WO1998013692A1

WO1998013692A1 - Microorganism separation system

Info

Publication number: WO1998013692A1
Application number: PCT/GB1997/002583
Authority: WO
Inventors: Pradip Dahyabhai Patel; John Haines
Original assignee: The Minister Of Agriculture Fisheries & Food In Her Britannic Majesty's Government Of The United Kingdom Of Great Britain & Northern Ireland
Priority date: 1996-09-28
Filing date: 1997-09-23
Publication date: 1998-04-02
Also published as: GB9620279D0; AU4313497A

Abstract

A method for detecting a microorganism, such as pathogenic bacteria, in a sample, which method comprises culturing said sample in a medium in which said microorganism can multiply in the presence of a surface which carries a binding member which binds said microorganism, and subsequently detecting the presence of said microorganism. The combination of the separation and culture steps enables time to be saved in the process. The method is suitably effected in a flask and the surface may be provided by an internal surface of the flask or by means of a membrane inserted into the flask. Kits for use in the method of the invention are also described and claimed.

Description

Microorganism Separation System

The present invention relates to a method for the separation, concentration and detection of microorganisms, in particular pathogenic microorganisms such as Salmonella which may be present in low levels in substrates such as foodstuffs, and to kits for carrying out the method.

Detection of microorganisms, such as bacteria, present in substrates such as consumer goods like food, medicaments or cosmetic preparations; or samples such as clinical samples, or samples collected for public health testing purposes is generally a very long process. However, it is clearly very important that even very low levels of pathogenic bacteria are identified in such substrates as their presence could, in many instances, have significant public health consequences.

Classical culture techniques require first a pre-enrichment step in which the substrate is mixed with a non-selective growth medium under culture conditions for a period of up to 24 hours during which time any damaged but still viable bacteria may repair. During this period however, all microorganisms, whether pathogenic or not, will multiply and may provide a high level of background "noise" which can hinder detection of a target species.

In order to reduce the background noise to an acceptable level, a selective growth amplification technique is used in which the broth from the pre-enrichment step is mixed with a selective growth medium which favours the target organism under culture conditions for a period of up to 48 hours. After this period, the mixture is plated onto selective diagnostic agar medium, in order to see if any potentially target colonies grow. If potential colonies can be identified after a suitable period of time, confirmation of the identity of the colony for example using biochemical identification techniques and ultimately serology, must be carried out.

This process can take anything up to 5 days to complete. Delays of this type are unacceptable in situations where, for example, the substrate comprises a degradable foodstuff which has a limited shelf life.

Alternative commercial techniques (e.g. ELISAs, DNA probes and impedance) can detect the presence of microorganism at levels as low as approximately 10⁵-10⁶ cfu per ml, which means that they still require at least 24 hours, but more often 48 hours, of cultural enrichment prior to rapid detection of the organism (Patel & Williams, (1994) "Evaluation of commercial kits and instruments for the detection of foodborne pathogens and bacterial toxins" in "Hapid .Analysis Techniques in Food Microbiology". Ed. P.D. Patel. Blackie Academic, Glasgow).

Despite extensive research, conventional and rapid methods for the detection of foodborne pathogens are still limited in terms of speed and reliability. Two significant problems exist that do not allow reliable results to be obtained within 24 hours. Firstly, the level of competing flora is generally higher than that of the pathogen. Following the pre-enrichment, the ratio of signal to noise is usually very low and the background flora, particularly the close phylogenetically related species, may show cross-reactions in the detection techniques. Secondly, the presence of food components can also interfere during the assay procedure.

Attempts have been made to separate target bacteria from other material prior to culture. For example, EP-A-0489920 describes a process in which antibodies are used to capture bacteria which are separated and subsequently cultured. Separation of target cells from a mixed population using magnetic beads of microspheres is also known, for example from US Patent No. 4,230,685, EP-A-605003 and P.D. Patel (1 94) Microbiological applications of Immunomagnetic techniques in "Rapid Analysis Techniques in Food Microbiology", Ed P.D. Patel, Blackie Academic & Professional. Glasgow, pp 104-131. Magnetic beads may be coated with antibodies which are specific for particular cell. When beads are added to a sample, any target cell present will be bound to the surface of the beads. The beads can then be removed from the remainder of the sample using magnetic separation. After separation, the cells are then cultured to allow them to reach measurable levels.

As outlined above, most detection methods require a threshold number of target cells of around, 10⁵-10⁶ organisms per ml. In addition, they can tolerate only certain numbers of competitor organisms. Failure to meet these requirements can lead to false-negative and false-positive results. US Patent No. 4,933,410 discloses a method for modifying a polystyrene substrate so that it allows for the covalent linking of a macromolecule such as a specific binding member. The substrate may then be used in diagnosis and therapy for separating cells from blood or other physiological fluids or dispersed tissue.

However, the use of such techniques in the separation of microorganisms, the cell walls of which have a quite different structure and constitution to the cell membranes found in multi- cellular organisms, has not hitherto been disclosed.

