The salmonella bacteria are known pathogens causing food poisoning. Being intestinal bacteria, they readily spread in animal spaces, contaminating animal products. A very problematic situation prevails in many Central European countries, where e. g. over 90 % of all broilers may be contaminated by salmonella. Animals bearing salmonella and animal products contaminated by salmonella may spread these bacteria to spaces where foodstuffs are handled, stored or sold and also to homes. In Finland, very small amounts of salmonella have been found in animals used for food production.

During 1989-94, less than 5 % of the broiler production batches were salmonella positive.

In may countries, antibiotics are used to curb salmonella infections. However, widespread use of anti- biotics has certain risks. Microbe populations resis- tant to antibiotics are readily generated, creating difficulties in the chemotherapy of infections in hu- mans as well. It is possible to reduce the need for an- tibiotics by improving animals'immunity to salmonella either by vaccination or by a so-called competitive ex- clusion (CE) method. The CE method is based on the fact that the intestines of a newly born animal are exposed to infections, whereas the varied normal flora of an adult animal prevents pathogenic colonisation. In Fin- land, the CE method has been used to prevent the devel- opment of salmonella in broilers. To promote the devel- opment of a protecting normal flora in the intestines of chickens, the chickens have been given a CE prepara- tion containing an adult hen's intestinal flora as early as possible after being hatched.

Besides good production hygiene, the CE method has proved to be an alternative worth consideration in

fighting salmonella in broilers. But for the control of microbial contamination of the environment, such as the surfaces and equipment in breeding spaces and other spaces and the media used in breeding spaces, there is still hardly any other expedient than careful hygiene and, after contamination has occurred, heavy and expen- sive washing and disinfecting treatment. For example, the sanitation costs of a typical raw material stock amounts to hundreds of thousands of marks (FIM). Be- sides, even if sanitation is carefully planned, its successful implementation is very uncertain.

The object of the present invention is to eliminate the drawbacks described above A specific object of the present invention is to disclose a method for control of salmonella in food industry plants and processes, using protecting mi- crobes and microbial protective substances.

As for the features characteristic of the in- vention, reference is made to the claims.

The invention is based on investigations in which it has been established that Pseudomonas bacteria and antimicrobial compounds produced by lactic acid bacteria can be utilised in control of salmonella.

In food industry, all substantially dry pro- duction and processing spaces and/or production and processing equipment can be treated with a Pseudomonas bacterial preparation to prevent salmonella contamina- tion. All production and/or processing spaces in food industry can be treated with a preparation of lactic acid bacteria to prevent salmonella contamination. Pro- duction and/or processing spaces comprise the surfaces in process spaces, storage spaces, transport spaces, animal breeding spaces and/or other spaces, such as e. g. large cookhouses. Production equipment comprises process materials, media and/or other objects. Substan- tially dry production and processing spaces and/or pro- duction and processing equipment comprise open, non-

watery spaces and equipment mentioned above. Substan- tially dry spaces does not refer e. g. to internal sur- faces of containers, pipes and the like used in food industry processes that come into contact with liquids, but dry spaces and equipment does refer e. g. to the outer surfaces of pipes and containers. Pseudomonas bacterial preparation refers to an antimicrobial prepa- ration inhibiting growth of salmonella, containing bac- teria and/or produced by bacteria, which has a low pro- tease and lipase activity. Lactic acid bacterium prepa- ration refers to an antimicrobial preparation inhibit- ing growth of salmonella, containing bacteria and/or produced by bacteria, which contains compounds having a molecular weight < 1000 Da.

In an embodiment, a Pseudomonas bacterial preparation, preferably containing above-mentioned bac- teria, which has an inhibiting effect on the growth of salmonella and a low protease and lipase activity, is applied to substantially dry surfaces. Pseudomonas bac- terial preparations are well suited for the prevention of salmonella growth on the surface of various surface materials used in above-mentioned spaces, such as metal (e. g. stainless steel), brick, tile, concrete, asphalt, rock material, plastic, wood and/or other materials.

The Pseudomonas bacterial preparation is well suited for the transparent of surfaces in storage spaces. The bacterial preparation is able to survive even in very harsh conditions, forming on the surface a biofilm inhibitive to salmonella. The protective effect of the bacterial preparation is particularly good e. g. on asphalt surfaces, which are favourable to growth of salmonella.

