Abstract
The public health risk posed by Listeria monocytogenes in ready-to-eat (RTE) foods depends on the effectiveness of its control at every stage of the production process and the strain involved. Analytical methods currently in use are limited to the identification/quantification of L. monocytogenes at the species level, without distinguishing virulent from hypovirulent strains. In these products, according to EU Regulation 2073/2005, L. monocytogenes is a mandatory criterion irrespective of strain virulence level. Indeed, this species encompasses a diversity of strains with various pathogenic potential, reflecting genetic heterogeneity of the species itself. Thus, the detection of specific L. monocytogenes virulence genes can be considered an important target in laboratory food analysis to assign different risk levels to foods contaminated by strains carrying different genes. In 2015–2016, a severe invasive listeriosis outbreak occurred in central Italy, leading to the intensification of routine surveillance and strain characterization for virulence genetic markers. A new multiplex real-time polymerase chain reaction targeting main virulence genes has been developed and validated against the enzyme-linked fluorescent assay (ELFA) culture-based method. Results of the improved surveillance program are now reported in this study.
Introduction
Listeria monocytogenes is the etiological agent of listeriosis, a foodborne disease with high case-fatality rate within well-defined risk groups. The ubiquity and the high capacity to survive and grow in a wide range of conditions and environments (pH 4.2–9.5, aw 0.92–0.99, from below 0°C to 44°C) make L. monocytogenes an important hazard in food production chain. L. monocytogenes persistence in food processing environments is still considered the major source of ready-to-eat (RTE) food contamination [EFSA Panel on Biological Hazards (BIOHAZ) et al., 2018; Guidi et al., 2021]. Many different RTE food types have been implicated in sporadic cases or outbreaks of human listeriosis in the EU (EFSA and ECDC, 2019). L. monocytogenes was detected in RTE pig meat products (2.5–1.7%), “milk and milk products” (2.8%) (EFSA and ECDC, 2018), “fish and fishery products” (3.18%), “cheeses and dairy products” (1.68%), “products of meat origin other than fermented sausages” (0.98%), and “RTE fruit and vegetables” (1.8%) (EFSA and ECDC, 2019). These and other international data (Buchanan et al., 2017) highlight the need to focus the attention on such food products, since all of them have the potential to contribute to the burden of disease when consumed by highly susceptible people [EFSA Panel on Biological Hazards (BIOHAZ) et al., 2018].
Microbiological monitoring for L. monocytogenes in food and food processing environments is mandatory in the EU (Regulation 2073/2005) (EU, 2005). The analytical methods currently in use (ISO 11290-1 and 11290-2) (ISO, 2017a, 2017b) are limited to the identification/quantification of L. monocytogenes at the species level, without distinguishing hypervirulent from hypovirulent strains; therefore, all food samples found to be positive should be considered a possible risk for the consumer health, regardless of the pathogenic potential of the contaminating strain. However, other criteria need to be considered to estimate consumers' risk, such as the microbial load, the shelf-life, and the capacity to grow in certain environments.
However, this species encompasses a diversity of strains with various virulence and pathogenicity levels (Maury et al., 2016). Listeriosis may appear either in noninvasive form, causing a self-limiting gastroenteritis in adult people, or in invasive form in subjects at risk, with meningoencephalitis and septicemia, sometimes with fatal outcome. Virulence diversity reflects the genetic heterogeneity of the species, which has a clonal population structure (Chenal-Francisque et al., 2011; Cantinelli et al., 2013). Recently, some less invasive strains with a strong epidemiological association with food, but not with clinical infection have been described (Maury et al., 2016). Distribution of InlA truncations and variations in virulence genes were the main features associated to the loss of virulence of hypovirulent clones (Maury et al., 2016). Therefore, the detection of specific L. monocytogenes virulence genes can be considered an important target in laboratory food analysis to assign different risk levels to foods contaminated by strains carrying different genes.
