Ascopyrone P, a novel antibacterial derived ... - Wiley Online Library

Journal of Applied Microbiology 2002, 93, 697–705

Ascopyrone P, a novel antibacterial derived from fungi L.V. Thomas1, S. Yu2, R.E. Ingram1, C. Refdahl2, D. Elsser3 and J. Delves-Broughton1 1

Danisco Innovation, Beaminster, UK, 2Danisco Innovation, Copenhagen, Denmark, and 3Danisco Niebu¨ll GmbH, Niebu¨ll, Germany

2002 ⁄ 103: received 15 March 2002, revised 24 June 2002 and accepted 24 July 2002

L.V. THOMAS, S. YU, R.E. INGRAM, C. REFDAHL, D. ELSSER AND J . D E L V E S - B R O U G H T O N . 2002.

Aims: To assess the antimicrobial efficacy of ascopyrone P (APP), a secondary metabolite formed by the fungi Anthracobia melaloma, Plicaria anthracina, Plic. leiocarpa and Peziza petersi belonging to the order Pezizales. Methods and Results: In vitro testing using a well diffusion procedure showed that APP at a high concentration (approximately 5%) inhibited the growth of Gram-positive and Gramnegative bacteria. Using an automated microbiology reader, growth curve analysis showed that 2000–4000 mg l)1 APP caused total or significant bacterial inhibition after incubation for 24 h at 30C. Against certain yeast strains, 1000– 2000 mg l)1 APP enhanced growth, although at higher concentrations inhibition of some yeasts was observed. Clostridium and fungal strains were not sensitive to 2000 mg l)1 APP. No significant cidal effect was observed after 2 h against Listeria monocytogenes or Escherichia coli. Results were identical whether the APP samples tested had been produced enzymatically or chemically. Conclusions: At a level of 2000 mg l)1, APP demonstrated growth inhibitory activity against a broad range of bacteria, but not yeasts or moulds. Significance and Impact of the Study: A possible application for this novel natural antimicrobial is in food preservation, to control the growth of Gram-negative and Grampositive bacteria in raw and cooked foods. Effective dosage levels would be 500–4000 mg kg)1, depending on food type. The efficacy, organoleptic and safety aspects of this compound in food still need to be assessed. INTRODUCTION Glycogen is the carbon storage polymer found in fungi and is degraded to form glucose when the growth medium is depleted of carbon sources. Under normal growth conditions, glucose enters the glycolysis pathway and provides energy and building blocks. Under biotic and abiotic stress conditions, however, the glycogenolysis is shifted, at least partly, from forming glucose to the formation of 1,5-anhydro-D-fructose (1,5-anhydro-D-arabino-hex-2-ulose, AF) and further, to different secondary metabolites (Baute et al. 1993), the so-called Anhydrofructose Pathway of glycogenolysis (Yu et al. 1995) (Fig. 1). The formation of AF from glycogen is catalysed by a-1,4glucan lyase (EC 4.2.2.13) (Baute et al. 1988; Yu et al. 1995, Correspondence to: Dr L. Thomas, Danisco, Innovation Department, 15 North Street, Beaminster, Dorset, UK, DT8 3DZ (e-mail: [email protected]).

ª 2002 The Society for Applied Microbiology

1997, 1999). The formation of ascopyrone P (1,5-anhydro-4deoxy-D-glycero-hex-1-en-3-ulose, APP) from AF is probably catalysed by a dehydratase (Baute et al. 1993). APP was first prepared as one of the pyrolysis products of amylopectin, amylose and cellulose in a yield under 3% (Shafizadeh et al. 1978). It has been further characterized and its crystal structure is known (Stevenson et al. 1981). The functionality and physiological role of APP is largely unknown, however. APP has only been tested for its activity against two strains of Staphylococcus and one strain of Escherichia coli with a minimal inhibitory concentration of 250 and 500 lg ml)1 (Baute et al. 1993). In the present study, the antimicrobial activity of APP was investigated against a wide range of micro-organisms: Gram-positive and Gram-negative bacteria, moulds and yeasts. The chosen test strains were representative of food-borne pathogens and food spoilage organisms. The possible inhibitory mechanism of APP against these organisms is discussed.