In another embodiment, a Pseudomonas bacterial preparation having an inhibitive effect on the growth of salmonella is applied to media. The bacterial prepa- rations are well suited for the treatment of desiccants

used in broiler breeding plants, such as peat, to fight salmonella.

In an embodiment, a preparation of lactic acid bacteria containing small-molecule compounds of a mo- lecular weight < 1000 Da which have an inhibitive ef- fect on the growth of salmonella is applied to the sur- faces to be treated. Preparations of lactic acid bacte- ria are well suited for the prevention of salmonella growth on various surface materials used in above- mentioned spaces, such as metal (e. g. stainless steel), brick, tile, concrete, asphalt, plastic, wood and/or other materials.

Especially the use of preparations of lactic acid bacteria is safe because it is generally allowed and approved in food industry. Preparations of lactic acid bacteria are particularly well suited for the treatment of desiccants, such as peat, used in broiler breeding plants.

Pseudomonas bacteria are gram-negative, aero- bic, mesophilic or psychophilic organisms. Their natu- ral environments are soil, water and e. g. plant sur- faces. Pseudomonas bacteria are not very demanding in respect of nutritives as they are able to synthetise most of their structural elements themselves. They are generally non-pathogenic, although certain Pseudomonas species and strains (e. g. P. aerucrinosa) may cause e. g. urinary tract infections. Some strains produce heat- resistant proteases and lipases, which may give rise to defects of structure and taste in pasteurised milk products.

Pseudomonas bacteria form compounds inhibiting the growth of other microbes. Examples of such com- pounds are phenazine antibiotics and siderophores. Bac- terial strains producing phenazine antibiotics are used as biopesticides in plant cultivation. Siderophores (iron carriers) are iron-specific organic chelates which form very permanent complexes with iron. A mi-

crobe that produces siderophores is able to take in iron into its cells in this chelate form, but by bind- ing the iron needed by other microbes it inhibits their growth. Siderophores have been used for control of pathogenic microbes in environments like process spaces containing liquids, e. g. in fermentation vessels or fish breeding environments (EP 360568). The publication Chemical Abstracts 126 (1997), 72473 describes the pre- vention of salmonella growth in aqueous environments using a certain Pseudomonas strain. The microbial ef- fect of Pseudomonas bacteria may also be based on other properties.

A characteristic inherent with Pseudomonas bacteria is that they form a biofilm. Biofilm means a covering formed in suitable conditions and consisting of microbes and/or exopolysaccharides and proteins se- creted by them, which can be used to cover a wide vari- ety of surfaces.

In the procedure of the invention, a biofilm based on Pseudomonas bacteria providing protection against salmonella is used. In the procedure, it is possible to use any generally available bacteria of the Pseudomonas species that has an inhibiting effect on the growth of salmonella as well as a low protease and lipase activity. The non-pathogenic strain used has a good capacity to produce siderophores and other com- pounds inhibiting the growth of salmonella. If desir- able, the sensitivity of the bacterial strains used in the procedure to the commonest disinfectants used by food industry and commerce can be determined. If neces- sary, it is possible to define disinfection conditions for the Pseudomonas strain for its elimination, e. g. in the case of a change of the purpose for which the prem- ises are used.

Usable Pseudomonas varieties have been de- scribed e. g. in an article by Palleroni, N. J. (1984) Genus I Pseudomonas. Gram Negative Aerobic Rods and

Cocci. Family I. PSEUDOMONADACEAE., in Bergey's Manual of Systematic Bacteriology, Vol 1. Ed. Krieg, N. R. and Holt, J. G., Williams & Wilkins, Baltimore, London. pp.

141-199. Usable strains are Pseudomonas marqinalis (e. g. VTT-E-94557), Pseudomonas aureofaciens, Pseudomo- nas chlororaphis (e. g. VTT-E91436 and VTT-E-95581T) and Pseudomonas fluorescens (e. g. VTT-E-94558 and VTT-E-93443T), preferably Pseudomonas marginalis.

