冷冻食品（英文第二版）：34990wp_08

8.1 Introduction The detection and enumeration of microorganisms either in foods or on food contact surfaces forms an integral part of any quality control or quality assurance plan. Microbiological tests done on foods can be divided into two types: (a) quantitative or enumerative, in which a group of microorganisms in the sample are counted and the result is expressed as the number of the organisms present per unit weight of sample; or (b) qualitative or presence/absence, in which the requirement is simply to detect whether a particular organism is present or absent in a known weight of sample. The basis of methods used for the testing of microorganisms in foods is very well established, and relies on the incorporation of a food sample into a nutrient medium in which microorganisms can replicate thus resulting in a visual indication of growth. Such methods are simple, adaptable, convenient and generally inexpensive. However, they have two drawbacks: firstly, the tests rely on the growth of organisms in media, which can take many days and result in a long test elapse time; and secondly, the methods are manually oriented and are thus labour intensive. Over recent years, there has been considerable research into rapid and automated microbiological methods. The aim of this work has been to reduce the test elapse time by using methods other than growth to detect and/or count microorganisms and to decrease the level of manual input into tests by automating methods as much as possible. These rapid and automated methods have gained some acceptance within the food industry and could form an important quality control tool in the chilled foods area. Positive release of chilled foods on the results of a rapid method could increase the shelf-life of a 8 Conventional and rapid analytical microbiology R. P. Betts, Campden and Chorleywood Food Research Association product by one or two days compared with a conventional microbiological technique. In addition, the availability of very rapid microbiological test methods indicates a potential for on-line control and the use of such systems in Hazard Analysis Critical Control Point (HACCP) procedures. 8.2 Sampling Although this chapter deals with the methodologies employed to test foods, it is important for the microbiologist to consider sampling. No matter how good a method is, if the sample has not been taken correctly and is not representative of the batch of food that it has been taken from, then the test result is meaningless. It is useful to devise a sampling plan in which results are interpreted from a number of analyses, rather than a single result. It is now common for microbiologists to use two or three class sampling plans, in which the number of individual samples to be tested from one batch are specified, together with the microbiological limits that apply. These type of sampling plans are fully described in Anon. 1986. Once a sampling plan has been devised then a representative portion must be taken for analysis. In order to do this the microbiologist must understand the food product and its microbiology in some detail. Many chilled products will not be homogeneous mixtures but will be made up of layers or sections, a good example would be a prepared sandwich. It must be decided if the microbiological result is needed for the whole sandwich (i.e. bread and filling), or just the bread, or just the filling, indeed in some cases one part of a mixed filling may need to be tested, when this has been decided then the sample for analyses can be taken, using appropriate aseptic technique and sterile sampling implements (Kyriakides et al., 1996). The sampling procedure having been developed, the microbiologist will have confidence that samples taken are representative of the foods being tested and test methods can be used with confidence. 8.3 Conventional microbiological techniques As outlined in the introduction, conventional microbiological techniques are based on the established method of incorporating food samples into nutrient media and incubating for a period of time to allow the microorganisms to grow. The detection or counting method is then a simple visual assessment of growth. These methods are thus technically simple and relatively inexpensive, requiring no complex instrumentation. The methods are however very adaptable, allowing the enumeration of different groups of microorganisms. Before testing, the food sample must be converted into a liquid form in order to allow mixing with the growth medium. This is usually done by accurately weighing the sample into a sterile container and adding a known volume of 188 Chilled foods sterile diluent (the sample to diluent ratio is usually 1:10); this mixture is then homogenised using a homogeniser (e.g. stomacher or pulsifier) that breaks the sample apart, releasing any organisms into the diluent. The correct choice of diluent is important. If the organisms in the sample are stressed by incorrect pH or low osmotic strength, then they could be injured or killed, thus affecting the final result obtained from the microbiological test. The diluent must be well buffered at a pH suitable for the food being tested and be osmotically balanced. When testing some foods (e.g. dried products) which may contain highly stressed microorganisms, then a suitable recovery period may be required before the test commences, in order to ensure cells are not killed during the initial phase of the test procedure (Davis and Jones 1997). 8.3.1 Conventional quantitative procedures The enumeration of organisms in samples is generally done by using plate count, or most probable number (MPN) methods. The former are the most widely used, whilst the latter tend to be used only for certain organisms (e.g. Escherichia coli) or groups (e.g. coliforms). Plate count method The plate count method is based on the deposition of the sample, in or on an agar layer in a Petri dish. Individual organisms or small groups of organisms will occupy a discrete site in the agar, and on incubation will grow to form discrete colonies that are counted visually. Various types of agar media can be used in this form to enumerate different types of microorganisms. The use of a non- selective nutrient medium that is incubated at 30oC aerobically will result in a total viable count or mesophilic aerobic count. By changing the conditions of incubation to anaerobic, a total anaerobe count will be obtained. Altering the incubation temperature will result in changes in the type of organism capable of growth, thus showing some of the flexibility in the conventional agar approach. If there is a requirement to enumerate a specific type of organism from the sample, then in most cases the composition of the medium will need to be adjusted to allow only that particular organism to grow. There are three approaches used in media design that allow a specific medium to be produced: the elective, selective and differential procedures. Elective procedures refer to the inclusion in the medium of reagents, or the use of growth conditions, that encourage the development of the target organisms, but do not inhibit the growth of other microorganisms. Such reagents may be sugars, amino acids or other growth factors. Selective procedures refer to the inclusion of reagents or the use of growth conditions that inhibit the development of non-target microorganisms. It should be noted that, in many cases, selective agents will also have a negative effect on the growth of the target microorganism, but this will be less great than the effect on non-target cells. Examples of selective procedures would be the inclusion of antibiotics in a medium or the use of anaerobic growth conditions. Finally, differential Conventional and rapid analytical microbiology 189 procedures allow organisms to be distinguished from each other by the reactions that their colonies cause in the medium. An example would be the inclusion of a pH indicator in a medium to differentiate acid-producing organisms. In most cases, media will utilise a multiple approach system, containing elective, selective and differential components in order to ensure that the user can identify and count the target organism. The number of types of agar currently available are far too numerous to list. For details of these, the manuals of media manufacturing companies (e.g. Oxoid, LabM, Difco, Merck) should be consulted. MPN method The second enumerative procedure mentioned earlier was the MPN method. This procedure allows the estimation of the number of viable organisms in a sample based on probability statistics. The estimate is obtained by preparing decimal (tenfold) dilutions of a sample, and transferring sub-samples of each dilution to (usually) three tubes of a broth medium. These tubes are incubated, and those that show any growth (turbidity) are recorded and compared to a standard table of results (Anon. 1986) that indicate the contamination level of the product. As indicated earlier, this method is used only for particular types of test and tends to be more labour and materials intensive than plate count methods. In addition, the confidence limits are large even if many replicates are studied at each dilution level. Thus the method tends to be less accurate than plate counting methods. 8.3.2 Conventional qualitative procedures Qualitative procedures are used when a count of the number of organisms in a sample is not required and only their presence or absence needs to be determined. Generally such methods are used to test for potentially pathogenic microorganisms such as Salmonella spp., Listeria spp., Yersinia spp. and Campylobacter spp. The technique requires an accurately weighed sample (usually 25g) to be homogenised in a primary enrichment broth and incubated for a stated time at a known temperature. In some cases, a sample of the primary enrichment may require transfer to a secondary enrichment broth and further incubation. The final enrichment is usually then streaked out onto a selective agar plate that allows the growth of the organisms under test. The long enrichment procedure is used because the sample may contain very low levels of the test organism in the presence of high numbers of background microorgan- isms. Also, in processed foods the target organisms themselves may be in an injured state. Thus the enrichment methods allow the resuscitation of injured cells followed by their selective growth in the presence of high numbers of competing organisms. The organism under test is usually indistinguishable in a broth culture, so the broth must be streaked onto a selective/differential agar plate. The microorgan- isms can then be identified by their colonial appearance. The formation of 190 Chilled foods colonies on the agar that are typical of the microorganism under test are described as presumptive colonies. In order to confirm that the colonies are composed of the test organism, further biochemical and serological tests are usually performed on pure cultures of the organism. This usually requires colonies from primary isolation plates being restreaked to ensure purity. The purified colonies are then tested biochemically by culturing in media that will indicate whether the organism produces particular enzymes or utilises certain sugars. At present a number of companies market miniaturised biochemical test systems that allow rapid or automated biochemical tests to be quickly and easily set up by microbiologists. Serological tests are done on pure cultures of some isolated organisms, e.g. Salmonella using commercially available antisera. 8.4 Rapid and automated methods The general interest in alternative microbiological methods has been stimulated in part by the increased output of food production sites. This has resulted in 1. Greater numbers of samples being stored prior to positive release – a reduction in analysis time would reduce storage and warehousing costs. 2. A greater sample throughput being required in laboratories – the only way that this can be achieved is by increased laboratory size and staff levels, or by using more rapid and automated methods. 3. A requirement for a longer shelf-life in the chilled foods sector – a reduction in analysis time could expedite product release thus increasing the shelf-life of the product. 