The properties of Bacillus coagulans and of other bacilli that contaminate paper and paperboard manufacturing processes were investigated under simulated industrial conditions. Nisin (0.05 to 0.125µg ml^-') blocked growth of indigenous bacilli that contaminate sizing starches. B. coagulans starch isolates, B. licheniformis, B. amylolique­faciens, and B. stearothermophilus grew at >50°C in industrial starch and produced a-glucosidase and cyclodextrins. The industrial isolates and reference strains of B. amyloliquefaciens, B. cereus, B. coagulans, B. flexus, B. licheniformis, B. pumilus, B. sporothermodurans, B. stearothermophilus and Alicyclobacillus acidoterres­tris were inhibited by <0.125 pg of nisin on agar. B. coagulans and B. stearothermophilus were similarly inhibited by <0.025 pg of nisin ml^-' and by 3 pg of the biocide DBNPA ml^-' in industrial starch. B. licheniformis and B. amyloliquefaciens strains were less sensitive. About 40% of nisin added to starch was retained after cooking. Fifty percent of the nisin remained active after 11 h of storage at 60°C. The results show that nisin has potential as a preservative for modified industrial starches. Journal of Industrial Microbiology& Biotechnology (2001) 26, 107-114.

Keywords: Bacillus; starch; biodeterioration; paper mill; nisin; thermotolerant

INTRODUCTION

Starches are bulk raw materials for the manufacture of food, paper, textiles, pharmaceuticals, and adhesives. The paper industry in North America uses over 2 million tons per year of modified starches for surface sizing, as wet-end additive and in coating [17,201. In Finland the annual use of starch is 250 000 tons by the paper industry and 20 000 tons by the food industry.

Microbially deteriorated sizing starch impairs the printing quality of office and journal papers and endangers the hygienic quality of food-packaging paper and board [27]. Although enzymatic liquefaction of starch has been studied intensively [ 5,9,26,43 ] , the microbiological events leading to hydrolysis and spoilage of industrial starches and their prevention are poorly understood [ 11,27,30,341. There are few published papers on the microbial species capable of spoiling starch in a paper machine environment [23,40-42]. The preservation strategies using biocides to prevent spoilage are therefore chosen by trial and error.

Biocides are toxic by definition [35] and may be sensitising [ 8,10 1. Surface - sizing starch is spread on paper or board at the dry end of the machine, representing the immediate interface to consumers' skin or to food in types of paper having no polyethylene extrusion coating. Understanding of microbial starch spoilage and the antimicrobial sensitivities of the spoilage organisms will facilitate the development of consumer-friendly starch-preserva­tion strategies.

This paper describes aerobic spoilage microorganisms in industrial starches and their abatement using nisin.

MATERIALS AND METHODS

Bacterial strains, inoculation and media

Aerobic starch-degrading isolates were picked from nutrient agar (LAB M, Bury, England) or from starch nutrient agar (nutrient agar with 0.2 wt.% of soluble starch, Merck, Darmstadt, Germany) incubated at 28 or 50°C. Origins of isolates and reference strains used in this study are listed in Table 1.

The bacterial density of the cultures was measured with a laser nephelometer (Minineph^®, The Binding Site, Birmingham, UK) calibrated using Polybead^® polystyrene 3 - µm (B. cereus) or 1-µm (other bacilli) microspheres (PolySciences, Washington). Indus­trial starch medium contained chemically oxidised surface-sizing potato starch (20 g 1^-' ), yeast extract ( 0.1 g 1^-' ), Na2HPO4 (16 mg 1^-' ), malic, citric and lactic acids (50 mg 1^-' each) and Tween 80 (0.1 vol.%), pH 6.1. All chemicals used were obtained from local sources and are of analytical quality unless otherwise stated.

Inhibition assays

The stock solution of commercial nisin (Sigma, St. Louis, MO), 50 µg (2000 IU) active substance ml^-' in 0.02 M HCl with 0.2% Tween 80, was stored at -20°C in the dark and neutralised with one volume of 0.02 M NaOH before dilution with water for use. 2.2 - Dibromo - 3 - nitrilopropionamide (DBNPA) was used as the tradeware Daracide 7819 (Betz-Dearborn, Helsinki, Finland) containing 9% to 12% w/vol of the active substance.

