Oxford Medium Research Articles

Evaluation of Several Modifications of an Ecometric Technique for Assessment of Media Performance

Recovery of Listeria monocytogenes 101M, Jonesia denitrificans, salmonellae, and Pediococcus sp. NRRL B-2354 across nine media was evaluated with three modified versions of an ecometric method. Two approaches involved the use of broth cultures (108 to 109 CFU/ml) of individual strains and either large (10-μl) or small (1-μl) presterilized plastic loops. The third approach involved precultured slants and the inoculation of media with presterilized plastic inoculating needles (104 CFU per needle). Absolute growth indices (AGIs) were compared. No significant differences (P < 0.05) between methods were found when tryptic soy agar supplemented with 0.6% yeast extract (TSAYE) was used for the recovery of L. monocytogenes, J. denitrificans, Pediococcus sp. NRRL B-2354, and Salmonella spp. However, the small loop–broth technique recovered significantly fewer Salmonella enterica Typhimurium DT104 and Salmonella Senftenberg 775W cells than the other two techniques did. The performance of each individual bacterial strain on each of nine media was assayed. The recovery of L. monocytogenes was excellent (AGI > 4.8) with TSAYE, PALCAM, modified Oxford medium (MOX), and Baird-Parker agar and slight with modified PRAB (AGI = 0.4) and deMan Rogosa Sharpe (MRS) agar (<0.1), and the organism was not recovered with the remaining media (modified lysine iron agar [MLIA], xylose lysine desoxycholate [XLD] agar, and xylose lysine tergitol 4 [XLT4] agar). The recovery of J. denitrificans with TSAYE and MOX was excellent, significantly better than that achieved with PALCAM (AGI = 3.0), but the organism was not recovered with Baird-Parker agar or with the other media tested. The recovery of Pediococcus sp. NRRL B-2354 was excellent with TSAYE and modified PRAB medium > Baird-Parker agar > acidified MRS agar, but the organism was not recovered with any of the other media tested. The best recovery of S. enterica Typhimurium DT104 was achieved with TSAYE > MLIA ≥ XLD agar ≥ XLT4 agar > Baird-Parker > PALCAM, MOX, acidified MRS agar, modified PRAB, and MRS agar. The best recovery of Salmonella Senftenberg 775W was achieved with TSAYE, MLIA, and XLD agar > XLT4 agar, but the organism was not recovered with the other media evaluated.

Inactivation of Listeria monocytogenes Scott A 49594 in Apple Juice Supplemented with Cinnamon

Normal (pH 3.7) and adjusted (pH 5.0) pasteurized apple juice containing cinnamon (0, 0.1, 0.2, and 0.3%) was inoculated with Listeria monocytogenes Scott A 49594 at 104 CFU/ml and stored at 5 and 20°C for 7 days. Counts on tryptic soy agar (TSA), modified Oxford (MOX) medium, and thin agar layer (TAL) were determined at 1 h and 1, 3, and 7 days. The TAL method (MOX medium overlaid with TSA) was used for the recovery of injured cells. In apple juice, both at normal and adjusted pH, with any doses of cinnamon, no L. monocytogenes (a 4.6-log CFU/ml reduction) was detected after 1 h of storage at both temperatures, and no growth occurred at any points of storage. Therefore, cinnamon by itself (regardless of pH) had a pronounced killing effect. A further enrichment step with brain heart infusion agar showed that L. monocytogenes was completely inactivated in apple juice stored at 20°C, except in pH 5.0 samples with 0.1% of cinnamon. The TAL method was as effective as TSA in recovering injured cells of L. monocytogenes. Cinnamon considerably inactivates L. monocytogenes in apple juice and thus enhances the safety of this product.

