Presence of Shiga Toxin-Producing Escherichia coli (STEC) in Fresh Beef Marketed in 13 Regions of ITALY (2017)

The aim of this study was to determine the prevalence of Shiga toxin-producing Escherichia coli in fresh beef marketed in 2017 in 13 regions of Italy, to evaluate the potential risk to human health. According to the ISO/TS 13136:2012 standard, 239 samples were analysed and nine were STEC positive, from which 20 strains were isolated. The STEC-positive samples were obtained from Calabria (n = 1), Campania (n = 1), Lazio (n = 2), Liguria (n = 1), Lombardia (n = 1) and Veneto (n = 3). All STEC strains were analysed for serogroups O26, O45, O55, O91, O103, O104, O111, O113, O121, O128, O145, O146 and O157, using Real-Time PCR. Three serogroups were identified amongst the 20 strains: O91 (n = 5), O113 (n = 2), and O157 (n = 1); the O-group for each of the 12 remaining STEC strains was not identified. Six stx subtypes were detected: stx1a, stx1c, stx2a, stx2b, stx2c and stx2d. Subtype stx2c was the most common, followed by stx2d and stx2b. Subtype stx2a was identified in only one eae-negative strain and occurred in combination with stx1a, stx1c and stx2b. The presence in meat of STEC strains being potentially harmful to human health shows the importance, during harvest, of implementing additional measures to reduce contamination risk.

The majority of STEC strains associated with disease in humans possess adherence factors that facilitate their attachment to the intestinal epithelial cells [20]. The principal adherence factor is the intimin protein encoded by the eae gene, and responsible for what is known as the "attaching and effacing" (A/E) lesion of the intestinal mucosa [4,[21][22][23]. The simultaneous presence of eae and stx2 genes is considered a reliable indicator of a particular STEC strain's ability to cause severe disease in humans [24]. However, STEC strains that lack the eae gene can also cause severe disease by utilising alternative adherence mechanisms, as evidenced recently during a large outbreak of HUS in Germany in 2011 and caused by an enteroaggregative haemorrhagic Escherichia coli (EAHEC) O104:H4 carrying the aggR and aaiC genes in combination with stx2a [25].
Serological identification, based on the somatic (O) and flagellar (H) antigens, has to date resulted in the identification of~470 STEC serotypes [8,26], all able to produce any one of the twelve known Stx subtypes or combinations of these subtypes [27]. The European Food and Safety Authority (EFSA) has identified STEC encoding the stx and eae genes that belong to serogroups O26, O103, O111, O145, O157, the so-called "big five", as those of major concern to human health in Europe [28]. Following the O104:H4 outbreak in Germany, this serotype was incorporated into the screening protocol for all eae-negative STEC isolated from food (Regulation EU No. 209/2013) [29].
STEC O157 is the most frequently reported serogroup worldwide [17,30,31]. The incidence of STEC O157 has however decreased in recent years, whereas the so-called non-O157 STEC serogroups are increasingly associated with haemorrhagic colitis (HC) and HUS in humans [27,31]; the most frequently encountered non-O157 serogroups are O26, O103, O111, O121, O145, O45, O118, O71 and O186 [31,32]. In 2015, as reported by EFSA and the European Centre for Disease Prevention (ECDC), in Europe, the STEC serogroups most commonly isolated from beef were O157 and O26, followed by O148, O145, O8, O113, O91, O130, O174 and O113. Many of these STEC serogroups were linked to human illnesses, confirming the epidemiological involvement of beef in STEC infections [33]. In 2016, in Europe, the STEC serogroup most frequently isolated from bovine meat was O157, followed by O113, O26, O145 and O174 [34]. In Italy, STEC O26 was the predominant serogroup in 2012 and responsible for about half of STEC cases in humans, followed by STEC O157 and STEC O111 [35]. In Europe, as a consequence of only a handful of countries doing any monitoring, few data exist on the isolation of STEC from beef [1, 21,34].
The main reservoirs of STEC are ruminants, including wildlife. STEC can colonize the gut asymptomatically, their excretion into the environment [21] serving as a significant route of infection in humans [3]. Other studies have demonstrated that the hides of cattle represent an important source of STEC, resulting in carcass contamination during harvest [21,36]. Transmission to cattle may take place on-farm or during transportation to the abattoir [5,36]. STEC prevalence in cattle appears to be influenced by the age of the animal, the season, and probably, also feed composition [5,17]. Pathways along which humans may become infected include faecal-oral contamination during harvest, direct contact with faeces, STEC cross-contamination and multiplication during the preparation and handling of animal-derived foodstuffs, and human-to-human transmission [21,37]. European legislation (Regulation EU No. 2073/2005 and its amendments Regulation EU No. 1441/2007) [38,39] did not include the screening of STEC from meat products because, originally, very few data were available on the health risks associated with STEC-contaminated food [1,21].
In 2012, the International Organisation for Standardisation (ISO) issued the ISO/TS 13136:2012 method for the detection of STEC with a focus on the stx1 and stx2 virulence genes and on the eae adhesion factor gene, as these are associated with the "big five" serogroups [40]. The method is based on the Real-Time PCR screening of enrichment cultures, followed by serogroup identification and characterisation of isolated strains. The initial enrichment step, by increasing concentrations of the target bacteria, not only enhances the sensitivity of the method but also ensures the viability of bacterial cells from which positive results are obtained [41]. To date, as stated in the ESFA and ECDC 2017 report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks, 91.5% of the samples tested during 2016 by the European Member States, were analysed using ISO/TS 13136:2012 [34]. Some studies have suggested that the culture conditions involving media formulations and incubation temperature, as currently recommended in the ISO, be modified to further enhance STEC growth [41][42][43][44]. While improvements to the current ISO standard are possible, food authorities will always promote the use of a standardized method so results from different countries remain comparable [34].
The aim of this study was to determine the prevalence of Shiga toxin-producing Escherichia coli in fresh beef marketed in 13 regions of Italy in 2017, to evaluate the potential risk to human health.