4. The increased application of HACCP procedures – rapid methods can be used in HACCP verification procedures. There are a number of different techniques referred to as rapid methods and most have little in common either with each other or with the conventional procedures that they replace. The methods can generally be divided into quantitative and qualitative tests, the former giving a measurement of the number of organisms in a sample, the latter indicating only presence or absence. Laboratories considering the use of rapid methods for routine testing must carefully consider their own requirements before purchasing such a system. Every new method will be unique, giving a slightly different result, in a different timescale with varying levels of automation and sample throughput. In addition, some methods may work poorly with certain types of food or may not be able to detect the specific organism or group that is required. All of these points must be considered before a method is adopted by a laboratory. It is also of importance to ensure that staff using new methods are aware of the principles of operation of the techniques and thus have the ability to troubleshoot if the method clearly shows erroneous results. Conventional and rapid analytical microbiology 191 8.4.1 Electrical methods The enumeration of microorganisms in solution can be achieved by one of two electrical methods, one measuring particle numbers and size, the other monitoring metabolic activity. Particle counting The counting and sizing of particles can be done with the ‘Coulter’ principle, using instruments such as the Coulter Counter (Coulter Electrics, Luton). The method is based on passing a current between two electrodes placed on either side of a small aperture. As particles or cells suspended in an electrolyte are drawn through the aperture they displace their own volume of electrolyte solution, causing a drop in d.c. conductance that is dependent on cell size. These changes in conductance are detected by the instrument and can be presented as a series of voltage pulses, the height of each pulse being proportional to the volume of the particle, and the number of pulses equivalent to the number of particles. The technique has been used extensively in research laboratories for experiments that require the determination of cell sizes or distribution. It has found use in the area of clinical microbiology where screening for bacteria is required (Alexander et al. 1981). In food microbiology however, little use has been made of the method. There are reports of the detection of cell numbers in milk (Dijkman et al., 1969) and yeast estimation in beer (MaCrae 1964), but little other work has been published. Any use of particle counting for food microbiology would probably be restricted to non-viscous liquid samples or particle-free fluids, since very small amounts of sample debris could cause significant interference, and cause aperture blockage. Metabolic activity Stewart (1899) first reported the use of electrical measurement to monitor microbial growth. This author used conductivity measurements to monitor the putrefaction of blood, and concluded that the electrical changes were caused by ions formed by the bacterial decomposition of blood constituents. After this initial report a number of workers examined the use of electrical measurement to monitor the growth of microorganisms. Most of the work was successful; however, the technique was not widely adopted until reliable instrumentation capable of monitoring the electrical changes in microbial cultures became available. There are currently four instruments commercially available for the detection of organisms by electrical measurement. The Malthus System (IDG, Bury, UK) based on the work of Richards et al. (1978) monitors conductance changes occurring in growth media as does the Rabit System (Don Whitley Scientific, Yorkshire), whilst the Bactometer (bioMerieux, Basingstoke, UK), and the Batrac (SyLab, Purkersdorf, Austria) (Bankes 1991) can monitor both conductance and capacitance signals. All of the instruments have similar basic components: (a) an incubator system to hold samples at a constant temperature 192 Chilled foods during the test; (b) a monitoring unit that measures the conductance and/or capacitance of every cell at regular frequent intervals (usually every 6 minutes); and (c) a computer-based data handling system that presents the results in usable format. The detection of microbial growth using electrical systems is based on the measurement of ionic changes occurring in media, caused by the metabolism of microorganisms. The changes caused by microbial metabolism and the detailed electrochemistry that is involved in these systems has been previously described in some depth (Eden and Eden 1984, Easter and Gibson 1989, Bolton and Gibson 1994). The principle underlying the system is that as bacteria grow and metabolise in a medium, the conductivity of that medium will increase. The electrical changes caused by low numbers of bacteria are impossible to detect using currently available instrumentation, approximately 106 organisms/ml must be present before a detectable change is registered. This is known as the threshold of detection, and the time taken to reach this point is the detection time. In order to use electrical systems to enumerate organisms in foods, the sample must initially be homogenised. The growth well or tube of the instrument containing medium is inoculated with the homogenised sample and connected to the monitoring unit within the incubation chamber or bath. The electrical properties of the growth medium are recorded throughout the incubation period. The sample container is usually in the form of a glass or plastic tube or cell, in which a pair of electrodes is sited. The tube is filled with a suitable microbial growth medium, and a homogenised food sample is added. The electrical changes occurring in the growth medium during microbial metabolism are monitored via the electrodes and recorded by the instrument. As microorganisms grow and metabolise they create new end-products in the medium. In general, uncharged or weekly charged substrates are transformed into highly charged end-products (Eden and Eden 1984), and thus the conductance of the medium increases. The growth of some organisms such as yeasts does not result in large increases in conductance. This is possibly due to the fact that these organisms do not produce ionised metabolites and this can result in a decrease in conductivity during growth. When an impedance instrument is in use, the electrical resistance of the growth medium is recorded automatically at regular intervals (e.g. 6 minutes) throughout the incubation period. When a change in the electrical parameter being monitored is detected, then the elapsed time since the test was started is calculated by a computer; this is usually displayed as the detection time. The complete curve of electrical parameter changes with time (Fig. 8.1) is similar to a bacterial growth curve, being sigmoidal and having three stages: (a) the inactive stage, where any electrical changes are below the threshold limit of detection of the instrument; (b) the active stage, where rapid electrical changes occur; and (c) the stationary or decline stage, that occurs at the end of the active stage and indicates a deceleration in electrical changes. The electrical response curve should not be interpreted as being similar to a microbial growth curve. It is accepted (Easter and Gibson 1989) that the lag and Conventional and rapid analytical microbiology 193 logarithmic phases of microbial growth occur in the inactive and active stages of the electrical response curve, up to and beyond the detection threshold of the instrument. The logarithmic and stationary phases of bacterial growth occur during the active and decline stages of electrical response curves. In order to use detection time data generated from electrical instruments to assess the microbiological quality of a food sample, calibrations must be done. The calibration consists of testing samples using both a conventional plating test and an electrical test. The results are presented graphically with the conventional result on the y-axis and the detection time on the x-axis (Fig. 8.2). The result is a negative line with data covering 4 to 5 log cycles of organisms and a correlation coefficient greater than 0.85 (Easter and Gibson, 1989), Calibrations must be done for every sample type to be tested using electrical methods; different samples will contain varying types of microbial flora with differing rates of growth. This can greatly affect electrical detection time and lead to incorrect results unless correct calibrations have been done. So far, the use of electrical instruments for total microbial assessment has been described. These systems, however, are based on the use of a growth medium and it is thus possible, using media engineering, to develop methods for the enumeration or detection of specific organisms or groups of organisms. Many examples of the use of electrical measurement for the detection/ enumeration of specific organisms have been published; these include: Fig. 8.1 A conductance curve generated by the growth of bacteria in a suitable medium. 194 Chilled foods Enterobacteriaceae (Cousins and Marlatt 1990, Petitt 1989), Pseudomonas (Banks et al. 1989), Yersinia enterocolitica (Walker 1989) and yeasts (Connolly et al., 1988), E.coli (Druggan et al. 1993), Campylobacter (Bolton and Powell 1993). In the future, the number of types of organism capable of being detected will undoubtedly increase. Considerable research is currently being done on media for the detection of Listeria, and media for other organisms will follow. Most of the electrical methods described above involve the use of direct measurement, i.e. the electrical changes are monitored by electrodes immersed in the culture medium. Some authors have indicated the potential for indirect conductance measurement (Owens et al. 1989) for the detection of microorgan- isms. This method involves the growth medium being in a separate compartment to the electrode within the culture cell. The liquid surrounding the electrode is a gas absorbent, e.g. potassium hydroxide for carbon dioxide. The growth medium is inoculated with the sample and, as the microorganisms grow, gas is released. This is absorbed by the liquid surrounding the electrode, causing a change in conductivity, which can be detected. This technique may solve the problem caused by microorganisms that produce only small conductance changes in conventional direct conductance cells. These organisms, e.g. many yeast species, are very difficult to detect using Fig. 8.2 Calibration curve showing changes in conductance detection time with bacterial total viable count (TVC). Conventional and rapid analytical microbiology 195 conventional direct conductance methods, but detection is made easy by the use of indirect conductance monitoring (Betts 1993). The increased use of indirect methods in the future could considerably enhance the ability of electrical systems to detect microorganisms that produce little electrical change in direct systems, thus increasing the number of applications of the technique within the food industry. 8.4.2 Adenosine triphosphate (ATP) bioluminescence The non-biological synthesis of ATP in the extracellular environment has been demonstrated (Ponnamperuma et al., 1963), but it is universally accepted that such sources of ATP are very rare (Huernnekens and Whiteley 1960). ATP is a high-energy compound found in all living cells (Huernnekens and Whiteley 1960), and it is an essential component in the initial biochemical steps of substrate utilisation and in the synthesis of cell material. McElroy (1947) first demonstrated that the emission of light in the bioluminescent reaction of the firefly, Photinus pyralis, was stimulated by ATP. The procedure for the determination of ATP concentrations utilising crude firefly extracts was described by McElroy and Streffier (1949) and has since been used in many fields as a sensitive and accurate measure of ATP. The light- yielding reaction is catalysed by the enzyme luciferase, this being the enzyme found in fireflies causing luminescence. Luciferase takes part in the following reaction: 1. Luciferase + Luciferin + ATP C33 Mg 2+ LuciferaseC0LuciferinC0AMP + PP The complex is then oxidised: 2. LuciferaseC0LuciferinC0AMP + O 2 C33 (LuciferaseC0LuciferinC0AMP = O) + H 2 O The oxidised complex is in an excited stage, and as it returns to its ground stage a photon of light is released: 3. LuciferaseC0LuciferinC0AMP = 0 C33 (LuciferaseC0LuciferinC0AMP = 0) + Light The light-yielding reaction is efficient, producing a single photon of light for every luciferin molecule oxidised and thus every ATP molecule used (Seliger and McElroy 1960). Levin et al. (1964) first described the use of the firefly bioluminescence assay of ATP for detecting the presence of viable microorganisms. Since this initial report considerable work has been done on the detection of viable organisms in environmental samples using a bioluminescence technique (Stalker 1984). As all viable organisms contain ATP, it could be considered simple to use a bioluminescence method to rapidly enumerate microorganisms. Research, however, has shown that the amount of ATP in different microbial cells varies depending on species, nutrient level, stress level and stage of growth (Stannard 196 Chilled foods 1989, Stalker 1984). Thus, when using bioluminescence it is important to consider: 1. the type of microorganism being analysed; generally, vegetative bacteria will contain 1 fg of ATP/cell (Karl 1980), yeasts will contain ten times this value (Stannard 1989), whilst spores will contain no ATP (Sharpe et al. 1970); 2. whether the cells have been subjected to stress, such as nutrient depletion, chilling or pH change. In these cases a short resuscitation may be required prior to testing; 3. whether the cells are in a relatively ATP-free environment, such as a growth medium, or are contained within a complex matrix, like food, that will have very high background ATP levels. When testing food samples one of the greatest problems is that noted in 3 above. All foods will contain ATP and the levels present in the food will generally be much higher than those found in microorganisms within the food. Data from Sharpe et al. (1970) indicated that the ratio of food ATP to bacterial ATP ranges from 40000:1 in ice-cream to 15:1 in milk. Thus, to be able to use ATP analysis as a rapid test for foodborne microorganisms, methods for the separation of microbial ATP were developed. The techniques that have been investigated fall into two categories: either to physically separate microorganisms from other sources of ATP, or to use specific extractants to remove and destroy non- microbial ATP. Filtration