Hypovirulent Listeria monocytogenes strains are less frequently recovered than virulent strains on PALCAM and Rapid' L. mono media

Several selective media have been developed to detect Listeria monocytogenes contaminated foodstuffs. Polymyxin-acriflavine-LiCl-ceftazidime-aesculin-mannitol (PALCAM) and Oxford media, required for the EN ISO method 11 290-1, are used for the detection of Listeria spp. in 2 days based on the expression of esculinase activity. Selective agar media such as Rapid' L. mono and Agar Listeria according to Ottaviani and Agosti (ALOA), based on the activity of phosphatidylinositol phospholipase C (PI-PLC) that allows the specific detection of L. monocytogenes in 2 days, are also used. However, no medium can assess the level of virulence of L. monocytogenes strains. Using a plaque-forming assay followed by subcutaneous footpad inoculation in mice, 15 virulent, 8 hypovirulent and 17 avirulent strains were discriminated among L. monocytogenes strains mainly originating from food (36/40). Their growth was tested on the four selective media. After 2 days, the number of colony forming units (cfu) of all the virulent strains was significantly superior to the number obtained with avirulent strains on all the four media tested, and superior to the number obtained with hypovirulent strains on PALCAM and Oxford media. These results showed a relationship between the level of virulence of L. monocytogenes strains and their growth on the selective agar media tested. Moreover, 1 out of 8 hypovirulent and 5 out of 17 avirulent strains did not grow on Rapid' L. mono medium, and 1 hypovirulent and 8 avirulent strains grew but did not express PI-PLC activity during the 7 days of incubation. The lack of detection of PI-PLC activity on Rapid' L. mono was not related to a gene mutation since these strains expressed enzymatic activity on ALOA medium, which detected up to 92% of the hypo- and avirulent strains. In contrast, some of these strains without growth or enzymatic activity expression would not be detected with PALCAM and Rapid' L. mono in foodstuffs on the second day.

Influence of sodium pyrophosphate on thermal inactivation of Listeria monocytogenes in pork slurry and ground pork

The thermal inactivation (55–62·5°C) of Listeria monocytogenes in pork slurry and ground pork that contained 0, 0·25 or 0·5% sodium pyrophosphate (SPP) was evaluated. Surviving cells were enumerated on Modified Oxford Medium. Decimal reduction (D)-values in pork slurry control (0% SPP) were 8·15, 2·57, 0·99, and 0·18 min, at 55, 57·5, 60 and 62·5°C, respectively; D-values in ground pork ranged from 15·72 min at 55°C to 0·83 min at 62·5°C. D-values in pork slurry that contained 0·25% SPP (w/v) were 4·75, 1·72, and 0·4 min, at 55, 57·5, and 60°C respectively; the values in ground pork ranged from 16·97 at 55°C to 0·80 min at 62·5°C. At 62·5°C,L. monocytogenes in slurry that contained SPP were killed too rapidly to allow determination of the D-value. Addition of 0·5% SPP further decreased (P<0·05) the heat resistance of L. monocytogenes in pork slurry but not in ground pork. The z-values in slurry ranged from 4·63 to 5·47°C whereas higher z-values (5·25 to 5·77°C) were obtained in ground pork. Degradation of SPP to orthophosphates in ground pork was two or three times greater than in pork slurry. Possible reasons for failure of SPP to reduce the thermal resistance of L. monocytogenes in ground pork are discussed.

Validation of ISO method 11290 Part 1 — Detection of Listeria monocytogenes in foods