The resazurin reduction assay [ 1 ] was performed on a microtiter plate (96 wells) with 330 µl of inoculated (10⁶ cells ml^-') industrial starch medium and 20 µl of resazurin (sodium salt, 0.05 mg ml -' in water, ICN Biomedicals, Ohio) per well supplemented with nisin, DBNPA, EDTA, or nil as indicated. The plate was incubated stationary at 45°C and the fluorescence of the emerging semireduced form ofresazurin (resorufin) was recorded at intervals of 15 min using the kinetic fluorometer Fluoroskan Ascent (Labsystems, Helsinki, Finland) and the filter pair 544/590 x4m.

Table 1 The bacterial strains used and their starch depolymerization abilities

For the kinetic turbidometric assay a 100-well microtiter plate was filled as above except that no resazurin was used. Optical density was measured at 45°C using Bioscreen C (Labsystems) and a wide-band filter (420-580 nm). The plate was shaken before each measurement, once an hour.

The inocula were grown and the agar diffusion assay was carried out at 37 or at 50°C on brain heart infusion (BHI) agar (Difco Laboratories, Detroit, MI) pH 7.0, except for A. acidoterrestris, which was grown on potato dextrose agar (PDA, Difco), pH 3.5. An overnight culture of the test bacterium, 10⁶ cells in 3.5 ml semisolid agar with 0.1 vol.% Tween 80, was spread on agar plates (90 mm) with 0.1 vol.% Tween 80. Nisin solution (5 pl) was placed on the inoculated agar surface and the plates were read after 24hat37orat50°C.

Analytical protocols

Whole-cell fatty acid analysis was performed as described by Vaisanen et al [411 except that the biomass was grown at 50°C. Sherlock Microbial Identification System with the aerobic TSBA library version 3.90 (MIDI, Newark, DE) was used to identify the fatty acids. Partial 16S rRNA gene sequences comprising 488 nucleotide positions at the 5^'-end of the gene were determined as described previously [ 311.

ATP content was measured using the ATP biomass kit (1243 - 118, BioOrbit, Turku, Finland) and a BioOrbit 1253 luminometer. Thermotolerant viable counting was done on plate count agar (PCA, Biokar Diagnostics, Beauvais, France) at 50°C; plates were read after 2 days.

To measure the a-gucosidase activity of the Bacillus strains, industrial starch medium was inoculated with 10⁶ bacteria ml^-1 and incubated for 18 h at 37 or at 50°C. a-Glucosidase activity of the cultures was measured using methylumbelliferyl-a­glucoside (MUF; Sigma) as substrate (1.4 mM final concentra­tion). The culture (50 µl) and the MUF-substrate stock (10 µl) were dispensed in each well of the 96-well microtiter plate and incubated at 55°C. After 15 min pH was adjusted to 8.9 by adding 10 µl of 1 M Tris-HCl buffer to each well, and the fluorescence was measured using the filter pair 355/460 run in the Fluoroskan Ascent. For the blank the fluorescence was measured without incubation.

Starch depolymerization was tested on agar containing 15 g soluble potato starch ( Sigma S -4251) , 2.5 g NaNO3, 1 g KzHPO4, 0.6 g MgSO4^.7H20, 0.1 g CaC12^.2H20, 0.1 g NaCl, 1 mg FeC13 and 10 g agar per liter, pH 7. The plates were incubated for 7 days at 37, 50 or 55°C and cooled to +4°C. Depolymerization of starch was visible as a clearing zone around the colony and cyclodextrin

accumulation as a zone of white precipitate inside the clearing zone (van der Veen and Dijkhuizen, personal communication).

Nisin activity in industrial starch medium (pH 5.5) was bioassayed as described by Wahlstrom and Saris [44] using L. lactic subsp. lactic LAC182. Starch was sampled in 5 mM HCl (1:1), stored (<24 h) at room temperature in the dark and analysed for residual raisin. A standard curve was prepared using commercial nisin (Sigma) in industrial starch medium.

Table 2. Inhibition zones of selected starch-deteriorating bacilli toward nisin

RESULTS

Twenty-six thermophilic starch-degrading isolates of Bacillus coagulans were isolated from cooked surface-sizing starches, sampled from 10 storage tanks and feed tanks in paper mills. They grew well at 50°C and poorly at 28°C. The isolates were identified as B. coagulans based on the whole-cell fatty acid compositions compared to the type strain (all strains grown at 50°C). The partial 16S rRNA gene sequences (488 nucleotides) determined for the strains S11, S29 and SB2b-50-1 shared 100% similarity with that determined for B. coagulans type strain DSM1^T.

