TECHNICAL FIELD

The present invention relates to methods and apparatus for the concentration and separation of microorganisms from complex sample materials, particularly from foodstuff derived materials. Bacteria and yeasts are preferred target microorganisms.

PRIOR ART

The detection of microorganisms in foods is of great interest to the food manufacturing and retailing industries, in particular in relation to avoiding instances of food poisoning, but also for quantifing desirable microorganisms in cultures etc.

Tranditional detection methods have tended to revolve around culture techniques to concentrate the organisms of interest followed by an identification step eg. as disclosed in EP-A- 0214 3^0 (BioControl syterns). Such methods are of necessity fairly slow and labour intensive. This reduces their suitability for proactive measures such as regular testing of samples of food products eg. before they leave the factory, in order to allow the bonding of contaminated batches.

More rapid techniques based have been proposed in the literature, for instance in "Rapid Analysis Techniques in Food Microbiology" (Patel, P.D. ed. Blackie) . However the practical application of eg. electrical and immunological techniques has been limited to a large extent by the complex and heterogenous nature of foodstuffs as samples. Thus it can be seen that a practical method for the rapid analysis of microorganisms in foodstuffs and other complex samples would provide a long awaited contribution to the art.

INVENTION

The present inventors have now developed a practical method for the

-analysis of microorganisms in complex samples, particularly food-derived samples, using dielectropohresis.

Dielectrophoresis is a phenomenon which is characterised, inter alia, by the motion induced by a non-uniform electric field on polarised uncharged particles UK Patent Application GB 2 071 843 (Pohl). It is a motion that is caused using alternating current fields rather than the direct current fields used in electrophoresis.

It is well known that the motion of microorganisms such as bacteria and yeasts may be manipulated using the dielectrophoretic principle; see for example Asencor et al (1990) Bioelectrochemistry and Bioenergetics 24, p203~214; Markx et al (1994) Microbiology 140, P585-5 1 and Hawkes et al (1993) Microbios 73. pδl-86, and that a variety of bacterial forms may be so controlled; see eg. Archer et al (1993) Microbios 73. pl65~172.

Additionally, the ability of dielectrophoretic apparatus to separate mixtures of particles, particularly bacteria, has been demonstrated using differential retardation of particles in a sample liquid flowing through a dielectrophoretic chamber (WO 93/20927 (BTG)). This work was focused upon the use of varying electric field to achieve separation of particular bacteria, but noted that pH and conductivity of the liquid and the frequency of the voltage affect the order and degree of retardation and thus separation.

However, the published work on electrophoresis to date has been generally been concerned with the use of relatively simple liquid samples, and in particular to the characterisation and resolution of bacteria having different dielectrophoretic characteristics. For instance in GB 2 071 843 A the sample was taken from pure cultures grown specifically for the purpose. In no instance has dielectrophoresis been practically applied in testing 'real world' food samples.

The present inventors have determined that direct positive dielectrophoretic aggregation of microorganisms cannot be achieved

from high conductivity high permittivity food and culture media broths using voltage settings of from 1 to 20 MHz at 1 to 10 volts and that dilution of broths by 1:10 does not remedy this situation. Dilution by between 1:100 and 1:100,000 was however found to produce some aggregation using these frequency and voltage parameters. For practical purposes however 1:1000 to 1:10000 dilutions gave the best results allowing up to 15 volts to be used above which point convection currents are produced by heating effects.

Dilution of samples will inevitably reduce the concentration of bacteria in a sample prior to performance of any method intended to concentrate them, and thus is in principle undesirable. This is particularly the case when dealing with samples eg. food samples, where the bacteria, if present at all, may be present only in low concentrations.

The present inventors have now provided a method for enabling complex samples, such as those derived from food, to be analysed by dielectrophoresis without first diluting them.

Thus in a first aspect of the invention there is disclosed a method of concentrating microorganisms in a food sample comprising the use of dielectrophoresis.

The concentration itself may be achieved by exposing the sample to the dielectrophoretic field to aggregate microorganisms and removing the remainder of the sample. The concentrated microorganisms may then be assayed, for instance as described below.

By 'food sample' is meant any liquid material derived from a human or animal foodstuff or potential foodstuff, including samples which have been cultured or otherwise pretreated eg. by suspension in a carrier liquid.

