Antibacterial Effects of Glucosinolate-Derived Hydrolysis Products Against Enterobacteriaceae and Enterococci Isolated from Pig Ileum Segments

Introduction

T here is increasing government and public concern about food safety and specifically contamination of animal and human foods by bacterial pathogens and bacterial toxins. Bacterial pathogens can arise in the foods during production (natural pathogens of the food materials) or during processing (contamination through industrial processes). Obviously, it is vital to control pathogenic bacteria at all stages of food production. Pigs are both a good model for studying bacterial pathogens and are a major food source in many European countries; in Portugal, there are large numbers of meat products derived from pigs. With the increase in multi-drug resistance (MDR) bacteria, it is very important to identify effective alternative means of controlling pathogenic bacteria during production and food processing (Livermore et al., 2008; Simões et al., 2009). The use of antibiotics to treat gastrointestinal infections is controversial and is recommended only for severe gastroenteritis cases (Phavichitr et al., 2003). Also, several resistance mechanisms have been detected, and therefore there is a growing recommendation to study other effective mechanisms to prevent and treat the occurrence of this pathogenic agent (Edwards et al., 2001; Saavedra et al., 2004). One line of research has been the evaluation of the antimicrobial effect of natural compounds against pathogenic and nonpathogenic bacteria (Aires et al., 2009). One of the most important groups of natural compounds is the glucosinolates and more specifically their enzymatic hydrolysis products (HP). These compounds are largely present in Brassicacea (Syn. Cruciferae) plants, comprising at least 120 compounds with well-defined structures (Bones and Rossiter, 2006). Glucosinolates occur in plants simultaneously with an endogenous β-thioglucosidase enzyme commonly named myrosinase (E.C. 3.2.1.147), which is responsible, after tissue disruption (cut, ground, or chewed), for the hydrolysis of glucosinolates into numerous biologically active products such as isothiocyanates (ITCs), nitriles, thiocyanates, and various indole compounds (Fahey et al., 2001; Bones and Rossiter, 2006). The types of compounds produced are very specific to the respective glucosinolate present in the tissue and conditions under hydrolysis occurs (Bones and Rossiter, 2006). The difference in chemical properties and biological activity between glucosinolates and their enzymatic HP is largely determined by the side-chain structure. In terms of antimicrobial action, the antimicrobial effect of ITCs is attributed to the binding to sulfhydyl groups on active sites of enzymes important to microbial growth and survival (Tang et al., 1974). In the past, Kolm et al. (1995) demonstrated that enzymes such as glutathione transferases promote addition of the thiol group of γ-glutamylcysteinylglycine to the electrophilic central carbon of the ITC group to form dithiocarbamates. Also, the ITCs can react with amino groups, forming dithioureas. These two types of reactions could be the major mode of antibacterial action of the isothiocyanates, i.e., modification of membrane or cellular proteins causing inactivation and also reductions in the cellular levels of important thiols (GSH and L-Cys), which could lead to formation of oxygen and other free-radicals (Aires et al., 2009). It has been shown that the biological effects are dependent on the levels and classes of glucosinolates present (Griffiths et al., 1998). Despite these results on the biological effect of HP derived from glucosinolates (Kuroiwa et al., 2006; Aires et al., 2009), no major studies have been conducted with bacteria from animal sources.

Thus, in this study we evaluate the antimicrobial properties of glucosinolate HP on Gram-positive and Gram-negative aerobic bacteria of porcine origin, and investigate the mode of action beneath the antimicrobial effect of glucosinolate HP.

Methods

Results

Various HP derived from common dietary glucosinolates were evaluated (Table 1). Because the pig gastrointestinal tract is predominated by strict anaerobes, there have been very few studies on the presence, if any, of aerobic species. There are reports of E. coli in pig intestines (as part of the natural flora and during certain infections, for example, post-weaning colibacillosis), which can grow aerobically (i.e., facultative anaerobes). In the current study, 13 aerobic species were isolated from the pig ileum.