Sampling
Between January and December 2017, 239 samples of refrigerated fresh beef were obtained from the retail market in 13 regions of Italy. The samples were collected originally to monitor antimicrobial resistance in zoonotic bacteria from food-producing animals and meat, under Decision 2013/652/EC [45]. The antimicrobial resistance aspects do not form part of this study, but they provided us with the opportunity to assay samples that were representative of most of Italy. The 13 regions account for >90% of the total animals harvested in Italy. The samples were arbitrarily chosen from supermarkets and traditional butcheries, and were obtained at least once monthly throughout the year to cover all four seasons. A single sample was collected from each lot of origin, either domestic or imported; frozen meat was excluded.

Screening of Enrichment Cultures
The samples were analysed following the ISO/TS 13136:2012 standard [40]. Twenty-five grams of meat homogenised with 225 mL of modified Tryptone Soya Broth (mTSB) (Biolife Italiana srl, Milan, Italy) supplemented with 16 mg/mL of novobiocin (Biolife Italiana srl, Milan, Italy) and incubated at 37 • C for 18-24 h. DNA was extracted from 1 mL of each enrichment culture, using an automated nucleic acid purification system (MagPurix ® 12S, Resnova, Rome, Italy), following the manufacturer's instructions. The extraction method provided approximately 100 ng/µL of DNA eluted in nuclease-free water at a final volume of 200 µL. DNA extracts were tested for the stx1, stx2 and eae genes by Real-Time PCR, following the ISO standard procedure given above. PCR amplifications were done maintaining a final volume of 20 µL that contained 3 µL of DNA template (standardized at a concentration of 20 ng/µL), 1× qPCR Master Mix (Kapa Biosystems, Resnova, Rome, Italy), 300 nM of each primer, and 125 nM of each probe (Eurofins Genomics, Milan, Italy). All the reactions included an internal amplification control (Exo IPC kit) (Eurogentec, Italy). PCR conditions comprised an enzyme activation step of 95 • C for 5 min, followed by 40 cycles: 95 • C for 3 s (denaturation) and 60 • C for 30 s (annealing/extension/data acquisition). All the reference material used as reaction positive controls were provided by the European Union Reference Laboratory for E. coli (EU-RL VTEC). All stx-positive and eae-positive enrichment broths were screened for serogroups O26, O111, O103, O145 and O157 [40], while the stx-positive but eae-negative broths were screened also for the O104 serogroup following an additional protocol provided by the EURL VTEC [46]. All the serogroup Real-Time PCRs were done using the same reagent formulas and PCR conditions described above for the stx and eae genes; only for serogroup O103 was the annealing/extension temperature lowered to 55 • C.

