Foodborne Listeria monocytogenes: A Real Challenge in Quality Control

Listeria monocytogenes is a foodborne pathogen, and the detection and differentiation of this bacterium from the nonpathogenic Listeria species are of great importance to the food industry. Differentiation of Listeria species is very difficult, even with the sophisticated MALDI-TOF MS technique because of the close genetic relationship of the species and the usual gene transfer. The present paper emphasizes the difficulties of the differentiation through the standardized detection and confirmation according to ISO 11290-1:1996 and basic available L. monocytogenes detection methods and tests (such as API Listeria test, MALDI-TOF MS analysis, and hly gene PCR). With the increase of reports on the pathogenesis of atypical Listeria strains in humans, the significance of species level determination has become questionable, especially in food quality control, and the detection of pathogenic characteristics seems to be more relevant.


Introduction
Out of the six closely related Listeria species (L. monocytogenes, L. ivanovii, L. seeligeri, L. innocua, L. grayi, and L. welshimeri) and the newly isolated nonpathogenic L. marthii [1], L. rocourtiae [2], L. weihenstephanensis, and L. fleischmannii [3], L. monocytogenes is generally regarded as the only one that is capable of producing human listeriosis [4] with a high mortality rate (greater than 20%). Therefore, laboratory diagnosis has focused primarily on the detection of L. monocytogenes and its differentiation from other Listeria species [5]. In food samples, the contamination level is usually low (10 CFU/g), but higher than 100 CFU/g can be detected in some cases, for example, in minced meat or smoked fish [6]. Previously, some practical approaches to differentiation tests included culturing with and without preenrichment or with direct analysis from the food sample with biochemical confirmation tests and/or nucleic acid based techniques (reviewed by Churchill et al. [7]), and these methods presented some success for the food industry. Others approached Listeria differentiation from a technical point of view by applying money-consuming methods for precise strain differentiation which is especially good for medical and/or research purposes (reviewed by Gasanov et al. [8]).
The standardized L. monocytogenes detection method in Europe requires a laborious testing scheme which is timeconsuming, taking five to six days [9,10]. Because of the high genome homology among Listeria species [11][12][13] the strains display similar morphological, biochemical, serological, and molecular characteristics [4,14] that causes difficulty of the identification.
In L. monocytogenes, the central virulence gene cluster (pathogenicity island, e.g., LIPI-1) codes for the main activities that have importance in the intracellular life of the microbe inside the host. According to some researchers, nonpathogenic L. innocua species have lost their pathogenicity island [4,11]. However, Johnson et al. [15] detected hly and mpl genes of the LIPI-1 pathogenicity island in atypical L. innocua and there are also reports on human infections caused by L. innocua [16], L. seeligeri [17], and L. ivanovii [18]. Here, a question is arising: what is more important in food hygiene? The species level determination of L. monocytogenes or the enumeration of possibly virulent Listeria species?
Here, ALOA medium, API Listeria biochemical test strip, MALDI-TOF MS analysis, and hly gene detection by PCR were compared to the standardized detection and confirmation according to ISO 11290-1:1996 on foodborne and reference strains.

Isolation of Foodborne Listeria.
All applied enrichment media, selective media, and supplements were identical in composition to the described media of the ISO 11290-1:1996 standard [9] and supplied in dehydrated form. All media were tested with the L. monocytogenes NCAIM B 01934 reference strain.
Thirty-one commercially available frankfurter products originated from different producers were sampled and analyzed. From each of them 25 g was measured into 225 mL preenrichment medium (half Fraser broth, Merck, Hungary) in sterile bags and was homogenized in a Stomacher homogenizer (IUL Instruments, Spain) for 1 minute [9].
The homogenized samples were incubated in Schott flasks with loosening caps for 24 hours at 30 ∘ C. From each culture 0.1 mL was inoculated into 10 mL enrichment Fraser broth and incubated at 37 ∘ C for 48 hours. Black color development, which is characteristic of Listeria species, was detected in some of the culture broths. All cultures were streaked onto Oxford and PALCAM selective agar plates (Merck, Hungary) and incubated at 37 ∘ C for 24-48 hours. All colonies showed homogeneous morphology on these media. Subcultures of colonies were tested on sheep blood agar plates (OXOID, UK) (24 ± 2 hours) and with the Christie-Atkins-Munch-Petersen (CAMP) test for enhanced beta-haemolysis; on ALOA plates (VWR BDH Prolabo, USA) for -glucosidase and phospholipase activities, catalase test and Gram staining were also performed. Gram-positive and catalase positive bacterial strains were stored in a Cryobank system (Mast Diagnostics, UK) at −75 ∘ C and were applied for further testing.

