TECHNICAL BACKGROUND AND PRIOR ART Micro-organisms play important roles in the food industry. Thus, so-called starter cul- tures, in particular of organisms belonging to the large group of lactic acid bacteria and strains of Propionibacterium species, Bifidobacterium species and fungal species are widely used in the manufacturing of a large variety of food products such as dairy products, meat products, bakery products and beverages including wine. Whereas the use of such desirable and useful organisms constitutes an integrated element of the food manufacturing process, other micro-organisms are highly undesirable in food products as they cause spoilage of the products or cause food poisoning and other food-borne diseases and the presence of such organisms must therefore be kept as low as possible.

In order to be able to provide food products not only having desirable and attractive sensory and dietary characteristics but also being microbiologically safe and having a long shelf life it is required that the food industry is able to draw upon methods of monitoring and controlling the development of desirable and undesirable micro-organ- isms during manufacturing, distribution and storage of food products.

A major challenge facing the food industry in that respect is the difficulties involved in selectively monitoring the growth and survival of particular organisms in the ecologi-

cally complex and varied environments of food starting materials in the process of manufacturing and in the finished food products. Thus, most food materials and prod- ucts comprise a large variety of micro-organisms of which some are added deliberately as starter cultures to confer desirable sensory and dietary characteristics to the food product, whereas others are contaminating spoilage or pathogenic organisms, the presence and activity of which is to be controlled to the largest possible extent.

Presently, a range of selective culturing methods are available to monitor the pres- ence of specific organisms or groups of organisms in food materials and products. In particular, such methods are available for detecting the presence and/or the numbers of pathogenic organisms such as e. g. Salmonella spp., toxinogenic E. coli, Shigella spp., Vibrio spp., Listeria spp., Campylobacter spp. and Staphylococcus spp.

Generally, such methods are based on the use of selective media that to a varying ex- tent inhibit or suppress the growth of organisms other that those searched for. Such methods are, however, time-consuming and the results of the testing are not available until at least one day after the sample has been collecte. Recently, other methods for monitoring the presence of pathogenic organisms based on the use of more or less specific DNA probes have become available which have made it possible to detect the presence of pathogenic or food poisoning organisms in food products within the same day. However, such methods do not generally permit the selective detection of a spe- cific strain of a microbial species.

A particularly difficult problem associated with the use of starter cultures as men- tioned above in food manufacturing is that generally applicable selective culturing methods are not available that can discriminate between individual strains of the same or related species or group of species. Thus, e. g. with respect to lactic acid bacteria there are relatively selective culturing methods available that permit the enumeration of viable lactic acid bacteria as a group based on counting the number of typical colo- nies of such bacteria on solid media. However, it is generally not possible by using such methods to distinguish between different species not to say individual strains of a species. The implication of this lack of methods to selectively monitor the presence of individual starter culture species and strains is that is not possible (i) to monitor the competitiveness of such individual strains or species in a pre-selected environment or (ii) to select such strains or species that in a given pre-selected environment is capable

of competing with other organisms with respect to growth, survival and/or metabolic activity. This problem is particularly pronounced in respect of starter culture composi- tions comprising two or more strains of lactic acid bacteria, Bifidobacterium and/or Propionibacterium as the morphology of colonies of such species are so similar that reliable selective enumerations of individual species or strains are not possible.

The above problem does not only exist in connection with the use of such organisms in manufacturing of food products, but also in relation to the manufacturing of com- mercial starter cultures comprising a multiplicity of individual strains which during manufacturing of the culture is propagated together. The distribution of the individual strains determines the sensory and dietary characteristics, which such a mixed starter culture confers to a food product. In order to provide such commercial cultures that have a reproducible performance it is required that the distribution of the individual strains be monitored and controlled in a reliable, rapid and simple manner. Presently, useful method for such a purpose are not available.

There is therefore a strong need for methods to selectively monitor the growth, sur- vival and/or metabolic activity of an individual microbial strain in pre-selected environ- ments including such environments where the individual strain is competing with other microorganisms.