Preferably the method is used in for the analysis of bacteria eg. Salmonella or Listeria spp. Thus the use of dieletrophoresis for the rapid analysis of food materials for bacterial presence will open the

way for proactive intervention in the food industry at an early stage of manufacture or distribution, hereby reducing the likelihood that contaminated foodstuffs will reach the ultimate consumer.

The method is particularly applicable to samples which have a high initial conductivity, eg. greater than 1000 μS/cm ² . For such samples the method is further characterised in that the sample is treated to reduce its conductivity, eg. to below 1000μS/cm ² , before dieletrophoresis is carried out. The preferred method of conductivity reduction is desalting.

The present inventors have found that the process of desalting allows for the rapidly dielectrophoretic separability of microorganisms while minimising interference from food particles and the disadvantages associated with dilution of samples.

In a preferred method of this aspect of the invention the conductivity is reduced to 500μS/cm ² or less, more preferably to between 1 and 450μS/cm ² and most preferably to about 100 - 200 μS/cm ²

For practical purposes a conductivity of from 50 to 450μS/cm ² will be acceptable. Clearly a very low conductivity may be desirable for efficient electrophoresis, but will also require more effort or a longer time to achieve, and may also lead to a loss of some sample organisms. The desalting technique and conductivity limit chosen must therefore be a balance of these factors. It should be noted that the present inventors have successfully used employed their invention at much higher conductivities than the preferred limit of ^• less than 10 μS/cm' taught in GB 2 071 8 3 A.

The particular desalting treatment used is not limited except in as much as it must not lose substantial quantities of bacteria from the sample.

Convenient methods envisaged by the inventors include use of desalting gels (eg Bio-Gel P-6DG, BioRad Laboratories and PD-10, Pharmacia) , dialysis membranes sized to retain bacteria, ion exchange

resins, reverse osmosis, ultrafiltration or ion specific balance changing. It is also possible merely to precipitate the sample solids content using centrifugation, eg. at 4000 rev/min for 15 minutes, and then resuspend the resulting pellet in, for example, distilled water or deionised water.

The ways in which materials such as gels might be used to desalt the sample may be varied. For example, exclusion gel methods may be applied to retain materials of different sizes; ion exchange resins may be used to directly remove ions from solution; quiescent, low pressure and countercurrent dialysis may be used to effectively wash away salts; pressure driven reverse osmosis, nanofiltration, counter -diffusion or specific ultrafiltration, diafiltration, linear crossflow or depth filter configurations may be applied, or charged systems such as those used in electrodialysis or ultraosmosis applied.

The use of gel filtration has particularly advantages in that it is rapid and the gel can act as a course filter reducing the particulate nature of the sample and thereby simplifying subsequent dielectrophoresis. Cross linked dextran gels appear to be particularly effective.

Dialysis cassettes and chambers have the advantage that they are simple and convenient to use, even by relatively unskilled personnel.

Continuous flow syterns may further simplify the desalting, and hence overall microorganism concentrating process.

It should be noted that the advantages accruing from the use of a desalting step, as discussed above, will pertain to the dielectrophoresis of all high conductivity (eg. >1000 μS/cm ² ) complex liquid samples, even those of non-food origin. Thus such methods, and apparatus employing them, form a further aspect of the present invention.

The parameters of the dielectrophoretic field used typically may

include frequencies of 1 to 20MHz at 1 to 15 volts; but this may be increased to up to 100 volts or more when more than 20MHz is used. This is an advantage over prior art dielectrophoresis techniques where the conductivity of the sample would not allow such high voltages to be used. It will be realised that once the remainder of the sample, i.e. that excluding the microorganisms, is removed from the field, the microorganisms themselves may be isolated in a carrier liquid by turning off the field and passing the carrier liquid through the area where the microorganisms were aggregated. The carrier liquid is conveniently a buffer solution, a saline or water, such as deionised or distilled water.

Using the preferred method of the invention direct positive dielectrophoresis of undiluted food and culture media derived samples, such as food broths and culture broths, may be carried out giving rapid concentration of whole microorganism (eg. bacterial) content or selected microorganism component populations.

A second aspect of the present invention provides a method for assay of microorganisms in a food sample comprising a method for concentrating the microorganisms as described above followed by the step of assaying the microorgansims.

