Given the impact of unsafe food traded on a global scale on economics and health, rapid detection and identification of pathogens in food has become a major task in food science. Molecular biological methods provide many opportunities for improved detection of micro-organisms in food based on fast and reliable technologies. The establishment of molecular diagnostics is directly associated with conventional and real-time PCR (Kaltenboeck and Wang 2005). The implementation of these methods for routine diagnostic purposes relies on careful validation of newly developed assays. Therefore, ISO-16140 provides a basis for comparison of alternative methods with the respective standard method (Anonymous 2003).
Facultative anaerobic Gram-positive Listeria monocytogenes is a foodborne pathogen widely distributed in nature. Infection with L. monocytogenes may result in severe illness and death. Several types of food have been reported to be contaminated with this pathogen, and a recent outbreak demonstrated the possible threat for an exposed population (Fretz et al. 2010).
As is the case for other bacterial foodborne pathogens, traditional methods for the detection of L. monocytogenes in food are time-consuming (Anonymous 1996; Rossmanith et al. 2006). Alternative methods have been described, including pre-enrichment procedures combined with PCR (Oravcováet al. 2006; Badosa et al. 2009). In a previous study by Rossmanith et al. (2006), a combined enrichment/real-time PCR method was presented based on an assay targeting the prfA locus of L. monocytogenes. After 24 h incubation in half Fraser (HF) broth, according to ISO-11290-1, an aliquot of the sample was subjected to DNA isolation and real-time PCR detection. The method was thoroughly tested. The underlying PCR amplification reaction and several DNA isolation methods have been established and compared. The resulting method protocol was validated, according to ISO-16140. Nevertheless, this validation resulted in a relative specificity of 76·9%, and the reason for this value is suspected to be based on the quality of the available samples and the low number of positives within the calculation basis.
In this study, we present the application of the combined enrichment/real-time PCR method to 44 acid curd cheese samples (Quargel) that caused a recent outbreak of L. monocytogenes (Fretz et al. 2010).
The 44 Quargel cheese samples from 18 different lots, which were provided by the regulatory authority, were stored at 4°C prior to analysis and were processed in duplicate. Qualitative examination of the samples was performed according to the standard method ISO-11290-1 (Anonymous 1996). Additionally, quantitative examination was carried out according to ISO-11290-2 (Anonymous 1998).
The combined enrichment/real-time PCR method was performed according to Rossmanith et al. (2006). Twenty-five grams of Quargel cheese was added to 225 ml of HF medium according to ISO-11290-1. After 24 h incubation at 37°C, a 9 ml aliquot was removed and centrifuged at 50 g for 2 min. The pellet was discarded, and the supernatant was centrifuged at 3220 g for 10 min. The resulting pellet was subjected to bacterial target DNA extraction by using the NucleoSpin® tissue kit (Macherey-Nagel, Düren, Germany) and the support protocol for Gram-positive bacteria, according to the manufacturer’s instructions.
Real-time PCR detection of L. monocytogenes by targeting a 274 bp fragment of the prfA gene was performed according to the previously published format (Rossmanith et al. 2006). All PCR were performed in an Mx3000p thermocycler (Stratagene, La Jolla, CA, USA). The 25 μl volume contained 5 μl of DNA template. Real-time PCR results were expressed as bacterial cell equivalents (BCE). PCR was performed in duplicate. For confirmation of Listeria species-suspected colonies on Palcam agar, according to ISO-11290, the colonies were subjected to PCR after a short Chelex-based DNA isolation method (Border et al. 1990; Bubert et al. 1999; Rossmanith et al. 2006). PCR products of conventional PCR were separated in 1·5% agarose gels at 90 V for 25 min and stained with 0·5 μg ml−1 ethidium bromide (Sigma-Aldrich GmbH, Steinheim, Germany).
Forty-four Quargel cheese samples were investigated immediately after recall from the retail market. Examination of the samples, according to ISO-11290-1, resulted in 39 L. monocytogenes-positive and five L. monocytogenes-negative samples. These results were confirmed by PCR examination, according to Border et al. (1990) and Bubert et al. (1999) and additionally by quantitative examination, according to ISO-11290-2 in all cases (Anonymous 1998). The results after 24 h incubation in HF medium and the results following incubation for 48 h in full Fraser medium were also consistent. The values of contamination obtained by quantitative ISO-11290-2 varied from ≤100 CFU g−1 to 6·6 × 107 CFU g−1 with a mean contamination of 1·2 × 107 CFU g−1.
Examination of the samples with the combined enrichment/real-time PCR method resulted in 39 L. monocytogenes-positive and five L. monocytogenes-negative samples corresponding to ISO-11290. The quantitative values obtained by real-time PCR varied from 1·1 × 105 to 2·4 × 109 BCE ml−1 independent from the distribution of the respective contamination values within the samples as obtained by ISO-11290-2 (Kolmogorow–Smirnov test; P ≪ 0·05).
The combined enrichment/real-time PCR method can be compared with the standard method (ISO-11290), according to the ‘Protocol for the Validation of Alternative Microbiological method’ (ISO-16140). The relative accuracy, relative specificity and relative sensitivity were each 100%, when analysing and comparing the 44 Quargel cheese samples.