SUMMARY OF THE INVENTION Accordingly, the present invention relates in a first aspect to a method of selectively monitoring the development and/or competitiveness of a microbial strain in a sample environment, the method comprising that a selectively detectable genetic label is in- serted into a replicon of said strain whose development and/or competitiveness is to be monitored, adding the thus labelled strain to the sample environment and isolating DNA or RNA of said strain from the sample environment and detecting the amount of the genetic label.

In a further aspect there is provided a method of selecting, among a multiplicity of mi- crobial strains, the strain that in a pre-selected environment and under pre-selected conditions has the highest capability to survive, multiply and/or express a pre-selected

gene, the method comprising providing in any one of said microbial strains a selec- tively detectable genetic label, introducing said labelled strains into the particular envi- ronment and keeping said environment under the pre-selected conditions for a period of time during which at least one sample is collecte and measuring in said sample the amount of cells of each individual microbial strain or the amount of gene expression as determined by the amount of their detectable genetic label, and selecting the strain of which the highest amount of cells or gene expression is present.

The invention pertains in a still further aspect to a microbial culture composition com- prising a multiplicity of microbial strains where at least one of said strains contains a selectively detectable genetic label permitting the development and/or competitiveness of said labelled strain (s) to be monitored selectively in a sample environment to which the composition is added.

In yet another aspect the invention relates to the use in the above methods of a PCR method that is capable of detecting a genetic label consisting of a mutational substitu- tion of one or more nucleotides in a coding sequence of a replicon, the PCR including the use of a probe that is capable of binding to a DNA or RNA template comprising the genetic label and which is labelled with at least two fluorescent labels, the probe is hydrolyse during each PCR cycle resulting in emission of a fluorescence signal, the amount of which is proportionate to the amount of genetic label detected, including such a use wherein the first PCR cycle involves the use of a reverse transcriptase whereby a labelled mRNA and/or cDNA template is generated.

DETAILED DISCLOSURE OF THE INVENTION It is one primary objective of the invention to provide a method of selectively moni- toring the development and/or competitiveness of a microbial strain in a sample envi- ronment. It is another objective of such a method to provide the means of identifying the proprietor of labelled microorganism, the quality of products comprising labelled microorganisms and/or the batch number of end products carrying labelled microbial strains. Whereas the method has been found to be particularly useful in the context of food manufacturing, surveillance of food quality and manufacturing of food starter cul-

ture strains, the method is generally applicable in any other context where there is a need to monitor the development or competitiveness of a specific microbial strain.

Thus, the method represents a generally applicable tool in microbial ecology. Thus, it is envisaged that the method is useful in studies of the ability of micro-organisms to grow, survive and compete in a variety of environments including soil, water, animal body habitats and industrial production facilities.

As used herein, the expression"selectively monitoring"indicates that at least one of the presence, number, physiological state or metabolic activity of a specific microbial strain can be determined selectively in a pre-selected environment including such an environment where other microbial strains are present. In the present context, the ex- pression"microbial strain"refers to a strain of any type of micro-organisms including prokaryotic and eukaryotic organisms such as bacteria, fungi including yeasts, viruses, bacteriophages, plant cells and animal cells. The expression"sample environment"is used herein to designate any environment that can be described in terms of space, time, physical and chemical parameters including chemical composition, temperature, pH, osmolality, water activity, presence/absence of preservatives, and presence/- absence of other micro-organisms. Thus, as a typical example, a food processing line containing the materials which form the starting material for the finished product constitutes a sample environment according to the invention.

The above method is based on the insertion into the microbial strain, the development and/or competitiveness of which is to be monitored a selectively detectable genetic label which permits that the cell number, physiological state and/or metabolic activity of the thus labelled microbial strain can be monitored in the sample environment by isolating DNA and/or RNA of said strain from the sample environment and detecting the presence and/or amount of the genetic label.