A statistically significant increasing trend of confirmed listeriosis cases in the EU member states, including Italy (p < 0.01), has been documented during the period 2009–2018. In Italy, 164 confirmed cases of listeriosis have been recently reported, with an incidence rate of 0.27/100.000 (ECDC, 2020). From January 2015 to March 2016, Central Italy experienced a severe invasive listeriosis outbreak with 24 confirmed clinical cases (mainly located in Marche region, although one clinical case was also identified in the neighboring Umbria region) caused by a strain belonging to serogroup 1/2a. The responsible food was hog head cheese produced by a small meat processing plant and contaminated by a pulsotype never isolated before in Europe (Duranti et al., 2018). In this epidemiological situation, routine monitoring of L. monocytogenes in RTE products was intensified, and pulsed-field gel electrophoresis (PFGE) and whole genome sequencing (WGS) were both used to identify outbreak strains (Duranti et al., 2018).
The aim of the study was to design a molecular test for fast strain characterization, to provide a sensitive and specific confirmation of the species, and to concurrently to give additional information about the presence of virulence genes in the contaminating strains. The real-time polymerase chain reaction (PCR) was validated and then applied on RTE foods, starting from enriched samples and colonies, to detect and characterize L. monocytogenes isolates to support the surveillance program.
Materials and Methods
Institutional review board approval was not required for this study.
RTE foods and L. monocytogenes detection
RTE products (n = 152 samples each in five sampling units for a total of 760), mainly dairy products and fermented meat (Table 1), were collected in years 2015–2016. The selected food items were as follows: RTE foods able to support the growth of L. monocytogenes placed on the market during their shelf-life or monitored before the food has left the immediate control of the food business operator and RTE foods unable to support the growth of L. monocytogenes and placed on the market during their shelf-life (EU, 2005, 2007).
Distribution of Listeria monocytogenes-Positive Ready-To-Eat Samples with Culture Method (ISO 11290-1 and 11290-2) and Multiplex Real-Time Polymerase Chain Reaction
Seasoning <30 d.
Typical Italian smoked and dry-cured sausage.
One seasoned loin sample, which tested positive with the culture method, resulted negative in m rt-PCR for all target, except the internal positive control (true negative); thus, species-specific identification of L. monocytogenes was not confirmed by the molecular test.
m rt-PCR, multiplex real-time polymerase chain reaction; RTE, ready to eat.
The samples were analyzed using the microbiological method based on the Enzyme-Linked Fluorescent Assay (ELFA, VIDAS L. monocytogenes II assay LMO2; bioMérieux, Durham, NC), equivalent to ISO 11290-1 in terms of precision, accuracy, specificity, and sensitivity (AFNOR BIO 12/11–03/04) and validated in compliance with ISO16140-2:2016 (ISO, 2016). An aliquot of enrichment culture (Fraser Broth) from each ELFA-positive sample was streaked on OXFORD agar and ALOA agar and incubated 48 h at 37°C. Typical colonies were tested for β-hemolysis on sheep blood agar, after a 24 h-incubation at 37°C (ISO, 2017a). Moreover, 200 μL of Fraser Broth from each sample unit was pooled together for a total of 1 mL, centrifuged 20 min at 6000 × g, and pellet stored at -20°C for the subsequent molecular testing by means of the multiplex real-time polymerase chain reaction (m rt-PCR) described in Multiplex Real-Time PCR section. L. monocytogenes colonies obtained after biochemical confirmation were tested with the same m rt-PCR as well.
Multiplex real-time PCR
A m rt-PCR assay was developed and validated for the simultaneous confirmation of L. monocytogenes species and detection of three virulence genetic markers from enriched samples collected as described in RTE Foods and L
Virulence target genes selection
As first attempt, a panel of five genes involved in L. monocytogenes virulence was selected for the development of a new on purpose “m rt-PCR”: inlJ, inlF, inlC, lapB, and lntA. The five target genes were selected based on bibliographic information about their association with strains able to cause disease: they were all present in some selected pathogenic L. monocytogenes strains (Camejo et al., 2011) and absent in hypovirulent L. monocytogenes strains M7 4a (Chen et al., 2011), L99 4a (Hain et al., 2012), and HCC23 4a (Paul et al., 2014); they were also absent in other Listeria species (Listeria innocua, Listeria welshimeri, and Listeria seeligeri) (Camejo et al., 2011).