Depolymerization of starch by selected B. coagulans strains and other bacilli relevant to the hygiene of food-grade paper­board is shown in Table 1. All strains that depolymerized starch

produced cyclodextrins. B. coagulans, B. licheniformis, B. stearothermophilus, and B. amyloliquefaciens reference strains and papermaking-related isolates grew well at 50T and most of them also had a-glucosidase activity.

Figure ] The metabolic inhibition by nisin, by 2,2-dibromo-3-nitrilopropionamide (DBNPA), and by EDTA of B. coagulans S11 (A) and of B. stearothermophilus DSM1550 (B). Cooked 2% industrial starch medium (pH 6.1) was inoculated with 10⁶ cells ml ^-1 and incubated without shaking at 45°C. The curves represent the time course of resazurin reduction as averages of four parallel measurements. Reduction of resazurin into resorufin was recorded by its fluorescence at 590 nm. The fluorescence curves obtained after treatments with 0.125 or with 0.5 µg of nisin ml^-1 (not shown), with or without EDTA, were similar to those with 0.025 µg of nisin ml -1 ( - ) No added inhibitor; (o) 0.0025 µg nisin ml^-1; (●) 0.0025 µg nisin ml ^-1 + 2 µg₁ EDTA ml - ¹• ( q) 0.025 µg nisin ml ^-1; (n) 0.025 µg nisin ml^- + 2 pg EDTA ml^{- 1}; (∆) 0.5 µg DBNPA ml^-(▲) 3 µg DBNPA ml ^-1; ( x ) 2 pg EDTA ml ^-1; ( .... ) no inoculum.

Table 2 records the presence or absence of inhibition zones by nisin of the studied bacilli from paper-industry -related sources, food and food-poisoning incidents. Starch isolates of B. coagulans (BHl agar, pH 5.8), B. coagulans type strain, B. flexus, B. sporothermodurans and A. acidoterrestris (PDA, pH 3.5) were very sensitive to nisin whereas B. cereus, B. pumilus,

Table 3 Metabolic delay of selected starch- degrading bacilli by nisin, DBNPA, and EDTA in cooked 2% industrial starch medium using resazurin reduction as an indicator (cf., Figure 1)

Figure 2 Growth inhibition of starch-degrading B. coagulans S11 by nisin, DBNPA and EDTA. The figure shows the time course of optical density increase of microliter plate cultures of the strain S11 in cooked 2% industrial starch (pH 6.1) at 45°C as averages of five parallel wells. Turbidity curves obtained after treatments with 0.125 or with 0.5 pg nisin ml^-1 (not shown), with and without EDTA, were similar to those with 0.025 µg of nisin ml^-1. Symbols are as in Figure 1.

B. licheniformis, and B. amyloliquefaciens were less sensitive. Three industrial contaminant strains of B. cereus and the type strain of B. licheniformis were only inhibited transiently by nisin.

Metabolic inhibition by raisin and by DBNPA, a commonly used biocide, was studied in thermotolerant starch-degrading strains of Bacillus. Figure 1 shows the metabolic sensitivities of an industrial starch contaminant, B. coagulans S11 (Figure IA), and of B. stearothermophilus DSM1550 (Figure 1B) in industrial starch medium. With no added inhibitor, B. coagulans S 11 showed a fluorescence peak, indicating reduction of resazurin, 4.5 h after inoculation with 10⁶ cells ml^-', a usual level of contamination in deteriorating starches in the process lines. Three micrograms of DBNPA ml -' was required to attenuate the metabolism of B. coagulans S 11 for > 18 h. The same effect was achieved with 0.0025 µg of nisin ml -' when 2 µg of EDTA ml^-' was present. In the absence of EDTA, 0.025 µg of raisin ml^-' was needed. EDTA alone (2 µg ml^-') postponed emergence of the fluorescence peak of B. coagulans S11 for 3 h. Treatments with >0.025 µg nisin ml^-', with or without EDTA, completely blocked the metabolism of B. coagulans S11 and of B. stearothermophilus DSM1550.