Isolation of STEC Strains
For STEC strain isolation, the stx-positive enrichment broths were cultured on Tryptone Bile X-Glucuronide (TBX) agar (Biolife Italiana srl, Milan, Italy) or, if screening of the enrichment broths indicated the presence of serogroup O26, were cultured also on Rhamnose MacConkey (RMAC) agar and incubated at 37 • C for 18-24 h. Then, of many colonies that phenotypically resembled E. coli, fifty were selected arbitrarily and re-analysed singly for the presence (or absence) of the stx and eae genes using Real-Time PCR. Based on the original enrichment broth results, STEC colonies were tested for one or more of the "big five" serogroups. Those stx-positive colonies that tested negative for the "big five", were then analysed for the O45, O55, O91, O113, O121, O128 and O146 serogroups, using a method provided by the EURL VTEC [46].

Stx Subtyping
The stx subtype of each STEC strain was identified using the PCR-based subtyping protocol of the Statens Serum Institut, WHO Collaborating Centre for Reference and Research on Escherichia and Klebsiella [9,11]. The PCR was done using the 2GFast Master mix (Resnova, Rome, Italy), 280 nM of each primer (Eurofins, Milan, Italy) and 5 µL of template DNA (20 ng/µL). Each reaction was adjusted to a final volume of 25 µL in nuclease-free water. The annealing temperature was 66 • C for subtyping stx1a-c, 62 • C for stx2a-c, and 64 • C for stx2d-g. Agarose gel electrophoresis was used to visualize the PCR products. A molecular weight marker (Euroclone S.p.a., Milan, Italy) was used to assign the molecular weight to amplicons produced. The samples were run in Tris-Borate-EDTA running buffer (VWR International Srl, Milan, Italy) at a constant voltage (100 V for the first ten minutes and 60 V until the end of electrophoresis). All Statens Serum Institut reference material used as reaction positive controls was provided through the EURL VTEC.