Biochemical Confirmation
Tests. The biochemical confirmation reactions were performed with the API Listeria test (bioMérieux, France) according to the instructions of the manufacturer using bacterial cultures cultivated freshly on sheep blood agar plates (OXOID, UK). Reading of the API test strips was evaluated with the mini-API instrument (bioMérieux, France).

MALDI-TOF MS. Bacterial isolates and
Listeria reference strains were grown overnight at 37 ∘ C on sheep blood agar (OXOID, UK). Cells of a whole colony were placed in a spot on the steel target plate. Each sample was overlaid with 1 L of matrix solution (saturated solution of -cyano-4-hydroxycinnamic acid in 50% acetonitrile, 2.5% trifluoroacetic acid) and air-dried at room temperature. MALDI-TOF MS was performed on a Microflex LT instrument (Bruker Daltonik GmbH, Germany). For automated data analysis, raw spectra were processed using the MALDI Biotyper 3.0 software (Bruker Daltonik GmbH, Germany) with default settings. Generated peak lists were used for matches against reference libraries directly using the integrated pattern matching algorithm of the MALDI Biotyper 3.0 software [20]. In the case of the hlyQF/hlyQR primers, PCR reaction conditions were determined by Rodríguez-Lázaro et al. [21]. The annealing temperature in the case of our hlyF/hlyR primers was 55 ∘ C. The amplification conditions were 3 min denaturation at 96 ∘ C, 35 cycles each consisting of 30 s at 94 ∘ C, 45 s at the corresponding annealing temperature, and 1.3 min at 72 ∘ C followed by a final extension for 10 min at 72 ∘ C. We applied L. monocytogenes NCAIM B 01934 genomic DNA template as a positive control in each PCR run.
Validation of the PCR protocol was carried out with artificially contaminated food samples. 25 g portions of commercial pasteurized frankfurter were suspended in 225 mL TSB (Tryptic Soy Broth; OXOID, UK) and homogenized in a Stomacher homogenizer (IUL Instruments, Spain) for 1 minute. The suspensions were inoculated with decimal dilutions of exponential phase culture of the L. monocytogenes NCAIM B 01934 reference strain suspension from 1 to 10 5 cells/mL concentrations, mixed well, and samples were taken immediately after inoculation. Following DNA isolation, hlyF/hlyR primers were applied in the PCR reactions. PCR products were sequenced in order to test their identity (data not shown). Reference sequences were gained from the National Center for Biotechnology Information Database (http://www.ncbi.nlm.nih.gov/) and aligned with the sequenced PCR products using the BioEdit Biological Sequence Alignment Editor Program [22].