The European and International Standard method for the detection of Listeria monocytogenes, described in EN ISO 11290 Part 1: 1997 (International Organisation for Standardisation, Geneva) was validated by order of the European Commission (Standards, Measurement and Testing Fourth Framework Programme Project SMT4-CT96-2098). Nineteen laboratories in 14 countries in Europe participated in a collaborative trial to determine the performance characteristics of the method, which are intended for publication in the corresponding standard. An additional objective of this project was to devise a new series of parameters to indicate the ‘precision’ of microbiological qualitative methods. The method was challenged with three food types, namely fresh cheese, minced beef and dried egg powder and a reference material. Inoculation levels ranged from 5 to 100 cfu/25 g. Each participant examined five replicates of each food type at three inoculum levels and five reference materials. Both PALCAM and Oxford media were assessed. All test materials were subjected to stringent homogeneity and stability testing before being used in the collaborative trial. The results demonstrated that the method prescribed in EN ISO 11290-1 had an overall sensitivity of 85.6% and a specificity of 97.4%. L. monocytogenes was detected in most cases after primary enrichment, although secondary enrichment often yielded further positives. However, a significant number of false-negative results were obtained with all food types when large numbers of L. innocua were present in the test materials. L. innocua tended to dominate L. monocytogenes during the selective enrichment stages and thus masked small numbers of colonies of L. monocytogenes on the isolation media. There was no evidence from this collaborative study to demonstrate a significant difference in performance between Oxford and PALCAM media. Due to the problem of false-negative results with this method as highlighted in this trial, recommendations have been made to ISO to launch a revision of the standard to improve the detection of low numbers of L. monocytogenes in foods. New statistical methods devised to advance the measurement of the performance of qualitative microbiological methods are also described.

Comparison of Different Enrichment Broths and Background Flora for Detection of Heat-Injured Listeria monocytogenes in Whole Milk

Various primary enrichment broths, including University of Vermont medium (UVM), Listeria enrichment broth (LEB), modified LEB, and aerobic and anaerobic L-PALCAMY, were compared with aerobic and anaerobic Pennsylvania State University (PSU) broths for the detection of severely heat-injured (62.8°C for 5, 10, or 15 min; no colony appearance after heat injury on aerobic Trypticase soy agar containing 0.6% yeast extract and modified Oxford medium) Listeria monocytogenes Scott A. Anaerobic conditions were produced by adding l-cysteine and then purging the headspace with N2. The effect of uninjured background flora (103 CFU/ml of Enterococcus faecium) on frequency of detection was examined. Anaerobic PSU broth resulted in the lowest false-positive rate and the highest frequency of detection of severely heat-injured L. monocytogenes compared with UVM, LEB, and modified LEB (P < 0.05). The presence of E. faecium significantly enhanced the detection of heat-injured (10 min at 62.8°C) L. monocytogenes in aerobic and anaerobic PSU and aerobic and anaerobic L-PALCAMY broths (P < 0.05). The highest concentration of uninjured E. faecium (>106 CFU/ml) inhibited the detection of heat-injured L. monocytogenes (P < 0.05). A heat-resistant, LiCl-tolerant Lactobacillus isolate from raw milk increased the rate of both false-positive and false-negative reactions.

A new rapid automated method for the detection of Listeria from environmental swabs and sponges

Many food and meat processors test environmental swabs and sponges to confirm the absence of Listeria spp. Spectral pattern changes in a liquid growth medium, resulting from esculin hydrolysis by Listeria in contaminated swabs and sponges, were automatically monitored by the BioSys instrument in a semifluid layer (SFL). The blackening of SFL in modified MOX broth resulted in sharply declining curves, which were easily detected by the instrument. The instrument detected all nine strains of Listeria monocytogenes tested. None of the gram negative organisms ( Proteus, Escherichia coli, Pseudomonas, Citrobacter and Yersinia) were detected by the system, nor were most gram positive organisms, including Bacillus, Streptococcus, and Lactobacillus strains. Staphylococcus aureus, Enterococcus faecium and E. faecalis, which hydrolyze esculin, produced black colonies on PALCAM and Oxford media and were also detected in the system. A total of 122 sponges and swabs collected at food processing plants were evaluated by this method. Of these, 99 were negative, and 11 were positive. L. innocua was the dominant Listeria species in these environmental samples. Good correlation was obtained between numbers of Listeria and detection times of esculin hydrolysis: 1000 CFU/swab were detected in 10–13 h, whereas 1–10 CFU/swab were detected in less than 22 h. The total assay time was 26 h.