The Pseudomonas bacterial preparation may con- sist of a culture solution with or without cells, cells, concentrated or diluted culture solution with or without cells, fractionated culture solution or a mi- crobicidal product completely or partially purified; e. g. in the form of an aqueous suspension.

In the procedure of the invention, the Pseu- domonas bacterial preparation is used in concentrations of e. g. about 102-10'pmy/cm2, preferably about 104- 106 pmy/cm2; if desirable, it is also possible to use considerably lower or higher concentrations, consider- ing the target and conditions of application, in order to fight salmonella.

Lactic acid bacteria are gram-positive micro- aerophilic rods or cocci, a typical characteristic of which is fermentative metabolism. Depending on the spe- cies, the main fermentation products besides lactic acid are acetate, carbon dioxide, ethanol, formiate and the like. Lactic acid bacteria are relatively demanding microbes and need a very rich culture medium to be able to grow. These bacteria and the fermentation phenomena caused by them are utilised in many traditional food- stuff processes. For this reason, they have a so-called GRAS (Generally Recognized As Safe) status. In most probiotic preparations, lactic acid bacteria are essen- tial.

Lactic acid bacteria often effectively prevent the growth of other microbes. The low pH resulting from the fermentation alone eliminates many rivals. In addi-

tion, the main fermentation product, lactic acid, is an antimicrobial compound. Many strains also produce mi- crobicidal peptides and proteins, so-called bacterio- sines. Most bacteriosines have a very narrow range of action, limiting their usability. An exception is the macromolecular nisin, which is effective against most gram-positive bacteria and which has been approved as a food additive in many countries, Finland among others.

Nisin is also used together with a chelating agent, which expands its range of action, to eliminate bacte- ria from surfaces that come into contact with food- stuffs (WO 96/39842).

The use of lactic acid bacteria in food and fodder industry is previously known. Lactic acid bacte- ria are a generally used in milk products, meat prod- ucts, fermentation of vegetables and in bakery products and fodder preservation. Lactic acid bacteria are used in foodstuffs to fight pathogenic microbes, such as Listeria and Salmonella bacteria. However, lactic acid bacteria have not been used to fight Salmonella bacte- ria on surfaces in spaces and in media.

In the procedure of the invention, any com- monly available lactic acid bacterium can be used which contains and/or produces micromolecular compounds hav- ing a molecular weight < 1000 Da and inhibiting the growth of salmonella. Usable genera of lactic acid bac- teria are Lactobacillus, Pediococcus and Lactococcus, which are described e. g. in the article Kandler, O. and Weiss, N. (1986) Genus Lactobacillus. Section 14. Regu- lar, Nonsporing Gram Positive Rods, un Bergey's Manual of Systematic Bacteriology, Vol 2. Ed. Krieg, N. R. and Holt, J. G., Williams & Wilkins, Baltimore, London. pp.

1209-1234, and in the articles Teuber et al., (1992) The Genus Lactococcus. Chapter 67 and Weiss, N. (1992) The Genera Pediococcus and Aerococcus. Chapter 68, in the book The Prokaryotes. Second Ed., Ed. Balows, A., Truper, Dworkin, M., Harder, W. and Schleifer, K-H.,

Springer-Verlag. New York. s. 1482-1501 ja 1502- 1504. Usable strains are Lactobacillus plantarum (e. g.

VTT-E-79098 ja VTT-E-78076), Pediococcus parvulus (e. g.

VTT-E-88315), Pediococcus acidilactici, Pediococcus damnosus (e. g. VTT-E-76065) or Lactococcus lactis. Par- ticularly advantageous strains are the Lactobacillus plantarum or Pediococcus parvulus strains.

A lactic acid bacteria preparation may consist of a culture solution with or without cells, cells, a concentrated or diluted culture solution with or with- out cells, a fractionated culture solution or a com- pletely or partially purified microbicidal product; e. g. in the form of an aqueous suspension.

In an embodiment of the procedure, the lactic acid bacteria preparation consists of a cell-free cul- ture solution, a cell-free concentrated culture solu- tion, a fractionated culture solution and/or a com- pletely or partially purified microbicidal product. Ac- tive compounds can be fractionated from the culture so- lution using gel filtration and ultrafiltration and different concentrating methods, such as evaporation or cold drying. According to a preferred embodiment, the treatment is carried out using an ultrafiltered super- natant of lactic acid bacteria which has been concen- trated to a 5-fold level by cold drying and/or using a diluted solution of it.