In the present context, the expression"genetic label"refers to any selectively detect- able modification of the DNA of the microbial strain to be labelled. Thus, in useful em- bodiments the modification consists in inserting into the selected microbial strain a nucleotide sequence including an isolated, naturally occurring sequence or a syntheti- cally made nucleotide sequence. The inserted nucleotide sequence label may comprise

naturally occurring nucleotides, non-naturally occurring nucleotides, a PNA sequence or a nucleotide sequence consisting of any mixture of such sequences.

In specific embodiments, the inserted nucleotide sequence comprises an identification code containing identifiable information such as e. g. on the origin of the strain or its batch number. Such an information carrying code is e. g. in the form of an alphanu- meric code.

Such sequences and the method of making such useful labels and of inserting them into living cells are disclosed in WO 96/17954 to which there is specifically referred in respect of how to make and insert such genetic labels.

The size of such labels may vary considerably, depending on the informational content which it is desirable to incorporate in the labels, a suitable size being in the range of 10 to 1000 nucleotides such as in the range of 50 to 500 nucleotides including the range of 100 to 250 nucleotides.

The above genetic labels can be inserted into the microbial strain according to any conventional technique for the insertion of nucleotide sequences. Thus, the insertion can be made by a site-specific recombination method, one particularly useful specific method being that disclosed in WO 94/19460 to which there is referred. This method is based on the use of bacteriophage oLC3 which comprises integration functions permitting that the bacteriophage vector is integrated at specific sites in a bacterial strain. This bacteriophage-based integration vector comprises a site for insertion of foreign DNA or other types of nucleotide sequences, permitting that a genetic label according to the present invention is inserted into the microbial strain to be labelled.

An other possible manner of inserting the above genetic label is by homologous re- combination implying that the labelling sequence is inserted by use of a DNA replace- ment vector being integrated through two homologous recombination events such as it is e. g. described by Biswas et al. 1993, J. Bact. 175: 3628.

It will be appreciated that the insertion of the genetic label can be made by integrating the label into the chromosome of the strain. This is generally considered to be advan-

tageous as the label will be maintained stably and passed on to the progeny of the cells. However, it is also within the scope of the invention that the label is introduced into an episomal replicon of the strain including a plasmid, a cosmid or a vector, in particular in cases where such extrachromosomal elements are modified to become stably maintained in their host cell.

An important aspect of the manner whereby the genetic label is inserted is that the insertion occurs at a site of the DNA of the microbial strain which does not result in any interruption or reduction in the expression of genes, the gene products of which is required for non-limited function of the cells of the microbial strain. It is possible to a large extent to avoid such an undesirable effect by using a site-specific integration method as referred to above, although it cannot be completely ruled out that the inser- tion of foreign nucleotide sequence may render the resulting recombinant strain slightly less competitive as the progeny of the initially labelled cells will be encum- bered by the synthesis of the labelling nucleotide sequence.

An alternative approach to avoiding any adverse effects on gene expression in the mi- crobial strain is to provide the genetic label according to the invention in the form of a substitution of one or more, preferably a few, nucleotides in a DNA sequence naturally occurring in the microbial strain. Such a labelling can, in contrast to the above inser- tion of a foreign sequence, even be made in a coding sequence of a gene without im- pairing the expression of the gene, in particular when the substitution is"silent", i. e. it does not result in a codon for a different amino acid. It will be appreciated by the per- son skilled in the art that substitutions of nucleotides in a DNA sequence of a micro- bial strain can be made by any conventional techniques for that purpose including site- directed mutagenesis using appropriate oligonucleotide primers. It will, however, be understood that the substitutional event may also be provided by using random muta- genesis followed by selecting cells where a substitution as defined above has occurred or by selecting spontaneously occurring mutants having such a substitution.

Having constructed the genetically labelled microbial strain it is added to the sample environment under conditions where the strain, if competitive, is capable of growth and metabolic activity for any appropriate period of time during which at least one

sample of the environment is collecte to determine to what extent the labelled strain has been capable of developing and competing in the selected environment.