The sensitivities of the metabolism of B. coagulans S 11 (Figure IA) and five other Bacillus strains to nisin and DBNPA in industrial starch medium are compiled in Table 3. It shows that 0.0025 to 0.025 µg nisin ml -' blocked the metabolism of all B. coagulans strains during the measurement period of 18 h, whereas B. licheniformis and B. amylolique­faciens were less sensitive. The industrial B. coagulans isolates were as sensitive as the type strain. The data in Table 3 and Figure 1 thus show that 0.0025 to 0.025 µg raisin ml^-' were equally or more effective than 3 µg of DBNPA ml^-' against all B. coagulans and B. stearothermophilus strains. Industrial contaminant strains of B. licheniformis and B. amyloliquefaciens were equally sensitive to 0.5 µg of nisin ml^-' and to 3 µg of DBNPA ml ^-1.

Figure 2 shows growth curves of B. coagulans S11 in industrial starch medium with or without added inhibitors. The observed increment of optical density in this medium was small, 0.1 O.D. units in 18 h, indicating that the medium, which is low

Table 4 Growth inhibition by nisin, DBNPA, and EDTA of selected bacilli in cooked 2% industrial starch medium (inoculurn 10⁶ cells ml^-', pH 6.1, 45°C)

Table 5 Viable counts of the Bacillus cultures (PCA 2 days, 50°C) 18 h after inoculation in industrial starch medium

in nutrients and already contained an unknown amount and type of biocide supplied by the starch manufacturer, supported only slow growth (µ<0.5 h^-1). The presence of >0.0025 µg nisin ml- ¹ blocked growth. Results of tests similar to those presented in Figure 2 were obtained also for other industrial contaminant strains and for reference strains of B. coagulans, B. licheniformis and B. amyloliquefaciens, as summarised in Table 4. B. stearothermophilus DSM1550 did not grow measurably under the conditions used. The areas beneath the growth curves of four B. coagulans strains were effectively diminished by 0.0025 to 0.025 µg nisin ml^-1, compared to uninhibited controls. B. licheniformis and B. amyloliquefaciens strains were inhibited by >0.125 µg nisin ml^-1. Inhibition by raisin of the B. coagulans type strain but not of the other strains was enhanced by 2 µg EDTA ml^-1.

The effects of nisin and DBNPA on the ATP contents and on cultivability of the contaminant bacilli were measured after 18 h incubation. Nisin (0.0025 to 0.025 µg ml ^-1) suppressed viable counts (Table 5) of the tested strains of B. coagulans and of B. stearothermophilus. For B. licheniformis and B. amyloliquefaciens strains, > 0.5 µg nisin ml ^-1 was required to obtain the same level of inhibition as by 3 µg of DBNPA ml^-1. Industrial contaminant B.

coagulans cultures (strains S11, S29, SB2b-50-1) lost >99% of ATP when 0.025 µg nisin ml^-1 was present. Paperboard contaminant B. licheniformis TSP 29a and B. amyloliquefaciens TSP55 lost 88% and 91% of ATP, respectively, when 0.5 µg nisin ml ^-1 was present. Nisin sensitivity of the strains measured by ATP assay thus supported results obtained by the fluorometric resazurin reduction test. This was expected as both methods reflect effects on energy metabolism.

The conclusion from the results presented above is that the starch-degrading thermotolerant bacilli were effectively inhibited by nisin under conditions simulating those prevailing in the paper industry. To see the effect of nisin on a mixed culture situation such as is likely to occur in the industry, cooked surface-sizing starches from storage tanks (80°C) and from feed tanks (46-65°C) were sampled (Table 6). These contained the biocides used by the mills at the time of sampling (information on the effective substances or concentrations not available). Three (A, B, C) of the 10 samples, taken during 4 months, contained 50-400 cfu of thermotolerant (50°C) bacteria ml^-1. Nisin (0.05 or 0.125 µg ml - ¹ ) effectively suppressed the thermotolerant colony counts of these during 24 h of incubation in the laboratory (Table 6). This is sufficient for industrial application, as the storage time of cooked starches at the mills is usually only a few hours.