PCR primers and probes (Table 2) were designed using the Primer Express software ver. 3.0.1 (Applied Biosystems–Thermo Fisher Scientific, Basingstoke, United Kingdom). Sequence specificity was assessed in silico (Basic Local Alignment Search Tool, BLAST,
Target Genes and Real-Time Polymerase Chain Reaction Oligonucleotides Used for Virulence Marker Detection
Amagliani et al. (2010).
Based on the results of the preliminary screening with the 38 clinical isolates described above, 2 virulence marker genes were chosen to be included in the m rt-PCR: inlJ and lapB.
Development of a m rt-PCR assay for the identification of L. monocytogenes and virulence genetic markers
A four-plex real-time PCR assay was developed and validated for the confirmation of L. monocytogenes species and detection of virulence genetic markers from enriched samples collected as described in RTE Foods and L
Assay sensitivity was assessed through amplification, both in singleplex and in multiplex conditions, of serial dilutions of L. monocytogenes DNA from clinical strain 490 (Lm 490) in the range 106–10 genomic units, g.u. (taken 3.21 fg as 1 L. monocytogenes g.u.) (Carloni et al., 2018). Each level was amplified in triplicate within three separate experiments.
Specificity had been previously assessed (Omiccioli et al., 2009) and additional tests were carried out with L. innocua ATCC 33090 DNA.
Assay validation with artificially contaminated RTE pork meat (cooked ham)
Since the m rt-PCR could be used as additional confirmation tool in routine analysis of RTE foods, its performances with artificially contaminated samples were investigated in a validation test compared to the reference method (ELFA L. monocytogenes assay confirmed by the ISO 11209-1 method). The clinical Lm 490 strain was exposed to chemical stress to simulate the same conditions of contaminated RTE fermented pork meat. Lm 490 was cultured in Tryptone Soya Broth Yeast Extract (TSBYE; Oxoid, Thermo Scientific) to mid-log phase. Bacteria were washed in phosphate-buffered saline and then resuspended in TSBYE 13.5% NaCl adjusted to pH 4.9 with lactic acid (Government of Canada, 2011). A control in TSBYE without chemical stress was also prepared. After incubation at room temperature for different time intervals, parallel cultures in selective media (ALOA; Oxoid) and nonselective Tryptone Soya Agar Yeast Extract (TSAYE; Oxoid) were prepared from bacterial dilutions. The stress procedure was considered appropriate when bacterial counts on ALOA were 20–50% of counts on TSAYE (50–80% injury level) as reported by Cloke et al. (2014).
The culture was diluted and used to contaminate 25 g test portions of cooked ham with 1 or 10 or 100 colony-forming unit (CFU); the negative control was inoculated with 100 CFU L. innocua ATCC 33090 (30 replicates for each contamination level). Contaminated food was analyzed with the ELFA method and the m rt-PCR assay. Samples for the molecular analysis were prepared from 1-mL aliquots of Fraser Broth (as reported in RTE Foods and L
Adhesivity and invasivity assays
With the objective to observe the behaviour of two distinct bacterial species, differing for the presence of the five target genes, in a cell culture model, the adhesive and invasive properties of the clinical strain Lm 490, possessing all the five virulence markers, were investigated through in vitro adhesivity and invasivity assays on Caco-2 cell cultures (Reddy and Austin, 2017).
L. innocua ATCC 33090, negative for the above-mentioned targets, was included as negative control. Data were statistically analyzed for significance by t-test.
Results
Multiplex real-time PCR
Virulence target genes selection and screening of clinical isolates of human origin
The in silico BLAST analysis indicated that oligonucleotide sequences for the m rt-PCR were sequence specific and conserved through different strains of the species.
All five virulence target genes (inlJ, inlF, inlC
Sensitivity and specificity of the m rt-PCR assay: validation with artificially contaminated RTE pork meat (cooked ham)
Assay sensitivity, assessed with dilutions of purified DNA both in singleplex and in multiplex conditions, was 10 g.u., and efficiency was in the range 96–109% for each target (Supplementary Table S2). No amplification product was obtained in specificity tests with L. innocua ATCC 33090.