methods have been successfully used to separate microorganisms from drinks (LaRocco et al. 1985, Littel and LaRocco 1986) and brewery samples (Hysert et al. 1976). These methods are, however, difficult to apply to particulate-containing solutions as filters rapidly become blocked. A potential way around this problem has been investigated by some workers and utilise a double filtration system/scheme (Littel et al. 1986), the first filter removing food debris but allowing microorganisms through, the second filter trapping microorganisms prior to lysis and bioluminescent analysis. Other workers (Baumgart et al. 1980, Stannard and Wood 1983) have utilised ion exchange resins to trap selectively either food debris or microorganisms before bioluminescent tests were done. The use of selective chemical extraction to separate microbial and non- microbial ATP has been extensively tested for both milk (Bossuyt 1981) and meat (Billte and Reuter 1985) and found to be successful. In general, this technique involves the lysis of somatic (food) cells followed by destruction of the released ATP with an apyrase (ATPase) enzyme. A more powerful extraction reagent can then be used to lyse microbial cells, which can then be tested with luciferase, thus enabling the detection of microbial ATP only. There are a number of commercially available instruments aimed specifically at the detection of microbial ATP; Lumac (Netherlands), Foss Electric (Denmark), Bio Orbit (Finland) and Biotrace (UK) all produce systems, including separation methods, specifically designed to detect microorganisms in foods. Generally, all of the systems perform well and have similar specifica- Conventional and rapid analytical microbiology 197 tions, including a minimum detection threshold of 10 4 bacteria (10 3 yeasts) and analysis times of under one hour. In addition to testing food samples for total viable microorganisms, there have been a number of reports concerning potential alternative uses of ATP bioluminescence within the food industry. The application of ATP analysis to rapid hygiene testing has been considered (Holah 1989), both as a method of rapidly assessing microbiological contamination, and as a procedure for measuring total surface cleanliness. It is in the latter area that ATP measurement can give a unique result. As described earlier, almost all foods contain very high levels of ATP, thus food debris left on a production line could be detected in minutes using a bioluminescence method, allowing a very rapid check of hygienic status to be done. The use of ATP bioluminescence to monitor surface hygiene has now been widely adopted by industry. The availability of relatively inexpensive, portable, easy to use luminometers has now enabled numerous food producers to implement rapid hygiene testing procedures that are ideal for HACCP monitoring applications where surface hygiene is a critical control point. Reports suggest (Griffiths 1995) that all companies surveyed who regularly use ATP hygiene monitoring techniques note improvements in cleanliness after initiation of the procedure. Such ATP based test systems can be applied to most types of food processing plant, food service and retail establishments and even assessing the cleanliness of transportation vehicles such as tankers. One area that ATP bioluminescence has not yet been able to address has been the detection of specific microorganisms. It may be possible to use selective enrichment media for particular microorganisms in order to allow selective growth prior to ATP analysis. This approach would, however, considerably increase analysis time and some false high counts would be expected. The use of specific lysis agents that release ATP only from the cells being analysed have been investigated (Stannard 1989) and shown to be successful. The number of these specific reagents is, however, small and thus the method is of only limited use. Perhaps the most promising method developed for the detection of specific organisms is the use of genetically engineered bacteriophages (Ulitzur and Kuhn, 1987, Ulitzur et al. 1989, Schutzbank et al. 1989). Bacteriophages are viruses that infect bacteria. Screening of bacteriophages has shown that some are very specific, infecting only a particular type of bacteria. Workers have shown it is possible to add into the bacteriophage the genetic information that causes the production of bacterial luciferase. Thus, when a bacteriophage infects its specific host bacterium, the latter produces luciferase and becomes luminescent. This method requires careful selection of the bacteriophage in order to ensure false positive or false negative results do not occur; it does however indicate that, in the future, luminescence-based methods could be used for the rapid detection of specific microorganisms (Stewart 1990). In conclusion, the use of ATP bioluminescence in the food industry has been developed to a stage at which it can be reliably used as a rapid test for viable microorganisms, as long as an effective separation technique for microbial ATP 198 Chilled foods is used. Its potential use in rapid hygiene testing has been realised and the technique is being used within the industry. Work has also shown that luminescence can allow the rapid detection of specific microorganisms but such a system would need to be commercialised before widespread use within the food industry. 8.4.3 Microscopy methods Microscopy is a well established and simple technique for the enumeration of microorganisms. One of the first descriptions of its use was for rapidly counting bacteria in films of milk stained with the dye methylene blue (Breed and Brew 1916). One of the main advantages of microscope methods is the speed with which individual analyses can be done; however, this must be balanced against the high manual workload and the potential for operator fatigue caused by constant microscopic counting. The use of fluorescent stains, instead of conventional coloured compounds, allows cells to be more easily counted and thus these stains have been the subject of considerable research. Microbial ecologists first made use of such compounds to visualise and count microorganisms in natural waters (Francisco et al. 1973, Jones and Simon 1975). Hobbies et al. (1977) first described the use of Nuclepore polycarbonate membrane filters to capture microorganisms before fluorescent staining, whilst enumeration was considered in depth by Pettipher et al. (1980), the method developed by the latter author being known as the direct epifluorescent filter technique (DEFT). The DEFT is a labour-intensive manual procedure and this has led to research into automated fluorescence microscope methods that offer both automated sample preparation and high sample throughput. The first fully automated instrument based on fluorescence microscopy was the Bactoscan (Foss Electric, Denmark), which was developed to count bacteria in milk and urine. Milk samples placed in the instrument are chemically treated to lyse somatic cells and dissolve casein micelles. Bacteria are then separated by continuous centrifuga- tion in a dextran/sucrose gradient. Microorganisms recovered from the gradient are incubated with a protease to remove residual protein, then stained with acridline orange and applied as a thin film to a disc rotating under a microscope. The fluorescent light from the microscope image is converted into electrical impulses and recorded. The Bactoscan has been used widely for raw milk testing in continental Europe, and correlations with conventional methods have reportedly been good (Kaereby and Asmussen 1989). The technique does, however, have a poor sensitivity (approximately 5C210 4 cells/ml) and this negates its use on samples with lower bacterial counts. An instrument-based fluorescence counting method, in which samples were spread onto a thin plastic tape, was developed for the food industry. The instrument (Autotrak) deposited samples onto the tape, which was then passed through staining and washing solutions, before travelling under a fluorescence microscope. The light pulses from the stained microorganisms were then Conventional and rapid analytical microbiology 199 enumerated by a photomultiplier unit. Tests on food samples using this instrument (Betts and Bankes 1988) indicated that the debris from food samples interfered with the staining and counting procedure and gave results that were significantly higher than corresponding total viable counts. Perhaps the most recent development in fluorescence microscope techniques to be used within the food industry for rapid counting is flow cytometry. In this technique the stained sample is passed under a fluorescence microscope system as a liquid in a flow cell. Light pulses caused by the light hitting a stained particle are transported to a photomultiplier unit and counted. This technique is automated, rapid and potentially very versatile. Of the microscope methods discussed here, the DEFT has perhaps the widest usage, whilst flow cytometry could offer significant advantages in the future. These two procedures will therefore be discussed in more detail. DEFT The DEFT was developed for rapidly counting the numbers of bacteria in raw milk samples (Pettipher et al. 1980, Pettipher and Rodrigues 1982). The method is based on the pretreatment of a milk sample in the presence of a proteolytic enzyme and surfactant at 50oC, followed by a membrane filtration step that captures the microorganisms. The pretreatment is designed to lyse somatic cells and solubilise fats that would otherwise block the membrane filter. After filtration the membrane is strained with the fluorescent nucleic acid binding dye, acridine orange, then rinsed and mounted on a microscope slide. The membrane is then viewed with an epifluorescent microscope. This illuminates the membrane with ultraviolet light, causing the stain to emit visible light that can be seen through the microscope. As the stain binds to nucleic acids it is concentrated within microbial cells by binding to DNA and RNA molecules; thus any organisms on the membrane can be easily visualised and counted. The complete pretreatment and counting procedure can take as little as 30 minutes. Although the DEFT was able to give a very rapid count, it was very labour intensive, as all of the pre-treatment and counting was done manually. This led to a very poor daily sample throughput for the method. The development of semi-automated counting methods based on image analysis (Pettipher and Rodridgues 1982) overcame some of the problems of manual counting and thus allowed the technique to be more user friendly. The early work on the uses of DEFT for enumerating cells in raw milk was followed with examinations of other types of foods. It was quickly recognised that the good correlations between DEFT count and conventional total viable counts that were obtained with raw milk samples did not occur when heat- treated milks were examined (Pettipher and Rodrigues 1981). Originally this was considered to be due to heat-inducing staining changes occurring in Gram- positive cocci (Pettipher and Rodrigues 1981); however, more recent work (Back and Kroll 1991) has shown similar changes occur in both Gram-positives and Gram-negatives. Similar staining phenomena have also been observed with heat-treated yeasts (Rodrigues and Kroll 1986) and in irradiated foods (Betts et 200 Chilled foods al. 1988). Thus the use of DEFT as a rapid indication of total viable count are mainly confined to raw foods. The types of food with which DEFT can be used has been expanded since the early work with raw milk. Reports have covered the use of the method with frozen meats and vegetables (Rodrigues and Kroll 1989), raw meats (Shaw et al. 1987), alcoholic beverages (Cootes and Johnson 1980, Shaw 1989), tomato paste (Pettipher et al. 1985), confectionery (Pettipher 1987), dried foods (Oppong and Snudden 1988) and hygiene testing (Holah et al. 1988). In addition, some workers (Rodrigues and Kroll 1988) have suggested that the method could be modified to detect and count specific groups of organisms. In conclusion, the DEFT is a very rapid method for the enumeration of total viable microorganisms in raw foods and has been used with success within the industry. The problems of the method are a lack of specificity and an inability to give a good estimate of viable microbial numbers in processed foods. The former could be solved by the use of short selective growth stages or fluorescently labelled antibodies; however, these solutions would have time and cost implications. The problem with processed foods can be eliminated only if alternating straining systems that mark viable cells are examined; preliminary work (Betts et al. 1989) has shown this approach to be successful, and the production and commercialisation of fluorescent viability stains could advance the technique. At present the high manual input and low sample throughput of DEFT procedures has limited the use of the procedure in the food industry. Flow cytometry Flow cytometry is a technique based on the rapid measurement of cells as they flow in a liquid stream past a sensing point (Carter and Meyer 1990). The cells under investigation are inoculated into the centre of a stream of fluid (known as the sheath fluid). This constrains them to pass individually past the sensor and enables measurements to be made on each particle in turn, rather than average values for the whole population. The sensing point consists of a beam of light (either ultraviolet or laser) that is aimed at the sample flow and one or more detectors that