Depending on the nature of the sample environment, it may be required, prior to de- tecting the amount of the genetic label, to include a step of isolating the microbial cells present in the sample. Such an isolation step is e. g. carried out by centrifugation, filtration or any other conventional cell separation technique. In accordance with the present method the sample material for determining the amount of labelled DNA is either the sample as such, a preparation of microbial cells separated from the sample or DNA preparations provided by isolating DNA from the sample itself or from more or less concentrated preparations of cells previously separated from the sample material.

When the genetic label that is inserted in the microbial strain is a nucleotide sequence or a nucleotide substitution as described above, the amount of the label is detected by any suitable PCR method which permits the amount and optionally, the sequence of the label to be determined e. g. using the methods as described in WO 96/17954.

It will, however, be understood that when using conventional PCR techniques, the de- termination of the amount of target nucleotide sequences, i. e. the genetic label, is at the best semi-quantitative, as it has to be based e. g. on a"most probable amount" concept, i. e. the amount is determined by serially diluting the sample containing the target sequence and determining the dilution at which the target DNA can no longer be detected. Based on this end point determination, the most probable amount of ge- netic label in the sample can be assessed at least semi-quantitatively.

Recently, a PCR technique has been developed which, in contrast to conventional techniques permits rapid and very specific quantitative detection of DNA and RNA se- quences. This technique is commercially available under the trade name TaqMan.

The TaqMan-based PCR differs from a standard PCR technique i. a. in that a DNA probe is used that hybridises selectively to the target DNA but which during the PCR cycle is hydrolyse by the 5'-3'nuclease activity of the Taq DNA polymerase. The probe is labelled with two different fluorogenic dyes, a reporter dye and a quencher dye and as long as the probe is intact, such that both dye moieties are attached to the same molecule, the quencher dye will absorb the excited state energy of the reporter

dye moiety. But as the probe anneals to a target sequence between the two PCR primers, and is subsequently cleaved by the 5'-3'nuclease activity of the Taq DNA polymerase, the quenching effect is eliminated as a result of the two dyes becoming separated. Accordingly, there will be an increase in reporter dye emission during am- plification that is proportionate to the amount of amplifie DNA produced (Holland et al., 1991, PNAS 88: 7276-7280). The extent of reporter dye emission is monitored continuously during the PCR cycling, e. g. by means of automated detection systems.

The present inventors have found that the above PCR technique is highly suitable in the methods of the invention as it permits rapid and reliable determination of the amount of genetic label to be made on a large number of samples simultaneously. Ad- ditionally, and importantly, the techniques are highly selective for specific sequences, as the labelled probe cannot be hydrolyse if just one single mismatch between the probe and its target sequence occurs.

The inventors have also recognised that the above technique permits that a distinction can be made between viable and dead cells of the labelled microbial strain. It will be appreciated that when using genetic label DNA as template in the TaqMan-based PCR the amount of genetic label determined will be from both viable and dead cells.

However, when using RNA as target template and carrying out the first amplification step using a reverse transcriptase whereby is generated a mRNA or a cDNA template, or by using a thermostable polymerase that can use both DNA and RNA as template, optionally after a treatment of the sample with DNase, only genetic labels of viable cells are detected. In the following, this approach is referred to as RT-PCR (RT: Re- verse Transcriptase).

In addition to the possibility of determining selectively the number of viable, geneti- cally labelled cells in a sample, the RT-PCR technique makes it possible to measure the level of translation of any pre-selected, genetically labelled gene by using the corre- sponding mRNA as template in the RT-PCR method. A significant aspect of the above use of the TaqMan-based PCR is the high degree of specificity. This implies that there can be made nucleotide substitutions as explained above in several genes of the same microbial strain or in the same gene in a number of different strains. Accord- ingly, it is not only possible to monitor quantitatively the number of viable and/or dead

cells of a labelled microbial strain, but also to monitor simultaneously the level of ex- pression of several genes in the same strain under the given sample environment con- ditions or to monitor the level of expression of the same gene in a range of microbial strains under identical environmental conditions. Any of these possibilities constitute important aspects of the invention.

