is a food-borne pathogen that causes disease mainly in vulnerable populations such as very young, old, pregnant or immunocompromised individuals. The high mortality rate of 15–30 deaths per 100 cases of listeriosis [1
] is mostly attributable to severe central nervous system infections, septicemia, abortions and neonatal listeriosis. In the United States alone, listeriosis is estimated to cause an annual loss of 8800 disease adjusted live years (DALY), of which the majority are premature deaths [5
Human cases of listeriosis often trace back to food products that were contaminated during production, with subsequent growth of L. monocytogenes
to high numbers. Ready-to-eat products such as salads or deli meat are of special concern due to the lack of a heating step prior to consumption. The frequent occurrence of L. monocytogenes
in the environment results in a high probability of introducing the organism into facilities, either on raw materials, through equipment or via employees. Once introduced, several factors increase the probability of a strain being able to establish long-lasting colonization of niches. L. monocytogenes
has a high tolerance against acid and salt stress, and is able to grow at refrigerating temperatures. In addition, the ability to form biofilms may enhance survival, especially in niches that are difficult to reach during cleaning procedures. Further, tolerance against commonly used disinfectants such as the quaternary ammonium chloride compound benzalkonium chloride were observed in L. monocytogenes
isolates from food processing environments [6
]. Accordingly, L. monocytogenes
presents a challenge to the food industry and has been shown to persist in food producing facilities for long periods of time, in some cases for more than a decade. For example, a meat factory in Texas harbored the same strain for at least 12 years, eventually causing an outbreak in 2000 [6
]. A smoked fish processing plant harbored the same strain of L. monocytogenes
for 11 years [8
]. A study in Ireland found that seven out of 48 food processors housed a persistent strain—defined as isolated at least six months apart [9
Here, we analyzed strains from a Swiss deli meat plant where L. monocytogenes strains were repeatedly isolated between 2015 and 2018. The aim was (i) to determine whether a clonal population of L. monocytogenes persisted in the facility or if L. monocytogenes was repeatedly reintroduced, and (ii), to characterize the resistance of the collected strains against benzalkonium chloride (BC) and peracetic acid (PAA) and their ability to form biofilms.
2.1. Subtyping of 124 Listeria Monocytogenes Isolated from a Swiss Meat Plant
A total of 124 strains of L. monocytogenes
were analyzed in this study. All strains were collected in the framework of a hygiene monitoring program in a meat processing facility in Switzerland between 2015 and 2018 (Table 1
). The collection contains four strains from 2015, three strains from 2016, 32 strains from 2017 and 85 strains from 2018. The samples originated from products (n = 8) and from the food production environment (n = 116).
In a first screening of the diversity of the 124 L. monocytogenes strains, their serogroup and multi locus sequence typing (MLST) profile was determined. The majority of the strains (n = 73, 58.9%) belonged to serogroup II (containing serotype 1/2c, 3c), 48 strains (38.7%) belonged to serogroup I (1/2a, 3a) and three strains (2.4%) belonged to serogroup IV (4b, 4d, 4e). Seven gene MLST revealed that all of the 73 serogroup II strains belonged to CC9 (n = 73, 58.9%). The strains belonging to serogroup I were more heterogenous with 31 strains assigned to CC204, seven strains assigned to CC121, four strains to CC20 and to CC29, and one to CC8 and to CC89. The three strains in serogroup IV all belonged to CC6. Only strains from CC9, CC204 and CC121 were consistently isolated over all four years. CC6 strains seem to have been introduced into the facility in July 2017 and persisted until the end of the sampling period in June 2018.
From these results we concluded that it was likely that clonal populations of CC9, CC204, CC121, and CC6 persisted in this facility, while strains from other complexes were sporadically introduced and deemed “presumably non-persistent”. We consciously use the term “presumably” because there is no reasonable way to determine if those strains would have persisted in different environments or under different growth conditions.
In a next step, the genomes of 20 strains were sequenced and a core genome MLST (cgMLST) comparison was performed (Figure 1
). The cgMLST revealed that most strains from the same CC differed in less than 10 alleles, a cutoff for strain clonality [10
2.2. Biofilm Formation
The capacity to form biofilms is an important mechanism for persistence [11
], and biofilm formation might have contributed to the maintained presence of clonal clusters of L. monocytogenes
strains in the facility for years. We therefore investigated the biofilm forming ability in a subset of strains of the collective (CC9, CC204, CC121 and CC6) at 22 and 8 °C.