Because the biocides used to preserve surface-sizing starches in paper mills gradually lose their activity, they are repeatedly dosed, i.e., into the starch slurry, into the storage tank of the cooked starch, and into the feed tank. To determine the stability of nisin, industrial starch medium containing 0.025 µg nisin ml^-1 was "cooked" (5 min, 90°C), stored at 60 or 80°C and the residual nisin activity measured. Calibration showed that the accuracy of the assay in industrial starch was optimal at assay dilutions of 0.001 to 0.01 N of nisin ml^-1 (Figure 3A). The persistence of active nisin in starch during cooking and storage at 60 or 80°C was measured. About 40% of the added nisin (1 N ml ^-1) remained active after cooking, 50% of that remained active after 11 h of storage at 60°C, and somewhat less after storage at 80°C (Figure 3B).

Table 6 Effects of nisin, DBNPA, and EDTA on the growth of indigenous bacteria in surface-sizing starches sampled from the feed tanks (46-65°C) of paperboard machines

Biocide-treated starches were sampled at the mills. The effective substances were routinely rotated at intervals of 2 to 4 weeks in the mills. Total thermotolerant colony counts (PCA 2 days, 50°C) were analysed in duplicate and the average is shown. The duplicates deviated <0.2 log units. ^aNot tested.

Figure 3 Assay and survival of nisin in cooked industrial starch. (A) Calibration curve for nisin obtained with L. lactic subsp. lactic LAC182 in industrial starch medium. The emitted luminescence by the strain LAC182 was measured after 3 h incubation with known amounts of nisin. 1 IU = 0.025 µg of active substance. (B) Residual nisin activity after storage at 60 or 80°C.

DISCUSSION

Bacterial contaminants are not evenly distributed inside paper and paperboard. Confocal laser scanning microscopy analyses have shown that the greatest number of microbes is located just underneath the polyethylene extrusion coating, and likely origi­nated from surface- sizing starch [ 38 1. The microbiological quality of surface- sizing starch is thus important for the hygienic quality of food-packaging paper and board.

In this study we used several parallel methods to test the sensitivity of paper machine and starch contaminants to antimicrobials. The fluorometric resazurin-resorufin method was superior to the other methods used. This method is targeted to detect metabolic activity, indicated by reduction of resazurin. Resazurin becomes fluorescent at the semireduced state, indicat­ing accumulation of NADH inside cells. The fluorometric resazurin-resorufin test measures bacterial metabolic activity rather than growth, and was therefore suitable for food-industry applications [1,191. Our results show that it was useful for detecting starch spoilage in the paper industry because it is possible to observe emergence of the fluorescence peak long before any significant increase in turbidity. Compared to plate counting, the fluorometric test was less laborious and shortened the time required for spoilage detection from several days to less than 1 day. ATP determination was about equal in sensitivity but more laborious.

Turbidometry has been used succesfally in inhibition studies of bacilli from paperboard in process waters [29] but its use for monitoring microbiological contamination in starch is seriously limited by the turbidity of most industrial starches. Similar problems were observed with paper industry mineral pigments [21 ]. The fluorometric resazurin reduction-based assay may also be an option for the microbiological quality control of other highly turbid raw materials like pulps, resin sizers, neutral sizers, and latexes.

The spoilers of starch in the paper industry are aerobic sporeforming bacteria [40,42]. Bacterial spores in starch survive cooking and storage at 60-80°C and may proliferate in the machine feed tank maintained at 45-65°C. B. amyloliquefaciens, B. coagulans, B. licheniformis, and B. stearothermophilus grow at 50-55°C and produce heat-stable starch-hydrolysing enzymes [9,28,40]. B. circulars and B. subtilis rarely grow at >50°C [28] but do produce heat-activable a-amylases and a-glucosidases with optimum temperatures of 50 to 90°C [9,40,43 1. Formation of cyclodextrins from starch by cyclodextrin glycosyltransferases (CGTases) and their subsequent uptake and intracellular degrada­tion by cyclodextrinases (CDases) is an alternative pathway to the combined actions of amylases and exo-glucosidases, especially among different bacilli [ 9,15,26,43 1.

B. coagulans was the most common contaminant in the surface - sizing starches analysed in this study. Cyclodextrins were produced by all starch depolymerizing isolates from paper industry -related materials, and also by the other tested strains belonging to the species B. amyloliquefaciens, B. cereus, B. coagulans, B. flexes, B. licheniformis, or B. stearothermophilus. The results indicate that cyclodextrin production is an important pathway for degradation of starch in the paper industry environment. a-Glucosidase activity was also commonly found.