The present invention makes it possible to prevent salmonella contamination in premises and proc- esses used in food industry. The procedure of the in- vention promotes elimination of salmonella from con- taminated spaces.

The procedure of the invention improves the overall hygienic level of premises and processes used in food industry in general.

In the following, the invention will be de- scribed in detail by referring to the attached drawings and examples.

Fig. 1 illustrates the effect of the Pseudomo- nas strain on the growth of salmonella on an asphalt surface, Fig. 2 illustrates the effect of the same Pseudomonas strain on the growth of the Salmonella ty- phimurium EELA 144 strain in a peat substrate, Fig. 3 illustrates the effect of the ultrafil- tered fraction of a lactic acid bacterial strain on the growth of the Salmonella strain in peat.

Fig. 4 illustrates the effect of the same Pseudomonas strain and a filtrate of the same lactic acid bacterial strain on the growth of the Salmonella strain in a peat substrate when one transmitter chicken has been given Salmonella.

Fig. 5 illustrates the effect of a sideropho- ric filtrate of the same Pseudomonas strain on the per- meabilisation of Salmonella.

EXAMPLE 1 This experiment was conducted to test the mi- crobicidal effect of the Pseudomonas bacterial strain on salmonella on different surface materials. The strains were selected on the basis of their antimicro- bial and enzymatic activity and their tolerance of dis- infectants.

In the experiment, the Pseudomonas marginais (VTT-E-94557 = VTT-E-557 = E-557) strain, which belongs to the collection of bacterial strains at the biologi- cal and foodstuff laboratory of VTT, the Technical Re- search Centre of Finland, was used. In a low-iron sub- strate, this strain has a high siderophoric production activity and an antimicrobial activity on the wild Sal- monella strains used. Moreover, the strain has a low lipolytic and proteolytic activity. It tolerates the Alcoran and Salmonkill disinfectants but is sensitive to disinfectants such as e. g. Virkon, Hygicept and Klo- roforte.

The test strains used were Salmonella infan- tis, Salmonella typhimurium (EELA 144) and Salmonella enteritidis, isolated and delivered by the Finnish In- stitute of Foodstuffs and Veterinary Medicine (EELA).

Contamination was effected using a (1: 1: 1) mixture pre- pared from the test strains.

Aqueous suspension of the Pseudomonas strain was applied in laboratory conditions to asphalt, con- crete and Leca brick surfaces in a concentration of 104 pmy/cm2. Part of the surface material was contaminated with a mixture of the test strains. In the tests, three different initial salmonella levels were used, a), b) and c) : < 10 pmy/cm2,102 pmy/cm2 and 104 pmy/cm2. The fitness of both the Pseudomonas strain and the salmo- nella strains on the surfaces was monitored using plate culture and enrichment methods. By the enrichment method, even a very slight salmonella growth could be discerned.

Surfaces with no Pseudomonas suspension ap- plied to them and a clean uninoculated surface were used as control surfaces. During the experiment, the sample surfaces were wetted with sterile water to pre- vent drying. Samples were taken from the surfaces by dabbing after 0,2 and 7 days from the beginning of the test.

Table 1 presents the results of the surface material tests. Inhibition of salmonella growth was very obvious with all the surfaces used in the test.

The inhibiting effect in the case of concrete and Leca bricks was partly due to the high pH of the sub- strate/material, whereas in the case of asphalt, which is favourable to salmonella growth, the growth inhibi- tion achieved by Pseudomonas inoculum was very dis- tinct. Fig. 1 illustrates the effect of the Pseudomonas marginais strain on the growth of salmonella on an as- phalt surface. On the control surfaces, salmonella grew well, reaching a level of 106 pmy/cm2 in a two-day sam-

ple. At a lower initial salmonella level <100 pmy/cm2, the growth inhibiting effect was more pronounced. In the samples treated with the cultivated Pseudomonas strain, salmonella appeared in a 7 d sample. During the growth period (0-2 days) of the Pseudomonas strain, the growth of salmonella was retarded. In a reference sample with enrichment at an inoculum level of 102 pmy/cm2, salmonella appeared in the sample for the starting day. From a surface with a corresponding sal- monella concentration and Pseudomonas treatment, en- richment did not reveal salmonella in a two-day sample.