Approximately double the biofilm mass was recovered after 96 h of growth at 22 °C compared to 168 h growth at 8 °C (p < 0.01) (Figure 2
). However, no individual strain had significantly different biofilm formation compared to the other tested strains (Supplementary File 1
At 22 °C, strains from CC9, CC204, CC121 and CC6 showed a biofilm forming capacity in the range of a high biofilm forming (HBF) control strain, and all strains had significantly higher biofilm formation than that of a low biofilm forming (LBF) control strain (p < 0.01). At 8 °C, all strains, including the HBF control strain, were impaired in their biofilm forming capacity and hence, no significant differences between the strains from different complexes, nor between the strains isolated from the meat processing facility vs. the control strains, were found (Figure 2
2.3. Tolerance to Benzalkonium Chloride
Benzalkonium chloride is a commonly used disinfectant in the food industry and resistance against BC might contribute to the persistence of a strain. One hundred and six out of 124 tested strains were resistant to 10 µg/mL BC (Figure 3
). None of the strains, however, were resistant to the cutoff for resistance—20 µg/mL BC. A cumulative link model [12
] revealed that CC29, CC89, CC8 and CC6 had a significantly lower tolerance to BC (combined average of 4.5 µg/mL BC) compared to the other CC in the dataset (combined average of 9.2 µg/mL BC, p < 0.05).
A genome-wide search in the 20 sequenced strains revealed the presence of either brcABC
] or qacH
] BC resistance genes in 16 (80%) of the strains (Table 1
genes were present in six strains of CC204 and CC9, respectively. qacH
was present in four CC121 strains. No strain carried the emrE
] BC resistance gene. All strains that carried resistance genes had the highest measured BC resistance at 10 µg/mL, except strain ILS AS1-0004 which was resistant to 7.5 µg/mL.
2.4. Minimal Inhibitory Concentration and Minimal Bactericidal Concentration of Peracetic Acid against L. monocytogenes
PAA is frequently used as a disinfectant in the food production environment with a recommended final concentration between 0.5% and 1%. The meat processing facility from which the strains originated used 0.8% PAA as a sanitizer; we therefore hypothesized that strains from this collective might have adapted to PAA over time.
The MIC was between 0.1% and 0.2% for all strains and the MBC was between 1.2% and 1.6% (Figure 4
, Supplementary File 1
). Since the strains were isolated up to four years apart, these data strongly suggest no adaptation to PAA over time.
All further analyses were performed in a subset of strains that were chosen to represent all four years of collection and were divided over the most frequent clonal complexes CC9 (n = 6), CC204 (n = 4), CC121 (n = 3), CC6 (n = 2) and CC20 (n = 1). To test whether the strains had an unusually high resistance to PAA, a comparison was made to L. monocytogenes
strains from unrelated sources that belonged to the same clonal complexes [CC9 (n = 4), CC204 (n = 3), CC121 (n = 1) and CC6 (n = 1)] (Table 1
). The minimal bactericidal concentration (MBC) of the strains was between 1.2% and 1.6% with no difference between the clonal complexes or between the strains isolated from the meat processing facility versus the control strains (Figure 4
a). Therefore, the strains from the meat processing facility showed no increased tolerance to PAA.
To test whether the availability of protein compounds in the test medium affected the outcome, the MBC was compared with PAA in tap water and PAA in brain heart infusion (BHI). The MBC of PAA in tap water was 0.1–0.4% and, therefore, lower compared to the PAA BHI (p < 0.05) (Figure 4
b), indicating that protein affects the effectiveness of PAA and that there is a protein error for PAA.
We further assessed whether the incubation temperature influenced the outcome, which would indicate that PAA has a cold error. While MBC values were generally a better indicator for disinfection, we used MIC measurements to address this question because they allowed us to test the effects of cold exposure over a longer time than MBC measurements. The MIC for PAA was 0.2% for all strains, regardless of incubation at 4 °C or 37 °C, suggesting no evidence for a cold error for PAA against L. monocytogenes.