US Patent No. 5,389,521 describes a technique whereby bacteria from clinical samples are cultured in the presence of a solid carrier in such a way that bacteria bind to the carrier and cause agglutination of the carrier which may be detected. This technique is applicable only to low volume samples and can be used only to detect particular bacteria which bind to particular carbohydrate compounds.

The applicants have derived a novel strategy for the real-time separation and concentration of viable microorganisms which allows detection and recovery of target microorganisms from the early stages of incubation in primary enrichment broths using a simple and robust system. This will improve the reliability and speed of testing such as pathogen testing.

The present invention provides a method for separating and concentrating a microorganism in a sample, which method comprises cultuπng said sample in a medium in which said microorganism can multiply in the presence of a surface to which is bound a binding member which binds said microorganism, and directly detecting the presence of microorganism on said surface.

The process of the invention is suitably employed in a pre-enrichment step and/or selective enrichment step in a process for detecting microorganisms, so that pathogen isolation if concomitant with growth, thus reducing the length of time taken for the process. Suitably, the method of the invention is carried out in a container, such as a flask, the surface of which has been modified so that it carries the binding member. Alternatively, a membrane which carries the binding member may be placed inside the flask prior to the introduction of the sample. Examples of such containers include small scale reagent flasks, for example of from 25-1000 ml volumes, or other commercially available culture-type vessels (e.g. tissue culture flasks and roller bottles), depending upon the volume of the sample for testing.

The volumes of sample in which the method may be employed may if required be relatively high for sample evaluation, for example from 20-500 ml, suitably for 25 - 225 ml samples.

The presence of microorganisms is detected directly on the surface, for example on the sides of the flask (which are preferably of a transparent material) or the membrane itself. This allows the process to be effected more rapidly and involves fewer processing operations.

Suitably, the surface is a substantially planar surface.

Preferably the surface is a hydrophilic affinity surface as this results in the specific agglomeration and growth of target organisms present at the surface, which is conducive to the culture process. Examples of such surfaces include modified polystyrene surfaces for example as described in US Patent No. 4,933,410. Containers comprising flasks of suitable material are sold as "Microcellector flasks" by Applied Immunosciences Inc., Menlo Park California.

Alternatively the surface may comprise a regenerated cellulose membrane. Although such membranes have been used hitherto for simple ultra filtration purposes or general filtration purposes, they have not previously been used as hydrophilic immunoaffinity support mediums for the immobilisation of specific viable microorganisms as in the method of the present invention. Particular membranes which are useful in the method of the invention include those having a pore size which allow proteins to pass through but which are too small to allow passage of the target bacterial species. In general therefore, membranes with pore sizes which do not exceed O.όmicrons may be preferred.

Binding members are suitably specific for the target microorganism. These may include primary monoclonal or polyclonal antibodies as well as binding fragments such as Fab or F(ab')2 fragments thereof, and other binding proteins such as lectins.

Immobilisation of binding members such as antibodies or binding fragments thereof on the said surface may be carried out in various way. These include (a) direct non-specific adsorption; (b) covalent coupling via a spacer chemical linkage such as a hydrocarbon chain and (c) by first binding an antibody binding protein such as Protein A or Protein G to the surface before application of the binding antibody. In a preferred embodiment, a protein comprising an antibody binding domain and a surface binding domain such as a cellulose binding domain, is applied to the surface, and the binding antibody applied subsequently. The said protein is suitably designed so as to ensure that the antibody against the target microorganism is orientated so as to ensure good binding to the target by way of the F(ab)₂ portion of the antibody and at an effective distance from the membrane surface. Even coverage of the surface is also preferred to avoid "patches" where target organisms may not bind.

A particularly preferred protein for use in attaching an antibody to a nitrocellulose membrane comprises a cellulose binding domain-Protein A conjugate obtainable from Sigma Chemical Co. under the trade name Cellulose binding domain Protein A fusion protein (CBD-Protein A).

Once the binding member is fixed to the surface, remaining binding sites are suitably blocked using a blocking agent such as casein, as is understood in the art.

Detection of the microorganisms can be effected using techniques known in the art. A diagnostic or selective detection system is preferable as this will eliminate any non-target organisms which become attached to the surface as a result of non-specific binding.