Any of the above techniques for monitoring the development or competitiveness of a particular microbial strains can be applied in any sample environment including food product starting materials or finished food products, feed product materials and feed compositions, pharmaceutical products, sample environments of animal bodies in- cluding human bodies, water bodies, soil and any production process environment where microorganisms are involved as production strains or contaminating pathogenic or spoilage organisms. In any of such environments there is a need to monitor the sur- vival, growth, gene expression and metabolic activity and interaction with other mi- crobial strains including strains that are indigenous for a particular environment. It will be appreciated that such monitoring can be made over a period of time by collecting samples from the environment at appropriate time intervals.

A particular useful application of the above method is for monitoring the growth and/or metabolic activity of a starter culture microbial strain during processing of food materials inoculated with the starter culture strain, typical examples being milk, vege- tables, meat products, wine musts or wine inoculated with a lactic acid bacterial strain to obtain a fermentation of the food product starting materials. Using a strain that is genetically labelled in accordance with the invention it is possible to collect samples of the inoculated food materials at appropriate time intervals and detect in these samples the amount of genetic label using any of the methods described above. The results of such detection tests can be provided in short time making it possible to adjust the processing conditions if the detection tests show that the growth or activity of the starter culture strain is not at an optimum level.

In one interesting embodiment, the genetically labelled microbial strain is a pathogenic organism. It is evident that there is a need to monitor the survival and competitiveness of such organisms in any of the above environments. As an example, the present method is useful for studying the survival and growth of pathogenic organisms in food

starting materials during manufacturing of food products and in the resulting products during distribution and storage. Evidently, the above RT-PCR technique e. g. provides the means for detecting under which environmental conditions genes coding for toxic gene products are expressed and under which conditions they are not. Another exam- ple of the usefulness of the above method of the invention is its application in study- ing the ability of a given pathogenic or beneficial microbial strain to colonise in an animal body, e. g. in the gastro-intestinal tract. In this connection"beneficial microbial strains"include so-called probiotically active strains of e. g. lactic acid bacteria that are administered to animals or humans in order to improve the general heath of the ani- mals or humans e. g. by inhibiting pathogenic organisms in the gastro-intestinal tract, by stimulating the immune system or by exerting a growth promoting effect. The abil- ity of such strains to become established in the body can be monitored by the present method.

The method can, as it is mentioned above, be used for monitoring any strain of pro- karyotic or eukaryotic micro-organisms including gram-positive bacterium species and gram-negative bacterial species. In this connection, interesting gram-positive species are lactic acid bacterial species, Propionibacterium species and Bifidobacterium spe- cies. Lactic acid bacterial species include Lactococcus species such as Lactococcus lactis, Streptococcus species such as Streptococcus thermophilus, Lactobacillus spe- cies including Lactobacilllus acidophilus, Leuconostoc species such as Leuconostoc oenos and Pediococcus species including Pediococcus acidilactici.

It is another objective of the invention to provide a method of selecting, among a mul- tiplicity of microbial strains, the strain that in a pre-selected environment and under pre-selected conditions has the highest capability to survive, multiply and/or express a pre-selected gene. Evidently, the availability of such a method provides an extremely useful tool in strain selection which presently is based on isolating a range of strains that are potentially useful and test any of these strains individually and in combination for their ability to grow and perform a certain activity in a pre-selected environment.

These procedures are very cumbersome as they involve not only measurements of physical and chemical changes in the environment to which the isolated strains are added but also a vast amount of viable cell counts using conventional culturing methods.

The method that is provided by the present invention represents a very convenient and cost saving alternative approach. As mentioned above, the method comprises provid- ing in any one of the microbial strains to be tested a selectively detectable genetic la- bel followed by introducing said labelled strains into the particular environment and keeping the environment under the pre-selected conditions for a period of time during which at least one sample is collecte and measuring in said sample or samples the amount of cells of each individual microbial strain or the amount of gene expression as determined by the amount of their detectable genetic label. Based on the results of these measurements, the strain of which the highest amount of cells and/or the high- est level of gene expression is present can be selected.