In this study, we showed that several clonal populations of L. monocytogenes
persisted in a meat processing facility in Switzerland for at least four years. The persistent strains belonged to CC9, CC204 and CC121. The dominant CC9 and CC121 have previously been shown to be overrepresented in food processing facilities in France [16
], Spain [17
], Switzerland [18
], and all over Europe [19
]. CC204 on the other hand has only been sporadically isolated from food processing facilities [18
] and human patients [20
Strains from the facility formed biofilm in the range of a high biofilm former. However, no difference was found between the individual strains isolated from the facility, nor between clonal complexes. Subtle differences between clonal complexes might be revealed by replacing the 96-well format biofilm screening protocol used in this study [21
] with more labor-intensive procedures, such as culturing biofilms on stainless steel coupons [22
], in a follow-up study. Contradictory results have been found by other authors: Some studies showed that persistent strains of L. monocytogenes
formed more biofilm than presumed non-persistent strains [11
], and persistent strains were more efficient in attaching to surfaces during a short contact time [23
]. Others did not observe a difference between persistent vs. presumed non-persistent strains [24
], and the ability of L. monocytogenes
to form true biofilms at all has been critically questioned by some authors [25
]. Given the high biofilm forming capacity we found in the strains of L. monocytogenes
isolated from this meat processing facility, biofilm formation may have contributed to persistence.
The overrepresentation of CC121 in food processing environments has often been explained with their higher resistance to BC due to the presence of the qacH
genes that encode efflux pumps [26
]. Confirming this, all CC121 strains that were sequenced in this study were positive for qacH
. Moreover, 80% of all sequenced strains carried either qacH
. Given the clonal structure of the CC121, CC9 and CC204 strains in this collection and their uniform resistance to 10 µg/mL BC, it is reasonable to assume that most strains in this collection carried either qacH
. However, tolerance to 10 µg/mL BC is below the typical in-use concentrations of BC (500–1000 µg/mL) [27
], indicating that BC should be active against the strains in this study.
In contrast, PAA was routinely used for disinfection in the facility, at a final concentration of 0.8%. Our data showed that the MBC (1.2–1.6%) and the MIC (0.1–0.2%) values were close to or above this concentration, to ensure efficient disinfection. Experiments without nutrients in the medium showed that even in the absence of proteins the MBC, between 0.2% and 0.4%, was still very close to the recommended concentration for use. Under real-life conditions, when disinfecting larger areas, it is likely that the concentration of the disinfectant becomes diluted through residual washing water and that residual organic matter is present. It is also likely that contact times differ due to run-off or delayed reach of hard to clean niches, which in this case would lead to ineffective concentrations of PAA.
Different strains of L. monocytogenes
also exhibited high MBCs for PAA in the range of up to 0.5% [28
]. In contrast, others concluded that PAA was able to significantly reduce L. monocytogenes
from multispecies biofilms at 0.15% [32
] and 0.3% [30
], respectively. The MIC of PAA for other organisms seemed to be much lower, in the range of 0.01–0.03% for Gram-positive and Gram-negative flora isolated from citrus fruit [33
], 0.0003% for Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa
], and 0.0001–0.001% for P. aeruginosa
, S. aureus
, E. coli
, and S. epidermidis
]. Given the high tolerance of L. monocytogenes
against PAA found in our data and by others, the ability of PAA to reduce L. monocytogenes
from biofilms may primarily have been due to the elimination of the supporting flora in multispecies biofilms.
Our data confirm the absence of adaptation of L. monocytogenes
to PAA found in other work that found no adaptation over several hundred generations when L. monocytogenes
were exposed to an industrial disinfectant containing PAA and hydrogen peroxide [36
Taken together, these experiments show that the concentration of 0.8% PAA used by the facility is insufficient to ensure safe production standards with regard to L. monocytogenes. However, PAA is a valuable option for disinfection due to its effectiveness against most bacteria, fungi and viruses, the absence of a cold error, its status as GRAS, and its colorless and odorless properties, but only when L. monocytogenes is not a major concern. Since PAA at concentrations above 1% is corrosive to equipment and irritating to the eyes, via fumes, an alternative disinfectant regimen should be considered in food production facilities that struggle to control L. monocytogenes in their environment.