Validation experiments with cooked ham samples, artificially contaminated with chemically stressed Lm490, showed complete agreement of positive and negative results between the ELFA method confirmed by ISO 11290 and m rt-PCR, even at the lowest contamination rate.
Adhesivity and invasivity assay
Adhesivity and invasivity analyses revealed a statistically significant difference between Lm490 (adhesivity 13% and invasivity 2.4%), possessing all the virulence genes investigated in this study, and L. innocua ATCC 33090 (adhesivity 0.9% and invasivity 0.05%; p < 0.001 for adhesivity and p < 0.01 for invasivity).
L. monocytogenes detection in RTE products
The microbiological analysis of RTE products performed in this study showed a prevalence of L. monocytogenes-positive samples of 31% (47/152). The most frequently contaminated food category was fermented meat products seasoned less than 30 d (fresh salami), followed by cheese, ciauscolo, and hog head cheese (typical Italian cured meats) (Table 1).
The molecular test, applied to the 47 DNA samples from selective enrichment cultures that tested positive with the culture method, showed almost complete agreement in the species-specific identification of L. monocytogenes (46/47 positive samples; 1 seasoned loin sample was negative for the 3 targets, including hlyA). Among these samples, 45/46 (98%) were also positive for lapB and 34/46 for inlJ (74%). If considering combinations of virulence genes, 33 strains (72%) carried all the 3 virulence markers; 12 (26%) tested positive for lapB and hlyA, but not for inlJ; and 1 (2%) tested positive for hlyA and inlJ, but not for lapB (Table 1). DNA amplification by m rt-PCR protocol of single colonies isolated from the same RTE foods confirmed the results obtained with total DNA from the respective enrichment cultures in all, but four strains (two from salami, one from ciauscolo, and one from ovine cheese), which tested negative for inlJ.
Discussion
In 2015–2016 a severe listeriosis outbreak associated with processed pork products has been reported in Central Italy (Duranti et al., 2018). In this scenario, a new molecular assay has been developed and in-house validated for the molecular detection of L. monocytogenes in RTE foods. The assay also included two additional amplification targets for the detection of genetic markers associated with virulence and reported in strains responsible for human disease. The application of this assay aimed to provide a sensitive and specific confirmation of the species and, concurrently, to give additional information about the L. monocytogenes strains, isolated from RTE food samples collected in the framework of the official control. However, the m rt-PCR did not claim to distinguish between strains with different levels of virulence.
Criteria for the selection of amplification targets relied at first on bibliographic information about their association with strains able to cause disease (Camejo et al., 2011; Chen et al., 2011; Hain et al., 2012; Paul et al., 2014); the subsequent experimental screening allowed to investigate their presence in strains responsible for human disease. It has been recently reported that inlF deletion in L. monocytogenes has no significant effect on infection or growth in cell culture models (Rupp et al., 2017), supporting the results of this study about the absence of this gene in many of the clinical strains examined.
Validation tests indicated that the developed m rt-PCR could potentially be applicable in routine testing of RTE foods. The presence of the IAC allowed to exclude false negative results, in accordance with the ISO 22174:2005 (ISO, 2005), which is of utmost importance in diagnostic applications.
The method proved to be sensitive and specific for L. monocytogenes species confirmation from secondary enrichment cultures, with almost complete agreement with the standard method. A very few differences regarding the presence of virulence genes were observed in single colony analyses, compared to enriched samples. However, as indicated by PFGE strain typing (Duranti et al., 2018), in some cases different co-contaminating strains have been isolated from the same food sample. For this reason, in a few cases, results deriving from the amplification of total DNA from enrichment cultures did not reflect the genetic content of each separate strain.
These data confirmed the genetic variability of L. monocytogenes strains, particularly in the distribution of virulence-related genes assessed in this study. Thus, the obtained results should suggest some differences in the pathogenic potential of isolates of food origin.