Preferably the microorganisms are transferred from the field before assaying them. The method used to assay the microorganisms is not limited to any particular group of techniques and may be as simple as direct visual observation using a microscope, including direct visual observation of the dielectrophoretic field area, eg. the ends of electrodes used to apply the field. Such techniques are well known to those skilled in the art.

A third aspect of the present invention provides a method of producing an approved batch of a foodstuff comprising producing a batch of a foodstuff in a conventional manner, taking a representative sample from the batch, assaying microorganisms in the representative sample using a dielectrophoretic method as described above, comparing the result of the assay with standard and approving

the batch in the event that the assay meets or exceeds the standard.

By 'representative sample' is meant a sample by which the quality of the whole can be judged at a statistically significant level. The standard used will depend on the nature of the foodstuff and the microoganisms assayed, but in many cases it may be require the complete absence of a given microorganism eg. Salmonella.

In a still further aspect of the invention there is provided an apparatus for performing concentration or separation of microorganisms from a complex liquid sample having conductivity over 1000μS/cm ² comprising a means for treating the sample such as to reduce its conductivity to a value of 1000μS/cm ² or less; a means for producing a dielectrophoretic field; and a means for passing the treated sample through the dielectrophoretic field. For separation purposes the apparatus will preferably include a means for controlling the electric field such as to allow it to be deenergised; a means for passing a carrier liquid through the means for producing the dielectrophoretic field in order to collect microorganisms aggregated therein, and a means for collecting the carrier liquid so passed.

The means for producing the dielectric field is not limited to any particular device other than it should allow passage of sample to and from a dielectric field sustaining area in use. The field is conveniently provided using electrodes in an electrode/field chamber; these being spaced as is known in the art such as to sustain a dielectrophoretic field rather than to produce electrophoresis.

The method and apparatus of the present invention will now be illustrated by way of example only by reference to the following non-limiting Examples and Figures. Further embodiments falling within the scope of the invention will occur to those skilled in the art in the light of these.

EIGJJRES

Figure 1: A diagrammatic representation of an apparatus of the invention wherein a microscope is used for directly observing the number of microorganisms separated from the sample liquid.

Figure 2: Graph of relative permittivity and conductivity of buffered peptone water (BPW) with dilution factor.

Figure 3: Graph illustrating the effect of dilution of BPW on the aggregation of a number of different bacteria when placed in a dielectrophoretic field.

Figure 4: Graph of relative permittivity and conductivity of Rappaport-Vassiliadis enrichment broth (RV) with dilution factor.

Figure : Graph illustrating the effect of dilution of RV on the aggregation of a number of different bacteria when placed in a dielectrophoretic field.

Figure 6: Graph of dielectric aggregation of the natural bacterial flora from homogenates of beef, chicken and cod as provided by centrifugation and resuspension as described in Example 2.

Figure 7: Graph illustrating the effect of column desalting as described in Example 3 on dielectrophoretic aggregation of bacterial cultures in TSB.

Figure 8: Graph illustrating the effect of column desalting on the aggregation, conductivity and recovery of total viable flora from beef homogenate using dieletrophoretic parameters of lOOKHz and 10 volts. as described in Example .

Figure 9: Graph illustrating the effect of column desalting on the aggregation, conductivity and recovery of total viable flora from chicken homogenate using dielectrophoretic parameters of lOOKHz and 10 volts as described in Example 4.

Figure 10: Graph illustrating the effect of column desalting on the

aggregation, conductivity and reef ery of total viable flora from cod homogenate using dielectrophoretic parameters of lOOKHz and 10 volts as described in Example 4.

Figure 11: Graph illustrating effect of dialysis cassette desalting of TSB on Salmonella count and conductivity.

Figure 12: Graph illustrating effect of dialysis chamber desalting of TSB on Listeria count and conductivity.

Figure 13- Graph illustrating effect of PD10 column desalting of S. typhi suspension on Salmonella count and conductivity.

Figure 14: Graph illustrating effect of Presto(TM) column desalting of S. typhi suspension on Salmonella count and conductivity.

Figure 15: Graph illustrating effect of Speedy(TM) column desalting of S. typhi suspension on Salmonella count and conductivity.

Figure 16: Graph illustrating effect of Speedy(TM) column desalting of S. enteritidis suspension on Salmonella count and conductivity.