The Enterobacteriaceae isolated and identified from the pigs were Enterobacter hormaechei, Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Serratia liquefaciens, and Serratia proteamaculans. AITC was only moderately effective against E. hormaechei at the highest dose tested (15.0 μmoles) and had no effect on any of the other Gram-negative bacteria (Table 2). SFN at the highest dose tested (3.0 μmoles) was moderately effective against E. hormaechei and K. pneumoniae, and more effective than gentamycin for all the other Gram-negative bacteria; inhibitory effects were observed at the 0.15 μmole SFN dose for all the bacteria, with the exception of E. hormaechei. E. hormaechei was also less affected by the two aromatic ITC (BITC and PEITC), and growth inhibition was only seen at the highest dose of 15.0 μmoles. All the other bacteria were susceptible to BITC and PEITC; at the highest dose of BITC (15.0 μmoles), E. coli, P. mirabilis, and S. proteamaculans were completely inhibited, and at the highest dose of PEITC (15.0 μmoles) S. liquefaciens was completely inhibited. The ITC mix (AITC, BITC, and PEITC) showed some synergistic effects against all of the bacteria. Bacterial inhibition by ITC, significantly more effective than gentamycin, was seen in E. hormaechei (ITC Mix), E. coli (SFN, BITC, PEITC, and ITC Mix), K. pneumoniae (ITC Mix), P. mirabilis (SFN, BITC, and ITC Mix), S. liquefaciens (SFN, BITC, PEITC, and ITC Mix), and S. proteamaculans (SFN, BITC, and ITC Mix). DIM and ASC had no effect on the Gram-negative bacteria at any of the concentrations, with the exception of ASC at the highest dose (15.0 μmoles), which had a slight inhibitory effect against P. mirabilis (Table 2). I3C had a slight effect against all of the Gram-negative bacteria but only at the highest dose (15.0 μmoles). IAN was moderately effective against all the Gram-negative bacteria at 3.0 and 15.0 μmole doses, but never achieved zones of inhibition greater than those of gentamycin.

Three Gram-positive bacteria, all enterococci, were isolated and identified from the pigs: Enterococcus faecalis, Enterococcus flavescens (syn. E. casseliflavus), and Enterococcus hirae. Of the ITC tested, AITC had no effect and SFN was consistently effective against all three Gram-positive bacteria (Table 2). At the highest dose tested (3.0 μmoles), SFN was more effective than gentamycin for all three Gram-positive bacteria. BITC and PEITC were effective against E. faecalis and E. hirae. BITC at the highest dose tested (15.0 μmoles) completely inhibited bacterial growth of E. faecalis and E. hirae. PEITC at the highest dose tested (15.0 μmoles) completely inhibited bacterial growth of E. hirae and at 1.5 μmoles was more effective than gentamycin against E. faecalis (Table 2). BITC was only slightly effective, at the highest dose, against E. flavescens, and PEITC had no effect against E. flavescens at any of the doses tested. The ITC mix (AITC, BITC, and PEITC in equimolar doses) showed some synergistic effects against all three bacteria at lower doses, and at the highest dose (15.0 μmoles) tested it completely inhibited bacterial growth of all three Gram-positive bacteria. Of the indoles tested, I3C was only effective against E. faecalis, IAN was slightly effective against E. flavescens and E. hirae, DIM was ineffective against all three bacteria, and ASC was moderately effective against all three bacteria. None of the indoles, even at the highest doses tested (15.0 μmoles), were as effective as gentamycin.

Discussion

Enterobacter, E. coli, Klebsiella, Proteus, Serratia, and Enterococcus species are known to produce various diseases in pigs, which can be easily transmitted to human beings (Hershberger et al., 2005; Shankar et al., 2006; Livermore et al., 2008; Hwang et al., 2009; Lucciano and Holey, 2009; Abatih et al., 2009; Duda et al., 2009; Luzzaro et al., 2009). The results of the present study show that, among the different compounds tested, the antibacterial activity was confined to a few of the glucosinolate HP, particularly ITCs. These results are in agreement with previous reports (Tierens et al., 2001; Aires et al., 2009), showing that some glucosinolate HP have very little, if any, antimicrobial activity. The antibacterial activities of the glucosinolate HP varied according to the compound, the concentration used, and the micro-organism tested.

Comparing these results with the previous studies (Aires et al., 2009) a similar pattern of sensitivity to the HP for the different bacteria (Gram-positive and Gram-negative) is observed. AITC was the least effective ITC against Gram-positive and Gram-negative bacteria isolated from pigs, although there are reports of AITC being effective at low doses and low pH against E. coli O157:H7 (Lucciano and Holey, 2009). The aromatic ITC (benzyl and 2-phenylethyl) had variable effects against Gram-positive and Gram-negative bacteria isolated from pigs, but generally BITC was more effective at the lower doses (up to 3.0 μmoles) compared with PEITC. It was also clear that the ITC mix (AITC, BITC, and PEITC) was often more effective than the single ITC against Gram-positive and Gram-negative bacteria, and was also consistently more effective than gentamycin (Aires et al., 2009). Of the ITC tested, SFN was consistently effective against both Gram-positive and Gram-negative bacteria (Aires et al., 2009). It is clear that SFN, possibly related to its physico-chemical and solubility properties, is a broad-range antimicrobial agent effective at low concentrations. It is not known if the structurally related dietary ITC such as iberverin (3-methylthiopropyl-ITC; derived from 3-methylthiopropylglucosinolate; glucoiberverin), iberin (3-methylsulfinylpropyl-ITC; derived from 3-methylsulfinylpropylglucosinolate; glucoiberin), and erucin (4-methylthiobutyl-ITC; derived from 4-methylthiobutyl-glucosinolate; glucoerucin) are as effective as SFN. The SFN has been shown to have antibacterial activity against a larger range of bacteria, and it is considered a powerful inducer of xenobiotic detoxification enzymes associated with anticarcinogenesis (Juge et al., 2007). Therefore SFN, provided in the diet from broccoli and other Brassica foods, could play an important role in antimicrobial activity.

