Evaluation of the mechanism of activity of extracts of Pterocarya fraxinifolia (Juglandaceae) trees used to replace antibiotic growth promoters in pig production in Europe

Abstract

The emergence of microorganisms resistant to antibiotics has led to the ban on many of the antibiotic feed additives used in animal production. The Swann Report and World Health Organisation (WHO) recommended that precautionary measures be established by governments to adopt a proactive approach to reduce the levels of antibiotic use in animals and put in place surveillance mechanisms to detect resistance. This led to the limitation and in some cases the total ban on the use of antibiotic growth promoters (AGP) in animal production. The first countries to enforce the ban included Denmark, Sweden and the United Kingdom. The main reason that led to the banning of antibiotics in Europe was the risk of antibiotic resistance transfer to human pathogens. Several research initiatives have been established to look into other alternative sources that can be used as antibiotics. Plants have received a lot of attention in this regard and phytogenic feed additives and/or antibiotics are now a focal point as new alternatives. In the European Commission funded research projects Rumen (QLK5-CT-2001-00992) and REPLACE (FOOD-CT-2004-506487), a catalogue/database of plant samples were established through the collection of plant material from six geographical locations in three European countries. Pterocarya fraxinifolia fed to pigs and poultry had good activity in replacing antibiotic growth promoters. Plants growing in Denmark, Germany and Poland had widely diverging activities in animal studies. Pterocarya fraxinifolia trees growing in locations in Denmark, Germany and Poland had activity levels ranging from high to medium and inactive according to in vivo results provided by Dr Ole Hojberg. It would make sense to isolate and characterize the antimicrobial compounds so that this can be used to select populations with a high activity. The original aim of this project, following the success achieved in the in vivo trials using unextracted plant material, was to isolate and characterize the antimicrobial compounds from P. fraxinifolia trees growing in different areas in Europe, active against relevant pathogens in order to facilitate the selection of tree populations with good activity without having to carry out further animal experiments. A secondary aim was to investigate alternative mechanism of activity to replace antibiotic feed additives in the REPLACE programme. A supplementary aim was to find out what in vitro method can be used to predict the in vivo activity of different tree populations. Antimicrobial activity was generally moderate for Pterocarya fraxinifolia extracts; the excellent activity was in the extracts from trees growing in Denmark and Germany (MIC 0.02 to 0.16 mg/ml). Activity against microaerophillic organisms (Clostridium perfringens and Campylobacter jejuni) was also good (MIC ranged = 0.02 to 0.63 mg/ml). The important aspect is that there were poor correlations between the antimicrobial activity of extracts from different populations and the in vivo activity of plants from different populations against the different pathogens. This indicates that antimicrobial activity may not be the mechanism of activity of plants included in the feed of the animals. In general the antimicrobial activities of the solvent-solvent derived fractions were lower than that of the crude extracts pointing to synergistic activities between different compounds in the crude extract. In considering the antimicrobial activity of the crude extracts, there was a low correlation between in vivo activity of the plant and in vitro activity of the pathogens that are usually considered to play an important role had very low R2 values, (E. coli K88 = 0.0047, C. jejuni = 0.0809, C. perfringens = 0.1092). Therefore other mechanisms of activity were also considered. Antioxidant compounds are known to disrupt bacterial cell membrane function and structure and may also influence the response of the host to the pathogen. There was excellent correlations between the antioxidant activity of the crude acetone extract and the in vivo activity of plants from different populations (R2= 0.8167). The correlation was even better in the case of the polar water/methanol fraction (R2 = 0.8746). Because the content of the gastrointestinal track consists of an aqueous solution, it is understandable that polar compounds may be readily solubilized from the plant material included in the diet. The anti-inflammatory activity was determined using the 15-lipoxygenase enzyme assay. The crude extracts and the solvent-solvent derived fractions of P. fraxinifolia had moderate activity for both leaf and fruit samples, with EC50 ranging from 4.47 ± 0.15 to 7.19 ± 0.05 mg/ml. There was not such a good correlation between anti-inflammatory activity and in vivo activity of the crude acetone extract as was the case for antioxidant activity (R2 =0.3685). It is interesting that there was a better correlation between in vivo activity and anti-inflammatory activity for the hexane fraction (R2 = 0.5536). This may indicate that non-polar compounds are involved with anti-inflammatory activity. The crude acetone extracts of P. fraxinifolia trees growing in Denmark, Germany and Poland were relatively toxic with cellular toxicity LD50 varying between 0.77 and 1.75 μg/ml. The lack of toxicity in the animal trials may possible be due to the lower solubility and/or lower bioavailability of toxic metabolites in the aqueous gastro intestinal fluid. Hence in vitro toxicity may not always equate to whole organism toxicity. The TLC bioautograms of plant extracts revealed the presence of several antimicrobial compounds in the plant extracts. The isolation and characterization of the compound(s) from P. fraxinifolia was performed using bioassay-guided procedures for antibacterial activity using open column chromatography with silica gel and Sephadex LH 20 as stationary phases. Five compounds were isolated and sufficient quantities of two were available for structure elucidation as β-sitosterol and pentacosanol. All five isolates had weak antimicrobial activity when compared to the crude extract. The isolation did not yield any novel compound of importance. It was also very disappointing that the compounds isolated had such low antimicrobial activity. Although bioassay guided fractionation was used up to a certain point the assumption that compound present in high concentration in active fractions were the active compound, may have been wrong. As has already been stated there were good correlations between in vivo activity and antioxidant activity, intermediate correlations with anti-inflammatory activity and poor correlations with antimicrobial activity. An alternative mechanism for in vivo activity could be anti-inflammatory activity or stimulation of the immune system (via antioxidant activity) based on good correlations obtained. Another positive aspect is that it appears that direct killing of pathogens by the plant extract is not a major factor responsible for the in vivo activity. If the factors above are indeed the main factors responsible for in vivo activity, there may be decreased development of resistance by pathogens against antimicrobial compounds present in the extracts. Additional animal and in vitro studies are required to confirm the results that antioxidant activities of extracts are the best predictor of activity in animal studies. This may facilitate the selection of plant material from different tree populations for replacing antibiotic feed additives in animal production.