measure light scatter or fluorescence as the particles pass under the light beam. The increasing use of flow cytometry in research laboratories has largely been due to the development of the reliable instrumentation and the numerous staining systems. The stains that can be used with flow cytometers allow a variety of measurements to be made. Fluorescent probes based on enzyme activity, nucleic acid content, membrane potential and pH all have been examined, whilst the use of antibody-conjugated fluorescent dyes confers specificity to the system. Flow cytometers have been used to study a range of eukaryotic and prokaryotic microorganisms. Work with eukaryotes has included the examina- tion of pathogenic amoeba (Muldrow et al. 1982) and yeast cultures (Hutter and Eipel 1979), whilst bacterial studies have included the growth of Escherichia coli (Steen et al. 1982), enumeration of cells in bacterial cultures (Pinder et al. 1990) and the detection of Legionella spp. in cooling tower waters (Tyndall et al. 1985). Conventional and rapid analytical microbiology 201 Flow cytometric methods for the food industry have been developed and have been reviewed by Veckert et al. (1995). Donnelly and Baigent (1986) explored the use of fluorescently labelled antibodies to detect Listeria monocytogenes in milk, and obtained encouraging results. The method used by these authors relied on the selective enrichment of the organisms for 24 hours, followed by staining with fluorescein isothiocyanate labelled polyvalent Listeria antibodies. The stained cells were then passed through a flow cytometer, and the L. monocytogens detected. The author suggested that the system could be used with other types of food. A similar approach was used by McCelland and Pinder (1994) to detect Salmonella typhimurium in dairy products. Patchett et al. (1991) investigated the use of a Skatron Argus flow cytometer to enumerate bacteria in pure cultures and foods. The results obtained with pure cultures showed that flow cytometer counts correlated well with plate counts down to 10 3 cells/g. With foods, however, conflicting results were obtained. Application of the technique to meat samples gave a good correlation with plate counts and enabled enumeration down to 10 5 cells/g. Results for milk and pate′ were poorer, the sensitivity of the system for pate′ being 10 6 cells/ml, whilst cells inoculated into milk were not detected at levels in excess of 10 7 ml. The poor sensitivity of this flow cytometer with foods was thought to be due to interference of the counting system caused by food debris and it was suggested that the application of separation methods to partition microbial cells from food debris would overcome the problem. Perhaps the most successful application of flow cytometric methods to food products has been the use of a Chemunex Chemflow system to detect contaminating yeast in dairy and fruit products (Bankes et al. 1991). The procedure used with this system calls for an incubation of the product for 16– 20 hours followed by centrifugation to separate and concentrate the cells. The stain is then added, and a sample is passed through the flow cytometer for analysis. An evaluation of the system by Pettipher (1991), using soft drinks inoculated with yeasts, showed that it was reliable and user friendly. The results obtained indicated that cytometer counts correlated well with DEFT counts, however, the author did not report how the system compared to plate counts. Investigations of the Chemflow system by Bankes et al. (1991) utilised a range of dairy and fruit-based products inoculated with yeast. Results indicated that yeast levels as low as 1 cell/25g could be detected in 24 hours in dairy products. In fruit juices a similar sensitivity was reported: however, a 48 hour- period was required to ensure that this was achieved. The system was found to be robust and easy to use. The Chemflow system has now been adapted to detect bacterial cells as well as yeasts, and applications are available for fermentor biomass and enumeration of total flora in vegetables. The Chemflow system has been fully evaluated in a factory environmental (Dumain et al. 1990) testing fermented dairy products. These authors report a very good correlation between cytometer count and plate count (r = 0.98), results being obtained in 24 hours, thus providing a time saving of three days over classical methods. 202 Chilled foods In conclusion, flow cytometry can provide a rapid and sensitive method for the rapid enumeration of microorganisms. The success of the system depends on the development and use of (a) suitable staining systems, and (b) protocols for the separation of microorganisms from food debris that would otherwise interfere with the detection system. In the future a flow cytometer fitted with a number of light detection systems could allow the analysis of samples for many parameters at once, thus considerably simplifying testing regimes. Solid phase cytometry A relatively new cytometric technique has been developed by Chemunex (Maisons-Alfort, France) based on solid phase cytometry. In this procedure samples are passed through a membrane filter which captures contaminating microorganisms. A stain is then applied to the filter to fluorescently mark metabolically active microbial cells. After staining, the membrane is then transferred to a Chemscan RDI instrument, which scans the whole membrane with a laser, counting fluorescing cells. The complete procedure takes around 90 minutes to perform and can detect single cells in the filtered sample. The Chemscan RDI solid phase cytometry system is an extremely powerful tool for rapidly counting low levels of organisms. It is ideally suited to the analysis of waters or other clear filterable fluids, and using specific labelling techniques could be used to detect particular organisms of interest. Foods containing particulate materials could, however, be problematic as organisms would need to be separated from the food material before filtration and analysis. 8.4.4 Immunological methods Antibodies and antigens Immunological methods are based on the specific binding reaction that occurs between an antibody and the antigen to which it is directed. Antibodies are protein molecules that are produced by animal white blood cells, in response to contact with a substance causing an immune response. The area to which an antibody attaches on a target molecule is known as the antigen. Antigens used in immunochemical methods are of two types. The first occurs when the analyte is of low molecular weight and thus does not stimulate an immune response on its own; these substances are described as haptens and must be bound to a larger carrier molecule to elicit an immune response and cause antibody production. The second type of antigen is immunogenic and is able to elicit an immune response on its own. Two types of antibody can be employed in immunological tests. These are known as monoclonal and polyclonal antibodies. Polyclonal antibodies are produced if large molecules such as proteins or whole bacterial cells are used to stimulate an immune response in an animal. The many antigenic sites result in numerous different antibodies being produced to the molecule or cell. Monoclonal antibodies are produced by tissue culture techniques and are derived from a single Conventional and rapid analytical microbiology 203 white blood cell; thus they are directed towards a single antigenic site. The binding of an antigen is highly specific. Immunological methods can therefore be used to detect particular specific microorganisms or proteins (e.g. toxins). In many cases, when using these methods a label is attached to the antibody, so that binding can be visualised more easily when it occurs. Labels The labels that can be used with antibodies are of many types and include radiolabels, fluorescent agents, luminescent chemicals and enzymes; in addition agglutination reactions can be used to detect the binding of antibody to antigen. Radioisotopes have been extensively used as labels, mainly because of the great sensitivity that can be achieved with these systems. They do however have some disadvantages, the main one being the hazardous nature of the reagents. This would negate their use in anything other than specialist laboratories, and certainly their use within the food industry would be questionned. Fluorescent labels have been widely used to study microorganisms. The most frequently used reagent has been fluorescein. However, others such as rhodamine and umbelliferone have also been utilised. The simplest use of fluorescent antibodies is in microscopic assays. Recent advances in this approach have been the use of flow cytometry for multiparameter flow analysis of stained preparations, and the development of enzyme-linked immunofluor- escent assays (ELIFA), some of which have been automated. Luminescent labels have been investigated as an alternative to the potentially hazardous radiolabels (Kricka and Whitehead 1984). The labels can be either chemiluminescent or bioluminescent, and have the advantage over radiolabels that they are easy to handle and measure using simple equipment, whilst maintaining a similar sensitivity (Rose and Stringer 1989). A number of research papers have reported the successful use of immunoluminometric assays (Lohneis et al. 1987); however, none have yet been commercialised. Antibodies have been used for the detection of antigens in precipitation and agglutination reactions. These assays tend to be more difficult to quantify than other forms of immunoassay and usually have only a qualitative application. The assays are quick and easy to perform and require little in the way of equipment. A number of agglutination reactions have been commercialised by manufacturers and have been successfully used within the food industry. These methods have tended to be used for the confirmation of microbial identity, rather than for the detection of the target organisms. They offer a relatively fast test time, are easy to use and usually require no specialist equipment, thus making ideal test systems for use in routine testing laboratories. Several latex agglutination test kits are available for the conformation of Salmonella from foods. These include the Oxoid Salmonella Latex Kit (Oxoid) designed to be used with the Oxoid Rapid Salmonella Test Kit (Holbrook et al. 1989); the Micro Screen Salmonella Latex Slide Agglutination Test (Mercia Diagnostics Ltd.); the Wellcolex Colour Salmonella Test (Wellcome Diag- nostics) (Hadfield et al. 1987a, b); and the Spectate Salmonella test (Rhone 204 Chilled foods Poulenc Diagnostics Ltd.) (Clark et al. 1989). The latter two kits use mixtures of coloured latex particles that allow not only detection but also serogrouping of Salmonella. Latex agglutination test kits are also available for Campylobacter (Microscreen, Mercia Diagnostics), Staphyloccocus aureus (Staphaurex, Well- come Diagnostics), Shigella (Wellcolex Colour Shigella Test, Wellcome Diagnostics) and Escherichia coli 0157:H7 (Oxoid). Agglutination kits have also been developed for the detection of microbial toxins, e.g. Oxoid Staphylococcal Enterotoxin Reverse Passive Latex Agglutination Test (Rose et al. 1989, Bankes and Rose 1989). Enzyme immunoassays have been extensively investigated as rapid detection methods for foodborne microorganisms. They have the advantage of specificity conferred by the use of a specific antibody, coupled with coloured or fluorescent end-points that are easy to detect either visually or with a spectrophotometer or fluorimeter. Most commercially available enzyme immunoassays use an antibody sandwich method in order initially to capture and then to detect specific microbial cells or toxins. The kits are supplied with two types of antibody: capture antibody and conjugated antibody. The capture antibody is attached to a solid support surface such as a microtitre plate well. An enriched food sample can be added to the well and the antigens from any target cells present will bind to the antibodies. The well is washed out, removing food debris and unbound microorganisms. The enzyme conjugated antibody can then be added to the well. This will bind to the target cell forming an antibody sandwich. Unbound antibodies can be washed from the well and the enzyme substrate added. The substrate will be converted by any enzyme present from a colourless form, into a coloured product. A typical microplate enzyme immunoassay takes between two and three hours to perform and will indicate the presumptive presence of the target bacterial cells. Thus positive samples should always be confirmed by biochemical or serological methods. There are a number of commercially available enzyme immunoassay test kits for the detection of Listeria, Salmonella, Escherichia coli 0157, Staphylococcal enterotoxins, and Bacillus diarrhoeal toxin, from food samples. The sensitivity of these systems is approximately 10 6 cells/ml, so that a suitable enrichment procedure must be used before analysis using the assay. Thus, results can be obtained in two to three days, rather than the three to five days required for a conventional test procedure. Over recent years a number of highly automated immunoassays have been developed, these add to the benefit of the rapid test result, by reducing the level of manual input required to do the test. Automation of enzyme immunoassays has taken a number of forms; a number of manufacturers market instruments which simply automate standard microplate ELISAs. These instruments hold reagent bottles and use a robotic pipetting arm, which