For example, the microorganisms may be made visible, for example by applying selective diagnostic growth media such as xylose lysine desoxycholate (XLD) agars which result in the production of visible colonies whose colour depends upon the nature of the organism. This media may be applied directly to the surface, for example by applying a thin coat to the surface of the reaction flask where this constitutes the surface, or by placing a porous membrane forming the surface directly onto an agar plate. Alternatively, the membrane may be blotted onto the plate as is conventional in the art. Alternatively, direct or indirect labels may be applied to the microorganisms using techniques such as ELISA Suitably the labels are administered by way of a binding element which attaches itself to the microorganisms In particular, the binding element will be specific for a particular organism and may comprise an antibody or antibody binding fragment The label means will be attached to the binding element Labels may be able to generate a visible signal directly such as particulate gold or latex labels or cheπuluminescent labels or bioluminescent labels such as the luciferase/lucifenn system Fluorescent labels which become visible when illuminated with light of a particular wavelength such as ultraviolet light, or radioactive labels may also be used Yet another alternative comprises an enzyme label such as horse radish peroxidase and phosphatase which acts as an indirect label by changing the colour of an applied substrate or the vitamin biotin which can be detected as a result of its reaction with enzyme-linked avidin or streptavidin

Other possible detection systems include those which detect cell nucleotides such as ATP, or enzymes such as adenylate kinase for example using biolummescence reagents such as the luciferase/lucifenn system One such assay is descnbed in WO 96/02666 In addition, the use of labelled DNA probes or amplification techniques such as the polymerase chain reaction (PCR) can be used to detect specific target microorganisms

Preferably however, the detection system used is one which requires minimal additional processing or manipulation For example, where the surface to which the microorganism is bound compπses the surface of a transparent flask, the use of a visible label or fluorescent label may be preferred In addition, the simple application of diagnostic growth mediums can produce visible colonies within the flask which can be detected readily Where the surface compnses a membrane contained m a flask, it will generally be easier to remove the membrane from the flask before the detection step

Specifically, the inventors have successfully produced anti-Salmonella antibody sensitised Microcellector flasks and assessed their efficiency in pure and mixed populations containing .Salmonella In preliminary studies, approximately 98% specific depletion was achieved from 4 ml of a pure culture (5 enter, tiώs at approximately 10⁵ cfu/ml) within a 6 hour penod A similar level of specific depletion of S enterxttdxs (approximately 97%) was achieved from 4 ml of a mixed culture (S. enteritidis and E coli in equal numbers at approximately 10⁵ cfu/ml) within the same period. The level of E. coli remained relatively constant during the separation procedure, indicating only a low level of non-specific binding at this concentration. Although there was no improvement in the specific depletion of Salmonella using the Protein G mediated antibody linkage, this procedure may be of value in reducing the level of non-specific binding of competing organisms .

The fact that high level of depletion of Salmonella over the 6 hour period was achieved in both pure and mixed cultures, indicates that the procedure has good specificity and also a high affinity between the antibody and the target organism.

The time required to specifically separate Salmonella may be reduced by further optimisation of the test conditions (e.g. antibody concentration, incubation temperature, use of mixing etc.).

However, these parameters will be readily determinable by the skilled person. For instance, antibody concentrations of from 0.1 to lOOμg/ml and temperatures in the range of from 20 to

37°C may be suitable.

Further enhancement of the specificity of Salmonella separation in accordance with the invention was achieved using a separation procedure involving a recombinant protein

(comprising Protein A-coupled to cellulose binding protein) bound to cellulose membranes.

Qualitatively, this technique has shown significant specific separation of S. enteritidis from a mixed culture containing S. enteritidis and E coli at equal levels (approximately 10⁵ cfu/ml) within a 1 hour period. A high signal-to-noise ratio is essential for the successful application of a rapid separation and concentration technique particularly in primary food enrichment broths. Therefore, this system may be used in combination with the Microcellector flasks described above to produce a simple, robust and rapid system for the isolation of Salmonella.

In addition however, the applicants have successfully separated Salmonella and a strain of E. coli, 0157 from pure and mixed cultures using a recombinant protein mediated antibody binding to cellulose membranes. The separation achieved in this instance was highly specific, with no significant binding of the competing organism. This high signal-to-noise ratio is important for the successful application of the invention during the growth of pathogens in primary food enrichment broths. The cellulose membrane, inserted in a normal tissue culture flask, allowed specific separation of the target pathogens {Salmonella and E. coli 0157) whilst minimising the level of non-specific binding of competing organisms. In this context, specific binding of Salmonella and E. coli 0157 was significant, as indicated by the high level of depletion (range 70-90%), with a corresponding reduction in the non-specific binding of competing organisms (indicated by a lower depletion ranging from 0-40%) from both pure and mixed cultures. The technique was shown to be highly sensitive as the separation procedure was carried out during concomitant growth of bacterial cells (starting level approximately 10 cfu/ml) in the pre-enrichment broth.

Kits for carrying out the method of the invention form a further aspect of invention. The kits may comprise an element which provides a suitable surface, such as a nitrocellulose membrane or a Microcellector flask, and a suitable binding member or range of binding members for various microorganisms which are to be detected. These binding member(s) may be fixed to the surface, or may be supplied separately with instructions for their administration. Other reagents which may be included in the kits include detection reagents such as labelled antibodies, diagnostic growth media or the like. In addition, it is possible that the kit will include growth media in which the target microorganisms will multiply, suitably growth media which favours the growth of the target microorganism.