The use of immuno-magnetic separation (IMS) as a tool in a sample preparation method for direct detection of L. monocytogenes in cheese

A sample preparation procedure was developed for direct detection of L. monocytogenes in cheese. The sample preparation protocol consisted of a 10-fold dilution and homogenization, a centrifugation step to precipitate large food particles, passage of the supernatant over a sieve and through a separatory funnel to further eliminate food particles and fat, a centrifugation step to recover the bacterial pellet and finally enzymatic digestion of the suspension to degrade the remaining small food particles. Recovery of L. monocytogenes was confirmed by plating on Oxford medium and confirmation of suspected colonies. This protocol enabled direct detection (without prior enrichment) of low numbers of L. monocytogenes (0.5–1.5 cfu/g cheese) from different types of cheese. The performance of Dynabeads ® Anti-Listeria (Dynal, Oslo, Norway) for selective recovery of L. monocytogenes and their applicability in the above mentioned procedure for direct detection of low numbers of L. monocytogenes from cheese was evaluated. IMS could not separate and recover L. monocytogenes from the food particles in the concentrated suspension. The use of IMS after a 24 h enrichment procedure (as recommended by the manufacturer) allowed for the detection of low numbers of L. monocytogenes (<10 cfu/g). However, experiments in broth cultures showed that although the detection limit of IMS with Dynabeads ® Anti-Listeria was 40–100 cfu/ml, the ratio of L. monocytogenes to non- Listeria flora was not increased. Thus, selective enrichment or concentration of L. monocytogenes was not obtained.

Development and Evaluation of a 24-Hour Method for the Detection and Quantification of Listeria monocytogenes in Meat Products

A 24-h filter monitor-based test, Listeria-SELeCT, has been developed to quantify Listeria monocytogenes organisms in meat samples with a sensitivity of ≤1.0 CFU/g. The technique comprises a filter monitor–based system and a colony lift immunoassay to identify and enumerate the target organism. Meat homogenates were centrifuged and the eluate was filtered to trap and immobilize the microorganisms on the filter. Fraser broth was then added to the filter apparatus to allow the organisms to become established overnight and to inhibit contaminants, after which the filters were transferred onto Modified Oxford medium agar, a selective medium for L. monocytogenes. After 10 to 12 h, a colony lift immunoassay was used to confirm and enumerate suspect colonies on the filter. A correlation study between the Listeria-SELeCT method and the most probable number technique showed the Listeria-SELeCT to be considerably more accurate than the most probable number for quantitatively determining the number of viable organisms in meat samples. Because of ease and speed of testing, the Listeria-SELeCT system also provided major advantages over the most probable number technology.

Thin Agar Layer Method for Recovery of Heat-Injured Listeria monocytogenes

A thin agar layer (TAL) method was developed to recover heat-injured Listeria monocytogenes. Modified Oxford medium (MOX), a selective plating medium, inhibits heat-injured L. monocytogenes from growing, whereas tryptic soy agar (TSA), a nonselective medium, does not. In order to facilitate recovery of heat-injured L. monocytogenes cells while providing selectivity of isolation of L. monocytogenes from other bacteria in the sample, a unique TAL procedure was developed by overlaying 5 ml of nonselective medium (TSA) onto prepoured and solidified MOX medium in an 8.5-cm–diameter petri dish. The injured L. monocytogenes repaired and started to grow in the TSA during the first few hours after incubation of the plate. During the resuscitation of injured cells, the selective agents from MOX diffused to the TSA top layer to inhibit other microorganisms. L. monocytogenes showed a typical reaction (black colonies) on TAL after 24 h of incubation at 37°C. The recovery rate for heat-injured L. monocytogenes with the TAL method was compared with those rates associated with TSA, MOX, and the traditional overlay method (OV; pouring selective agar on top of resuscitated cells on TSA agar after 3 h incubation). Milk and 0.1% peptone water that were inoculated with L. monocytogenes (4 to 5 log CFU/ml) were heated for 15 min at 55°C. L. monocytogenes was enumerated on TSA, MOX, OV, and TAL media and procedures. No significant difference occurred among TSA, OV, and TAL (P > 0.05) in terms of enumeration of heat-injured L. monocytogenes, but these media recovered significantly higher numbers than did MOX agar (P < 0.05)—in both samples. The TAL method involves only one step, whereas OV is a more cumbersome two-step procedure.