698 L . V . T H O M A S ET AL.

Fig. 1 The anhydrofructose pathway of glycogenolysis (Yu et al. 1995)

M A T E R I A LS A N D M E T H O D S Culture of the APP–producing fungus Anthracobia melaloma CBS239.54 was obtained from Centraalbureau Voor Schimmelcultures (the Netherlands). It was cultured at 24C on Potato Dextrose agar medium (PDA, Difco), containing per litre: 4 g potato extract, 20 g dextrose, 15 g agar. The final pH of the medium was pH 5Æ6 ± 0Æ2. The fungal culture was harvested after 20 days of growth. APP preparation The harvested biomass was subjected to abiotic stress by freezing at )20C overnight and thawing at ambient temperature (24C) for 1–2 h. Three volumes of 50 mmol l)1 Mops-NaOH (pH 6Æ2) containing 5 mmol l)1 CaCl2, 1% toluene (v ⁄ v) and 10% dextrin-10 (w ⁄ v) (Fluka Chemie AG) were added and thoroughly mixed to ensure even distribution and cell disruption. The mixture was left at ambient temperature for 2–3 days with stirring. APP formation was monitored continually at 289 nm spectrophotometrically and by HPLC by taking 5–20 ll samples from the reaction mixture and mixing them with 1 ml water. At the end of the reaction, the mixture was centrifuged at 30 000 g for 30 min and the supernatant fluid was freezedried. APP was selectively extracted from the freeze-dried material with 80% acetonitrile (v ⁄ v). Acetonitrile was evaporated under reduced pressure and the APP syrup was extracted with water, filtered, freeze-dried and used for antimicrobial activity testing. By this method, two APP samples were prepared: APP20010213 (169 mg ml)1) and APP20010215 (138 mg ml)1). A third APP sample, E002012 (49Æ3 mg ml)1), was prepared by chemical synthesis using glucose as the starting material (Andersen et al. unpublished results). All samples were kept at )20C until use. Stock solutions were made up in deionized water and filter-sterilized (0Æ2 lm filter). The structure of APP was confirmed using NMR as described earlier (Baute et al. 1993). In certain cases, APP was further separated from AF, glucose, maltose and maltodextrins by interaction ion exchange chromatography on monosphere resin 99 Ca ⁄ 320 (Dow Chemical Co. Ltd, Midland, MI, USA).

Aqueous APP with a pH in the range of 1Æ5–5Æ5 was quantified using an E (mM) of 6Æ1 (Stevenson et al. 1981) at 289 nm. APP was analysed on a Waters HPLC instrument (model WISP 710B, Waters Corporation, Milford, MA, USA) equipped with a differential refractometer (model 410) and a u.v. monitor (Lambda-Max model 481 LC spectrophotometer) set at 289 nm. The column used was a carbohydrate Ca2+ column (6Æ5 · 300 mm, Interaction Chromatography Inc., San Jose, CA, USA) and a symmetry shield C18 column (3Æ9 · 150 mm, Waters Corporation, Milford, MA, USA). Chromatographic separation was performed at 24C and isocratically at 0Æ5 ml min)1, using water as the eluent, and monitored at both 289 nm and by reflex index. APP was also analysed by thin layer chromatography (TLC). The solvent system was composed of ethylacetate, HAC, methanol and water (volume ratio 12:3:3:2). An aluminium silica gel 60 TLC plate (0Æ2 mm thickness) of 20 · 20 cm from Merck (Darmstadt, Germany) was used. APP, AF, glucose and maltodextrins on the TLC plates were visualized by spraying a reagent consisting of 25 ml acetic acid, 0Æ5 ml concentrated sulphuric acid and 0Æ25 ml anisadehyde, and then warmed in a thermostatically controlled oven at a temperature of 105C. AF and glucose, the by-products formed in APP preparation, were analysed as described earlier (Yu et al. 1998). APP test samples Sample E002012 (49Æ3 mg ml)1) was dissolved in sterile deionized water and used, without filtration, in the well diffusion tests, for growth curve analysis and investigation of cidal activity. Sample APP20010213 (169 mg APP ml)1) and APP sample APP20010215 (138 mg APP ml)1) were used for growth curve analysis testing. Sample APP20010215 was also used in tests for fungal and Clostridium sensitivity. Culture and preparation of test strains for antimicrobial spectrum testing The following test strains were all from the Danisco culture collection: Aspergillus niger CBS733.88; A. versicolor CBS08959; Bacillus cereus strains 204; Campden; 3.046; B. subtilis Campden; Brochothrix thermosphacta CRA7883; Clostridium sporogenes strains 1.221; Campden; ABC20; 4.440; Cl. tyrobutyricum 2753; Enterobacter aerogenes 10102; E. coli strains CRA109; S15 (serovar O157); 2.083; CRA1593; CRA161; CRA92; Lactobacillus sake A10; Listeria monocytogenes strains 272; 358; S23; Scott A; F6861; Micrococcus luteus NCIMB8166; Penicillium commune ABC118; P. discolor CBS547.95; P. roquefortii S44; Pseudomonas fluorescens strains 3.756; 10460; 1331; 327; Saccharomyces carlsbergensis CRA6413; S. cerevisiae strains

ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 697–705

ASCOPYRONE P, A NOVEL ANTIBACTERIAL

699

ATCC9763; H78; S. cerevisiae var. paradoxus H103; Salmonella sp. S19; Salm. Typhimurium S29; Staphylococcus aureus 5.720; Yersinia enterocolitica S16. The strains were derived either from commercial or private culture collections, or isolated directly from food. All test micro-organisms were taken from storage at )80C. Most were tested as vegetative cell suspensions from overnight broth culture. Bacillus and Clostridium species were tested as endospore suspensions prepared earlier and stored at 4C. Spore suspensions were prepared by growing Bacillus strains on Brain Heart Infusion agar (BHIA, Oxoid) at 30C overnight and Clostridium strains on Reinforced Clostridial agar (RCA, Oxoid) anaerobically overnight at 37C. The plates were then kept at ambient temperature for up to 5 days. When spore production had been confirmed by microscopic examination, suspensions were made in Maximum Recovery Diluent (MRD, Oxoid), heat-treated at 80C for 20 min to kill vegetative cells, and the spores enumerated by viable count. Fungal strains were tested as spore suspensions, stored at )20C. These were prepared by growing the fungal strains for 7 days on Malt Extract agar (MEA, Oxoid) at 25C, after which time they were examined macroscopically for spore production. The spores from one plate were removed into 10 ml 15% glycerol ⁄ water suspension containing 0Æ01% Tween 20. The spores were enumerated microscopically using a haemocytometer. For antimicrobial inhibition testing, most bacteria were grown in Brain Heart Infusion broth (BHI, Oxoid). Lactobacillus sake A10 was grown in de Man, Rogosa, Sharpe medium (MRS, Oxoid). Yeasts were grown in Sabouraud Liquid medium (SLM, Oxoid). Unless otherwise stated, bacteria were cultured at 30C. Lactic acid bacteria were grown on solid medium in enriched CO2 atmosphere using CO2 generating packs in anaerobic jars (GENbox CO2, bioMe´rieux, l’Etoile, France). Clostridium species were grown in Reinforced Clostridial medium (RCM, Oxoid) at 37C anaerobically. Brochothrix thermosphacta, moulds and yeasts were grown at 25C. Fungi were cultured on Malt Extract agar (MEA, Oxoid).

An automated Microbiology Reader Bioscreen C (Labsystems Oy, Helsinki, Finland) was used to measure growth curves of the strains in the presence and absence of APP. The Bioscreen C simultaneously measures kinetically, by vertical photometry, the development of turbidity (i.e. growth) in 200 wells of a honeycomb microtitre plate. The optical changes in liquid medium are correlated with microbial counts in the samples. The system consists of a Bioscreen C analyser, which is an incubator and measurement unit, integrated with a PC, software (BioLink v 5.30), printer and a ‘Honeycomb 2’ cuvette multiwell plate. Growth curve data can be analysed within the BioLink software or exported to programmes such as Excel. Broth culture media were dispensed in 240 ll volumes into the wells. Serial dilutions of a filter-sterilized APP stock solution were then dispensed into the same wells as appropriate. The wells were inoculated with 30 ll of an appropriately diluted overnight broth culture or spore suspension to give a final inoculum level of about 103 cfu ml)1. The tests were incubated in the Bioscreen C for either 24 h at 30C (most bacteria), or 72 h at 25C (yeasts and B. thermosphacta), with readings taken every 10 or 20 min after the trays were shaken. After the incubation period was finished, the data were exported to Excel for analysis to generate graphs. Inhibition was analysed using the BioLink programme to measure the area under the curve as an indication of microbial growth. The area reduction percentage values were estimated to demonstrate the inhibitory effects of APP (Skytta et al. 1993). An average was taken of two or more readings from different APP samples. For comparative purposes, stock solutions of APP and potassium sorbate (Sigma) were prepared and tested at 2000 mg l)1 in the Bioscreen at pH 4 (for yeasts), pH 5 (for bacteria) and pH 6 (for all strains).