EXΔMELES

Dielectrophoretic apparatus: the apparatus used for all the Examples below included a number of electrodes of different diameters and spacings. Those having narrow radii of curvature (diameter) generated the highest fields highlighted by increased aggregation adjacent the narrowest electrodes. Extent of aggregation was highest at the electrodes of diameters 1 to llμm and spacings of 30-135um, and least at those of 42- 3μm and spacing of 500μm. The aggregation of microorganisms was observed using an apparatus as shown in Fig 1 wherein an image analysis system (1) received images of the dielectrophoretic field on a microelectrode array (4) from a microscope (3) via a camera (2) and in turn this fed the switching matrix (5) of the electrodes of the array with control signals via a function generator (6) .

Example 1: The effect of dilution on dielectrophoretic samples Figs

2 & 3 show the effect of dilution on the conductivity and dielectrophoretic properties of Buffered Peptone Water (BPW) . Figs 4 & 5 show the corresponding results for Rappaport Vassiliadis (RV) medium.

Example 2: Reduction of conductivity and dielectrophoresis of food pre-enrichment samples bv centrifueation and resuspension: Samples (10 g) of beef, chicken or cod were added to 90ml volumes of sterile BPW, stomached for 1 minute (Colworth Stomacher, A J Steward Ltd) and then incubated for 18 hours at 37°C. After incubation 1ml aliquots of the homogenates were centrifuged for 15 minutes at 4000 rev/min and the pellet resuspended in distilled water. The dielectrophoretic aggregation of the natural bacterial flora in the suspensions was then measured at lOKHz, lOOKHz, 1MHz and 10MHz. Results are shown in Fig 6.

Example . Reduction of conductivity and dielectrophoresis of bacterial cultures bv use of desalting columns: Aliquots (1ml) of bacterial cultures in Tryptone Soya Broth (TSB) containing approximately 10 ⁷ to 10 ⁸ organisms/ml were passed through commercially available desalting columns (Bio-Gel PβDG, BioRad Laboratories) that had been sterilised by autoclaving at 121°C for 15 minutes. Eluates were collected in lOOμl fractions and the dielectrophoretic aggregation of bacteria in each fraction was measured at lOOKHz and 10 volts. Results are shown in Fig 7•

Example 4: Reduction of conductivity and dielectrophoresis of food pre-enrichment samples bv use of desalting columns: Samples (lOg) of beef, chicken and cod were added to 90ml volumes of sterile BPW and incubated for 18 hours at 37°C. After incubation 1ml aliquots of the enrichment broths were passed through columns containing 2ml of hydrated desalting gel (Bio-Gel P-6DG, Biospin) and the eluates collected in lOOμl fractions. Each fraction was analysed for conductivity, total viable count (using a spread plate method on Tryptone Soya Agar (TSA)) and for dielectrophoretic aggregation at

100 kHz and 10 volts. The results are shown in Figs 8, 9 & 10.

Example 5: Reduction in conductivity of bacterial cultures and food pre-enrichment samples: comparison of desalting columns, dialysis cassettes and dialysis chambers:

METHODS

Strains of S typhimurium (55698) and S enteriditis (P48120) resistant to streptomycin and nalidixic acid were cultured overnight in TSB. Portions (2.5 ml) of the cultures were applied to columns (1.5 cm x 1 cm) of desalting gel (PD10, Pharmacia) and eluted with distilled water. Aliquots (0.5 μl) of the eluate were collected and the conductivity determined using a Compact Conductivity Meter (Horiba Ltd) . The salmonellae in each fraction were enumerated by serial dilution in 1/4 strength Ringers solution and surface plating of 0.1 ml onto TSA containing streptomycin and nalidixic acid. The plates were incubated at 37°C for 24 hours.

The desalting columns were also evaluated using pre-enrichment cultures of coconut and chicken inoculated with the two antibiotic-resistant strains of Salmonella. 25 g samples of each food were added to 225 ml of BPW. The pre-enrichment cultures were then inoculated with 100 μl from the 10 ^"5 dilution of an overnight culture of either S enteriditis or S typhimurium. The pre-enrichments were incubated overnight at 37°C. Samples (0.5 ml) of the incubated pre-enrichments were applied to 1 cm-deep columns and fractions collected for conductivity measurement and enumeration of Salmonella, as before.