The method in accordance with the invention may be used to detect a range of microorganisms including bacteria (both gram positive and gram negative bacteria), bacterial spores, yeasts, moulds, fungal spores, viruses, protozoan cells such as Cryptosporidium, and oocysts. It is particularly useful in the detection of organisms which cause diseases. In particular, pathogenic bacteria such as Salmonella, Lister ia, Campylobacter, pathogenic strains of E. coli such as VTEC, and Staphylococcus aureus may be detected.

By combining the enrichment and separation and concentration stages, the time taken to obtain a result from an analysis can be substantially reduced. The method may be applicable in a wide range of industrial applications, for example, in the food and beverage industries, water, agriculture, medicare and pharmaceutical industries as well as in public health monitoring. The invention will now be particularly described by way of example with reference to the accompanying diagrammatic drawings in which:

Figure 1 shows flasks which have contained Salmonella samples, one of which is antibody sensitised to Salmonella and the other which has not;

Figure 2 shows the binding of S. enteritidis from a mixed culture containing E. coli on to cellulose membranes treated with and without antibody to Salmonella, and

Figure 3 shows the binding of S. enteritidis from a mixed culture(top), and of £. coli from a pure culture (bottom) on to cellulose membranes treated with antibody to Salmonella.

EXAMPLE 1

Microcellector Separation Technique for Salmonella

A commercially available polyclonal antibody against Salmonella was immobilised directly or indirectly (via a Protein G linker) onto the Microcellector flask. The coated flask was used to specifically separate and concentrate Salmonella from pure and mixed cultures. Cell depletion was monitored by sampling the nonadherent cell population in suspension and plating on to appropriate microbiological media. The bound cell population was detected by adding a thin layer of a diagnostic agar to the flask and monitoring the characteristic colony appearance.

Specifically 5 ml of antibody solution (lOμg/ml, Bactrace goat anti -Salmonella CSA- 1, Kirkegaard and Perry Laboratories) was added to a Microcellector surface activated T-25 cell culture flask (Applied Immune Sciences). This surface of this flask has an activated surface suitable for the covalent immobilisation of antibodies and other protein ligands.

Alternatively, similar activated flasks were first coated with Protein G or non-viable Protein G containing Streptococci by adding 5 ml of Protein G (10 μg/ml Sigma) or Protein G cell suspension (10 μg/ml Sigma) to the flask and incubating at ambient temperature for 2 hours. The above-mentioned antibody was then immobilised to the flask via the Protein G-antibody linkage. Each flask was gently rocked to completely wet the surface and incubated at ambient temperature for 2 hours on a flat, vibration-free surface to allow binding of the antibody The flask was washed with 3 x 10 ml portions of 0 01 M phosphate buffered saline (PBS, pH 7 4), pπor to the addition of 5 ml of blocking solution (approximately 10⁸ cfu/ml heat-killed E colt R6 reconstituted in 0 5 % Bovine serum albumin) The flasks were incubated at ambient temperature for 2 hours and then washed with 3 x 10 ml portions of PBS The coated flasks were stored in PBS at 4°C pπor to the cell selection procedure Each flask was aspirated and 4 ml of a bacterial cell suspension added (either S enteritidis RI or E colt R6 or a mixture of both m stenle distilled water at approximately 10⁵cfu/ml) The flasks were incubated at ambient temperature for 24 hours Controls were set-up in non-treated glass vials (l e without the antibody) in order to monitor the levels of cells in the test samples over the separation penod At time intervals of 1, 6 and 24 hours, the flasks were agitated gently in order to resuspend non-adhered cells, and a portion (10 μl) was removed for enumeration on a selective diagnostic agar (Xylose Lysine Desoxycholate, (XLD) Oxoid) Each flask was washed with 6 x 10 ml portions of PBS, and the binding surface of the flask overlaid with tempered diagnostic agar (XLD at 48°C) Both the agar plates and flasks were incubated at 37 °C for 24 h The colonies on the agar plates were counted and the extent of depletion calculated The flasks were observed for bactenal growth and characteristic colony appearance

Table 1 shows the extent of depletion of pure cultures of S enteritidis RI and E coli R6 in Microcellector flasks coated with polyclonal antibody to Salmonella

TABLE I

Depletion of S enteritidis- and E coli in antibody coated Microcellector flasks Time -Antibody S ent Depletion E coli Depletion

00 linkage (cfu/ml) (%) (cfu/ml) (%)