Discussion
Cattle are considered a major reservoir for virulent strains of Shiga toxin-producing Esherichia coli (STEC) and the most important source of human infections through the consumption of contaminated beef products. The aim of this study was to identify and characterise the STEC strains found to occur in fresh beef obtained from 13 regions in Italy. A culture method involving selective and non-selective media, and following an initial enrichment step, was used to isolate STEC strains [1,47]. Specific PCR assays were used to identify pathogenicity factors (eae and stx genes), serogroups, and stx subtypes [1, 17,40,48,49]. Initial enrichment yielded 20 (8.4%) stx-positive cultures, while STEC strains were only isolated from nine cultures. The failure to isolate STEC from a stx-positive enrichment culture has been reported upon previously [21,[50][51][52]. To isolate STEC from food can be challenging because the number of STEC cells are likely to be low; other hurdles include sublethal cell injury, or cell growth suppressed in the presence of a large population of competing microflora [4,21,53]. For these reasons, enrichment cultures are essential to augment sensitivity, thereby promoting the isolation of STEC strains needed to confirm the presence of the stx genes in the live cell, while excluding the presence of free DNA or free prophages in the cultures [28]. Recently, various authors have reported the reduced sensitivity of mTSB enrichment broths supplemented with novobiocin (16 mg/L), suggesting that a decrease in novobiocin concentration might improve detection of O111 and other non-O157 serogroups [41][42][43][44]. While it is possible that reduced concentrations of novobiocin facilitate the isolation of non-O157 serogroups, 19 of the 20 isolates obtained represented non-O157 STEC strains. In this study, overall STEC contamination in beef was 3.8%, a prevalence rate that agrees only partly with rates obtained previously in Italy. A frequency rate of 0.42% for STEC O157 matches that obtained during a nationwide survey conducted by Conedera et al. [54] and who reported STEC in four (0.43%) of 931 minced beef samples. These were screened only for serogroup O157. In the region of Piemonte, Rantsiou et al. [55] found six (5.9%) STEC strains in 101 mixed meat products using a method developed in-house. In the Emilia Romagna region, Bardasi et al. [56], following the ISO/TS 13136:2012 protocol, demonstrated an STEC presence in four (0.6%) of 689 meat samples (representing pork, bovine and poultry). In a more recent study, the same ISO protocol was used to test 675 pork samples (comprising both fresh and dried products) collected in the Umbria and Marche regions of Italy [57]; these authors reported the presumptive presence of stx-genes in 2.8% of the products, but were unable to isolate any STEC strains. The discrepant STEC prevalence rates obtained may find causes in various factors, including geographic compartmentalization of the E. coli population amongst food animals, laboratory techniques and protocols employed [54,55], and the wide range in meat products analysed. According to EFSA, in Europe, the overall presence of STEC in 18,975 food samples assayed was 2.5%, the highest proportion found in meat, particularly that from small ruminants [34]. In Switzerland, Fantelli et al. reported the presence of STEC in 2.3% of 211 minced beef samples tested [58]; in France, 4% of 411 beef samples were STEC positive [59]. Our STEC prevalence rates are comparable to some of those obtained for beef previously in Europe [34,58,59].
Real-Time PCR, based on the O-antigen synthesis genes (wzx and wzy), is widely used to serogroup STEC strains [60]. However, the Real-Time PCR methods currently used do not cover all known serogroups, hence many serogroups to which a STEC strain may belong remain unidentified [33]; consequently, 12 of our 20 STEC strains could not be identified to serogroup. While serogroups and serotypes are not virulence factors and not predictive of a virulence profile, they nevertheless remain useful for conducting surveillance and for investigating outbreaks [61]. Serogroups O26, O103, O104, O111 and O157, along with the eae and stx2 virulence genes, were detected in 11 enrichment broths; the failure to isolate STEC strains from these 11 broths is the reason why the corresponding samples were classified as "presumptive" positive. The potential risk to human health that "presumptives" represent, means the responsible authorities must continue to monitor for STEC.
An association between Stx subtype and severity of disease in humans has been observed [51,62]. In this study, 20 STEC strains were isolated from nine beef samples and carried stx subtypes in various combinations. The stx1 subtypes detected were stx1a and stx1c. The Stx1a toxin subtype is often produced by strains that are eae-positive and known to cause severe disease in man [32]. In this study, nine STEC strains, either eae-positive or eae-negative, had stx1a alone or in combination with stx1c or stx2a, stx2b, stx2c and stx2d. The Stx1c toxin subtype is reported mainly in eae-negative strains causing mild infections [63]; we found stx1c both in identified (serogroup O91) and unidentified serogroups that were eae-positive (Table 2). With regard to toxin type Stx2, the subtypes stx2a, stx2c and stx2d have been linked to HC and HUS in humans [11,16]. In this study STEC isolates carrying stx2c were the stx subtypes most commonly found, followed by stx2d and stx2b. Subtype stx2a was identified in only one eae-negative strain and occurred in combination with stx1a, stx1c and stx2b. Of the eight eae-positive STEC strains obtained, four belonged to serogroup O91, one to O157, while three represented unidentified serogroups. Two strains of O113 and one of O91 were eae-negative; the nine remaining strains were not identified to serogroup (Table 2). Contrary to other reports [64], we found four O91 STEC strains to be eae-positive; this finding is unusual and we cannot explain it satisfactorily. Subtype stx2d was found in eight STEC strains, of which four were eae-negative, while the other four were eae-positive; the latter group included a strain O157 that was also stx1aand stx2c-positive. The stx2d subtype is usually associated with eae-negative strains and severe disease in humans [65]; recently, in Spain, Sanchez et al. [66] reported upon a O157:H7 strain that was eaeand stx2d-positive and isolated from a 2-year-old child with BD. This unusual virulence combination, though rare, has been reported also from several HUS-affected patients in France and separately involving STEC O26:H11 [67] and STEC O80:H2 [68].

Conclusions
In Italy, the isolation of STEC strains from fresh meat samples signals the recurring threat that beef, consumed either undercooked or raw, poses to human health. The variety of stx types and subtypes and multiple STEC serogroups detected, are amongst those found elsewhere in the world and where, in humans, they have been demonstrated to be involved in severe diseases, such as BD, HC and HUS. The presence in meat of potentially harmful STEC strains emphasizes the importance, during harvest, of implementing additional measures to reduce contamination risk. Linked to this, an efficient surveillance strategy for STECs in retail foodstuffs, remains a national priority. The laboratory diagnostic protocols needed to isolate and accurately identify STEC strains are laborious, expensive, and time-consuming. However, they continue to remain pivotal to assessing the strain of pathogenic E. coli involved, and for identifying the possible source of infection. This knowledge is needed to enable the competent authorities to respond precisely and rapidly. Improvements to current isolation techniques, and the validation and standardization of molecular protocols, remain a matter of urgency. It is foreseen that in the future new high-power methodologies, such as Next-Generation Sequencing (NGS), will become more widely utilised and that these will lead to further improvements in the currently used standards for diagnosing STEC in foods.