Dialysis cassettes (Slide-A-Lyzer, Pierce & Warriner) were filled with 1 ml of TSB inoculated with S typhimurium and dialysed against 500 ml of distilled water at room temperature (RT) and 4°C, with stirring. Aliquots (100 μl) were removed at intervals for conductivity measurements. Salmonella were enumerated before and after dialysis, as before.

Double-sided dialysis chambers (Spin Biodialyser, Sialomed Inc) , fitted with 0.1 μm or 0.45 μm polycarbonate membranes, were filled with 1 ml of TSB inoculated with L monocytogenes and dialysed against 500 ml of distilled water RT, with stirring. Listeria were enumerated, before and after dialysis, by serial dilution in 1/4 strength Ringers solution and surface plating of 0.1 ml of appropriate dilutions onto PALCAM agar (Oxoid) .

RESULTS

Using the PD10 desalting gel, the conductivities of TSB cultures containing 9-6 x 10 ⁷ cfu/ml of S. enteriditis and 1.2 x 10 ⁸ cfu/ml of S typhimurium were reduced from >2,000 to 84 μS/cm ² and 117 μS/cm ² , respectively, in the second fraction from the column, within 5 min. The corresponding recoveries of S enteriditis and S typhimurium in fraction 2 were 100% and 17%. respectively.

In the case of pre-enrichment cultures of coconut and chicken inoculated with the two serotypes of Salmonella, the conductivity of the enrichments was reduced from >2,000 μS/cm ² before desalting to between 10 and 14 μS/cm ² in fraction 2, after desalting. For coconut and chicken pre-enrichments containing 5-5 x 10 ⁵ cfu/ml S. typhimurium, the recoveries after desalting were 6% and 10%, respectively. For coconut and chicken pre-enrichments containing 2.1x 10 ⁵ S enteriditis, the recoveries were 7 and 20%, respectively.

At RT, the Slide-A-Lyzer dialysis cassette reduced the conductivity of TSB from >2,000 to 530 μS/cm ² after 1 hour, and to 155 μS/cm ² after 6 hours. The log ₁₀ Salmonella count increased slightly, from 8.6 initially, to 9-0 after 6 h. At 4'C, the rate of desalting was slightly lower and the final conductivity slightly higher (Fig 11). The Spin Biodialyser gave a similar rate of desalting to the Slide-a-Lyzer (Fig 12) , but the final conductivity was considerably lower (36 μS/cm ² ). There was little difference between the 0.1 μm and 0.45 μm membranes in terms of the rate of desalting and final conductivity. The Listeria count decreased slightly, from 8.6

initially, to 8.1 after 6 hours.

Thus desalting columns are effective as a means of rapidly reducing the ionic concentration of food enrichments to a level that would permit positive dielectrophoresis (< 500 μS/cm ² ). The recovery of Salmonella after passage through the columns varied with the serotype. Dialysis cassettes and chambers are potentially a simple and convenient method for desalting food pre-enrichments, but the rate of desalting of TSB with the Slide-A-Lyzer and Spin Biodialyser was much slower than that using the gel column system. The double-sided dialysis chambers with built-in magnets for use with a magnetic stirrer gave a greater reduction in conductivity than the dialysis cassettes.

Example 6: Reduction in conductivity of food pre-enrichment samples using 3 ml capacity dialysis cassettes:

METHOD

Samples (25 g) of chicken and coconut were weighed into 225-ml volumes of BPW and incubated overnight at 37 ^* C. After incubation, the enrichment cultures were inoculuted with 1 ml of a 1/1000 dilution of an overniqht culture of S typhimurium in TSB. Samples (1 ml) of the enrichment cultures were transferred to dialysis cassettes and dialysed against 1 1 of distilled water. Aliquots (100 μl) of enrichment culture were removed from the dialysis cassettes at hourly intervals for conductivity measurements using a card type-conductivity meter (Horiba) . Salmonella counts were determined before and after dialysis by preparing serial decimal dilutions of the suspension in Maximum Recovery Diluent (MRD) and surface plating 0.1 ml onto Xylose Lysine Desoxycholate agar (XLD) .

RESULTS

The cassettes reduced the conductivity of chicken and coconut enrichment cultures from >2000 to 260 μS/cm ² after 1 hour at room emperature (RT) . Thereafter the rate of desalting decreased the

conductivity of the chicken and coconut enrichments falling to 160 and 179 μS/cm ² respectively, after 2 hour and finally reaching around 100 μS/cm ² after 6 hour. The numbers of S typhimurium in the dialysed chicken and coconut enrichments did not alter significantly, increasing by 0.6 and 0.3 log cycles, respectively, after 6 hour at RT.