During the experiment, condensed water and aquatic aerosols were formed in the containers. By the agency of the water vapour aerosol, even the uninocu- lated surface was contaminated with salmonella. The surface treated with the Pseudomonas strain was con- taminated two days later.

Table 1. Salmonella concentrations in surface material samples. Sampling @@@@ 0 2 7 Sample surface treatment Asphalt Con- Leca Asphalt Con- Leca Asphalt crete brick crete brick pH 8,9 pH 11,6 pH 9,9 pH 8,9 pH 11,6 pH 9,9 pH 8,9 Cleanness control* -1) - - + - - ++ P.marginalis 104 pmy cm-2* - - - - - - + S.typhimurium < 10 pmy cm-2 - - - ++++5) - - ++++ P.marginalis + S.typhimurium - - - + - - ++++ < 10 pmy cm-2 S.typhimurium 102 pmy cm-2 +2) + - ++++ + - ++++ P.magrginalis + S.typhimurium - - - + - - ++++ 102 pmy cm-2 S.typhimurium 104 pmy cm-2 ++3) + + ++++ - - ++++ P.marginalis + S.typhlmrium++ + - +++4) - - +++ 104 pmy cm-2 1) - = result of enrichment culture Salmonella-negative, result of culture < 10 pmy cm@<BR> 2) + = result of enrichment culture Salmonella-posotive, result of culture < 10 pmy cm@<BR> 3) ++ = result of enrichment culture Salmonella-positive, result of culture >10 - 103 pm@<BR> 4) +++ = result of enrichment culture Salmonella-positive, result of culture 103 - 106 pm@<BR> 5)++++ = result of enrichment culture Salmonella-positive, result of culture > 106 pmy cm@<BR> * = no Salmonella added to surface material

EXAMPLE 2 In this experiment, the microbicidal effect of the Pseudomonas strain on salmonella in peat substrate with different pH values as used in broiler breeding plants. It has been established that the pH value of peat changes during broiler breeding from pH 3.4 to pH 8 within a week.

In the test, the Pseudomonas marginais VTT-E-94557 strain was used. The test strain used was Salmonella typhimurium (EELA-144). In a series of tests, an autoclaved sterile peat sample was inoculated with both Pseudomonas and Salmonella strains. The pH of the peat substrate was adjusted using ammonia.

The Pseudomonas strain retarded the growth of the Salmonella strain in peat at a neutral pH. In acid peat with pH<5, the Pseudomonas strain did not survive.

Fig. 2 illustrates the effect of the Pseudomo- nas marginais strain on the growth of the Salmonella typhimurium strain in a peat substrate (pH 7,4). At the 102 starting level of Salmonella contamination, the ad- dition of Pseudomonas retarded the growth of salmonella during 5 days so that the microbial content of the Sal- monella strain in the sample due to the Pseudomonas ad- dition was 100 times lower than in a reference sample to which only salmonella had been added.

EXAMPLE 3 This experiment was carried out to test the effect of the substances produced by lactic acid bacte- ria on salmonella in a peat substrate as used in broiler breeding plants.

The production strains selected for the test were Lactobacillus plantarum (VTT-E-79098 = VTT-E-98 = E-98) and Pediococcus parvulus (VTT-E-88315), which were known to have an antimicrobial effect on salmo- nella. The salmonella inhibiting efficiency of the ac- tive compounds produced in the breeding experiments was tested using a culture solution filtrate purified by

ultrafiltration (cut-off 1000 D) and concentrated (5x) by cold drying. The reference sample used was lactic acid (10 % stock solution).

In a series of tests, an autoclaved sterile peat sample, whose pH had been adjusted to the value 6 using ammonia, was inoculated with the Salmonella ty- phimurium strain and concentrated ultrafiltrate was added to it. The starting levels of the salmonella in- oculum were 102 and 104 pmy/g. Samples were taken after 0,2,4 and 7 days from the beginning of the test.