dispenses the different reagents required in the correct sequence. Automated washing and reading completes the assay with little manual input needed. At least two manufacturers have designed immunoassay kits around an automated instrument, to produce very novel systems. Conventional and rapid analytical microbiology 205 The Vidas system (bioMerieux, Basingstoke) uses a test strip, containing all of the reagents necessary to do an ELISA test, the first well of the strip is inoculated with an enriched food sample, and placed into the Vidas instrument, together with a pipette tip internally coated with capture antibody. The instrument then uses the pipette tip to transfer the test sample into the other cells in the strip containing various reagents needed to carry out the ELISA test. All of the transfers are completely automatic, as is the reading of the final test result. Vidas ELISA tests are available for a range of organisms including, Salmonella, Listeria, Listeria monocytogenes, E.coli 0157, Campylobacter and staplylococ- cal enterotoxin. Evaluations of a number of these methods have been done (Blackburn et al 1994, Bobbitt and Betts 1993) and indicated that results were at least equivalent to conventional test methods. The EIAFOSS (Foss Electric, Denmark) is another fully automated ELISA system, in this case the instrument transfers all of the reagents into sample containing tubes, in which all of the reactions occur. The EIAFOSS procedure is novel as it uses antibody coated magnetic beads as a solid phase. During the assay these beads are immobilised using a magnet mounted below the sample tube. EIAFOSS test kits are available for Salmonella, Listeria, E.coli 0157 and Campylobacter, and evaluations have indicated that these methods operate well (Jones and Betts 1994). The newest immunoassay procedure that has been developed into a commercial format is arguably the simplest to use. Immunochromatography operates on a dipstick, composed of an absorbent filter material which contains coloured particles coated with antibodies to a specific organism. The particles are on the base of the dipstick and when dipped into a microbiological enrichment broth, they move up the filter as the liquid is moved by capillary action. At a defined point along the filter material lies a line of immobilised specific antibodies. In the presence of the target organism, binding of that organism to the coloured particles will occur. This cell/particle conjugate moves up the filter dipstick by capillary action until it meets the immobilised antibodies where it will stick. The build up of coloured particles results in a clearly visible coloured line, indicating a positive test result. A number of commercial kits are based around this procedure including the Oxoid Listeria Rapid Test (Jones et al. 1995a) and the Celsis Lumac Pathstik (Jones et al. 1995b.) have been developed and appear to give good results. The immunochromatography techniques require an enrichment in the same way as other immunoassays; they do not however require any equipment or instrumentation, and once the dipstick is inoculated, need only minutes to indicate a positive or negative result. Immunoassay conclusions In conclusion, immunological methods have been extensively researched and developed. There are now a range of systems that allow the rapid detection of the specific organism to which they are directed. Numerous evaluations of commercially available immuno-based methods have indicated that the results 206 Chilled foods generally correlate well with conventional microbiological methods. Enzyme immunoassays in particular appear to offer a simple way of reducing analysis times by one or two days; automation or miniaturisation of these kits has reduced the amount of person time required to do the test and simplified the manual procedures considerably. The main problem with the immunological systems is their low sensitivity. The minimum number of organisms required in an enzyme immunoassay system to obtain a positive result is approximately 10 5 /ml. As the food microbiologist will want to analyse for the presence or absence of a single target organism in 25g of food, an enrichment phase is always necessary. The inclusion of enrichment will always add 24–48 hours to the total analysis time. 8.4.5 Nucleic acid hybridisation Nucleic acids The specific characteristics of any organism depend on the particular sequence of the nucleic acids contained in its genome. The nucleic acids themselves are made up of a chain of units each consisting of a sugar (deoxyribose or ribose, depending on whether the nucleic acid is DNA or RNA), a phosphorus- containing group and one of four organic purine or pyrimidine bases. DNA is constructed from two of these chains arranged in a double helix and held together by bonds between the organic bases. The bases specifically bind adenine to thymine and guanine to cytocine. It is the sequence of bases that make different organisms unique. The development of nucleic acid probes Nucleic acid probes are small segments of single-stranded nucleic acid that can be used to detect specific genetic sequences in test samples. Probes can be developed against DNA or RNA sequences. The attraction of the use of gene probes in the problem of microbial detection is that a probe consisting of only 20 nucleotide sequences is unique and can be used to identify an organism accurately (Gutteridge and Arnott 1989). In order to be able to detect the binding of a nucleic acid probe to DNA or RNA from a target organism, it must be attached to a label of some sort that can easily be detected. Early work was done with radioisotope labels such as phosphorus ( 32 P) that could be detected by autoradiography or scintillation counting. Radiolabels, however, have inherent handling, safety and disposal problems that make them unsuitable for use in food laboratories doing routine testing. Thus the acceptance of widespread use of nucleic acid probes required the development of alternative labels. A considerable amount of work has been done on the labelling of probes with an avidin-biotin link system. This is based on a very high binding specificity between avidin and biotin. The probe sequence of nucleic acid is labelled with biotin and reacted with target DNA. Avidin is then added, linked to a suitable Conventional and rapid analytical microbiology 207 detector, e.g. avidin-alkaline phosphatase, and binding is detected by the formation of a coloured product from a colourless substrate. These alternative labelling systems proved that non-radiolabelled probes could be used for the detection of microorganisms. However, the system was much less sensitive than isotopic procedures, requiring as much as a 100-fold increase in cell numbers for detection to occur, compared to isotope labels. In order to develop non-isotopic probes with a sensitivity approaching that of isotope labels, it was necessary to consider alternative probe targets within cells. Probes directed toward cell DNA attach to only a few sites on the chromosome of the target cell. By considering areas of cell nucleic acid that are present in relatively high copy number in each cell and directing probes toward these sites, it is possible to increase the sensitivity of non-isotopic probes considerably. Work on increasing probe sensitivity centred on the use of RNA as a target. RNA is a single-stranded nucleic acid that is present in a number of forms in cells. In one form it is found within parts of the cell protein synthesis system called ribosomes. Such RNA is known as ribosomal RNA (rRNA), and is present in very high copy numbers within cells. By directing nucleic acid probes to ribosomal RNA it is possible to increase the sensitivity of the assay system considerably. Probes for organisms in food Nucleic acid hybridisation procedures for the detection of pathogenic bacteria in foods have been described for Salmonella spp. (Fitts 1985 Curiale et al. 1986), Listeria spp. (Klinger et al. 1988, Klinger and Johnson 1988), Yersinia enterocolitica (Hill et al. 1983b, Jagow and Hill 1986), Listeria monocytogenes (Datta et al. 1988), enterotoxigenic Escherichia coli (Hill et al. 1983a, 1986), Vibrio vulnificus (Morris et al. 1987), enterotoxigenic Staphylococcus aureus (Notermans et al. 1988), Clostridium perfringens (Wernars and Notermans 1990) and Clostridium botulinum (Wernars and Notermans 1990). The first commercially available nucleic-probe-based assay system for food analysis was introduced by Gene Trak Systems (Framingham, MA, USA) in 1985 (Fitts 1985). This test used Salmonella-specific DNA probes directed against chromosomal DNA to detect Salmonella in enriched food samples. The format of the test involved hybridisation between target DNA bound to a membrane filter and phosphorous 32-labelled probes. The total analysis time for the test was 40–44 hours of sample enrichment in non-selective and selective media, followed by the hybridisation procedure lasting 4–5 hours. Thus the total analysis time was approximately 48 hour. The Salmonella test was evaluated in collaborative studies in the USA and appeared to be at least equivalent to standard culture methods (Flowers et al. 1987). Gene Trak also produced a hybridisation assay for Listeria spp., based on a similar format (Klinger and Johnson 1988). The Gene Trak probe kits gained acceptance within the United States and a number of laboratories began using them. In Europe, however, there was a reluctance among food laboratories to use radioisotopes within the laboratory. In 208 Chilled foods addition, 32 P has a short half-life, which caused difficulties when transporting kits to distant sites. In 1988 Gene Trak began marketing non-isotopically labelled probes for Salmonella, Listeria and Escherichia coli. The detection system for the probes was colorimetric. In order to overcome the reduction in sensitivity caused by the use of non-isotopic labels, the target nucleic acid within the cell was ribosomal RNA. This nucleic acid is present in an estimated 500 to 20,000 copies per cell. The colorimetric hybridisation assay is based on a liquid hybridisation reaction between the target rRNA and two separate DNA oligonucleotide probes (the capture probe and the reporter probe) that are specific for the organism of interest. The capture probe molecules are extended enzymatically with a polymer of approximately 100 deoxyadenosine monophosphate residues. The reporter probe molecules are labelled chemically with the hapten fluorescein. Following a suitable enrichment of the food under investigation, a test sample is transferred to a tube and the organisms lysed, releasing rRNA targets. The capture and detector probes are then added and hybridisation is allowed to proceed. If target rRNA is present in the sample, hybridisation takes place between the probes and the target 16s rRNA. The solution containing the target probe complex is then brought into contact with a solid support dipstick, containing bound deoxythymidine homopolymer, under conditions that will allow hybridisation between the poly-deoxyadenosine polymer of the capture probe and the poly-deoxythymidine on the dipstick. Unhybridised nucleic acids and cellular debris are then washed away, leaving the capture DNA-RNA complex attached to the surface of the dipstick. The bound fluoresceinated reporter probe is detected by the addition of an antifluorescein antibody conjugated to the enzyme horseradish peroxidase. Subsequent addition of a chromogenic substrate for the enzyme results in colour development that can be measured spectrophotometrically. Results of the colorimetric assays (Mozola et al. 1991) have indicated a good comparison between the probe methods and conventional cultural procedures for both Salmonella and Listeria. The sensitivity of the kits appeared to be between 10 5 and 10 6 target organisms/ml, and thus the enrichment procedure is a critical step in the methodology. Since the introduction of the three kits previously mentioned, Gene Trak have begun to market systems for Staphylococcus aureus, Campylobacter spp. and Yersinia enterocolitica. Commercially available nucleic acid probes for the confirmation of Campylobacter, Staphylococcus aureus and Listeria are available from Gen- probe (Gen Probe Inc., San Diego, USA). These kits are based on a single- stranded DNA probe that is complementary to the ribosomal RNA of the target organism. After the ribosomal RNA is released from the organism, the labelled DNA probe combines with it to form a stable DNA:RNA hybrid. The hybridised probe can be detected by its luminescence. The assay method used is termed a hybridisation protection assay and is based on the use of a chemiluminescent acridinium ester. This ester reacts with hydrogen peroxide under basic conditions to produce light that can be measured Conventional and rapid analytical microbiology 209 in a luminometer. The acridinium esters are covalently attached to the synthetic DNA probes through an alkylamine arm. The assay format is based on differential chemical hydrolysis of the ester bond. Hydrolysis of the bond renders the acridinium permanently non-chemiluminescent. When the DNA probe which the ester is attached, hybridises to the target RNA, the acridinium is protected from hydrolysis and can thus be rendered luminescent. The test kits for Campylobacter and Listeria utilise a freeze-dried probe reagent. The Campylobacter probe reacts with C. jejuni, C. coli and C. pylori; the Listeria probe reacts with L. monocytogenes. In both cases a full cultural enrichment protocol is necessary prior to using