Direct 1 7xl0⁴ 61 3 OxlO⁵ 80 Protein G 2 lxlO⁴ 52 5 OxlO⁵ 67

Protein G Cell Suspension 1 OxlO⁴ 77 5 OxlO⁵ 67

Control 44xl0⁴ 15xl0⁶

Direct 5 OxlO³ 90 2 OxlO⁶ 35 Protein G 6 OxlO³ 88 1 6x10⁶ 48

Protein G Cell Suspension 6 Ox 10 88 1 9xl0⁶ 39 Control 5 OxlO⁴ 3 lxlO⁶

24 Direct <10³ >99 1 6xl0⁶ 83

Protein G 1 OxlO³ 99 4 9xl0⁶ 48 Protein G Cell Suspension 1 OxlO³ 99 2 8x 10° 71

Control 7 6xl0⁴ 9 5x10 -,6°

No significant difference in performance was found between the three immobilisation methods (direct, Protein G and Protem G cells), with approximately 90 and 99% depletion of Salmonella occumng after 6 and 24 hours respectively A degree of non-specific binding was observed for E coli (range 35-83%) depending on the length of incubation in the Microcellector flask

Table II shows the results of the repeat expenment using Microcellector flask directly coated with the polyclonal antibody to Salmonella The control vials did not contain the antibody TABLE II

Depletion of S. enteritidis and E. coli in antibody coated Microcellector flasks

Time Sample S. ent Depletion E.coli Depletion

<h) (cfu/ml) (%) (cfu/ml) (%)

1 Test 4.4xt0⁴ 55 2.0xl0⁶ 67

Control 9.8xl0⁴ 6.0x10⁶

6 Test 1.4xl0⁴ 84 l. lxlO⁶ 83

Control 8.9xl0⁴ 6.4x10⁶

24 Test <10³ 98 1.7xl0⁶ 70

Control 6.4x 10⁴ 5.7xl0⁶

As previous results (Table I), progressive depletion of Salmonella was obtained over the incubation period, with up to 98% depletion after 24 h. The level of E coli remained relatively constant over the same period of time but showed a degree of non-specific binding ranging from 67-83%.

Table III shows results of further experiments using pure and mixed cultures containing S. enteritidis and E. coli, and Microcellector flasks directly coated with the polyclonal antibody to Salmonella.

TABLE III

Depletion of Salmonella from pure and mixed cultures

Time Sample B '. ent Depletion E. coli Depletion

(h) (cfii/ml) (cfu/ml) (%)

Pure culture

Test 2.6x10⁴ 64 l.OxlO⁵ 0

Control 7.3xl0⁴ 1.0X10⁵

Test 1.2xl0⁴ 98 1.8xl0⁵ 10

Control 5. OxlO⁵ 2.0xl0⁵

Mixed culture

Test 4.3xl0⁴ 34 .OxlO⁵ 17

Control 6.5xl0⁴ l.2xl0⁵

Test 2.7x 10⁴ 97 1.3x10^s 58 Control 9.0x10⁵ 3.0xl0⁵

A similar rate of depletion of Salmonella was observed from both the pure and mixed cultures, with up to 98% depletion after a 6 hour period. In contrast, the levels of E. coli remained relatively constant, indicating minimal non-specific binding to the flask (up to 58% in 6 hours).

Further specificity of the Microcellector flasks, directly immobilised with the anti-Salmonella antibody to Salmonella, for S. enteritidis is shown in Table IV and Figure 1. TABLE IV

Depletion of Salmonella with and without antibody

Time Sample S. ent Depletion

(h) (cfu/ml) (%)

+ .Antibody 5.5x 1C 15

Control 6.5x 1C

- Antibody 7.2x 1C

Control 7.1x10'

+ Antibody 6.0x10- 90

Control 6.1x10'

- Antibody 6.1xl0⁴ 31

Control 8.9xl0⁴

In the antibody coated flask, the depletion of Salmonella was 90% after 6 hours compared with 31% in the control flask without the antibody. In the control flask, however, reduction in actual cell numbers was insignificant over the same 6 hour period.

Example 2

Cellulose Membrane Separation Technique for Salmonella The following cellulose membrane-based separation technique was developed with a view to enhancing specific separation of pathogens using uncoated Microcellector flasks, or indeed any other flask. The technique involved the use of a recombinant protein for the immobilisation of antibodies on the surface of a cellulose membrane. 5 ml of cellulose binding domain-Protein A solution (lOOμg, Sigma) was added to a petπ dish containing a cellulose membrane (Spectra/Por 3, MWCO 3500, Spectrum) and incubated at ambient temperature for 1 hour The membrane was washed with 3 x 10 ml portions of PBS prior to the addition of 5 ml of rabbit anti-Salmonella polyclonal antibody (1 200 dilution, Biogenesis) The membrane was incubated at ambient temperature for 1 hour, washed with 3 x 10ml portions of PBS, and the remaining binding sites blocked with 10 ml of 0 5% casein solution The membrane was incubated at ambient temperature for 1 hour and then washed with 3 x 10ml portions of PBS Cell separation was earned out by adding 5 ml of a bactenal cell suspension (S enteritidis RI or £ coli R6 or a mixture of both in steπle distilled water) at approximately 10⁵ cfu/ml The membrane was incubated at ambient temperature for 1 hour

After this penod the membrane was πgorously washed with 6 x 10ml portions of PBS, and the moist membrane transferred to the surface of a diagnostic agar (XLD) plate The plate was incubated at 37°C for 24 hours and the bactenal growth and colony appearance monitored