A validation of the combined enrichment/real-time PCR method by Rossmanith et al. (2006) resulted in a relative accuracy of 96·0%, a relative specificity of 100% and a relative sensitivity of 76·9%. As presented in Table 1, the then investigated and positive tested samples included samples mainly examined after long-term storage at −80°C (n = 42; 55%), in addition to examination after storage at +4°C (n = 9; 12%) and immediate examination (n = 25; 33%). The fraction of immediately investigated foodstuffs exclusively included negative tested samples. Together with the relative low percentage of positive samples (reference method: n = 13; 17%), the deviating results of the combined enrichment/real-time PCR method resulted in a comparatively poor relative sensitivity of 76·9%.
|Sample number||Flavour||Real-time PCR BCE ml−1*||ISO 11290-2 CFU g−1*||ISO 11290-1|
|Half Fraser†||Full Fraser‡|
|2||Natural||2·2 × 107||3·0 × 107||+||+|
|3||Natural||3·9 × 106||1·3 × 104||+||+|
|4||Natural||1·2 × 107||1·4 × 106||+||+|
|5||Natural||1·0 × 108||3·4 × 106||+||+|
|6||Natural||4·8 × 105||2·5 × 104||+||+|
|7||Natural||2·9 × 106||5·0 × 103||+||+|
|8||Natural||9·0 × 107||1·4 × 106||+||+|
|9||Natural||1·1 × 105||≤1·0 × 102||+||+|
|10||Natural||1·7 × 108||3·7 × 103||+||+|
|11||Natural||1·6 × 108||7·0 × 103||+||+|
|12||Natural||9·9 × 105||7·2 × 103||+||+|
|13||Natural||1·2 × 106||≤1·0 × 102||+||+|
|14||Natural||2·7 × 106||7·2 × 105||+||+|
|15||Natural||3·4 × 106||1·4 × 104||+||+|
|16||Natural||1·2 × 107||1·1 × 107||+||+|
|20||Natural||6·9 × 105||7·2 × 103||+||+|
|22||Natural||1·6 × 107||7·5 × 105||+||+|
|23||Natural||1·5 × 107||5·2 × 106||+||+|
|24||Natural||5·4 × 106||1·1 × 107||+||+|
|25||Natural||6·8 × 106||1·0 × 105||+||+|
|26||Natural||1·0 × 107||8·3 × 106||+||+|
|27||Natural||5·4 × 107||4·7 × 106||+||+|
|28||Natural||1·0 × 107||3·0 × 107||+||+|
|29||Natural||3·7 × 107||1·3 × 105||+||+|
|30||Natural||1·4 × 107||4·4 × 105||+||+|
|31||Natural||6·8 × 106||1·5 × 105||+||+|
|32||Natural||1·4 × 108||4·7 × 106||+||+|
|33||Natural||2·8 × 107||6·1 × 107||+||+|
|34||Natural||9·6 × 106||7·0 × 106||+||+|
|35||Natural||7·0 × 107||1·2 × 107||+||+|
|36||Natural||2·3 × 108||6·1 × 107||+||+|
|37||Natural||2·8 × 108||6·6 × 107||+||+|
|38||Natural||1·7 × 108||2·1 × 107||+||+|
|39||Natural||8·5 × 107||5·7 × 107||+||+|
|40||Natural||3·5 × 107||2·3 × 107||+||+|
|41||Natural||1·1 × 109||2·0 × 105||+||+|
|42||Natural||2·4 × 109||4·5 × 106||+||+|
|43||Natural||8·1 × 108||8·4 × 106||+||+|
|44||Natural||1·6 × 109||6·6 × 106||+||+|
|Rossmanith et al. (2006)||This study||Total|
|No. of positives (Ref.¶/Alt.**)||13/10||39/39||52/49|
|No. of negatives (Ref.¶/Alt.**)||63/66||5/5||68/71|
|No. of samples stored at −80°C||42||0||42|
|No. of samples stored at +4°C||9||44||53|
|No. of samples immediately examined||25||0||25|
|False negatives −80°C††||3||0||3|
|False negatives +4°C††||0||0||0|
|Relative accuracy (%)||96·0||100||97·5|
|Relative sensitivity (%)||76·9||100||94·2|
|Relative specificity (%)||100||100||100|
However, examination of the samples that were stored at +4°C resulted in 100% concordance between the combined enrichment/real-time PCR method and the reference method. These results are in good concordance with the data presented in this study and suggest that the storage conditions rather than the effective performance of the method was responsible for the deviating results of the combined enrichment/real-time PCR and the reference methods in the previous study. This is mainly because of poor reproduction of the target bacteria after long-term storage at −80°C during the first methodical step, the incubation in HF medium. Accordingly, all false-negative samples of the previous study displayed no detectable bacterial growth after 24 h of incubation in HF medium (results have not been shown) and negative results with ISO-11290-2.
Nevertheless, comparison of the different storage conditions would not have been significant because of the low number of samples stored at +4°C (n = 9) and the low total number (n = 13; 17%) of positive samples that were included in the previous work. However, together with the results of this study, it becomes obvious that the long-term storage of the frozen samples at −80°C has a strong influence on the performance of both ISO-11290-2 and the combined enrichment/real-time PCR method. This is based on the resulting low reproduction rate and/or the prolonged lag-phase of the bacterial targets (Golden et al. 1988; El-Kest and Marth 1992). The performance of both methods thus reflects the effect of storage at −80°C on the L. monocytogenes target cells.
In conclusion, the combined enrichment/real-time PCR method provides excellent performance when applied to fresh samples or examination after storage at +4°C. Because food samples for routine surveillance, or from actual outbreaks, should preferably be processed as rapidly as possible, the combined enrichment/real-time PCR method provides 100% relative sensitivity for these applications. When regarded as a single study, the summarized results of this study and that recently by Rossmanith et al. (2006), which includes the data of samples stored at −80°C, result in a relative accuracy of 97·5%, a relative sensitivity of 94·2% and a relative specificity of 100%.
Retrospective studies, which are based on frozen samples, would benefit from direct pathogen detection to reflect the original pathogen contamination independently of bacterial growth (Brehm-Stecher et al. 2009; Mester et al. 2010).