Such a method is particularly useful in the selection of a food starter culture strain e. g. of a lactic acid bacterial species, a Propionibacterium species or a Bifidobacterium species which are widely used in the manufacturing of fermented food products such as dairy products including cheese, yoghurt and butter. It is recognized in the art that within the same species there are considerable variation among individual strains with respect to their ability to perform a food fermentation process and hence, it is required when improved starter culture strains are to be selected to subject a relative large number of strain of the selected species to a comprehensive and costly screening scheme including testing the performance of the strains under industrial fermentation conditions.

In accordance with the above method, the pre-selected strains can now be genetically labelled, using for each individual strain a unique selectively detectable label as de- scribed above. This individual labelling then permits the growth and metabolic activity of each strain to be determined selectively. It will be appreciated that it is possible to add several of such individually labelled strains to the same environment and test their performance simultaneously.

In such a selection method it is particularly useful to apply a label as described above consisting of a mutational substitution of one or more nucleotides in a coding se- quence of the strain, as this makes it possible to monitor not only the number of vi- able and/or dead cells at a given point in time but also the level of expression of a la-

belled gene by measuring the amount of labelled mRNA or cDNA in accordance with the RT-PCR method as also described above.

Although one presently preferred use of the above selection method based on the use of genetically labelled strains is for the selection of food starter culture strains, it will readily be appreciated by the person skilled in the art that the method is generally ap- plicable in the selection of suitable strains of any industrially used microbial species.

Accordingly, the use of the method in the selection of recombinant strains of microbial species used in the production of desired gene products is envisaged. Such species include as examples E. coli, Bacillus species such as Bacillus subtilis, Streptomyces species, Actinomycetes species, species of filamentous fungi including Aspergillus species, Rhizomucor species, Trichoderma species, Penicillium species, yeast species such as e. g. Saccharomyces species, Torula species, Pichia species and Kluyveromy- ces species.

Yet a further objective of the invention is to provide a microbial culture composition comprising a multiplicity of microbial strains where at least one of said strains con- tains a selectively detectable genetic label as described above. As also described above, this labelling will permit the identification and origin of said labelled strain (s) to be determined or the development and/or competitiveness of said labelled strain (s) to be monitored selectively in a sample environment to which the composition is added.

Thus, in one embodiment, the strain/strains is/are labelled with a genetic label that is selected from the group consisting of a DNA sequence comprising naturally occurring nucleotides, a RNA sequence comprising naturally occurring nucleotides, a DNA se- quence comprising non-naturally occurring nucleotides, a RNA sequence comprising non-naturally occurring nucleotides, a PNA sequence and a nucleotide sequence con- sisting of any mixture of such sequences.

In another embodiment, the genetic label consists of a mutational substitution, in- cluding a silent substitution, of one or more nucleotides in a coding sequence of the labelled strain (s).

In one presently preferred embodiment, the culture composition according to the in- vention is a food starter culture composition such as a dairy starter culture composi- tion, comprising two strains or more of a lactic acid bacterium, a Propionibacterium species and a Bifidobacterium species.

Commercial food starter culture compositions including dairy starter culture composi- tions are commonly mixed strain compositions comprising two to ten different strains of the same or different species. Thus, as an example, a starter culture composition for the manufacturing of yoghurt may comprise a mixture of Lactobacillus bulgaricus and Streptococcus thermophilus. Another example is a mixed strain culture for acidi- fying cheese milk that comprises strains of Lactococcus lactis subspecies cremoris, Lactococcus lactis subspecies lactis, Leuconostoc mesenteroides subsp. cremoris and Lactococcus lactis subspecies diacetylactis. Other useful species, which are used in dairy starter cultures, include Propionibacterium species, Lactobacillus helveticus and Lactobacillus delbrueckii subspecies bulgaricus.

