 8.0-8.6 min). For the purification of F7 (20 mg), a 0.1% acetic acid: methanol (45:55) solution at 4.5 mL/min was used as mobile phase, resulting in one active fraction (3,5-dicaffeoylquinic acid, 7.4 mg, 7.7-8.5 min). NMR and MS analysis were employed to identify both substances.

ANTIBACTERIAL ASSAYS
Antibacterial activity was evaluated in duplicates, with four standard bacterial strains acquired from the American Type of Culture Collection (ATCC, USA): Aeromonas hydrophila ATCC 7966, Bacillus subtilis ATCC 6633, Pseudomonas aeruginosa ATCC 27853 and Staphylococcus aureus ATCC 25923. Bacteria were grown in tryptic soy agar (TSA, Acumedia, USA), during 24 h at 37 • C. From each culture, a cell suspension was prepared with an aqueous 0.85% NaCl solution and adjusted to 0.5 MacFarland turbidity. In the next step, a swab was used to inoculate bacteria on the surface of Müeller-Hinton agar (Merck, Germany) plates (95 × 15 mm). Subsequently, 40μL from each sample (dissolved in ethanol/water 7:3) were deposited in 6 mm diameter holes made on the agar medium. All plates were incubated at 37 • C for 24 h. After this period, those fractions affording inhibition zones around the holes were considered active. Chloramphenicol (Sigma, USA) and etanol/water (7:3) were employed as positive and negative controls, respectively.
To determine minimal inhibitory and minimal bactericidal concentrations (MIC and MBC), a broth microdilution assay was employed, using Müeller-Hinton broth (MHB, Biolife, Italy) supplemented with calcium and magnesium cations (Alderman and Smith 2001) and standard bacterial inoculums (7.5 × 10 4 CFU/well). The crude extract was dissolved in an aqueous 1% (g/mL) Tween 80 solution (10 mg/mL) and filtered through a 0.22μm membrane (GV Durapore PVDF, Milipore, USA). Ten twofold serial dilutions were prepared to final concentrations ranging from 5,000 to 9.7μg/mL. The isolated substances were dissolved in DMSO (2.0 mg/ 100μL) and diluted in MHB, resulting in a 2.0 mg/mL solution. Twofold serial dilutions were prepared to final concentrations ranging from 1,000 to 1.95μg/mL. Chloramphenicol (Sigma, USA) and DMSO were used as positive and negative controls, respectively. After 24 h at 37 • C, the experiment was evaluated and 10μL were withdrawn from the content of each well with no visible bacterial growth and subcultured in TSA. MIC was defined as the lowest concentration of the tested substance that prevented a visible bacterial growth and MBC was defined as the lowest concentration yielding no subcultures during 24 h at 37 • C.

ANTIFUNGAL ASSAYS
To direct fractionation steps, antifungal activity was evaluated with Aspergillus ochraceus isolated from coffee beans as described in the literature (Kulwant et al. 1991). Suspension A was made from such fungus' spores (40μL − 3.69 × 10 5 spores) and 200μL of an aqueous 1% (g/mL) Tween 80 solution containing the sample to be evaluated at approximately 4.0 mg/mL. Czapek yeast extract agar (CYA) (Kulwant et al. 1991) was sterilized at 120 • C during 15 min and deposited (200μL) into each well of a polypropylene 96 wells plate. After CYA solidification, suspension A (20μL) was poured into the wells and the plate was kept at 25 • C, with a 12 h photoperiod, during 48 h. Assays were carried out with three repetitions, employing benzalkonium chloride and DMSO as positive and negative controls, respectively. Those samples which did not allow fungal growth were considered active.
Minimal inhibitory concentrations were obtained by a broth microdilution assay, using Czapek yeast extract without agar (CYB) and standard fungal inoculums (1 × 10 4 spores). The crude extract and the isolated substances were dissolved in DMSO (96.0 mg/mL and 12.5 mg/mL, respectively) and diluted with CYB (16 mg/mL and 1.14 mg/mL, respectively). Ten twofold serial dilutions were prepared to final concentrations ranging from 8,012.0 to 15.6μg/mL (crude extract) and 568.0 to 1.1μg/mL (isolated substances). After 48 h at 25 • C, with a 12 h photoperiod, MIC was the lowest concentration of the tested substance that prevented fungal growth.

RESULTS AND DISCUSSION
Up to now only a very brief report on the S. palinacanthum antimicrobial activity has been published (Alves et al. 2006 (Kambizi and Afolayan 2001). Similarly, in this study it was observed that S. palinacanthum prevented the growth of A. aeruginosa, B. subtilis, S. aureus and the fungus A. ochraceus, but P. aeruginosa, a resistant bacterium to several antimicrobial agents, was not affected by the methanol extract of that plant (Table I).
The fractionation of such extract guided by antimicrobial assays yielded two substances. One of them, isolated as a pale yellow residue, presented NMR spectra in perfect agreement with data previously reported   Rest et al. 2006). However, to the best of author's knowledge, this is the first report on the isolation of such substance from S. palinacanthum. When submitted to broth microdilution assays (Table II), it presented low antibacterial activity. Nevertheless, MIC value against A. ochraceus was close to the one observed for benzalkonium chloride, the only fungicide used in Brazil to control such fungus in coffee beans (Ministério da Agri-cultura, Pecuária e Abastecimento. Sistema de Agrotóxicos Fitossanitários 2006). Although the antimicrobial properties of rutin have already been described (Cushnie and Lamb 2005, Van der Watt and Pretorius 2001), its potential to control A. ochraceus, a food-borne pathogen in stored food, especially coffee beans, has not been previously reported.
The second active substance was isolated as a light yellow residue whose NMR spectra were in perfect agreement with data (Carnat et al. 2000, Beninger et al. 2004, Clifford et al. 2005) registered in the literature for 3,5-dicaffeoylquinic acid (Fig. 1). In the positive mode, MS spectra presented peaks at m/z 555 [M+K] + , 539 [M+Na] + and 517 [M+H] + , while in the negative mode a peak at m/z 515 [M−H] − was observed. Although caffeoylquinic acid derivatives are known as coffee components, they are widespread in plants (Guerrero et al. 2001), including those of the Solanum genus. Concerning 3,5-dicaffeoylquinic acid specifically, it has been isolated from S. melongena (Whitaker and Stommel 2003),   et al. 2003), antiproliferative, tirosinase inhibitory and antihypertensive activity were also attributed to this compound (Iwai et al. 2004, Mishima et al. 2005. Summarizing, the substances responsible for the antimicrobial properties of the S. palinacanthum leaves extract have been isolated and identified as rutin and 3,5dicaffeoylquinic acid. Although their in vitro antibacterial activities were not as pronounced as expected, rutin was very promising to control the fungus A. ochraceus.