the probe for confirmation testing. An evaluation of the L. monocytogenes probe kit (Bobbit and Betts 1991) indicated that it was totally specific for the target organism. The sensitivity required approximately 10 6 L. monocytogenes to be present in order for a positive response to be obtained. The kit appeared to offer a fast reliable culture confirmation test and had the potential to be used directly on enrichment broth, thus reducing test times even further. Probes – the future The development and use of probes in the food industry has advanced little in recent years. The kits that are currently available show great promise but are not as widely used as immunoassays. Microbiologists must always consider the usefulness of analysing the genetic information within cells, for example to detect the presence of genes coding for toxins could be detected, even when not expressed, and screening methods could be devised for pathogenicity plasmids, such as that in Yersinia enterocolitica. It may be however, that the advances in molecular biology mean that the best way to test for such information is by using nucleic acid amplification methods such as the Polymerase Chain Reaction (PCR). Nucleic acid amplification techniques In recent years, several genetic amplification techniques have been developed and refined. The methods usually rely on the biochemical amplification of cellular nucleic acid and can result in a 10 7 -fold amplification in two to three hours. The very rapid increase in target that can be gained with nucleic acid amplification methods, makes them ideal candidates for development of very rapid microbial detection systems. A number of amplification methods have been developed and applied to the detection of microorganisms: ? Polymerase Chain Reaction (PCR) and variations, including nested PCR, reverse transcriptase (RT) PCR and multiplex PCR. ? Q Beta Replicase. ? Ligase Amplification Reaction (LAR). ? Transcript Amplification System (TAS), also known as Self Sustained Sequence Replication (3SR) or Nucleic Acid Sequence Based Amplification (NASBA). 210 Chilled foods Of these amplification methods only PCR has been commercialised as a kit- based procedure for the detection of food-borne microorganisms. Much research has been done with NASBA and there are a number of research papers outlining its use for detecting food pathogens but as yet, no commercially available kits are on the market. Polymerase chain reaction (PCR) PCR is a method used for the repeated in vitro enzymic synthesis of specific DNA sequences. The method uses two short oligonucleotide primers that hybridise to opposite strands of a DNA molecule and flank the region of interest in the target DNA. PCR proceeds via series of repeated cycles, involving DNA denaturation, primer annealing and primer extension by the action of DNA polymerase. The three stages of each cycle are controlled by changing the temperature of the reaction, as each stage will occur only at particular defined temperatures. These temperature changes are accomplished by using a specialised instrument known as a thermocycler. The products of primer extension from one cycle, act as templates for the next cycle, thus the number of target DNA copies doubles at every cycle. Reverse transcriptase – PCR This involves the use of an RNA target for the PCR reaction. The PCR must work on a DNA molecule; thus initially reverse transcriptase is used to produce copy DNA (cDNA). The latter is then used in a conventional PCR reaction. The RT-PCR reaction is particularly applicable to certain microbiological tests. Some food-borne viruses contain RNA as their genetic material; thus RT-PCR must be used if amplification and thus detection of these viruses is necessary. A second use of RT-PCR is in the detection of viable microorganisms. One of the problems associated with PCR is its great sensitivity and ability to amplify very low concentrations of a target nucleic acid. Thus, if using PCR to detect the presence or absence of a certain microorganism in a food, PCR could ‘detect’ the organism, even if it had been previously rendered inactive by a suitable food process. This could result in a false positive detection. A way to overcome this problem is to use an RT-PCR targeted against cellular messenger RNA, which is only produced by active cells and once produced has a short half-life. Thus a detection of specific mRNA by an RT-PCR procedure is indicative of the presence of a viable microorganism. NASBA NASBA is a multi-enzyme, multicycle amplification procedure, requiring more enzymes and reagents than standard PCR. It does however, have the advantage of being isothermal, therefore all stages of the reaction occur at a single temperature and a thermocycler is not required. Various research papers have been published which use NASBA to detect foodborne pathogeses (e.g. Uyttendaele et al. 1996), however, the procedure has yet to be commercialised. Conventional and rapid analytical microbiology 211 Commercial PCR-based kits Currently there are three manufacturers producing kits based on PCR for the detection of food-borne microorganisms. BAX (Qualicon, USA) utilises tabletted reagents and a conventional thermocycler, geL electrophoresis-based approach. Positive samples are visualised as bands on an electrophoresis gel. BAX kits are available for Salmonella (Bennett et al. 1998), Listeria Genus Listeria monocytogenes, E. coli 0157:H7; the tests for Salmonella and E.coli 0157 have been through an Association of Official Analytical Chemists Research Institute (AOACRI) testing procedure and have gained AOACRI Performance Tested Status. The second of the commercially available PCR kits is the Probelia kit (Sanofi, France); this uses conventional PCR followed by an immunoassay and colorimetric detection system. Kits are available for Salmonella and Listeria. The final commercial PCR system is the TaqMan system (Perkin Elmer, USA). This uses a novel probe system incorporating a TaqMan Label. This is non-fluorescent in its native form, but once the probe is bound between the primers of the PCR reaction, it can be acted upon by the DNA polymerase enzyme used in PCR to yield a fluorescent end product. This fluorescence is detected by a specific fluorescence detection system. TaqMan kits are available for Salmonella and under development for Listeria and E. coli 0157. Perhaps one of the most interesting future aspects of TaqMan is its potential to quantify an analyte. Currently PCR-based systems are all based on presence/absence determinations, TaqMan procedures and instrumentation can give information on actual numbers. Therefore the potential for using PCR for rapidly counting microorganisms could now be achieved. Separation and concentration of microorganisms from foods In recent years there has been considerable interest in the potential for separating microorganisms from food materials and subsequently concentrating them to yield a higher number per unit volume. The reason for this interest is that many of the currently available rapid test methods have a defined sensitivity, examples are: 10 4 /ml for ATP luminescence, 10 6 /ml for electrical measurement, 10 5 C010 6 / ml for immunoassay and DNA probes and approximately 10 3 /ml for current PCR based kits. These sensitivity levels mean that a growth period is usually required before the rapid method can be applied and this growth period may significantly increase the total test time. One way in which this problem can be addressed is by separating and concentrating microorganisms from the foods, in order to present them to the analytical procedure in a higher concentration. An additional advantage being that the microbial cells may be removed from the food matrix, which in some cases may contain materials which interfere with the test itself. A simple example of the use of concentration, is in the analysis of clear fluids (water, clear soft drinks, wines, beers, etc.). Here contamination levels are usually very low, thus large volumes are membrane filtered to concentrate the microorgan- isms onto a small area. These captured organisms can then be analysed. A 212 Chilled foods thorough review of separation concentration methods has been given by Betts (1994). They broadly fall into five categories: 1. filtration 2. centrifugation 3. phase separation 4. electrophoresis 5. immuno-methods. Of these categories only one has reached commercialisation for use in solid foods, these are the immuno-methods. Immunomagnetic separation relies on coating small magnetic particles with specific antibodies for a known cell. The coated particles can be added into a food suspension or enrichment, and if present, target cells will attach to the antibodies on the particles. Application of a magnetic field retains the particles and attached cells allowing food debris and excess liquid to be poured away, thus separating the cells from the food matrix and concentrating them. This type of system has been commercialised by Dynal (Norway), LabM (England), and Denka (Japan), an automated system incorporating the procedure is produced by Foss Electric (EIAFOSS). The various companies produce kits for Salmonella, Listeria, E. coli 0157, other verocytotoxin producing E. coli and Campylobacter. Immunomagnetic separation systems for detecting the presence of E. coli 0157 have been very widely used and become accepted standard reference methods in many parts of the world. Identification and characterisation of microorganisms Once an organism has been isolated from a food product it is often necessary to identify it, this is particularly relevant if the organism is considered to be a pathogen. Traditionally, identification methods have involved biochemical or immunological analyses of purified organisms. With the major advances now taken in molecular biology, it is now possible to identify organisms by reference to their DNA structure. The sensitivity of DNA based methods will in fact allow identification to a level below that of species (generally referred to as characterisation or sub-typing). Sub-typing is a powerful new tool that can be used by food microbiologists not just to name an organism, but also to find out its origin. Therefore it is possible in some cases to isolate an organism in a finished product, and then through a structured series of tests find whether its origin was a particular raw material, the environment within a production area or a poorly cleaned piece of equipment. A number of DNA-based analysis techniques have been developed that allow sub-typing, many of these have been reviewed by Betts et al. (1995). There is, however, only one technique that has been fully automated, and made available to food microbiologist on a large scale, and that is Ribotyping through use of the Qualicon RiboPrinter (Qualicon, USA). This fully automated instrument accepts isolated purified colonies of bacteria, and produces DNA band images (RiboPrint patterns), that are automatically compared to a database to allow Conventional and rapid analytical microbiology 213 identification and characterisation. The technique has successfully been used within the food industry to identify contaminants, indicate the sources and routes of contamination and check for culture authenticity (Betts 1998). 8.5 Microbiological methods – the future Conventional microbiological methods have remained little changed for many decades. Microbiologists generally continue to use lengthy enrichment and agar- growth-based methods to enumerate, detect and identify organisms in samples. As the technology of food production and distribution has developed, there has been an increasing requirement to obtain microbiological results in shorter time periods. The rapid growth of the chilled foods market, producing relatively short shelf-life products, has led this move into rapid and automated methods, as the use of such systems allows: (a) testing of raw materials before use; (b) monitoring of the hygiene of the production line in real time; and (c) testing of final products over a reduced time period. All of these points will lead to better quality food products with an increased shelf-life. All of the methods considered in this chapter are currently in use in Quality Control laboratories within the food industry. Some (e.g. electric methods) have been developed, established and used for a considerable time period, whilst others (e.g. Polymerase Chain Reaction), are a much more recent development. The future of all of these methods is good; they are now being accepted as standard and routine, rather than novel. Some users are beginning to see the benefits of linking different rapid methods together to gain an even greater test rapidity, e.g. using an enzyme immunoassay to detect the presence of Listeria spp., then using a species-specific nucleic acid probe to confirm the presence or absence of L. monocytogenes. One of the problems of many of the rapid methods is a lack of sensitivity. This does in many cases mean that lengthy enrichments are required prior to using rapid methods. Research on methods for the separation and concentration of microorganism from food samples would enable microorganisms to be removed from the background of food debris and concentrated, thus removing the need for long incubation procedures. The developments in DNA-based methods for both detection and identification/characterisation have given new tools to the food microbiologist, there is no doubt that these developments will continue in the future giving significant analytical possibilities that are currently difficult to imagine. 