Growth inhibition test: well diffusion

Sensitivity testing of Clostridium species

A small sample of APP was available for this preliminary antimicrobial screening and therefore, the method was adapted for low volume testing. Agar pour plates (10 ml) were prepared, which had been seeded with various test organisms by inoculation with 20 ll of either a spore suspension or overnight broth. This gave an inoculum of approximately 105)106 cfu ml)1. After the agar plates had set, small wells were cut, into which 20 ll of the APP sample were loaded. The plates were incubated overnight at a temperature appropriate to the test micro-organism, and

Stock solutions of APP at concentrations of 0Æ5, 1 and 2% were prepared and filter-sterilized. Cooked Meat medium (CMM, Oxoid) was prepared by distributing 1 g of the medium to individual test tubes, which were then filled with 8Æ8 ml water. After autoclaving, 100 ll of the APP stock solutions were added to give final concentrations of 0, 500, 1000 and 2000 mg l)1 APP. The tests were inoculated with 100 ll of Clostridium spores, to give an inoculum level of about 103 cfu ml)1. The tubes were overlaid with 3 ml 2% agar to maintain anaerobiosis. The tests were incubated at

examined after 1–2 days (when microbial growth was clearly visible) for zones of inhibition. Analysis of growth curves of yeast and bacteria in the presence or absence of APP



37C and examined daily for signs of growth (turbidity, gas production).

conditions were varied. The conditions for optimum production of APP by surface culture were also investigated. The optimal pH of APP preparation was pH 6Æ2 and the optimal temperature was 24C in a reaction time of 2–3 days. APP was found to be unstable in organic solvent due to isomerization. APP was, however, found to be more stable in an organic solvent–water mixture, in particular in 80% acetonitrile. Using this aqueous acetonitrile mixture, APP was selectively extracted from the reaction mixture containing AF, glucose, dextrins and the cellular components. Using the current method, approximately 10 g APP were prepared routinely at laboratory scale.

Sensitivity testing of fungal species A 2% (w ⁄ v) APP solution was used to prepare 25 ml MEA plates containing 0, 1000 and 2000 mg l)1 APP. The plates were inoculated in the centre with 10 ll of a fungal spore suspension (106 spores ml)1). The plates were incubated at 25C and examined after 5 days. The colony size and morphology were then recorded. Investigation of APP cidal activity

APP stability

Sterile deionized water or APP sample E002012 (100 ll) was added to 890 ll 10 mmol l)1 HEPES buffer (pH 7, Sigma). Overnight culture of either L. monocytogenes S23 or E. coli S15 (10 ll) was added to the test, which was incubated at ambient temperature for 2 h. After this time, the bacteria were enumerated by viable count. The concentration of APP in the test was 4Æ93 mg ml)1.

At room temperature, APP was stable for approximately 2 weeks in aqueous solution at a pH value < 6Æ5. APP was unstable at elevated temperature. At 100C an aqueous water solution of 500 mg l)1 APP had a half-life of approximately 15 min. Growth inhibitory activity of APP

RESULTS

All the strains tested against 4Æ93% APP in the well diffusion test showed clear zones of inhibition. These strains were: B. cereus 204; Cl. sporogenes Campden; L. monocytogenes S23; M. luteus NCIMB8166; Lact. sake A10; B. thermosphacta CRA7883; E. coli S15; Ps. fluorescens 327; S. carlsbergensis CRA6413; S. cerevisiae ATCC9763. Growth curve data obtained using the Bioscreen C were calculated as area reduction percentage and averaged (Table 1). APP demonstrated inhibitory activity against all

Preparation of APP In this study, the activation for APP production was optimized for the surface-cultured fungus, Anthracobia melaloma, by combining the abiotic stress conditions of freezing ⁄ thawing, homogenization and exposure to 1% toluene. The production of APP using submerged culture has, however, proved unsuccessful, even when the culture

Table 1 Average growth area reduction percentage in the presence of 250–4000 mg l)1 APP. Comparison is with the growth area of a control test containing no APP Average growth area reduction percentage APP level (mg l)1)