Thus these cassettes provide a simple, convenient method of rapidly (eg. 2 - 3 hours) desalting food enrichment cultures prior to dielectrophoresis with no evidence of loss of bacterial numbers during the process.

Example 7: Reduction in conductivity and dielectrophoresis of food pre-enrichment samples using I ml capacity dialysis cassettes:

METHOD

Enrichment cultures of minced beef, chicken, skimmed milk powder and coconut were prepared and inoculated with an antibiotic-resistant strain of S typhimurium (S5968) as described in Example 6. Large-volume cassettes (10,000) MW cutoff 15 ml capacity. Pierce & Warriner) were filled with 15 ml portions of the enrichment cultures. The cultures were dialysed against 3 litres of distilled water with changes at hourly intervals. Aliquots (100 μl) of suspension were removed at intervals for conductivity measurements and for Salmonella counts on Xylose Lysine agar containing streptomycin (1 mg/ml) and nalidixic acid (50 μg/ml) (XLNS) before and after dialysis. At the end of dialysis an aliquot of desalted enrichment culture was transferred to a dielectrophoresis cell and observed for dielectrophoretic separation of the total microbial flora in the SMP at 20 MHz 20 V, using a light microscope and x 80 objective.

RESULTS

With 15 ml capacity dialysis cassettes the conductivities of chicken, minced beef (MB), skimmed milk powder (SMP) and coconut enrichments fell from >2000 before dialysis to 520, 65O, 58O and 240 μS/cm ² ,

respectively, after 2 hours; 240. 740, 310 and 143 μS/cm ² , respectively, after 3 hours; and reached 104, 210 143 and 7 μS/cm ² , respectively, after 5 hours. After 5 hours dialysis Salmonella counts had increased slightly (0.1 log cycle) in the chicken enrichment an decreased slightly in the MB, SMP and coconut enrichments (0.5. 0.3 & 0.9 log cycles, respectively). After dialysis, positive electrophoresis of the SMP enrichment was successfully demonstrated.

Thus these cassettes provide a convenient method of rapidly desalting food relatively large volumes of enrichment cultures prior to dielectrophoresis with no evidence of a loss of bacterial numbers during the desalting process. The results suggest that there is a reasonable expectation that a continuous-flow dialysis system, such as are available commercially could be successfully used in the methods of the present invention.

Example 8; Reduction in conductivity and dielectrophoresis of bacterial cultures and food pre-enrichment samples using dialysis ηhwmhfi fl

METHODS

S typhimurium was cultured overnight at 37°C in TSB and enumerated by preparing serial decimal dilutions in MRD and surface plating 0.1 ml of appropriate dilutions onto XLD. The XLD plates were incubated at 37°C for 24h. The conductivity of the 10 ^"3 dilution was then measured and portions of this dilution were transferred to Spin Biodialyser chambers (Sialomed Inc) fitted with 0.45 μm or 0.6 μm polycarbonate membranes. The culture dilution was dialysed against 1 1 of distilled water for 5 hours on a magnetic stirrer and 100 μl samples were removed at hourly intervals for conductivity readings. Salmonella were enumerated at the end of dialysis, as described in Example 7• The experiment was repeated with L monocytogenes using 0.45 μm membranes only. Listeria were enumerated by plating the dilutions onto PALCAM agar and incubating at 30°C for 72 hours.

Food enrichment cultures were inoculated with antibiotic-resistant S typhimurium were prepared as in Example 6. Samples of the enrichment culture were transferred to Biodialysers and dialysed against 1 1 volumes of distilled water. Conductivity measurements were taken at the start of dialysis and after 30 min, 1, 2 and 3 hours. Salmonella were enumerated on XLNS agar before and after dialysis, as described in Example 7-

At the end of dialysis samples of the desalted enrichment cultures were transferred to a dielectrophoretic cell and examined as described in Example 7•

RESULTS

A Spin Biodialyser fitted with a 0.45 urn polycarbonate membrane reduced the conductivity of a suspension of S typhimurium from >2000 to 470, 104, 62 and 39 μS/cm ² , and with a 0.6 μm Polycarbonate membrane from >2000 to 490. 152, 72 & 40 μS/cm ² , after 1,2,3 and 4 hours, respectively. Salmonella counts showed slight fluctuation during dialysis, with an overall decrease of approximately 0.7 and 0.5 log cycles for the 0.45 μm and 0.6 μm membranes, respectively. Similar results were obtained with a suspension of L monocytogenes, although in this case there was no significant overall change in the Listeria count.