High-Pressure Destruction Kinetics of Listeria monocytogenes on Pork

Packaged fresh pork chops (30-g samples) containing an indigenous bacterial population of ∼107 CFU/g were inoculated with 107 CFU of Listeria monocytogenes Scott A per g, heat sealed, and subjected to high-pressure processing at 200 to 400 MPa for up to 90 min. Total counts and the number of surviving L. monocytogenes cells were determined by a spread plate technique on tryptic soy agar and modified Oxford medium, respectively. The pressure destruction was characterized by a dual-behavior, consisting of a step change in the number of survivors (Pko) with the application of a pressure pulse and a first-order rate drop in the number of survivors during the pressure hold period. Higher pressures resulted in higher rates of microbial inactivation, as indicated by their associated lower D values (and higher k values). The pressure sensitivities of the kinetic parameters were evaluated on the basis of Arrhenius and pressure death time (PDT)-type models. The results suggested that L. monocytogenes was more resistant to pressure inactivation than the indigenous microflora (the volume change of activation, ΔV≠ [Arrhenius model]), and zp values (PDT model) were −4.17 × 10−5 m3 mole−1 and 134 MPa for indigenous microflora and −3.43 × 10−5 m3 mole−1 and 163 MPa for L. monocytogenes respectively.

NEW METHODS FOR SCREENING LACTIC ACID BACTERIA THAT PRODUCE BACTERIOCIN ACTIVE AGAINST LISTERIA MONOCYTOGENES AND FOR DETERMINING BACTERIOCIN ACTIVITY 1

Abstract New methods using a microtiter plate for screening lactic acid bacteria (LAB) that produce bacteriocin active against Listera monocytogenes, and determination of bacteriocin activity using Omnispec 4000 were developed. In Modified Oxford (MOX, Oxoid) medium, L. monocytogenes produces a black color around the growth. In this study, inhibition of L. monocytogenes by bacteriocin was determined using the color change of the medium in a microtiter plate. Because each well of a microtiter plate limits the spread of microbial growth activity, this method can overcome the disadvantage of the agar overlay method in which a large clear zone can cover other microorganisms that are not producing bacteriocin. This new method is rapid, simple, and economical. Among the 96 strains of LAB isolated from meat, 25 strains produced bacteriocin against L. monocytogenes. The Omnispec 4000 was used to automatically record color changes due to the growth of L. monocytogenes in MOX medium containing six different concentrations of Nisin (bacteriocin) in wells of a microtiter plate. Depending on bacteriocin activities (IU). the slopes of the color changes recorded were different. The bacteriocin activities were determined easily and accurately by using the slopes to construct a calibration curve (r2= 0.96).

Survival and Injury of Listeria monocytogenes, Listeria innocua and Listeria ivanovii in Ground Pork Following Electron Beam Irradiation

The sensitivity of five strains of Listeria to electron beam irradiation in ground pork as well as the extent of sublethal radiation injury exhibited by each were investigated. Ground pork was inoculated with one of five strains of Listeria and irradiated with from 0 to 1.25 kGy at 0.25 kGy intervals. Listeria innocua NADC 2841 was more radiation-resistant (D10 = 0.638 kGy) than L. monocytogenes NADC 2045 Scott A (D10 = 0.447 kGy), L. monocytogenes NADC 2783 (a hamburger isolate) (D10 = 0.424 kGy), L. monocytogenes ATCC 15313 (D10 = 0.445 kGy), and L. ivanovii NADC 3518 (D10 = 0.372 kGy), when recovered on tryptic soy agar supplemented with 0.6% yeast extract. D10 values for L. innocua, L. ivanovii, and L. monocytogenes ATCC 15313 were lower when cells were recovered on modified Oxford medium. These three strains were susceptible to radiation-induced sublethal injury, with the numbers of injured organisms increasing with irradiation dose. The two pathogenic strains of L. monocytogenes were not injured significantly at the dose levels used. The results show that the dose range currently being considered by the Food and Drug Administration for the irradiation of beef and pork (1.5 to 4.5 kGy) is adequate for the elimination of L. monocytogenes from pork.