Bacillus cereus 204 B. cereus Campden Listeria monocytogenes S23 L. monocytogenes 272 Lactobacillus sake A10 Brochothrix thermosphacta CRA7883 Staphylococcus aureus 5Æ720 Escherichia coli CRA109 E. coli S15 Pseudomonas fluorescens 3Æ756 Salmonella Typhimurium S29 Saccharomyces carlsbergensis CRA6413 S. cerevisiae ATCC9763 S. cerevisiae H78 S. cerevisiae var. paradoxus H103

4000

2000

1000

500

250

100 94Æ9 100 98Æ5 100 100 100 87Æ2 82Æ9 100 97Æ2 6Æ1 ) 6Æ9 28Æ2 70Æ2

88Æ1 67Æ0 90Æ4 69Æ2 98Æ9 75Æ7 97Æ4 38Æ3 39Æ2 100 57Æ7 ) 9Æ5 ) 60Æ5 2Æ9 46Æ7

51Æ1 39Æ2 54Æ4 37Æ1 97Æ2 26Æ3 63Æ5 18Æ1 18Æ8 98Æ5 27Æ7 ) 58Æ8 ) 80Æ3 ) 1Æ5 7Æ2





the bacteria tested. At a level of 4000 mg l)1, APP caused total inhibition of certain strains of Bacillus and Listeria, as well as Lact. sake, B. thermosphacta, Staph. aureus and Ps. fluorescens. Inhibition was observed mainly as an extension of the lag phase, but reduction of growth rate and ⁄ or final numbers also occurred (examples shown in Figs 2–5). In another Bioscreen run (data not shown in Table 1), 1000 mg l)1 APP caused total inhibition for 24 h at 30C of Ps. fluorescens strains 1331 and 327. A level of 2000 mg l)1 APP caused total inhibition of Ps. fluorescens 1460 and Y. enterocolitica S16. APP at 3000 mg l)1 caused total inhibition of B. cereus 3.046, B. subtilis Campden, and L. monocytogenes strains 358, Scott A and F6861. A level of 4000 mg l)1 APP caused total inhibition of Ent. aerogenes 10102, E. coli strains 2.083, CRA1593, CRA161 and CRA92, and Salmonella sp. S19. At the highest level tested (4000 mg l)1), APP inhibited the growth of three of the four yeasts tested. At lower levels, however, APP promoted the growth of certain of the test yeast strains, resulting in a greater growth area compared with the control. This was due to a reduction in lag time and increased growth rate, with higher final cell numbers (Fig. 6). No growth inhibition was observed at the highest level tested (2000 mg l)1 APP) against any of the Clostridium species tested. In the presence of 0, 1000 and 2000 mg l)1 APP, there was no reduction in colony size or difference in colonial morphology observed for the four fungal strains tested.

701

Investigation of cidal activity After 2 h incubation at ambient temperature in buffer with or without 4Æ93 mg ml)1 APP, there was no evidence of a cidal effect caused by APP against L. monocytogenes; the counts in the presence of APP were not significantly lower than those of the control test. The E. coli counts in the presence of APP dropped by 1Æ5 logs compared with the control test, indicating a possible mild cidal effect caused by high levels of APP. DISCUSSION This study showed that APP had a regulatory effect on the growth of a range of micro-organisms. APP samples prepared either enzymatically or chemically gave identical results in their effect towards bacteria, yeasts and moulds. Chemical analysis by NMR, TLC and HPLC further showed their identical structure. The improved APP preparation method involved a direct incubation of the Anthracobia cell-free extract that had all the APP-forming enzymes with dextrins and selective extraction of APP with 80% acetonitrile at the end of the reaction. As chemical synthesis of APP involved nine steps from glucose as the starting material with low yield (Andersen et al., unpublished results), enzyme-prepared APP was used for the majority of the tests reported here.

Fig. 2 Growth curves of spores of Bacillus cereus Campden at 30C for 24 h in the presence of 0 ppm APP (·); 500 ppm APP (h); 1000 ppm APP (…); 2000 ppm (r); 4000 ppm (—). Turbidometric data were measured using the Bioscreen C ª 2002 The Society for Applied Microbiology, Journal of Applied Microbiology, 93, 697–705


Fig. 3 Growth curves of Listeria monocytogenes S23 at 30C for 24 h in the presence of 0 ppm APP (·); 500 ppm APP (h); 1000 ppm APP (…); 2000 ppm (r); 4000 ppm (—). Turbidometric data were measured using the Bioscreen C