8.6 References and further reading ALEXANDER M K, KHAN M S and DOW C S, (1981) Rapid screening for bacteriuria using a particle counter, pulse-height analyser and computer. Journal of Clinical Pathology 34, 194–8. 214 Chilled foods ANON, (1986) Microorganisms in Foods 2. Sampling for microbiological analysis: Principles and specific applications, ICMSF. Blackwell Scien- tific, Oxford. BACK J P and KROLL R G, (1991) The differential fluorescence of bacteria stained with acridline orange and the effect of heat, Journal of Applied Bacteriology, 71 51–8. BANKES P, (1991) An evaluation of the Bactrac 4100. Campden Food & Drink Research Association Technical Memorandum 628, Campden Food & Drink Research Association, UK. BANKES P and ROSE S A, (1989) Rapid detection of Staphylococcal enterotoxins in foods with a modification of the reversed passive latex agglutination assay, Journal of Applied Bacteriology, 67 395–9. BANKES P, ROWE D and BETTS R P, (1991) The rapid detection of yeast spoilage using the Chemflow system. Campden Food & Drink Research Association Technical Memorandum 621, Campden Food & Drink Research Associa- tion, UK. BANKS J G, ROSSITER L M and CLARK A E, (1989) Selective detection of Pseudomonas in foods by a conductance technique. In: Rapid Methods and Automation in Microbiology and Immunology Florence 1987. Balows, A., Tilton, R.C. and Turano, A. (eds) Brixia Academic Press, Brescia, pp. 725– 7. BAUMGART J, FRICKLE K and KUY C, (1980) Quick determination of surface bacterial content of fresh meat using a bioluminescence method to determine adenosine triphosphate, Fleischwirtschaft, 60 266–70. BENNETT A R, GREENWOOD D, TENNANT C, BANKS J G and BETTS R P, (1998). Rapid and Definitive detection of Salmonella in foods by PCR, Letters in Applied Microbiology, 26 (6) 437–41. BETTS R P, (1993) Rapid electrical methods for the detection and enumeration of food spoilage yeasts, International Biodeterioration and Biodegradation, 32 19–32. BETTS R P, (1994) The separation and rapid detection of microorganisms In: Rapid Methods and Automation in Microbiology and Immunology Spencer R.C., Wright E.P. and Newsom S.W.B., (eds) pp. 107–120, Intercept Press, Hampshire, England. BETTS R P, (1998) Foodborne bacteria in the spotlight, Laboratory News September p. A6–7. BETTS R P and BANKES P, (1988) An evaluation of the Autotrak – A rapid method for the enumeration of microorganisms. CampdenFoodandDrink Research Association Technical Memorandum 491, Campden and Chorley- wood Food Research Association, UK. BETTS R P, BANKES P, FARR L and STRINGER M F, (1988) The detection of irradiated foods using the Direct Epifluorescent Filter Technique, Journal of Applied Bacteriology, 64 329–35. BETTS R P, BANKES P and BANKS J G, (1989) Rapid enumeration of viable microorganisms by staining and direct microscopy. Letters in Applied Conventional and rapid analytical microbiology 215 Microbiology, 9 199–202. BETTS R P, BANKES P and GREEN J, (1991) An evaluation of the Tecra enzyme immunoassay for the detection of Listeria in foods. In: Food Safety and Quality Assurance: Applications of Immunoassays Systems. Morgan M R A, Smith C J and Williams P A (eds) Elsevier, London, pp. 283–98. BETTS R P, STRINGER M, BANKS J G and DENNIS C, (1995) Molecular Methods in Food Microbiology. Food Australia, 47 (7) 319–22. BILLTE M and REUTER G, (1985) The bioluminescence technique as a rapid method for the determination of the microflora of meat, International Journal of Food Microbiology, 2 371–81. BLACKBURN C DE W and STANNARD C J, (1989) Immunological detection methods for Salmonella in foods. In: Rapid Microbiological Methods for Foods, Beverages and Pharmaceuticals. Society for Applied Bacteriology Technical Series 25, Stannard, C.L., Pettit, S.B. and Skinner, F.A. (eds) Blackwell Scientific, Oxford. BLACKBURN C DE W, CURTIS L M, HUMPHESON L and PETITT S, (1994) Evaluation of the Vitek Immunodiagnostic Assay System (VIDAS) for the detection of Salmonella in foods, Letters in Applied Microbiology, 19 (1) 32. BOBBITT J A and BETTS R P, (1991) Evaluation of Accuprobe Culture Confirmation Test for Listeria monocytogenes. Campden Food & Drink Research Association Technical Memorandum 630, Campden Food & Drink Research Association, UK. BOBBITT J A and BETTS R P, (1993) Evaluation of the VIDAS Listeria system for the detection of Listeria in foods. Campden Food and Drink Research Association Technical Memorandum 674, Campden and Chorleywood Food Research Association, UK. BOLTON F J and GIBSON D M, (1994) Automated electrical techniques in microbiological analysis. In Rapid Analysis Techniques in Food Micro- biology, ed. Patel P., Blackie Academic, London. BOLTON F J and POWELL S J, (1993) Rapid methods for the detection of Salmonella and Campylobacter in meat products, European Food and Drink Review, Autumn 73–81. BOSSUYT R, (1981) Bacteriological quality determination of raw milk by an ATP assay technique, Milchwissenschaft, 36 257–60. BREED R S, and BREW J D, (1916) Counting bacteria by means of the microscope. Technical Bulletin 49. New York Agricultural Experimental Station, Albany, New York. CANDLISH A, (1992) Rapid testing methods for Salmonella using immunodiag- nostic kits. Presented at: Complying with Food Legislation through Immunoassay, February 1992, Glasgow. CARTER N P and MEYER E W, (1990) Introduction to the principles of flow cytometry. In: Flow cytometry, a practical approach, Ormerod, M.G. (ed.) IRL Press, Oxford, pp. 1–28. CLARK C, CANDLISH A A G and STEELL W, (1989) Detection of Salmonella in foods using a novel coloured latex test. Food and Agricultural Immunology, 216 Chilled foods 1 3–9. CLAYDEN J A, ALCOCK S J and STRINGER M F, (1987) Enzyme linked immunoabsorbant assays for the detection of Salmonella in foods. In: Immunological Techniques in Microbiology. Society for Applied Bacteriol- ogy Technical Series 24 Grange, J.M., Fox, A. and Morgan, N.L. (eds). Blackwell Scientific, Oxford, pp. 217–30. CONNOLLY P, LEWIS S J and CORRY J E L, (1988) A medium for the detection of yeasts using a conductimetric method, International Journal of Food Microbiology, 7 31–40. COOMBES P, (1990) Detecting Salmonella by conductance. Food Technology International Europe, No. 4229, 241–6. COOTES R L and JOHNSON R, (1980) A fluorescent staining technique for determination of viable and non-viable yeast and bacteria in wines, Food Technology in Australia, 32 522–24. COUSINS D L and MARLATT F, (1990) An evaluation of a conductance method for the enumeration of enterobacteriaceae in milk, Journal of Food Protection, 53 568–70. CURIALE M S, FLOWERS R S, MOZOLA M A and SMITH A E, (1986) A commercial DNA probe based diagnostic for the detection of Salmonella in food samples: In: DNA Probes: Applications in Genetic and Infectious Disease and Cancer, Lerman, L. S. (ed.) Cold Spring Harbour Laboratory, New York, pp. 143–8. CURIALE M S, KLATT M L, ROBINSON B J and BECK L T, (1990a) Comparison of colorimetric monoclonal enzyme immunoassay screening methods for detection of Salmonella in foods, Journal of the Association of Official Analytical Chemists, 73 43–50. CURIALE M S, MCIVER D, WEATHERSBY S and PLANER C, (1990b) Detection of Salmonella and other enterobacteriaceae by commercial deoxyribonucleic acid hybridization and enzyme immunoassay kits, JournalofFood Protection, 53 1037–46. DATTA A R, WENTZ B A, SHOOK D and TRUCKSESS M W, (1988) Synthetic oligonucleotide probes for detection of Listeria monocytogenes, Applied and Environmental Microbiology, 54 2933–7. DAVIS S C and JONES K L, (1997) A study of the recovery of microorganisms from flour. Campden and Chorleywood Food Research Association R&D Report No. 51, Campden and Chorleywood Food Research Association, UK. DIJKMAN, A.J., SCHIPPER, C.J., BOOY, C.J. and POSTHUMUS, G. (1969) The estimation of the number of cells in farm milk, Netherlands Milk and Dairy Journal, 23 168–81. DONNELLY C W and BAIGENT G T, (1986) Method for flow cytometric detection of Listeria monocytogenes in milk, Applied and Environmental Microbiology, 52 689–95. DRUGGAN P, FORSYTHE S J and SILLEY P, (1993) Indirect impedance for microbial screening in the food and beverage industries. In: New Techniques in Food Conventional and rapid analytical microbiology 217 and Beverage Microbiology, (eds R G Kroll and A Gilmour) SAB Technical Series 31, London, Blackwell. DUMAIN P P, DESNOUVEAUZ R, BLOC’H L, LECONTE C, FUHRMANN B, DE COLOMBEL E, PLESSIS M C and VALERY S, (1990) Use of flow cytometry for yeast and mould detection in process control of fermented milk products: The Chemflow System – A Factory Study, Biotech Forum Europe, 7 (3) 224–9. EASTER M C and GIBSON D M, (1985) Rapid and automated detection of Salmonella by electrical measurement, Journal of Hygiene, Cambridge, 94 245–62. EASTER M C and GIBSON D M, (1989) Detection of microorganism by electrical measurements. In: Rapid Methods in Food Microbiology. Progress in Industrial Microbiology Vol. 26 Adams, M.R. and Hope, C.F.A. (eds) Elsevier pp. 57–100. EDEN R and EDEN G, (1984) Impedance Microbiology, Research Studies Press. Letchworth, Herts, England. FITTS R, (1985) Development of a DNA-DNA hybridization test for the presence of Salmonella in foods, Food Technology, 39 95–102. FLOWERS R S, KLATT M J and KEELAN S L, (1988) Visual immunoassay for the detection of Salmonella in foods. A collaborative study. Journal of the Association of Official Analytical Chemists, 71 973–80. FLOWERS R S, MOZOLA M A, CURIALE M S, GABIS D A and SILLIKER J H, (1987) Comparative study of DNA hybridization method and the conventional cultural procedure for detection of Salmonella in foods, Journal of Food Science, 52 842–5. FRANCISCO D D, MAH R A and RABIN A C, (1973) Acridine orange epifluorescent technique for counting bacteria in natural waters, Transactions of the American Microscopy Society, 92 416–21. GRIFFITHS M W, (1995) Bioluminescence and the food industry. Journal of Rapid Methods and Automation in Microbiology, 4 65–75. GUTTERIDGE C S and ARNOTT M L, (1989) Rapid Methods: An Over the Horizon View. In: Rapid Methods in Microbiology. Progress in Industrial Microbiology Vol. 26, Adams, M.R. and Hope, C.F.A. (eds) Elsevier, pp. 297–319. HADFIELD S G, JOVY N F and MCILLMURRAY M B, (1987b) The application of a novel coloured latex test to the detection of Salmonella. In: Immunological Techniques in Microbiology. Society for Applied Bacteriology Technical Series 24. Grange, J.M., Fox, A. and Morgan, N.L. (eds) Blackwell Scientific, Oxford, pp. 145–52. HADFIELD S G, LANE A and MCILLMURRAY M B, (1987a) A novel coloured latex test for the detection and identification of more than one antigen, Journal of Immunological Methods, 97 153–8. HILL W E, MADDEN J M, MCCARDELL B A, SHAH D B, JAGOW J A, PAYNE W L and BOUTIN B K, (1983a) Foodborne entertoxigenic Esch coli: detection and enumeration by DNA colony hybridisation Applied and Environmental Microbiology, 46 636–41. 218 Chilled foods HILL W E, PAYNE W L and AULISIO C C G, (1983b) Detection and enumeration of virulent Yersinia enterocolitica in food by colony hybridisation, Applied and Environmental Microbiology, 46 636–41. HILL W E, WENTZ B A, JAGOW J A, PAYNE W L and ZON G, (1986) DNA colony hybridisation method using synthetic oligonucleotides to detect enterotoxi- genic Esch. Coli: a collaborative study, Journal of the Association of Official Analytical Chemists, 69 531–6. HOBBIES J F, DALEY R J and JASPER S, (1977) Use of nucleoport filters for counting bacteria by fluorescence microscopy, Applied and Environmental Microbiology, 33 1225–8. HOLAH J T, (1989) Monitoring the Hygienic Status of Surfaces. In: Hygiene – The issues for the 90’, Symposium Proceedings, Campden and Chorley- wood Food Research Association, UK. HOLAH J T, BETTS R P and THORPE R H, (1988) The use of direct epifluorescent microscopy (DEM) and the direct epifluorescent filter technique (DEFT) to assess microbial populations on food contact surfaces, Journal of Applied Bacteriology, 65 215–21. HOLBROOK R, ANDERSON J M, BAIRD-PARKER A C and STUCHBURY S H, (1989) Comparative evaluation of the Oxoid Salmonella Rapid Test with three other rapid Salmonella methods, Letters in Applied Microbiology, 9 161–4. HUERNNEKENS F M and WHITELEY H R, (1960) In: Florkin, M. and Mason, H.S. (eds) Comparative Biochemistry, Volume 1. Academic Press, New York, p. 129. HUGHES D, SUTHERLAND P S, KELLY G and DAVEY G R, (1987) Comparison of the Tecra Salmonella tests against cultural methods for the detection of Salmonella in foods, Food Technology in Australia, 39 446–54. HUTTER K J and EIPEL H E, (1979) Rapid determination of the purity of yeast cultures by immunofluorescence and flow cytometry, Journal of the Institute of Brewing, 85 21–2. HYSERT D W, LOVECSES F and MORRISON N M, (1976) A firefly bioluminescence ATP assay method for rapid detection and enumeration of brewery micro- organisms, Journal of the American Society of Brewing Chemistry, 34 145–50. JAGOW J and HILL W E, (1986) Enumeration by DNA dolony hybridisation of virulent Yersinia enterocolitica colonies in artificially contaminated food, Applied and Environmental Microbiology, 51 411–43. JONES J G and SIMON B M, (1975) An investigation of errors in direct counts of aquatic bacteria by epifluorescence microscopy with reference to a new method of dyeing membrane filters, Journal of Applied Bacteriology, 39 1– 13. JONES K L and BETTS R P, (1994) The EIAFOSS system for rapid screening of Salmonella from foods. Campden Food & Drink Research Association Technical Memorandum 709, Campden and Chorleywood Food Research Association, UK. JONES K L, MACPHEE