Bacillus species are difficult to eliminate from the industrial process environment using biocides [361. Nisin (E234) is an antimicrobial peptide produced by Lactococcus lactic subsp. lactic. It is authorised for food preservation in more than 50 countries [61 and it has been used to prevent spore outgrowth, e.g., in processed cheese, cheese spreads, canned vegetables and other heat­processed foods [ 6,7 1. Heating sensitizes spores to nisin [ 6,7 1.

Nisin in concentrations of 0.0025-2.5 µg ml ^-1 (=0.1-100 N ml^-1) is useful for inhibiting different bacilli [Refs. [2,16,32], this paper]. Spores of thermophilic bacteria such as Bacillus stearothermophilus and B. coagulans have been reported to be particularly sensitive [ 6,13,33 1. Both vegetative cells and spores of bacilli are sensitive to nisin [7,32].

Our results show that 0.05-0.125 µg of nisin ml^-1 (2-5 N ml - ¹ ) effectively preserved surface- sizing starches sampled from paperboard processes. Pure cultures of B. coagulans and B. stearothermophilus (industrial starch contaminants and reference strains) were even more sensitive. The concentration of nisin required to inhibit B. licheniformis and B. amyloliquefaciens in industrial starch (0.5 µg ml - 1) was close to that reported to be inhibitory to B. licheniformis spores in milk (25 N ml - ¹) [ 18 1. Our results show that a major amount of nisin remained active under conditions simulating cooking and storage of preserved starch in the industry.

Although nisin inhibited all bacilli tested in the agar diffusion assay, B. cereus strains (type strain, industrial, environmental and food poisoning isolates) were among the least sensitive. Maintain­ing high temperature during storage of cooked starches ( > 50°C) is thus important for controlling growth of B. cereus and other mesophilic bacilli.

A. acidoterrestris and B. sporothermodurans are relatively new food contaminant organisms and have not yet been reported in packaging materials. B. sporothermodurans produces exceptionally heat-resistant endospores [22] and the acid-tolerant and heat­resistant A. acidoterrestris is a worldwide spoilage problem in pasteurized and heat-treated juices [4,14]. We show here that these species were also effectively inhibited by raisin, analogous to the findings for A. acidoterrestris in fruit juices [ 16 1.

Blackburn [31 observed that the inhibitory effect of nisin against vegetative Gram-positive bacteria at pH >5 was improved by the presence of 0.01-3 mM EDTA (= 3 -1000 µg ml - ¹) . We found synergism with 2 µg EDTA ml^-1 under paper machine­simulating conditions with one Bacillus strain out of seven strains tested. Higher EDTA concentrations were not tested because they are not technically applicable in paper mills.

Our results show that nisin has potential for preservation of starch in industrial processes (Salkinoja-Salonen and Pirttijarvi, patent applied for).

ACKNOWLEDGEMENTS

We thank T. Ali-Vehmas for advice with the resazurin reduction tests, L. Dijkhuizen and B. van der Veen (University of Groningen) with the cyclodextrin production test and M. Kahkonen with the a-glucosidase assay. We thank H. Ketola (Raisio Chemicals) for help in acquiring the figures of starch usage in Finland. The ABS Graduate School and the Academy of Finland (project 38632) gave financial support.

REFERENCES

1 Ali-Vehmas T, M Louhi and M Sandhohn. 1991. Automation of the resazurin reduction test using fluorometry of microtitration trays. J Vet Med B 38: 358-372.

2 Beuchat LR, MRS Clavero and CB Jaquette. 1997. Effects of nisin and temperature on survival, growth and enterotoxin production character­istics of psychrotrophic Bacillus cereus in beef gravy. Appl Environ Microbio163: 1953-1958.

3 Blackburn P. 1997. Nisin compositions for use as enhanced, broad range bacteriocides. US Patent 5,691,301.

4 Brown. 1995. New microbiological spoilage challenges in aseptics: Alicyclobacillus acidoterrestris spoilage in aseptically packed fruit juices. In: Ohlsson T (Ed), Proceedings of the International Symposium "Advances in Aseptic Processing and Packaging Technologies," Copenhagen, Denmark, September 11-12, 1995.