Comparative study of two plating media (PALCAM and Oxford) for detection of Listeria species in a range of meat products following a variety of enrichment procedures

Fifty‐two different varieties of sausage, salami and pâté were examined using Oxford (Oxoid) and PALCAM (Merck) listeria‐selective agars. Seven (13%) samples were positive for Listeria monocytogenes and 14 (27%) samples were positive for other Listeria species, while 31 (59%) samples were negative for Listeria species. The effectiveness of PALCAM and Oxford medium for the isolation of L. monocytogenes from the meat samples was compared. PALCAM medium was consistently more effective in suppressing other micro‐organisms thus enhancing the possibility of detecting Listeria species present in low numbers. The isolation of Listeria species for identification was also easier using PALCAM medium.

A 20–24 H MICROCOLONY‐IMMUNOBLOT TECHNIQUE to DIRECTLY DETECT and ENUMERATE LISTERIA MONOCYTOGENES INOCULATED INTO FOODS

A microcolony‐immunoblot technique (MCIBI) was developed to directly enumerate, in less than 24 h, very low numbers of Listeria monocytogenes (8–12 colony forming units: CFU/g or mL) inoculated into foods. Four meat and poultry and two dairy products were artificially inoculated with L. monocytogenes V7 diluted and plated on Oxford agar medium. Each plate was overlaid with an Immobilon‐P membrane and incubated for 18–20 h at 37C. Blot‐transferred colonies on these membranes were probed with C11E9 monoclonal antibody (MAb) and developed using peroxidase conjugated goat antimo use Ig G and a water insoluble substrate (3,3‐diaminobenzidin tetrahydmchloride; (DAB‐HCI), Nickel chloride and H2O2). the MCIBT gave L. monocytogenes counts that were not significantly lower than direct colony counts on selective agars. This technique allowed the recovery of 94–100% of L. monocytogenes cells inoculated into foods containing natural background flora counts of 3 × 104 to 8 × 106 CFU/g or mL. Using a 2 h resuscitation period on nonselective agar before overlay with Oxford media, the MCIBT allowed detection of sublethally heat injured cells of strain V7.

Isolation and Identification of Listeria monocytogenes from Retail Meats in Beijing

To determine the contamination of retail meats by Listeria monocytogenes in Beijing, 70 meat samples (25 pork, 10 beef, 14 lamb, and 21 chicken) were analyzed between January and March in 1990. Eight (11%) samples were positive for L. monocytogenes, and 39 (56%) samples were found to contain other Listeria spp. Seven pork and one chicken sample contained L. monocytogenes, whereas all beef and lamb were free of L. monocytogenes. Meanwhile, 15 (60%) pork, 11 (52%) chicken, 7 (70%) beef, and 6 (43%) lamb samples were positive for other Listeria spp. The eight confirmed isolates of L. monocytogenes were serologically typed. Six (5 from pork and 1 from chicken) were serotype l/2a, and the other two isolates (2 pork) were serotype l/2c and l/2b. No isolates were of serotypes 3 and 4. All positive samples for L. monocytogenes were frozen meats, and the incidence of Listeria spp. was greater for frozen samples than for fresh. Sixty of 70 samples (22 pork, 8 beef, 11 lamb, and 19 chicken) were used to compare the effectiveness of Oxford medium and Palcam medium for the isolation of L. monocytogenes from meat samples. L. monocytogenes was isolated from seven samples by Palcam medium, while five samples were positive by Oxford medium, indicating that Palcam medium was more effective for recovering L. monocytogenes than Oxford medium.