Figure 2 shows the specific binding of Salmonella (black colonies) from a mixed culture (S enteritidis and E coli at approximately 10⁵ cfu/ml) using a cellulose membrane coated with antibody to Salmonella Figure 2 shows a low degree of non-specific binding (black and yellow colonies Salmonella and E coli respectively) from the same mixed culture using a cellulose membrane that was not treated with antibody Figure 3 shows the specific binding of Salmonella (black colonies) from a mixed culture (S enteritidis and E coli at approximately 10⁵ cfu/ml), and the extent of non-specific binding of E coli (pure culture at approximately 10⁵ cfu/ml), using anti-Salmonella antibody- coated cellulose membranes From the mixed culture, there was significant binding and subsequent growth of Salmonella, with no visual observation of any growth of £ coli The non- specific binding of £ coli from the pure culture was patchy, mainly around the outer edge of the membrane The agar medium colour changed from red to yellow due to acid production by E coli Example 3

Separation of Salmonella from Tissue Culture Flasks containing Cellulose Membrane Inserts

A modified technique to that described in Example 2, which allows scale-up of the procedure for use with larger sample volumes was effected as follows. Cellulose binding domain-Protein A solution (4 ml of lOOμg/ml, Sigma) were added to an uncoated tissue culture flask containing a cellulose membrane insert (Spectra/Por 3, MWCO 3500, Spectrum) and incubated at ambient temperature for 1 hour. The membrane was washed with 3 x 10 ml portions of PBS prior to the addition of rabbit anti-Salmonella polyclonal antibody (4ml, 1:200 dilution, Biogenesis). The flask was incubated at ambient temperature for 1 hour, washed with 3 x 10ml portions of PBS, and the remaining binding sites blocked with 10 ml of 0.5% casein solution. The flask was incubated at ambient temperature for 1 hour and then washed with 3 x 10ml portions of PBS.

Cell separation was assessed by adding either 5 ml or 25ml of a bacterial cell suspension in various media at approximately lOcfu/ml and 10 or 10⁵ cfu/ml for target cells and competing cells, respectively. The inoculated flask was incubated at 37°C for 6 hours with gentle mixing (incubator shaker, new Brunswick Scientific), Controls were set up in non-treated glass vials (i.e. without the antibody sensitised membrane) in order to monitor the levels of cells in the test samples over the separation period. After 6 hours, the flask was agitated in order to resuspend the non-adhered cells and a portion (lOOμl) was removed for enumeration on selective diagnostic agar (XLD). The plate was incubated at 37°C for 24 hours and the resulting colonies counted. The extent of depletion was then calculated.

Table V compares the extent of depletion of S. enteritidis RI and E. coli R6 from static cultures (5ml starting level of approximately 10⁵ cfu/ml in sterile distilled water) and during concomitant growth (also 5ml starting level of approximately 10 cfu/ml in TSB) . TABLE V

Separation of Salmonella from static culture and during concomitant growth in TSB

Time Sampl le Growth Flask Control Depletion

(h) (cfu/ml) (cfii/ml) (%)

0 S ent Static 1 4xl0⁵ -

£ coli Static 1 9xl0⁵ -

S ent Concomitant 1 4x10' -

£ coli Concomitant 1 9x10' -

6 Sent Static 1 3xl0⁵ 2 2xl0⁵ 40

E coli Static 3 4xl0⁵ 3 5xl0⁵ 4

S ent Concomitant 2 3xl0⁴ 1 3xl0⁵ 82

E coli Concomitant 3 lxlO⁵ 3 1x10 1

A significantly greater depletion of Salmonella (82%) occuned dunng concomitant growth in the flask compared with the static culture (40%) in the flask The levels of £ coli depletion remained very low (<4%)

Table VI shows the separation of 5 enteritidis RI and £ coli R6 during concomitant growth over 6 hours using 5 and 25 ml volumes of TSB inoculated with the starting level of approximately 10 cfu/ml bacteria Once again, the flasks contained membrane sensitised with antibody to Salmonella

TABLE VI

Separation of Salmonella dunng concomitant growth in 5 and 25 ml samples

lme Sample Sample Flask Control Depletion

(h) Volume (cfu/ml) (cfu/ml) (%) (ml)

0 S ent 5 1 lxlO¹ .

25 -

- 25 _

S ent 5 3 2x10" 1 lxlO⁵ 71

25 3 6xl0⁴ 67

Ecoli 5 2 2xl0⁵ 3 2xl0⁵ 27

25 2 2xl0⁵ 27

Similar rates of depletion of Salmonella were obtained in both the 5 and 25 ml samples, with no change in the level of non-specific binding of £ coli