5 Crabb WD and JK Shetty. 1999. Commodity scale production of sugars from starches. Curr Opin Microbiol 2: 252-256.

6 Delves-Broughton J, P Blackburn, RJ Evans and J Hugenholtz. 1996. Applications of the bacteriocin, nisin. Antonie van Leeuwenhoek 69: 193 -202.

7 De Vuyst, L and EJ Vandamme. 1994. Nisin, a lantibiotic produced by Lactococcus lactic subsp. lactic: properties, biosynthesis, fermenta­tion and applications. In: De Vuyst L and EJ Vandamme (Eds), Bacteriocins of Lactic Acid Bacteria: Microbiology, Generics and Applications. Blackie Academic and Professional, London, pp. 151-221.

8 Eriksson U, A Johanson and M Torlund. 1995. Risk assessment of slimicides, KEMI Report 9/95, The Swedish National Chemical Inspectorate, Solna, Sweden. 113 pp.

9 Fogarty WM and CT Kelly. 1990. Recent advances in microbial amylases. In: Fogarty WM and CT Kelly (Eds), Microbial Enzymes and Biotechnology, Elsevier Applied Science, London, pp. 71-132.

10 Fransway AE 1988. Sensitivity to Kathon CG: findings in 365 consecutive patients. Contact Dermatitis 19: 342-347.

11 Frieden A. 1940. The relation of starch to microbiological problems and methods for controlling bacterial growth. Paper Trade J 110: 233-236.

12 Hallaksela A-M, O Vaisanen and M Salkinoja-Salonen. 1991. Identification of Bacillus species isolated from Picea abies by physiological tests, phage typing and fatty acid analysis. Scand J Forest Res 6: 365-377.

13 Hurst A and DG Hoover. 1993. Nisin. In: Davidson PM and AL Branen (Eds), Antimicrobials in Foods, Marcel Dekker, New York, pp. 369-394.

14 Jensen N. 1999. Alicyclobacillus - a new challenge for the food industry. Food Aust 51: 33-36.

15 Kitahata S, M Taniguchi, SD Beltran, T Sugimoto and S Okada. 1983. Purification and some properties of cyclodextrinase from Bacillus coagulans. Agric Biol Chem 47: 1441-1447.

16 Komitopoulou E, IS Boziaris, EA Davies, J Delves -Broughton and MR Adams. 1999. Alicyclobacillus acidoterrestris in fruit juices and its control by nisin. Int J Food Sci Technol 34: 81- 85.

17 Maher SL and CW Cremer. 1987. Use of modified starches: paper industry. In: Wurzburg OB (Ed), Modified Starches: Properties and Uses. CRC Press, Boca Raton, Florida, pp. 213-227.

18 Mansour M, D Amri, A Bouttefroy, M Linder and JB Milliere. 1999. Inhibition of Bacillus licheniformis spore growth in milk by nisin, monolaurin and pH combinations. J Appl Microbio186: 311-324.

19 Mattila-Sandhohn T, T Ali-Vehmas, G Wirtanen, U Ronner and M Sandholm. 1991. Automated fluorimetry in quality control of pasteurized and ultra-high temperature -treated starch soup. IntJFood Sci Technol 26: 325-336.

20 Maurer HW and RL Kearney. 1998. Opportunities and challenges for starch in the paper industry. Starch/Starke 50: 396-402.

21 Mentu J, T Pirttijarvi, H Lindell and M Salkinoja-Salonen. 1997. Microbiological control of pigments and fillers in paper industry. In: Baker CF (Ed), The Fundamentals of Papermaking Materials. Transactions of the 11th Fundamental Research Symposium held at Cambridge, September 1997, Vol. 2. Pira International, Leatherhead, UK, pp 955-993.

22 Pettersson B, F Lembke, P Hammer, E Stackebrandt and FG Priest. 1996. Bacillus sporothermodurans, a new species producing highly heat-resistant endospores. Int J Syst Bacterio146: 759-764.

23 Pirttijiirvi TSM, TH Graeffe and MS Salkinoja-Salonen. 1996. Bacterial contaminants in liquid packaging boards: assessment of potential for food spoilage. JAppl Bacteriol 81: 445-458.