Comparison of Three Selective Isolation Media for the Detection of L. monocytogenes in Foods

391 different foods were examined in order to compare three selective isolation media for Listeria: blood agar with nalidixic acid, Palcam agar, Listeria selective Oxford medium (Oxford agar). The percentage of positive samples (presence of Listeria in 25 g) obtained on the three media after an enrichment procedure is identical: 15.8%. The species L. monocytogenes is found in respectively 8.43%, 8.43% and 8.18% of the foods examined on Palcam agar, Oxford agar and nalidixic acid agar; this represents more than 50% of the isolates of the genus Listeria. However, Palcam agar and Oxford agar offer the advantages of a great reduction of the development of the contaminating microflora and a clearly less fastidious reading. The technique used for the direct counting of L. monocytogenes on these two selective media does not allow the detection of a low contamination of foods (theoretical positivity threshold = 100 Listeria/g): 65.2% of the positive samples would not have been detected without prior enrichment. 41.1% of the isolated L. monocytogenes belong to serovar 1/2a, 5.8% to serovar 1/2b, 20.58% to serovar 1/2c, 5.8% to serovar 3b and 24.47% to serovar 4b.

Performance of enhanced haemolysis agar compared to Oxford medium for the isolation of Listeria monocytogenes from food enrichment broths

Combinations of tryptone soya broth plus yeast extract trypaflavine and nalidixic acid (TSByeAB), a modified Listeria enrichment broth (MLEB), Oxford medium and Enhanced Haemolysis Agar (EHA) were used to analyse 99 food samples from supermarkets in continental Europe. A total of 48 samples were positive for Listeria spp. using all combinations of enrichment broth, enrichment time and isolation medium. TSByeAB enrichments tended to result in loss of detection of positives over 48 h enrichment, whereas MLEB showed an improved detection. Oxford medium detected 33 (68·7%) of the 48 Listeria spp. positive samples, compared to 46 (95·8%) of the 48 detected by EHA. Listeria monocytogenes was detected in 36 of the 48 samples, 15 (41·6%) being detected on Oxford medium compared to 35 (97%) on EHA. Similar tendencies were observed for both isolation media irrespective of the enrichment broth used, EHA being the superior medium. In 16 enrichment broth/time combinations, Listeria innocua and L. monocytogenes were detected together on the same plate. This occurred 13 times on EHA but only once on Oxford medium. In 8 of the 16 cases, L. innocua or nothing was detected on the Oxford plate when EHA was positive for both organisms. On only one occasion was Oxford medium positive for both organisms when L. innocua or nothing was present on the EHA plate. These results show EHA to have advantages for both the isolation of Listeria spp. and the discrimination between haemolytic and non-haemolytic species of this genus.

Enhanced haemolysis agar (EHA)—an improved selective and differential medium for isolation of Listeria monocytogenes

Enhanced Haemolysis Agar (EHA) medium has been developed for the selective isolation of Listeria monocytogenes from food enrichments. It is based on the selective agents of the Oxford medium (Curtis et al. 1989a), with the addition of sheep blood (5%), 4-methylumbelliferyl-β- d-glucoside (50 mg l −1) and sphingomyelinase (10 units/l −1). The medium contains no esculin or ferric citrate and the lithium chloride concentration is reduced to 1%. L. monocytogenes colonies are distinguished by their light blue fluorescence and yellow centre (in UV 366nm), colony morphology and the presence of a distinct zone of haemolysis around the colony, indicating the production of the CAMP factor. Listeria innocua does not produce haemolysis on this medium and therefore cannot be confused with L. monocytogenes. Haemolysis positive Listeria seeligeri, although found relatively rarely in foods, can be distinguished by performing a xylose fermentation test after isolation on EHA. Strains of Listeria ivanovii from our collection failed to grow on this medium. Although haemolysis-negative and CAMP-negative L. monocytogenes may not be detected by this medium, their apparent lack of pathogenicity makes them of limited interest to food microbiologists.