Fig. 3 illustrates the effect of a L. Plantarum filtrate on the growth of the S. typhimurium 144 strain.

The effect of the filtrate is apparent from the sample taken after 4 days. Salmonella was found in all samples by the enrichment method, but direct plate culture did not reveal salmonella in the sample with a concentrated ultrafiltrate addition. The inhibition of growth was not solely due to the effect of lactic acid. Even at a higher initial contamination level (104 pmy/g, Fig.

3a), the inhibiting effect was visible as a retardation of salmonella growth. The ultrafiltrate of the P. parvulus strain was not as effective. Inhibition of growth is based on micromolecular compounds, whose ef- fect becomes more potent through the action of the lac- tic acid produced by the strain.

EXAMPLE 4 This experiment was carried out to test the effect of a protective culture or microbicidal prepara- tion added to peat moss used as substrate material for chickens on salmonella growth in the peat substrate and on the transmission of salmonella contamination from one chicken to another. In the test, a 105 pmy g-1 addi- tion of Pseudomonas marqinalis VTT-E-94557 a 5 % (v/w) addition of L. plantarum E-98 filtrate were used. In each test group (8 newly hatched chickens/group), one of the chickens was given 103 pmy of Salmonella infan- tis VTT E-97871 bacteria (nalidixic acid resistant

strain) directly into the crop. The pH of the peat sub- strate was adjusted to the value 6 by adding CaCO3.

Some of the test groups were given Broilact protective culture. The salmonella content of the peat substrate was monitored by the plate culture technique (samples taken on days 1,4 and 6). At the end of the test, the chickens were killed and the Salmonella content of their caecum was determined quantitatively and by en- riching.

The effect of the L. plantarum E-98 filtrate addition on the salmonella content in peat substrate is illustrated by Fig. 4a. In the first-day (1 d) sample, the salmonella quantity in the peat substrate in the reference group was 105 pmy g-1. Lactic acid bacterium filtrate E-98 reduced salmonella growth to the level of 103 pmy g~16 Broilact effectively inhibited salmonella growth and the salmonella quantity in the peat sub- strate remained low, at the level of about 3x102 pmy g-1 in the 1 d sample and 103 pmy g~1 in the 5 d sample. Ta- ble 2 presents the salmonella quantity in the caeca of the chickens at the end of the test. The salmonella quantity in the reference samples was 109 pmy g~1, 107 pmy g~1 in the E-98 filtrate groups and 101 pmy g-1 in the Broilact groups.

Table 2. In vivo test, Salmonella quantity in chickens' caeca at end of test.

Group Treatment IF 1 Broilact 0 2 Broilact 0.7 3 Broilact + L. plantarum E98 0.8 4 Broilact + L. plantarum E98 1.6 5 L. plantarum E98 6.7 6 L. plantarum E98 6.4 7 Reference sample 8.3 8 Reference sample 8.9

IF = geometric mean value of the salmonella quantities found in the caeca of the chickens in the group. If no salmonella is found in the caecum, then IF=0. If salmonella is only found by the enrichment method, then IF=1. Transmitter chickens are not in- cluded in the IF values.

Fig. 4b illustrates the effect of the Pseudo- monas marginais VTT-E-94557 addition on salmonella content in peat substrate. The first-day (1 d) sample of the peat substrate of the reference group had a sal- monella content of 105 pmy g-1 and 5x107 pmy g~1 at the end (6 d) of the test. Pseudomonas VTT-E-94557 pro- tective culture reduced salmonella growth in the 1 d sample to the level of 104 pmy g-1 and in the 6 d sample to 1 x 1x106 pmy g-1. In the Broilact groups, the salmo- nella quantity in the 1 d sample was 1x104 pmy g-1 and 1x105 pmy g-1 in the 6 d sample. Table 3 presents the salmonella quantities in the chickens'caeca at the end of the test. The salmonella quantity in the reference samples was 5x108 pmy g~1/106 pmy g~1 in the Pseudomonas E557 protective culture groups and 102 pmy g-1 in the Broilact groups.

Table 3. Salmonella quantity in chickens'caeca at end of test.