Fig. 4 Growth curves of Salmonella Typhimurium S29 at 30C for 24 h in the presence of 0 ppm APP (·); 500 ppm APP (h); 1000 ppm APP (…); 2000 ppm (r); 4000 ppm (—). Turbidometric data were measured using the Bioscreen C

APP is a very weak acid with a pKa of 9Æ5 (Ka ¼ 3Æ162 · 10)10) (Fig. 7) (Shafizadeh et al. 1978). This immediately suggests that its mechanism of antimicrobial activity could be similar to other organic acids, such as acetic, benzoic, lactic, sorbic and propionic acid, all of which are used in food preservation (Eklund 1989). Generally, these acids are inhibitory towards both bacteria and fungi, but acetic and lactic acids have much less

reported activity against yeasts and moulds (Bogaert and Naidu 2000). In the present study, it was shown that APP is generally inactive against fungi except at high levels (>4000 mg l)1). The pKa of organic acids used in food preservation is much lower than APP. These acids are usually more active in their undissociated state, due to their ability to dissociate within the cytoplasm disrupting the proton motive force leading to cell starvation. These



703

Fig. 5 Growth curves of Escherichia coli S15 at 30C for 24 h in the presence of 0 ppm APP (·); 500 ppm APP (h); 1000 ppm APP (…); 2000 ppm (r); 4000 ppm (—). Turbidometric data were measured using the Bioscreen C

Fig. 6 Growth curves of Saccharomyces cerevisiae ATCC9763 at 25C for 72 h in the presence of 0 ppm APP (·); 2000 ppm (r); 4000 ppm (—). Turbidometric data were measured using the Bioscreen C

molecules also have several other specific mechanisms of activity (Bogaert and Naidu 2000; Marshall et al. 2000; Sofos 2000). Structurally, APP has a keto and an enol function. These functional groups are unstable and are prone to hydration to

form the stable isomer ascopyrone T (Fig. 7). Ascopyrone T, which has been shown to be inefficient as an antimicrobial (Baute et al. 1993), was reported to accumulate in significant amount in certain truffles, such as Terfezia sp. (Baute et al. 1993). The conversion of APP to ascopyrone T



Fig. 7 The acidic nature of ascopyrone P and the possible mechanism of ascopyrone P inactivation by hydration to ascopyrone T

could be an inactivation mechanism, especially at elevated temperature. APP could be regarded as a fungal metabolite involved in defence. This hypothesis is supported by the evidence of the present study, in which APP activity was evident against a wide range of Gram-negative and Gram-positive bacteria. APP may also have a role in regulating fungal growth. This study showed inhibition of certain fungal strains at high APP levels, but lower APP levels sometimes promoted yeast growth. APP is produced under biotic and abiotic stress (Baute et al. 1993). Clearly, under such conditions, the fungi are more vulnerable to bacterial attack. It is still, however, unclear how such stress factors induce the formation of APP. The regulation can be at the level of glucan lyase or dehydratase or both, but under certain conditions, AF production was promoted but not APP formation. It is still not understood why surface culture of A. melaloma enhanced APP production while submerged culture was unproductive, even under various culture conditions (different medium composition and different culture time). APP is an interesting antimicrobial as it can easily be prepared industrially from renewable raw material; in this study it was prepared by incubating dextrins with a cell-free extract of A. melaloma that had both glucan lyase and dehydratase activity. This is in contrast with many other fungal and plant secondary metabolites, whose synthesis routes are either unclear or often involve multiple enzymecatalysed steps. Based on the results described above, APP can also be classified as a natural antimicrobial. Such compounds are of great interest in food preservation. There are few preservatives that fit the description of natural

