Five pork meat-processing plants were selected with the assistance of the Local Health Authority. The plants were selected in order to account for the whole range of plants production size of the Veneto region of Italy (northeast of the Country) being representative of small, lower intermediate, intermediate, upper intermediate and large production plants (1 plant for each size category, according to expert opinion). The dimension of the processing space was used as proxy of the size of the production plants.

Hazard Analysis and Critical Control Points (HACCP) manuals were evaluated for their coherence with International Life Sciences Institute Research guidelines (ILSI, 2005). A checklist for each plant was created, taking into consideration critical procedures. The customized checklist was used to assess the cleaning and disinfection procedures (C&D) performed after each production cycle. The entire checklists are available under request.

Each selected plant was sampled before and after the C&D procedures. Sampling was performed to assess the presence of L. monocytogenes before and after C&D procedures. To fulfill this purpose, the following surfaces/environments were sampled: broom, the containers for processed raw meat, the water-drainage border surface, the corner between the wall, and the floor of the processing room, the drying space, the floor, the moving hooks, the kart containing meat, the meat kneader, the knives, the drainage wells of the refrigerating, and processing rooms, the needle meat aerator, the meat mincer, the pallet containing bacon, the refrigerating room, the rope for hook mobility, the salami tying, the sausage maker, the sharpening steel, the sinks, spatulas, tables (upper and lower surfaces), the processing room walls, and the warehouse. The sampling was performed using Whirlbag sponge, rehydrated before use with 9 ml of an acqueous solution of Sodium Chloride (0.9%) and kept at refrigeration temperature (4°C) until analysis. All the samples were processed within 24 h from the sampling.

The isolation and characterization of L. monocytogenes were carried out according to EN ISO 16140 and UNI EN ISO 11290-1:2005. Briefly, the sponges were resuspended in 100 ml of Half Fraser Broth (HFB) and incubated at 30°C. After 24 h, 1.5 ml of each sample was collected and used for L. monocytogenes detection using the Bio-Rad IQ-Check- L. monocytogenes II kit (Biorad, USA). Positive samples were plated on ALOA agar and on Oxford agar media. A number of five colonies with the typical L. monocytogenes appearance were tested for haemolysis on Blood agar medium and finally tested by Christie Atkins Munch-Petersen test (CAMP test). Positive colonies were confirmed by Bio-Rad IQ-Check- L. monocytogenes II kit (Biorad, USA). Confirmed colonies were stored at −80°C in Microbank^TM (Pro-Lab Diagnostics, US).

A subsample of L. monocytogenes strains (one colony for each tested surface) was tested for antibiotic susceptibility by using a commercial microdilution test (Sensititre® Streptococcus panel ITST4F) against a panel of 15 antimicrobials: ampicillin, cefotaxime, ceftriaxone, clindamicin, daptomicin, doxiciclin, eritromicin, levofloxacin, linezolid, meropenem, moxifloxacin, penicillin, teicoplanin, trimetoprim/sulfametoxoazol, and vancomycin, according to the manufacturer's recommendations. Results were assessed after 48 h of incubation at 37°C. The minimum inhibitory concentration (MIC) was defined as the lowest concentration of the antimicrobial that completely inhibited visible growth. The results were thus analyzed according to the cut-offs set by EUCAST (www.eucast.org).

Bacterial DNA extracted from the selected collection of strains was examined for the presence of the following panel of genes of resistance to biocide and heavy metals: cadA1-Tn5422 (Mullapudi et al., 2010); cadA2-pLM80 (Ratani et al., 2012); cadA3-EGDe (Ratani et al., 2012); LMOSA_2330; LMOSA_2220; pLI37; F2365_2257; mdrL, lde. The primer sequences are listed in Table S1.

Listeria monocytogenes isolates were typed by means of Multi Locus Sequence Type (MLST) (Salcedo et al., 2003) and enterobacterial repetitive intergenic consensus (ERIC). MLST was performed according to the Pasteur Institute guidance (http://bigsdb.web.pasteur.fr/listeria/primers_used.html). PCR- amplicons were visualized on agar gel electrophoresis and sequenced for allele identification. The MLST profile was then identified using the Pasteur Institute database.

For ERIC-PCR typing, 2.5 μl of DNA template was added to a reaction mix containing 5 μl of Green Buffer (Promega), 0.3 mM each PCR Nucleotide Mix 1 μM of each primer and 0.3 μl of GoTaq; the reaction volume was made up to 25 μl with autoclaved and filtered Milli Q water (Millipore, Billerica, MA, USA). PCRs were carried out using a My Cylcer thermocycler (Bio- Rad). Primer sequences ERIC- 1R (ATGTAAGCTCCGGGGATTCAC) and ERIC2 (AAGTAAGTGACTGGGGTGAGCG) were designed by Versalovic et al. (1991). The ERIC- PCR profiles were obtained by electrophoresis of the different amplicons for 6 h at 60 V, in 2% agarose Tris borate- EDTA (TBE) gel stained with Midori Green (Bulldog Bio). A 100 bp DNA ladder H3 RTU (Nippon Genetics, Germany) was used as the PCR fragment size marker.