There are many reports on the effectiveness of indole-based compounds against bacteria (Sung and Lee, 2007). For the indole HP (I3C, IAN, DIM, and ASC), the results for pig bacteria were similar to those for human bacteria (Aires et al., 2009); that is, in general, the DIM has no antimicrobial activity, the antimicrobial activity of ASC was restricted to the Gram-positive bacteria, and the I3C and IAN antibacterial effect was more pronounced in Gram-negative bacteria. Nevertheless, we noted that their effect was strong only for higher doses, even though it was less than those obtained for antibiotic. It seems that only IAN, ASC, and I3C have antimicrobial activity, probably because of their similar chemical structure (Sung and Lee, 2007). Conversely, IAN had no major effects against the Gram-positive bacteria, perhaps because of efficient metabolism of this indole, but had moderate antimicrobial activity against many of the Gram-negative bacteria. Cheng et al. (2004) have shown that certain strains of human intestinal Bifidobacterium can convert intact glucosinolates (sinigrin and glucotropaeolin), using a myrosinase-like enzyme, to their corresponding nitriles. In addition, many bacteria have been reported to have effective nitrile (alkyl, aliphatic, and aromatic) metabolizing pathways involving nitrilases, converting nitriles to their corresponding carboxylic acids and ammonia (Podar et al., 2005). Certain bacteria can efficiently metabolize amines (alkyl and aryl) to aldehydes catalyzed by amino oxidases and consequently quench any potential antimicrobial activity (Van Ginkel et al., 2008).

Based on these results, it is clear that there is a strong relationship between the chemical structure and the antimicrobial activity of glucosinolates HP. ITCs can react, nonenzymatically, with thiol groups (e.g., in glutathione or redox-active proteins) to form dithiocarbamates and with amino groups (e.g., in proteins) to form thioureas (Juge et al., 2007). These reactions may form part of the antimicrobial activities seen with the ITC, specifically the increase in oxidation and inhibition of essential proteins or enzymes from membranes, leading to bacterial cell death. It has also been proposed that ITC may be oxidized by interactions with cytochrome P-450 enzymes, producing much more reactive isocyanates, even more chemically and biologically reactive than the parent compounds (Davidson and Botting, 1997).

This in vitro study provides data demonstrating the potential for these natural dietary chemicals for treating infectious diseases. It was also clear that the ITC mix (AITC, BITC, and PEITC) was often more effective than the single ITC against Gram-positive and Gram-negative bacteria, and was also consistently more effective than gentamycin (Aires et al., 2009). As for the many antibacterial phytochemicals, the mode of action of ITC is not currently well known (Simões et al., 2009). Nevertheless, it seems that ITCs can be more effective because they can affect the redox status of cells by binding to glutathione (forming dithiocarmbamates), which can potentially lead to cellular oxidation and the formation of toxic free radicals. ITC can also modulate cell signaling mechanisms by reacting with redox-active thiols in signal protein and inhibit enzymes (by forming dithiocarbamates with thiol groups and/or forming thioureas with amine groups). Therefore, it is possible that ITC inhibit bacterial growth by affecting cell redox and inhibition of key bacterial enzymes.

Clearly, more studies are needed to evaluate their effectiveness in vitro and in vivo (i.e., determining MIC values and evaluating synergistic effects of HP mixtures as found in the consumption of the brassica plants). In vivo studies need to be performed to see how effective the ITC are in the gastrointestinal environment and specifically how rapidly their effects are lost due to uptake and metabolism in the enterocyte cells lining the intestinal tract; this rapid metabolism (primarily conjugation to glutathione) has been shown in other animal and human cell models and in vivo intervention studies. Long-term low-dose studies with pure compounds and Brassica-related plant materials incorporated into the pig feed should also be performed to evaluate how dietary intervention modifies the gastrointestinal tract bacteria and the physiological status of the pigs. Finally, the potential toxicity of HP to pigs must be studied.