Results and Discussion
From 31 frankfurter samples, 11 Listeria strains were isolated with the standardized method of L. monocytogenes detection. The isolates produced typical colonies on PALCAM agar (except Isolate 2 on Table 1) and/or Oxford selective agar media were Gram-positive and catalase positive.
L. monocytogenes should show haemolysis on sheep blood agar plates or with Staphylococcus aureus with the CAMP test and utilize rhamnose but not xylose [9,10]. However, our foodborne strains showed the absence of haemolysis after 24 hours of incubation on sheep blood agar contrary to the reference L. monocytogenes NCAIM B 01934 and presented it weakly only after 48 hours of incubation. The ISO standard refers to the possibility of the isolation of L. monocytogenes strains with weak haemolysis or negative CAMP test [10,23]. Foodborne L. monocytogenes strains with either gene alteration or deletion of the hly gene or its regulatory protein PrfA are frequently detected, which cause weak haemolytic or nonhaemolytic phenotype in the atypical, aberrant, or nonhaemolytic strains [24,25]. This phenomenon is highly probable for bacteria isolated from processed food or natural sources. The absence of haemolysis at 24 hours of incubation makes species identification hardly possible because only rhamnose utilization can be applied as a confirmation test by the standard method. At least four Listeria species (L. monocytogenes, L. innocua, L. welshimeri, and L. grayi ssp. murray) are known to utilize rhamnose and not xylose [10]. Therefore, a Listeria strain can be determined only at the genus level.
ALOA medium, which was inserted into the standard method [10], differentiates phospholipase positive Listeria colonies from another phosphatidylinositol negative Listeria [26]. On ALOA plates, all foodborne Listeria isolates produced blue-green colonies with an opaque halo. However, other strains than L. monocytogenes (e.g., L. seeligeri and L. ivanovii) may also possess plcA gene coded phospholipase (PI-PLC) activity. Moreover, other bacteria species (e.g., Bacillus circulans [27]) also possess phospholipase andglucosidase activities that resulted in halo and blue-green color of the colonies, respectively. Therefore, application of ALOA medium independently from the application of preliminary selective enrichment or other selective media (which also needs, at least, two days of incubation) could lead to false positive results.
Results of the biochemical analysis with the API Listeria test were compared with the reference profiles given by the manufacturer (bioMérieux, France). All of our foodborne Listeria isolates showed negative arylamidase (DIM), Dxylose (DXYL), and D-tagatose (TAG) tests (API Listeria numeric profile: 6510) ( Table 1) and were suspected to be L. monocytogenes. It is generally regarded that the API test fails in some (10-15%) percent [23,24,28,29]. The main reason of the misidentifications is hidden in the doubtfulness of the color determination of the weak arylamidase (DIM) test. The negative DIM test gives the basis for the differentiation of L. monocytogenes from the closely related L. innocua [30]. A weak positive DIM reaction can be evaluated as a negative result and can increase the number of the false positive determinations. Furthermore, false negative identification of atypical strains is also frequent [29,31,32].
A quite new technique for species identification is the Matrix Assisted Laser Desorption Time of Flight (MALDI-TOF) analysis. The isolates are differentiated on the basis of spectral peaks that typically correspond to highly abundant conserved proteins (e.g., ribosomal, cold shock, and nucleic acid binding proteins) that are so-called "housekeeping" proteins [33]. The results of the pattern matching process of the fingerprints obtained in MALDI-TOF MS analysis are expressed in score values (0-3.000). More than half of the samples ( = 18) were determined by 2.000-2.299 score values that mean secure identification only at genus level while only seven strains were given a highly probable species identification with higher than 2.300 scores. The identification revealed L. monocytogenes-L. innocua misidentification of the reference strains (L. monocytogenes ATCC 35152 and L. innocua ATCC 33090) with scores of high probability species determination of L. innocua in case of the L. monocytogenes ATCC 35152 atypical strain and for L. seeligeri ATCC 35967 gave L. monocytogenes or L. innocua (Table 1).
Our PCR primers were designed for conserved regions of the hly gene of LIPI-1 in L. monocytogenes and amplified a 438 bp fragment, while primer pair hlyQF/hlyQR was constructed for RT-PCR and amplified a 64 bp fragment from the L. monocytogenes hly gene [21]. Considering that only one band appeared with agarose gel electrophoresis, the tested hlyQF/hlyQR primer pair [21] was much reliable than, for example, that of Border et al. [34] with more amplificants. Interestingly, the hly gene detection with hlyQF/hlyQR primer pair revealed atypical strain determination for L. seeligeri ATCC 35967 (Table 1). It is well known that nonhaemolytic L. seeligeri possess a distinct LIPI-1 structure [4,14]. Moreover, this primer pair also gave a reaction with atypical L. monocytogenes NCAIM B 01966. The specificity of the primer pair constructed in this work was higher than hlyQF-hlyQR or the primers of Aznar and Alarcón [29], because they received positive PCR results with their hly gene probes for other Listeria species (L. ivanovii and L. seeligeri). Furthermore, better than that of Cooray et al. [35], which failed to detect the hly gene from L. monocytogenes ATCC 15313 strain. With our primer pair, six out of eleven of the foodborne bacterial isolates gave positive reaction (Table 1), and we did not get a reaction with atypical or nonpathogenic Listeria strains.