Dairy starter culture compositions and starter culture compositions for other food fer- mentation purposes such as fermentation of doughs, meat products, vegetables, fruit musts and wine are generally provided as concentrates of the culture in freeze-dried or frozen form. Typically, such compositions comprises a high number of viable cells, calculated as colony forming units (CFU) e. g. in the range of 106 to 10'2 CFU per g.

The compositions may, in addition to the bacterial cultures, comprise further compo- nents, including nutrient components and cryoprotectants. Such compositions com- prising one or more strains that are genetically labelled in accordance with the inven- tion constitutes a further aspect of the invention.

In a further aspect, the invention provides the use of a PCR method that is capable of detecting a genetic label consisting of a mutational substitution of one or more nucleo- tides in a coding sequence of a replicon wherein the PCR includes the use of a probe that is capable of binding to a DNA or RNA template comprising the genetic label and which is labelled with at least two fluorescent labels, the probe is hydrolyse during each PCR cycle resulting in emission of a fluorescence signal, the amount of which is proportionate to the amount of genetic label in any of the method according to the

invention, including such use wherein the first PCR cycle involves the use of a reverse transcriptase whereby a labelled mRNA and/or cDNA template is generated.

The invention is further illustrated in the following non-limiting example and the draw- ing wherein Fig. 1 illustrates the construction of the replacement vector pTAG1, and Fig. 2 shows the integration of a genetic label DNA sequence into the Bacillus subtilis chromosome; the cloned chromosomal DNA fragment (open box) is interrupted by the label sequence (filled box).

EXAMPLE Genetic labelling of Bacillus subtilis Two 1 kb DNA fragments homologous to a 2 kb Bacillus subtilis chromosomal se- quence were produced by PCR using primers containing the recognition sites for Xbal, EcoRl, BamHI and Kpnl, respectively at their 5'ends. After digesting these PCR-gener- ated fragments and a genetic label fragment with the appropriate endonucleases, the three fragments were mixed and joined by ligation. The recombining cassette com- prising the 2 kb B. subtilis chromosomal sequence interrupted by the label sequence was amplifie from the ligase reaction mixture as shown in Fig. 1. The cassette was then ligated between the Xbal and Salol restriction sites of the thermosensitive plasmid pG+host4 (Appligene, lllkirch, France) to give the replacement vector pTAG1.

Following introduction of the pTAG1 vector into B. subtilis, a Campbell-like single crossing-over event between the chromosome and the homologous region in pTAG1 will result in plasmid integration and duplication of the homologous sequence at the vector-chromosome junction regions. Subsequently, recombination between these DNA repeats will lead to excision of the plasmid. Integration through a followed by excision through b (or vice versa) will result in integration of the label sequence (Fig.

2).

pTAG1 was introduced into B. subtilis cells by electrotransformation as described by McDonald et al., J. Appl. Bacteriol. 79: 213-218,1995. Electroporated cells were di- luted 10-fold in 2xLB medium and incubated with shaking for 2.5 hours at 30°C and subsequently mixed with 10 ml of LB containing 10 ug of erythromycin per ml. For curing of plasmid and selection for single crossing-over integrants, the temperature was raised to 37.5°C (a temperature that is non-permissive for plasmid-directed repli- cation) and the culture was incubated overnight to obtain a population of integrants.

The culture was then diluted 1: 105 in LB medium without antibiotic and the tempera- ture shifted to 28°C to stimulate homologous recombination and excision of the inte- grated plasmid through plasmid-directed replication. After 12-15 hours of incubation at this temperature, the cell culture was plated at various cell concentrations at 37.5°C without erythromycin selection. Colonies were transferred by use of tooth sticks to plates containing 10 ug of erythromycin per ml. Erythromycin sensitive colonies were analysed for presence of genetic label sequence by PCR, using PCR primers hybridising to the chromosomal DNA sequence on each side of the label sequence. Finally, the integrated state of the genetic label sequence was verified by Southern hybridisation.