Tables VII and VIII show the separation of Salmonella dunng concomitant growth of pure and mixed cultures, respectively in 25 ml buffered peptone water (BPW) and the level of nonspecific separation of a range of competing organisms The flasks contained membrane sensitised with antibody to Salmonella Table VII Separation of S Enteritidis RI, S. typhimurium R5, £ coli R6, C.freundii G2 and L monocytogenes B2 dunng concomitant growth of pure cultures in BPW

Time Sample Flask Control Depletion

(h) (cfu/ml) (cfu/ml) (%)

0 Sent 1 8x10' -

S typh 5 2x10° -

Cfreundn 1 7x10' -

L mono 1.7x10' -

6 S.ent 1.9xl0⁴ 1 9xl0⁵ 90

S.typh 4.0x10³ 3.6xl0⁴ 89

E.coli l. lxlO⁵ 1 3xl0⁵ 15

C.freundii 5.7xl0⁴ 4 8xl0⁴ 0

L mono <10³ <10³ -

The specific separation of Salmonella (89-90%) was significantly greater than the non-specific depletion of the competing organisms (0-15%)

TABLE VIII

Separation of S enteritidis- RI, S typhimurium R5, £ coli R6, Cfreundii G2 and

L monocytogenes B2 dunng concomitant growth of mixed cultures in BPW

Time Sample Flask Control Depletion

(h) (target/competitor) (cfu/ml) (cfu/ml) (%) target competitor target competitor target competitor

0 S ent/E coli R6 1 5x10' 1 1x10' 1 5x10' 1 1x10' - -

S ent/E coli R6 1 2x10' 1 8xl0⁵ 1 2x10' 1 8xl0⁵ - -

S typh/E coli R8 1 3x10' 1 2x10' 1 3x10' 1 2x10' - -

S typh/E cob R8 1 2x10' 1 6xl0⁵ 1 2x10' 1 6x10' - -

6 S ent/E coli R6 8 OxlO³ 9 9xl0⁴ 2 9x1 ⁴ 1 OxlO⁵ 72 1

S ent/E coli R6 5 9xl0³ 1 OxlO⁸ 2 lxlO⁴ 1 0x10^s 72 0

5 ι τ-Λ/ co/i R8 4 OxlO¹ 1 5x10" 4 lxlO⁴ 2 3xl0⁴ 90 35

S typh/E coli R8 3 OxlO³ 4 OxlO⁷ 1 3xl0⁴ 3 OxlO⁷ 77 0

In mixed cultures, the specific separation of Salmonella (72-90%) was achieved after 6 hours incubation in BPW even in the presence of 10⁸ cfu/ml competing organisms

Example 5

Separation of £ coli

The method of Example 4 was repeated using vanous cultures of £ coli The membranes introduced into the flask in this instance had been sensitised in a similar manner but using an antibody to E coli 0157

Table IX shows the separation of £ coli 0157 R121 and £ coli R6 dunng concomitant growth over 6 hours using 5 and 25 ml volumes of TSB inoculated with approximately 10 cfu/ml bacteria TABLE IX

Separation of £. coli 0157 during concomitant growth in 5 and 25 ml samples

Time Sample Sample Flask Control Depletion

(h) Volume (cfu/ml) (cfu/ml) (ml)

0 E coli 0157 5 1.2x10'

25 1.3x10'

E.coli 5 1.4x10' " 25 1.6x10

6 E coli ' 0157 5 2.1xl0⁴ l . lxlO⁵ 81

25 3.7x10" 3.9xl0⁵ 91

E.coli 5 1.9xl0⁵ 3.4xl0⁵ 44

- 25 4.8xl0⁵ 5.0xl0⁵ 4

As expected, significantly greater separation of E.coli 0157 (up to 91%) was obtained in both the 5 and 25ml samples, than the separation of the non-verotoxigenic £. coli (4-44%).

Tables X and XI show the separation of £. coli 0157 during concomitant growth of pure and mixed cultures, respectively, in 25ml BPW and the level of non-specific separation of a range of competing organisms.

TABLE X

Separation of £. coli 0157 R121, £ coli 0157 R104, £ coli R6, S typhimurium R5, and

L monocytogenes B2 dunng concomitant growth of pure cultures in BPW

Time Sample Flask Control Depletion

(h) (cfu/ml) (cfu/ml) (%)

0 £. coh 0157 R121 1 1x10' -

£ co// 0157 R104 1.2x10' -

E.coli 1.3x10' -

S. typhimurium 5 2x10° -

L. monocytogenes 1 7x10' -

6 E. coli 0157 R121 1.5xl0⁴ 8.6xl0⁴ 83

E. coh 0157 R104 1.0x10" 7 OxlO⁴ 86

£.co/ι 1 1x10^s 1.3xl0⁵ 15

5. typhimurium 2 0x10" 3 6xl0⁴ 44

L. monocytogenes <10³ <10³ -

The separation of E.coli 0157 (83-86%) was significantly greater than the non-specific depletion of the competing organisms ( 15-44%)

TABLE XI

Separation of £. coli 0157 R121, £. coli 0157 R104, £. coli R6 and £. coli R8 during concomitant growth of pure cultures in BPW.