antimicrobials; nisin and natamycin are two that are authorized for use in food (Delves-Broughton 1990; Davidson and Doan 1993; Thomas et al. 2000). The continuing trend for ready-prepared or processed foods, together with a ‘time-starved’ working population of men and women, have resulted in a huge demand for natural and convenient foodstuffs including soups, snacks and meals with good shelf-lives at chill or ambient temperature storage. There has also been a steady increase in incidents of food poisoning associated with this fast food trend. Pathogens such as Salmonella, E. coli and Listeria are of particular concern. The present study indicates that APP may have potential as a food preservative in raw or cooked foods to counter the growth of Gram-negative and Gram-positive bacteria. Based on the in vitro results described above, concentrations of 500–4000 mg kg)1 would be required. For an APP dosage of 2000 mg kg)1, the added cost per kilogram of food product would be approximately 0Æ2 euro cents. APP has a production cost of approximately 10 euros. Further tests are necessary to determine effective levels in food models and food products. Food components may reduce effectiveness, necessitating higher addition levels; conversely, additional preservation hurdles such as acidity, salt content and low temperature could increase APP efficacy. Before APP can be considered for food use, its effect on the organoleptic properties of food products needs to be investigated and, most importantly, its safety for human consumption must be established by toxicological testing. ACKNOWLEDGEMENTS The authors are thankful for the excellent technical assistance of Mr Rene Brylle Svensson and Mr Ole Orum Kristiansen. REFERENCES Baute, M.A., Baute, R. and Deffieux, G. (1988) Fungal enzymic activity degrading 1,4-a-D-glucans to 1,5-D-anhydrofructose. Phytochemistry 27, 3401–3404. Baute, M.A., Deffieux, G., Vercauteren, J., Baute, R. and Badoc, A. (1993) Enzymic activity degrading 1,4-a-glucans to ascopyrones P and T. Pezizales ad Tuberales. Phytochemistry 33, 41–45. Bogaert, J.C. and Naidu, A.S. (2000) Lactic acid. In Natural Food Antimicrobial Systems ed. Naidu, A.S. pp. 613–636. Boca Raton: CRC Press. Davidson, P.M. and Doan, C.H. (1993) Natamycin. In Antimicrobials in Foods ed. Davidson, P.M. and Branen, A.L. pp. 395–407. New York: Marcel Dekker. Delves-Broughton, J. (1990) Nisin and its uses as a food preservative. Food Technology 44, 100–117. Eklund, T. (1989) Organic acids and esters. In Mechanisms of Action of Food Preservation Procedures ed. Gould, G.W. pp. 161–200. London: Elsevier Applied Science.



Marshall, D.L., Cotton, L.N. and Bal’a, F.A. (2000) Acetic acid. In Natural Food Antimicrobial Systems ed. Naidu, A.S. pp. 661–688. Boca Raton: CRC Press. Shafizadeh, F., Furneaux, R.H., Stevenson, T.T. and Cochran, T.G. (1978) 1,5-anhydro-4-deoxy-D-glycero-hex-1-en-3-ulose and other pyrolysis products of cellulose. Carbohydrate Research 67, 433–447. Skytta, E., Haikara, A. and Mattila-Sandholm, T. (1993) Production and characterization of antibacterial compounds produced by Pediocococcus damnosus and Pediococcus pentosaceus. Journal of Applied Bacteriology 74, 134–142. Sofos, J.N. (2000) Sorbic acid. In Natural Food Antimicrobial Systems ed. Naidu, A.S. pp. 637–659. Boca Raton: CRC Press. Stevenson, T.T., Stenkmap, R.E., Jensen, L.H., Cochran, T.T., Shafizadeh, F. and Furneaux, R.H. (1981) The crystal structure of 1,5-anhydro-4-deoxy-D-glycero-hex-1-en-3-ulose. Carbohydrate Research 90, 319–325. Thomas, L.V., Clarkson, M.R. and Delves-Broughton, J. (2000) Nisin. In Natural Food Antimicrobial Systems ed. Naidu, A.S. pp. 463–524. Boca Raton: CRC Press.

705

Yu, S., Ahmad, T., Pederse´n, M. and Kenne, L. (1995) a-1,4-Glucan lyase, a new class of starch ⁄ glycogen degrading enzyme. III. Substrate specificity, mode of action, and cleavage mechanism. Biochimica et Biophysica Acta 1244, 1–9. Yu, S., Bojsen, K., Svensson, B. and Marcussen, J. (1999) a-1,4Glucan lyases producing 1,5-anhydro-D-fructose from starch and glycogen have sequence similarity to a-glucosidases. Biochimica et Biophysica Acta 1433, 1–15. Yu, S., Christensen, T.M.I.E., Kragh, K.M., Bojsen, K. and Marcussen, J. (1997) Efficient purification, characterization and partial amino acid sequencing of two a-1,4-glucan lyases from fungi. Biochimica et Biophysica Acta 1339, 311–320. Yu, S., Olsen, C.E. and Marcussen, J. (1998) Methods for the assay of 1,5-anhydro-D-fructose and a-1,4-glucan lyase. Carbohydrate Research 305, 73–82.