4. Material and Methods
4.1. Bacterial Strains and Preparation of Bacterial Cultures
The hygiene-monitoring program of the meat-producing facility entailed product samples and swabs of the production environment such as floor drains, trolleys, elevators, cold room floors, scales, production and packaging lines, toilet drains and toilet floors or doors between production sections. All samples were tested for L. monocytogenes using the “Assurance Genetic Detection System” (GDS, Biocontrol, Nieuwerkerk aan den IJssel, The Netherlands) according to the protocol. In short, swabs or samples were incubated in Half Frazer Broth (HFB, BioRad, Marne-la-Coquette, France) at 30 °C for 48 h. L. monocytogenes were further enriched on magnetic beads and then identified via a kit-specific PCR in an “Assurance GDS Rotor-Gene” cycler. To obtain single colonies, the enriched HFB was streaked on Oxoid chromogenic Listeria agar (OCLA) plates (Oxoid, Pratteln, Switzerland) and incubated at 37 °C for 24 h. All strains were kept in 15% glycerol stocks at –80 °C.
To obtain overnight cultures for experiments, the strains were streaked on BHI agar (Oxoid, Pratteln, Switzerland) and incubated overnight at 37 °C. A single colony was inoculated into 5 mL BHI and incubated for 18 h overnight at 37 °C with shaking at 200 RPM in a shaking incubator (Edmund Buehler SM30/TH30 combination, Huber AG, Reinach, Switzerland). To obtain exponential phase cultures with an OD590 of 0.4, 50 µL of the overnight culture was subcultivated into 5 mL of fresh BHI and incubated for 3 h at 37 °C with shaking at 200 rpm.
4.2. Serogrouping by qPCR
DNA was isolated from 1 mL overnight cultures using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany) and eluted in 10 mM Tris pH 7.3. The DNA concentration was measured with a Nanodrop 1000 (Thermo Fisher Scientific, Basel, Switzerland) and PCR templates were standardized to 105
copies per 1 µL by dilution in fresh 10 mM Tris. Multiplex qPCRs were performed according to Vitullo et al. [37
], with the following modifications: Instead of a triplex-PCR we performed the qPCR as duplex with PUC19 [38
] as an internal control on the third channel. Primers and probes according to Vitullo et al. (37) (Table 2
) were obtained from Sigma-Aldrich (Buchs, Switzerland), and used at a final concentration of 0.4 μM for the primer and 0.2 μM for the probe.
Cycling conditions for the two-step PCR on a LightCycler 2.0 (Roche Science, Rotkreuz, Switzerland) were as follows: 5 s at 95 °C, followed by 40 cycles of 45 s at 95 °C and 45 s at 60 °C.
4.3. Multi Locus Sequence Typing (MLST)
MLST was performed on all 124 strains according to Ragon et al. [39
]. All primers (Table 2
) were ordered from Microsynth (Balgach, Switzerland). The fragment sizes were confirmed by gel-electrophoresis, the products were sequenced by Microsynth (Balgach Switzerland), assembled in Geneious (Version 11.1.4, Biomatters, Newark, NJ, USA) and analyzed using the website of the Institute Pasteur (http://bigsdb.pasteur.fr/listeria/listeria.html
4.4. Whole Genome Sequencing
Based on the MLST results, a selection of 20 strains (Table 1
) were Illumina sequenced. DNA was extracted, as for the MLST. Sequencing libraries were prepared using the Illumina Nextera DNA Flex chemistry and sequenced on an Illumina MiniSeq (Illlumina, San Diego, CA, USA) with a minimal coverage of 30 ×. After quality control with FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/
), the reads were assembled with Spades 3.12.0 [40
]. Core-genome multilocus sequence typing (cgMLST) was performed in the software package SeqSphere 4.1.9 (Ridom, Münster, Germany). Assembled genome sequences were imported and blasted against 1701 genes of the seed genome EGD-e, using the standard settings [10
]. A minimal spanning tree was produced in SeqSphere with the options “ignore missing values pairwise” and “discard genomes with >3% missing genes”. Strains with less than 10 different alleles were considered to belong to the same complex.
4.5. Biofilm Formation
The 15 strains that represented the four clonal complexes that seemed to persist in the factory (CC9 (n = 9), CC204 (n = 2), CC121 (n = 2)), (CC6 (n = 2)) were chosen for biofilm assays, according to the protocol published by Harvey et al. with minor changes [21
]. A single colony was inoculated into 5 mL of tryptone soy broth (TSB, from Fluka, obtained from Sigma-Aldrich, Buchs, Switzerland), incubated for 20 h at 30 °C with shaking at 200 rpm, subcultured 1:250 into fresh TSB, and incubated for an additional 20 h at 30 °C with shaking at 200 rpm. The resulting cultures were adjusted to an OD600
of 1.0, diluted 1:80 in TSB and added to 96-well plates, which were incubated for 96 h at 22 °C, or for 168 h at 8 °C, respectively. Biofilms were then washed three times with distilled water, stained with crystal violet, and washed five times with distilled water. The remaining crystal violet was dissolved in ethanol and the OD600
was measured in a Synergy plate reader (BioTek, Lucern, Switzerland). Control strains that were high and low biofilm formers (Institute for Food Safety and Hygiene, Zurich; unpublished results) (Table 1
) were included in each experiment.