S, TURNER A and BETTS R P, (1995a) An evaluation of the Oxoid Listeria Rapid Test incorporating clearview for the detection of Conventional and rapid analytical microbiology 219 Listeria from foods. Campden and Chorleywood Food Research Associa- tion R&D Report No 19, Campden & Chorleywood Food Research Association, UK. JONES K L, MACPHEE S, TURNER A and BETTS R P, (1995b) An evaluation of Pathstick for the detection of Salmonella from foods. Campden & Chorleywood Food Research Association R&D Report No 11, Campden and Chorleywood Food Research Association, UK. KAEREBY F and ASMUSSEN B, (1989) Bactoscan – Five years of experiences. In: Rapid methods and Automation in Microbiology and Immunology, Balows, A., Tilton, R.C. and Turano, A. (eds) Florence 1987. Brixia Academic Press, Brescia, pp. 739–45. KARL D M, (1980) Cellular nucleotide measurements and applications in microbial ecology, Microbiological Review, 44 739–96. KLINGER J D and JOHNSON A R, (1988) A rapid nucleic acid hybridisation assay for Listeria in foods, Food Technology, 42 66–70. KLINGER J D, JOHNSON A, CROAN D, FLYNN P, WHIPPIE K, KIMBALL M, LAWNE J and CURIALE M, (1988) Comparative studies of a nucleic acid hybridisation assay for Listeria in foods, Journal of the Association of Official Analytical Chemists, 73 669–73. KRICKA L J and WHITEHEAD T P, (1984) Luminescent immunoassays: new labels for an established technique, Diagnostic Medicine, May, 1–8. KYRIAKUDES A, BELL C and JONES K L, (eds) (1996) A Code of Practice for Microbiology Laboratories Handling Food Samples. Campden & Chorley- wood Food Research Association Guideline No. 9, Campden and Chorley- wood Food Research Association, UK. LAROCCO K A, GALLIGAN P, LITTLE K J and SPURGASH A, (1985) A rapid bioluminescent ATP method for determining yeast concentration in a carbonated beverage, Food Technology, 39 49–52. LEVIN G V, GLENDENNING J R, CHAPELLE E W, HEIM A H and ROCECK E, (1964) A rapid method for the detection of microorganisms by ATP assay: its possible application in cancer and virus studies, Bioscience, 14 37–8. LITTEL K J and LAROCCO K A, (1986) ATP screening method for presumptive detection of microbiologically contaminated carbonated beverages, Journal of Food Science, 51 474–6. LITTEL K J, PIKELIS S and SPURGASH A, (1986) Bioluminescent ATP assay for rapid estimation of microbial numbers in fresh meat, Journal of Food Protection, 49 18–22. LOHNEIS M, JASCHKE K H and TERPLAN G, (1987) An immunoluminomtric assay (ILMA) for the detection of staphylococcal enterotoxins, International Journal of Food Microbiology, 5 117–27. MCCLELLAND R G and PINDER A C, (1994) Detection of Salmonella typhimurium in dairy products with flow cytometry and monoclorical antibodies Applied and Environmental microbiology, 60 (12), 4255. MACRAE R M, (1964) Rapid Yeast Estimations, in: Proceedings of European Brewing Convention Brussels 1963. Elsevier Publishing, Amsterdam, pp. 220 Chilled foods 510–12. MCELROY W D, (1947) The energy source for bioluminescence in an isolated system, Proceedings of the National Academy of Science, USA, 33 342–5. MCELROY W D. and STREFFIER B L, (1949) Factors influencing the response of the bioluminescent reaction to adenosine triphosphate, Archives of Biochem- istry, 22 420–33. MATTINGLY J A, BUTMAN B T, PLANK M, DURHAM R J and ROBINSON B J (1988) Rapid monoclonal antibody based enzyme-linked immunosorbant assay for the detection of Listeria in food products, Journal of the Association of Official Analytical Chemists, 71 679–81. MORRIS J G, WRIGHT A C, ROBERTS D M, WOOD P K, SIMPSON L M and OLIVER J D, (1987) Identification of environmental Bibrio vulnificus isolates with a DNA probe for the cytotoxin-hemolysin gene, Applied and Environmental Microbiology, 53 193–5. MOZOLA M, HALBERT D, CHAN S, HSU H Y, JOHNSON A, KING W, WILSON S, BETTS R P, BANKES P and BANKS J G, (1991) Detection of foodborne bacterial pathogens using a colorimetric DNA hybridisation method. In: Genetic Manipulation Techniques and Applications. Society for Applied Bacteriol- ogy Technical Series No. 28 Grange, J. M., Fox, A. and Morgan, N. L. (eds) Blackwell Scientific, Oxford, pp. 203–16. MULDROW L L, TYNDALL R L and FLIERMANS C B, (1982) Application of flow cytometry to studies of free-living amoebae, Applied and Environmental Microbiology, 44 (6) 1258–69. NEAVES P, WADDELL M J and PRENTICE G A, (1988) A medium for the detection of Lancefield group D cocci in skimmed milk powder by measurement of conductivity changes, Journal of Applied Bacteriology, 65,437–48. NOTERMANS S, HEUVELMAN K J and WERNARS K, (1988) Synthetic enterotoxin B, DNA probes for the detection of enterotoxigenic Staphylococcus aureus, Applied and Environmental Microbiology, 54 531–3. OPPONG D and SNUDDEN B H, (1988) Comparison of acridine orange staining using fluorescence microscopy with traditional methods for microbiological examination of selected dry food products, Journal of Food Protection, 51 485–8. OWENS J D, THOMAS D S, THOMPSON P S and TIMMERMAN J W, (1989) Indirect conductimetry: A novel approach to the conductimetric enumeration of microbial populations, Letters in Applied Microbiology, 9 245–50. PATCHETT R A, BACK J P, PINDER A C and KROLL R G, (1991) Enumeration of bacteria in pure cultures and foods using a commercial flow cytometer Food Microbiology, 8 119–25. PETTIT S B, (1989) A conductance screen for enterobacteriaceae in foods, In: Rapid Microbiological Methods for Foods Beverages and Pharmaceuticals, Stannard, C J, Petitt, S B, Skinner, F A (eds) Blackwell Scientific, Oxford, pp. 131–42. PETTIPHER G L, (1987) Detection of low numbers of osmophilic yeasts in cre`me fondant within 24 h using a pre-incubated DEFT count, Letters in Applied Conventional and rapid analytical microbiology 221 Microbiology, 4 95–8. PETTIPHER G L, (1991) Preliminary evaluation of flow cytometry for the detection of yeasts in soft drinks, Letters in Applied Microbiology, 12 109– 12. PETTIPHER G L and RODRIGUES U M, (1981) Rapid enumeration of bacteria in heat treatment milk and milk products using a membrane filtration-epifluores- cence microscopy technique, Journal of Applied Bacteriology, 50 157–66. PETTIPHER G L and RODRIGUES U M, (1982) Semi-automated counting of bacteria and somatic cells in milk using epifluorescence microscopy and television image analysis, Journal of Applied Bacteriology, 53 323–9. PETTIPHER G L, MANSELL R, MCKINNON C H and COUSINS C M, (1980) Rapid membrane filtration-epifluorescence microscopy technique for the direct enumeration of bacteria in raw milk, Applied and Environmental Microbiology, 39 423–9. PETTIPHER G L, WILLIAMS R A and GUTTERIDGE C S, (1985) An evaluation of possible alternative methods to the Howard Mould Count, Letters in Applied Microbiology, 1 49–51. PHILLIPS A P and MARTIN K L, (1988) Limitations of flow cytometry for the specific detection of bacteria in mixed populations, Journal of Immunolo- gical Methods, 106 109–17. PINDER A C, PURDY P W, POULTER S A G and CLARK D C, (1990) Validation of flow cytometry for rapid enumeration of bacterial concentrations in pure cultures, Journal of Applied Bacteriology, 69 92–100. PONNAMPERUMA C, SAGAN C and MARINER R, (1963) Synthesis of adenosine triphosphate under possible primitive earth conditions, Nature, 199 222–6. RICHARDS J C S, JASON A C, HOBBS G, GIBSON D M and CHRISTIE R H, (1978) Electronic measurement of bacterial growth, Journal of Physics, E: Scientific Instruments, 11 560–8. RODRIGUES U M and KROLL R G, (1986) Use of the direct epifluorescent filter technique for the enumeration of yeasts, Journal of Applied Bacteriology, 61 139–44. RODRIGUES U M and KROLL R G, (1988) Rapid selective enumeration of bacteria in foods using a microcolony epifluorescence microscopy technique. Journal of Applied Bacteriology, 64 65–78. RODRIGUES U M and KROLL R G, (1989) Microcolony epifluorescence microscopy for selective enumeration of injured bacteria in frozen and heat treated foods, Applied and Environmental Microbiology, 55 778–87. ROSE S A and STRINGER M F, (1989) Immunological Methods. In: Rapid Methods in Food Microbiology. Progress in Industrial Microbiology Vol. 26, Adams, M.R. and Hope, C F A (eds) Elsevier, pp. 121–67. ROSE S A, BANKES P and STRINGER M F, (1989) Detection of staphylococcal enterotoxins in dairy products by the reversed passive latex agglutination (Setrpla) kit, International Journal of Food Microbiology, 8 65–72. SCHUTZBANK T E, TEVERE V and CUPO A, (1989) The use of bioluminescent bacteriophages for the detection and identification of Salmonella in foods. 222 Chilled foods In: Rapid methods and Automation in Microbiology and Immunology, Florence 1987. Balows, A, Tilcon, R C and Turano, A (eds) Brixia Academic Press, Brescia, pp. 241–51. SELIGER H and MCELROY M D, (1960) Spectral emission and quantum yield of firefly bioluminescence, Archives of Biochemistry and Biophysics, 88 136– 41. SHARPE A N, WOODROW M N and JACKSON A K, (1970) Adenosine triphosphate levels in foods contaminated by bacteria, Journal of Applied Bacteriology, 33 758–67. SHAW B G, HARDING C D, HUDSON W H and FARR L, (1987) Rapid estimation of micobial numbers in meat and poultry by the direct epiflourescent filter technique, Journal of Food Protection, 50 652–7. SHAW S., (1989) Rapid methods and quality assurance in the brewing industry. In: Rapid Methods in Food Microbiology, Progress in Industrial Microbiology Vol. 26. Adams, M.R. and Hope, C F A (eds) Elsevier, pp. 273–96. SKARSTAD K, STEEN H B and BOYE E, (1983) Cell cycle parameters of growing Escherichia coli. B/r studied by flow cytometry, Journal of Bacteriology, 154 652–6. SKJERVE E and OLSVIK O, (1991) Immunomagnetic separation of Salmonella from foods, International Journal of Food Microbiology, 14 11–18. STALKER R M, (1984) Bioluminescent ATP analysis for the rapid enumeration of bacteria-principles, practice, potential areas of inaccuracy and prospects for the Food Industry. Campden Food and Drink Research Association Technical Bulletin 57, Campden & Chorleywood Food Research Associa- tion, UK. STANNARD C J, (1989) ATP Estimation. In: Rapid Methods in Food Microbiology. Progress in Industrial Microbiology Vol. 26. Adams, M.R. and Hope, C.F.A. (eds) Elsevier. STANNARD C J, and WOOD, J.M. (1983) The rapid estimation of microbial contamination of raw meat by measurement of adenosine triphosphate (ATP), Journal of Applied Bacteriology, 55 429–38. STEEN H B and BOYE E, (1980) Escherichia coli growth studied by dual parameter flow cytophotometry, Journal of Bacteriology, 145 145–50. STEEN H B, BOYE E, SKARSTAD K, BLOOM B, GODAL T and MUSTAFA S (1982) Applications of flow cytometry on bacteria: Cell cycle kinetics, drug effects and quantification of antibody binding, Cytometry, 2 (4) 249–56. STEWART G N, (1899) The change produced by the growth of bacteria in the molecular concentration and electrical conductivity of culture media, Journal of Experimental Medicine, 4 23S–24S. STEWART G S A B, (1990) A review: In vivo bioluminescence: New Potentials for Microbiology, Letters in Applied Microbiology, 10 1–9. TYNDALL R L, HAND R E Jr, MANN R C, EVAN C and JERNIGAN R, (1985) Application of flow cytometry to detection and characterisation of Legionella spp., Applied and Environmental Microbiology, 49 852–7. Conventional and rapid analytical microbiology 223 ULITZUR S and KUHN J, (1987) Introduction of lux genes into bacteria, a new approach for specific determination of bacteria and their antibiotic susceptibility. In: Proceedings of Xth International Symposium on Bioluminescence and Chemoluminescence, Freiburg, Germany 1986. Scholmerich, J., Andreesen, A., Kapp, A., Earnst, M., Woods, E.G. (eds) J. Wiley. pp. 463–72. ULITZUR S, SUISSA M and KUHN J C, (1989) A new approach for the specific determination of bacteria and their antimicrobial susceptibilities. In: Rapid methods and automation in microbiology and immunology Balows, A., Tilton, R.C. and Turano, A. (eds). Florence 1987. Brixia Academic Press, Brescia, pp. 235–40. UYTTENDAELE M, SCHUKKINK A, VAN GEMEN B and DEBEVERE J, (1996) Comparison of the nucleic acid amplifications system NASBA and agar isolation for the detection of pathogenic campylobacters in naturally contaminated poultry, Journal of Food Protection, 59 (7) 683–9. VANDERLINDE P B and GRAU F H, (1991) Detection of Listeria spp. in meat and environmental samples by an enzyme-linked immunosorbant assay (ELISA), Journal of Food Protection, 54 230–1. VECKERT J, BREEUWER P, ABEE T, STEPHENS P and NEBE VON CARON G, (1995) Flow Cytometry Applications in physiological study and detection of foodborne organisms, International Journal of Food Microbiology, 28 (2) 317–26. VERMUNT A E M, FRANKEN A A J M, and BEUMER R R, (1992) Isolation of Salmonellas by immunomagnetic separation, Journal of Applied Bacteriol- ogy, 72 112–8. WALKER S J, (1989) Development of an impedimetric medium for the detection of Yersinia enterocolitica in pasteurised milk. In: Modern Microbiological Methods for Dairy Products, Proceedings of a seminar organised by the International Dairy Federation. IDF, pp. 288–91. WALKER S J, ARCHER P and APPLEYARD J, (1990) Comparison of Listeria-Tek ELISA kit with cultural procedures for the detection of Listeria species in foods, Food Microbiology, 7 335–42. WERNARS K and NOTERMANS S, (1990) Gene probes for detection of foodborne pathogens. In: Gene probes for bacteria Macario, A.J.L. and de Macario, E.C. (eds). Academic Press, London, pp. 355–88. WOLCOTT M J, (1991) DNA-based rapid methods for the detection of foodborne pathogens, Journal of Food Protection, 54 387–401. 224 Chilled foods