Group Treatment IF 1 Broilact 2.0 2 Broilact 2.8 3 E557 6.3 4 E557 6.5 5 Broilact + E557 2.8 6 Broilact + E557 3.4 7 Control 8.0 8 Control 9.0 The use of Broilact protective culture for chickens inhibits the adherence of salmonella to the chickens'intestines even if the peat substrate has a

high salmonella content. Lactic acid bacterium filtrate E-98 and Pseudomonas marginais E-94557 protective cul- ture retard salmonella growth in siccative peat, and the salmonella quantities in caecal samples of the chickens in the corresponding groups were lower than in the reference group. No additive effect was detected between Broilact and the filtrate.

EXAMPLE 5 This experiment was carried out to test the mechanisms of action and bioactivity of the Pseudomonas strain toward different Salmonella strains. A cell-free filtrate of the Pseudomonas marginais VTT-E-94557 bac- terium was used in the test.

In a low-iron culture medium, Pseudomonas bac- teria produce siderophores, which have a chelating ef- fect on iron. In growth inhibition tests performed us- ing a Bioscreen device, the growth inhibiting effect of the cell-free filtrate on Salmonella was evident. The results of the growth inhibition tests are presented in Table 4. An iron addition, 0.1 mM Fers04, almost com- pletely cancelled the growth inhibiting effect.

Table 4. Growth inhibition % age, determined by a turbi- dometric method.

Filtrates P. marqinalis E-94557 Target organism 0 mM FeSO4 0.1 mM FeSO4 S. cholerasuis ATCC 10708 38.3 25.3 S. tvphimurium ATCC 13311 44.7 6.6 S. tvphimurium ATCC 14028 49.1 3.46 S. enteritidis ATCC 13076 54.6 18.8 S. infantis ELI 5 51.9 1.5

The effect of the same Pseudomonas bacterium on the permeabilisation of the cell membrane Salmonella bacteria was tested using a fluorescent colouring agent, N-phenyl naphthyl amine (NPN). NPN is a highly hydrophobic compound which is intensively fluorescent in a lipid environment but not in an aqueous environ- ment. The normal outer membrane of salmonella has a hy- drophilic outer surface and does not bind any signifi- cant amounts of NPN. However, if the outer membrane has been damaged by the action of an external agent, NPN can enter into the lipid layers and further into the cell membrane while emitting an intensive fluorescence, which can be measured photometrically. The test strain used was S. typhimurium ATCC 13311. The test was carried out using a raw filtrate of Pseudomonas marginais VTT-E-94557 culture 191,192 and 198 (with only cells removed), so in addition to siderophores it also con- tained all the other metabolic products (e. g. pro- teins). Therefore, the results are based on the com- bined effect of a plurality of factors. The tests were performed on three independent samples and their re- spective substrates.

Fig. 5 illustrates the effect of an addition of siderophore filtrates on the permeabilisation of Salmonella. All three siderophore supernatants caused a weakening of the outer membrane of the S. tvphimurium ATCC 13311 strain. This appears as a rise in the NPN uptake factor in samples treated with filtrates as com- pared with mere substrate treatment.

The effect of the filtrates on the detergent and lysozyme sensitivity (1 % Triton X-100,0,1 % SDS, lysozyme) of Salmonella cells was tested using cell ly- sis measurement. In the measurement, the effect of cell-free growth supernatants on the lysability of Sal- monella cells was tested by turbidometrically measuring the turbidity of the cells using a 405 nm wavelength after various treatments. Table 3. Cell lysis sensitivity of Salmonella to lyso- zyme and detergents. Percentage of unlysed cells () Lysing Concentra-Untreated Cells 192-treated agent tion cells treated on cells R-sub- strate Lysozyme 10 yg ml-1 97-3#1. 8 98.81.8 TX-100 0.1 % 104.22.1 TX-100 1 % 93.4#2. 3 SDS 0.1 % 97.5#1. 6 81.96 55.712 SDS 1 % 78. 7 35.5 21.1

Supernatant 192 renders Salmonella clearly more sensitive to the lysing effect of SDS but not to the effect of neutral detergent TX-100.

The invention is not limited to the examples of its embodiments described above, but instead many variations are possible within the scope of the inven- tive idea defined by the claims.