4.6. Tolerance to Benzalkonium Chloride
MICs for BC were determined for all 124 strains included in this study, as per the protocol of Meier et al. with minor changes [26
]. Five µL of an exponential phase culture were spotted on BHI plates containing BC at 0, 2.5, 5.0, 7.5, 10.0, 15.0, 20.0, 25.0 and 30.0 µg/mL (Sigma Aldrich, Buchs, Switzerland). The plates were incubated for 48 h at 37 °C. Strains were considered resistant to the concentration of BC on which confluent growth was observed. The cutoff for resistance was set, according to Langsrud et al. [41
], as double the value of the lowest concentration that inhibited growth in >50% of the tested strains.
All sequenced genomes were searched for the qacH
BC resistance genes using BLASTP 2.7.1+ [42
] with the standard settings and an e-value cutoff of 10−20
4.7. Minimal Inhibitory Concentration and Minimal Bactericidal Concentration of Peracetic Acid against L. monocytogenes
These assays were performed in PAA diluted in BHI to mimic a worst case scenario that assumed incomplete cleaning of organic matter from surfaces before disinfection, and to assess the overall effect of PAA on bacteria within an otherwise favorable environment. To determine the influence of the protein in BHI on the outcome, some of the assays were additionally performed in PAA diluted in tap water (City of Zürich, supplementary file 2
). First, the MIC and MBC for PAA were determined in a screening of all 124 strains. Given the clonal nature of much of the strain collection, we then compared the resistance to PAA in detail in a subset of the strains: CC9 (n = 6), CC204 (n = 4), CC121 (n = 3), CC6 (n = 2) and CC20 (n = 1). A control dataset contained strains from unrelated sources: CC9 (n = 4), CC204 (n = 3), CC121 (n = 1) and CC6 (n = 1) (Table 1
To obtain the MBC values, serial dilutions were prepared to obtain final PAA concentrations of 2.8%, 2.4%, 2.0%, 1.6%, 1.2%, 0.8%, 0.4%, 0.2%, 0.1% and 0.05%. Next, 190 µL of these dilutions were added to 96-well plates and cooled to 4 °C. Then, exponential phase cultures were diluted in 0.9% NaCl and 10 µL were added to each well to achieve an inoculum of approximately 5 × 102
CFU/well for the dilution rows in BHI and 5 × 104
CFU/well for the dilution rows in water. The plates were incubated for 20 min at 4 °C. After incubation, the wells were mixed by pipetting and 20 µL were washed in 180 µL 0.9% NaCl in a fresh 96-well plate. These plates were centrifuged at 3220 g for 5 min, the supernatant was discarded, and the cells were resuspended in 20 µL 0.9% NaCl. Next, 10 µL of each well was spotted on the edge of a BHI agar plate and run down the plate by tilting [43
]. These plates were incubated for either 7 days at 8 °C, to mimic the conditions in a food processing plant, or for 48 h at 37 °C, to provide more favorable growth conditions. Surviving bacteria were enumerated by direct colony count and the MBC was defined as the concentration of PAA that produced no colonies. To determine the MIC, the serial dilution plates were incubated at 37 °C for 48 h and at 8 °C for 7 days. The MIC was defined as the concentration of PAA that allowed for no visible growth [44
4.8. Statistical Analysis
All of the experiments were performed in triplicate unless otherwise indicated.
The results were analyzed in RStudio Version 1.1.456 (RStudio, Boston, MA, USA)) and all statistical analyses are provided as a supplementary file
(Supplementary File 1
). In short, a linear mixed effects model, using lmer in a LmerTest [45
], was modelled to the biofilm data and lsmeans was used to create contrasts [46
]. A cumulative link model was calculated for the BC and PAA data using polR in MASS [47
], and model selection was done with stepAIC in MASS [47
]. All graphics were done using ggplot2 [48