Time Sample Flask Control Depletion (h) (target/competitor) (cfu/ml) (cfu/ml) (%) target competitor target competitor target competitor

0 £.co//0157R121/£.co//R6 1.6x10' 1.7x10' 1.6x10' 1.7xlθ' -

£. CΌ//0157R121/£. CO//R6 1.6X 10' 1.5X 10⁵ 1.6x10' 1.5x10^s -

£.cσ//0157R104/£.co//R8 1.6x10' 1.7xlθ' 1.6x10' 1.7xlθ' - £.co//0157R104/£.cσ/;R8 1.6xlθ" 1.5xl0⁵ 1.6xlθ' 1.5xl0⁵ -

6 £.co//0157R121/£.cσ//R6 3.0xl0⁴ 4.0xl0⁵ l .OxlO⁵ 3.0xl0⁵ 70 0

£.CO//O157R121/£.CO//R6 3.0xl0⁴ 7.6xl0⁸ 9.8xlθ" 7.5xl0⁸ 69 0

£.co/.O157R104/£.co/;R8 3.0xl0⁵ 2.2x10" 1.1x10^s 2.0x10" 73 0 £.cσ/.0157R104/£.cσ/ιR8 l .OxlO⁵ 3 8xl0⁸ 5.0xl0⁵ 2.9xl0⁸ 80 0

Specific separation of E. coli 0157 (69-89%) was achieved even in the presence of 10⁸ cfu/ml competing organism.

Example 6

Direct Detection of Salmonella on membrane by ELISA

The direct detection of Salmonella bound to the membrane-lined flask treated in accordance with Example 4 after the 6 hour separation during concomitant growth of pure cultures using ELISA was carried out. After the 6 hour period, the flask was rigorously washed with 5 x 10 ml portions of PBS, and portions of peroxidase-labelled KPL antibody (4 ml, 1 : 1000 dilution) added. The flask was incubated at ambient temperature for 1 hour and washed as before. A soluble peroxidase substrate solution (4ml, OPD, Sigma) was then added to the flask. After 10 minutes, portions (200μl) of the test solutions were transferred to a microtitre plate and the chromogenic product measured at 450 ran in an ELISA reader (MR590, Dynatech).

Table XII shows the results obtained. TABLE XII

Direct Detection of Bound S enteritidis RI and L monocytogenes B2 by the ELISA for

Salmonella

Time Absorbance at 450nm (mm) S ent L mono

0 0 0

5 0 12 0.04 10 0 20 0 07

The preliminary results for the ELISA showed a degree of specificity for Salmonella, as compared with the non-specific detection of Listeria

Claims

1. A method for detecting a microorganism in a sample, which method comprises culturing said sample in a medium in which said microorganism can multiply, in the presence of a surface which carries a binding member which binds said microorganism, and subsequently detecting said microorganism on said surface.

2. A method according to claim 1 wherein the surface is hydrophilic.

3. A method according to claim 1 or claim 2 wherein the surface comprises a polymeric material.

4. A method according to claim 3 wherein the surface comprises a polystyrene.

5. A method according to claim 3 wherein the surface comprises a cellulose membrane.

6. A method according to any one of claims 1 to 5 wherein the binding member comprises an antibody or a binding fragment thereof, or a lectin which is specific for said microorganism.

7. A method according to any one of the preceding claims wherein the binding member is attached to the surface by means of a conjugating protein.

8. A method according to claim 7 wherein the binding member comprises a cellulose-binding domain attached to an antibody binding protein.

9. A method according to claim 8 wherein the antibody binding protein comprises Protein A, or Protein G.

10. A method according to any one of the preceding claims wherein a blocking agent is applied to free binding sites on the surface.

11. A method according to claim 10 wherein the blocking agent is casein.

12 A method according to any one of the preceding claims wherein the microorganism is detected using a fluorescent label or an ELISA.

13. A method according to any one of the preceding claims wherein the surface is applied to a diagnostic culture medium and the cultures detected visibly.

14. A method according to claim 13 wherein the diagnostic culture medium comprises xylose lysine desoxycholate (XLD)

15. A method according to any one of the preceding claims wherein the presence of microorganisms is detected directly on the surface.

16. A method according to any one of the preceding claims for the detection of viable pathogenic bacteria.

17. A method according to claim 16 where the bacteria is Salmonella or £. coli.

18. A kit for use in a method according to any one of the preceding claims, which comprises a surface and a binding member which specifically bind a microorganism which is capable of fixing on said surface.

19. A kit according to claim 18 wherein said surface comprises a cellulose membrane.

20. A kit according to claim 18 or claim 19 which further comprises container means capable of holding a sample.

21. A kit according to any one of claims 18 to 20 which comprises a Microcellector flask.

22. A kit according to any one of the claims 18 to 21 wherein the binding member is fixed to said surface.