24 Pirttijarvi TSM, LM Ahonen, LM Maunuksela and MS Salkinoja-Salonen. 1998. Bacillus cereus in a whey process. Int JFood Microbiol 44: 31-41.

25 Pirttijarvi TSM, MA Andersson, AC Scoging and MS Salkinoja-Salonen. 1999. Evaluation of methods for recognising strains of the Bacillus cereus group with food poisoning potential among industrial and environmental contaminants. System Appl Microbiol 22: 133-144.

26 P6csi 1. 1999. Physiological and ecological evaluation of bacterial cyclodextrin glycosylhwisferases (CGTases). Biologia (Bratislava) 54: 603-616.

27 Poock SJ. 1985. Microbial contamination in the use of starch. Tappi J 68: 78-80.

28 Priest FG. 1989. Isolation and identification of aerobic endospore­forming bacteria. In: Harwood CR (Ed), Bacillus. Plenum, New York, pp. 27-56.

29 Raaska L, H-L Alakomi, M Salkinoja-Salonen and T Mattila-Sandhohn. 1999. Antagonistic activity of Staphylococcus siderophores and chemical biocides against Bacillus subtilis in a paper-machine environment. JInd Microbiol Biotechnol 22: 27-32.

30 Radley JA. 1976. Antiseptics and preservatives used in starch products. In: Radley JA (Ed), Starch Production Technology. Applied Science Publishers, London, pp. 99-108.

31 Rainey FA, N Ward-Rainey, RM Kroppenstedt and E Stackebrandt. 1996. The genus Nocardiopsis represents a phylogenetically coherent taxon and a distinct actinomycete lineage: proposal of Nocardiopsaceae fam. nov. IntJSystBacterio146: 1088-1092.

32 Ray B. 1992. Nisin of Lactococcus lactic ssp. lactic as a food preservative. In: Ray B and M Daeschel (Eds), Food Biopreserva­tives of Microbial Origin. CRC Press, Boca Raton, Florida, pp. 207-264.

33 Roberts CM and DG Hoover. 1996. Sensitivity of Bacillus coagulans spores to combinations of high hydrostatic pressure, heat, acidity and nisin. JAppl Bacterio181: 363-368.

34 Robertson L. 1994. Microbial growth in starch. PIMA Mag 76: 74-76.

35 Rossmoore HW 1995. Introduction to biocide use. In: Rossmoore HW (Ed), Handbook of Biocide and Preservative Use, Blackie Academic and Professional, London, pp. 1-18.

36 Russell AD. 1998. Assessment of sporicidal efficacy. Int Biodeterior Biodegrad 41: 281-287.

37 Salkinoja-Salonen MS, R Vuorio, MA Andersson, P Kampfer, MC Andersson, T Honkanen-Buzalski and AC Scoging. 1999. Toxigenic strains of Bacillus licheniformis related to food poisoning. Appl Environ Microbio165: 4637-4645.

38 Suominen 1, M-L Suihko and M Salkinoja-Salonen. 1997. Micro­scopic study of migration of microbes in food-packaging paper and board. JInd Microbiol Biotechnol 19: 104-113.

39 Turnbull PCB, JM Kramer, K Jorgensen, RJ Gilbert and J Melling. 1979. Properties and production characteristics of vomiting, diarrheal, and necrotizing toxins of Bacillus cereus. Am J Clin Nutr 32: 219-228.

40 Vaisanen O, S Elo, S Marano and M Salkinoja-Salonen. 1989. Enzymatic characterization of Bacilli from food packaging paper and board machines. JInd Microbiol 4: 419-428.

41 Vaisanen OM, E-L Nurmiaho-Lassila, SA Marano, MS Salkinoja-Salonen. 1994. Structure and composition of biological slimes on paper and board machines. Appl Environ Microbio160: 641-653.

42 Vaisanen OM, A Weber, A Bennasar, FA Rainey, H-J Busse and MS Salkinoja-Salonen. 1998. Microbial communities of printing paper machines. JAppl Microbiol 84: 1069-1084.

43 Vihinen M and P Mantsala. 1989. Microbial amylolytic enzymes. Crit Rev Biochem Mol Bio124: 329-418.

44 Wahlstrom G and PEJ Saris. 1999. A nisin bioassay based on bioluminescence. Appl Environ Microbio165: 3742-3745.

(order Full Text from publisher)