The presence of virulence genes in L. monocytogenes isolates obtained from food and food processing facilities has already been described in a number of studies, demonstrating their relevance for the characterization of this pathogen's virulence potential (Lomonaco et al., 2012; Jamali et al., 2013; Camargo et al., 2015; Haubert et al., 2015; Silva et al., 2016). A large number of genes potentially involved in virulence have been recognized by Maury et al. (2016). More recently, Sereno et al. (2019) investigated the pathogenic potential of isolates carrying virulence-related genes responsible for the main pathogenic pathways of L. monocytogenes (Liu et al., 2007; De las Heras et al., 2011; Pizarro-Cerda et al., 2012; Camargo et al., 2016). However, a recent investigation (Lee et al., 2019), in contrast with several studies about intraspecies genetic variability, focuses on transcriptomic analysis and highlights the role of complex genetic regulation in pathogenicity, describing the major contribution of transcription factors with key roles in virulence.
It has been reported that some steps in the infectious process, such as invasiveness, can be measured in an in vitro bioassay using the intestinal epithelial cell line Caco-2, and the ability to invade cells correlates with bacterial virulence (Lorentzen et al., 2011); nevertheless, referring to our study, a real correlation between genotype and phenotype cannot be ascertained by testing a such limited strain number, and was out of the scope of this investigation.
While WGS and the subsequent bioinformatics analysis are of paramount importance for the molecular-based monitoring of L. monocytogenes with epidemiologic purposes, due to completeness of genetic information and the very high discriminatory power, PCR-based methods are instead more suitable for routine analysis by local laboratories other than national reference centers. Real-time PCR still represents an affordable tool, more easily applicable to routine screening of food samples, along with culture-based tests (Rivera et al., 2018).
An infection results from the interplay between host and bacterial factors. Virulence of L. monocytogenes is conditioned by several factors: the different nature of the food, the condition of the host in which the bacterium penetrates, and finally the presence of specific virulence genes. Therefore, the latter can be considered important targets in laboratory diagnosis, and amplification-based methods are also more convenient than in vivo bioassays or in vitro cell assays. Nevertheless, given the complexity of the pathogenetic mechanism of L. monocytogenes, it is impossible to predict the capacity of a strain to cause human disease only referring to a PCR assay targeting a limited panel of virulence markers.
However, the optimized method could be used in routine food analysis, along with the culture-based approach, to confirm species identification and give additional information about the contaminating strains and the potential risk for consumers' health.
Conclusions
New trends in consumers' needs, going toward an increasing demand for RTE foods, suggest raising awareness on listeriosis and the risk associated with certain types of food and consumption patterns/habits (EFSA and ECDC, 2019). The clonal structure of L. monocytogenes, composed by hypervirulent and hypovirulent clones, should be taken into consideration for the monitoring of this major foodborne pathogen, which accounts for nearly half of the deaths associated with foodborne infections in Western countries (Maury et al., 2016).
Technologies such as WGS bring new insight, as they allow for additional information on the characterization of L. monocytogenes isolates, for example, about virulence, persistence, and clonal tracing, with an unprecedented detail compared with previous techniques such as PFGE and real-time PCR.
However, amplification assays, such as the one described in this study, still appear to be useful/could still be used for routine monitoring and typing in diagnostic laboratories at local level.
Employment
There is no recent (within the past 5 years), current, or anticipated employment by an organization that may gain or lose financially from publication of the article.
Footnotes
Authors' Contributions
Conceptualization and data analysis: G.A. and A.P. Investigation, data collection, and methodology: G.F.S., G.Bl., S.S., and S.F. Experimental work: C.G., S.D.L., and D.S. Resource provision (clinical strains): A.G. and M.V.G. Writing—original draft preparation: G.A. and A.P. Writing—review and editing: F.P., A.D., F.T., and G.Br.
Acknowledgment
The authors would like to thank Dr. Mauro De Santi for his assistance with statistical analysis.
Disclosure Statement
No competing financial interests exist.
Funding Information
This work was supported by the Italian Ministry of Health, Department of Veterinary Public Health, Nutrition and Food Safety (IZS UM 05/13 RC). Research did not receive any support by organizations that may gain or lose financially from publication of the article.
Supplementary Material
Supplementary Table S1
Supplementary Table S2
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