Chicken, SMP and coconut enrichment cultures all reached conductivities of <100 μS/cm ² within 2 h. MB reached <300 μS/cm ² within this time. Positive dielectrophoretic aggregation of the total microbial flora from the food enrichment cultures was successfully carried out after desalting with the Spin Biodialyser.

Thus these chambers provide a convenient method of rapidly desalting food enrichment cultures prior to dielectrophoresis with no evidence of a loss of bacterial numbers during the desalting process. No significant differences were noted between the rates using the 0.45 μm and 0.6 μm membranes.

Example 8: Reduction in conductivity bacterial cultures using dextran and polvacrvlamide desalting columns

METHODS

Mini-columns containing cross-linked dextran (Presto (TM) 5ml Pierce & Warriner and PDIO Pharmacia) and beaded polyacrylamide (Speedy (TM) 5ml Pierce and Warriner) were used to desalt culture suspensions and food enrichment cultures. The columns were first decontamined by washing with 3 ml of 95% ethanol solution and then washed with distilled water until the conductivity of the effluent stabilised at a low level (approximately 20 μS/cm ² ). A portion (1 ml) of the effluent was retained for a sterility check. A strain of S typhimurium (S5698) resistant to streptomycin and nalidixic acid was cultured overnight at 37°C in TSB. The Salmonella were then enumerated by serial decimal dilution in MRD and surface plating as in Example 7- A portion (1.5 ml) of the 10 ^"7 dilution was then applied to the column (3 ml of 10 ^"3 dilution for PDIO columns) and allowed to displace the distilled water. Ten 200 μl (1 ml for PDIO columns) fractions were collected in Eppendorf tubes and their conductivity measured. Portions of each fraction (100 μl) were diluted in MRD and the Salmonella count determined as stated before.

RESULTS

PD10 desalting columns reduced the conductivity of a Salmonella suspension from >2000 to 166 μS/cm ² in the first fraction. Thereafter, conductivity increased to 430 in fraction 2 and returned to > 2000 μS/cm ² in fractions 3 to 6, before falling again in fractions 7 to 10 (Fig 13). Similar results were obtained with a suspension of L monocytogenes with the conductivity falling to 210 μS/cm ² in the first fraction, increasing to 76O μS/cm ² in fraction 2 and >2000 μS/cm ² in fractions 3 to 6, before falling again in fractions 7 to 10. Maximum recovery of Salmonella and Listeria was obtained in fraction 2 (2.5% & 8.5% respectively).

Fig 14 shows the mean conductivity and recovery of S typhimurium for

duplicate experiments using Presto (TM) dextran desalting columns. These columns reduced the mean conductivity of the Salmonella suspension from >2000 μS/cm ² before treatment to 18 μS/cm ² in fraction 1. In subsequent fractions the conductivity gradually increased reaching 169 μS/cm ² in the tenth fraction. The maximum recovery of Salmonella was obtained in fraction 8 (69%).

Fig 15 and 16 show the results of similar desalting experiments using Speedy(TM) polyacrylamide desalting columns. In the case of S typhimurium (Fig 15), the conductivity fell from >2000 to 980 uS/cm ² in fraction 1 and then increased to 1088 μS/cm ² in fraction 2, returning to >2000 μS/cm ² in fractions 8.9 & 10. With S enteritidis (Fig 16), the conductivity fell from 2000 to 720 μS/cm ² in fraction 2, returning to >2000 μS/cm ² in fractions 8, 9 & 10. The maximum recovery of Salmonella was obtained in fraction 10, ranging from 40% for S typhimurium, (Fig 15) to 64% for S enteritidis (Fig 16) . The desalting procedure took about five minutes.

As demonstrated in Examples 3. 4, and 5. desalting columns provide a very rapid method of desalting food enrichment cultures. Recovery of bacteria was, however, varied. Presto (TM) dextran columns appeared to be superior in this respect.