The gene expression of two different strains collected in the same plant before and after the cleaning and disinfectant procedures was assessed, which belonged to the same MLST group and with the same genotyping profile. The strains were incubated at 37°C in the presence of 0, 10, and 100 ppm of Benzalkonium Chloride in Mueller Hinton Broth (MHB). The expression of two broad specificity drug efflux pumps, mdrL and lde, that confer resistance to a wide spectrum of biocides (Ortega Morente et al., 2013), and of tufA as housekeeping gene, was evaluated in a time course experiment after 0, 1, 4, and 24 h of incubation. At every time point, the total RNA was extracted by a QIAGEN RNAeasy kit plus RNA Protect Bacteria as suggested by the manufacturer (QIAGEN). Extracted RNA was quantified by Nanodrop, and then immediately retrotranscripted using the QuantiTect-Reverse Transcription (QIAGEN). The obtained cDNA was stored until use at −20°C. The qPCRs were performed according to Di Cesare et al. (2015), except for the expression of results. The Non Reverse Transcripted sample (NoRT) was analyzed along with the retrotranscripted samples. Test specificity was tested by means of melting profile and gel electrophoresis. The adjusted Ct (aCt) was calculated as follows:

where Ct was the cycle threshold of the retrotranscripted sample and NoRTCt was the cycle threshold of the Non Reverse Transcripted sample. The resulted value was used for the data analysis. Each experiment was repeated twice.

The adjusted resulted Cts were averaged and used to obtain a ΔΔCt value according to Livak and Schmittgen (2001).

The hygiene assessment highlighted a series of critical action disregarded during C&D procedures. As it can be noticed from Table 1, the actions neglected by three or more processing plants were: the correct use of disinfectant concentration, the correct exposure time to cleaning agents, the control of rinsing water temperature, the appropriate use of cleaning nozzles, avoidance of aerosol formation, changing of work clothes between cleaning, and disinfection procedures, avoidance of raw meat debris persistence, the suitability of the dimension of drainage holes for debris, all the surfaces of the plants cleaned and drainage wells used appropriately. Interestingly four of the five investigated plants did not use the correct biocide concentration during C&D procedures, nor avoided the persistence of raw meat debris.

A total of 169 environmental swabs were collected before and after the C&D procedures on 28 different surfaces. A number of 98 samples were gathered before and 71 after the C&D procedures. The number of samples collected in each plant varied according to the plant typology and to the number of the sampleable surfaces.

Overall 45 L. monocytogenes strains were isolated from the environmental swabs, 36 of them were retrieved before and nine after the C&D procedures.

In Table 2 all the sampled surfaces and the quantity of positive strains collected before or after the C&D procedures were reported.

The surfaces revealed to be contaminated with L. monocytogenes before the procedures were the following: corner wall-floor, refrigerating room, warehouse, knife, meat container, kneader, sausage maker, salami tying, drying room, floor, corner floor-drainage, drainage of the working room, container of raw materials.

The surfaces found positive to L. monocytogenes after the C&D were: meat aerator, kneader, corner wall-floor, corner floor-drainage, drainage of the refrigerating room, broom, table, drainage of the working room.

A subsample of 19 L. monocytogenes isolates, selected among the 45 positive samples were characterized. This subsample was selected to be representative of the surfaces where positivity for L. monocytogenes was found both before and after C&D (5 couple of isolates). Moreover, 4 and 5 strains isolated before and after C&D respectively, but being unrelated and representative of the whole variety of surfaces, were included.

The entire selected subsample of L. monocytogenes was tested for its susceptibility to a panel of antibiotics. Results are described in Table 3. All investigated isolates resulted susceptible to the majority of the tested molecules.

The assessed genetic resistance determinants to biocide and heavy metals are reported in Table 4. Results showed the panpresence of lde in the entire tested sample of strains and a widespread presence of mdrL and cadA1-Tn5422 found in 15 and 11 out of 19 L. monocytogenes strains respectively. Less than half of the sample harbored F2365_2257 (7/19) and three strains LMOSA_2330. A total of 8 strains displayed the presence of three resistance determinants (lde, mdrL and cadA1-Tn5422), being the most frequent genotype among the investigated sub-sample (Table 4).

Table 5 reported the number of strains of each Sequence Type, according to MLST protocol. The obtained sequence types were ST 1 (1/19), ST 9 (12/19), ST 26 (1/19), ST 121 (3/19), ST 489 (1/19), ST 517 (1/19). Among the five couple of isolates found, both before and after C&D, only two couples displayed the same sequence type (ST9). The rest belonged to different sequence types suggesting their unrelatedness. Regarding the isolated tested for their representativeness of the sampled surfaces, the most widespread type was ST9, found in six samples, of which four belonging to the post C&D isolates.

In order to obtain a deeper insight into the differences among the two couples of strains (1a/1b and 5a/5b) found to be related in terms of sequence type, their ERIC profile was investigated. The results showed a complete overlap in the ERIC profiles of the two couples of tested strains (Figure 1) thus suggesting the potential for their identity. One couple (1a/1b) was selected for the gene expression assay.

The lde and mdrL activities of two strains, 1a and 1b, collected before and after C&D procedures from the same table, were further studied using gene expression profiling. Genotypic and phenotypic characteristics of the investigated strains were reported in Table S2. Gene expression was evaluated calculating the increase or decrease in expression over time of the two target genes in the absence (control) and presence (cases) of benzalkonium chloride (BC) treatment (10 and 100 ppm). The results were normalized against the expression of tufA gene (housekeeping) both in the treatment and control groups. As can be noticed from Figure 2, the two tested strains, 1a and 1b, displayed different expression profiles for the two investigated genes.

Lde was expressed mainly by 1a strain (isolated before the C&D procedures), in the presence of BC exposure, with a dose-dependent trend. On the contrary, strain 1b (isolated after the C&D procedures) displayed a low and constant trend of expression with a slight increase in the presence of the maximum BC tested dose (100 ppm; Figures 2A–C). In the case of mdrL, a different scenario was observed. The two strains presented a similar expression profile with the maximum expression displayed in the absence of BC and a strong reduction of the gene expression showed under exposure to BC at 10 and 100 ppm (Figures 2A–C). A slight increase in the gene expression was observed in the presence of the maximum dose of BC, if compared with the medium used concentration (Figures 2B,C).

Results were described in Figure S1 and Figure 2.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.