Staphylococcus aureus is a well-known opportunistic foodborne pathogen, and is involved in numerous nosocomial and community-associated (CA) outbreaks worldwide (Paterson et al., 2014). The widespread use of antibiotics, and particularly inappropriate use or overuse, has facilitated the emergence of pathogens resistant to antibiotics, such as methicillin-resistant S. aureus (MRSA). In addition to human clinical applications, antibiotics are also used in veterinary medicine and animal feeding, contributing to the substantial appearance of antibiotic-resistant strains (Valsangiacomo et al., 2000; Grema et al., 2015). The first reported nosocomial episode involving MRSA was in 1960, and MRSA has become an emergent pathogen (Bonten and Weinstein, 2016; Oniciuc et al., 2017), affecting patients in hospitals and people in community settings such as nursing homes and nurseries (Lo et al., 2007; Murphy et al., 2012; Blumental et al., 2013). MRSA has been found in livestock, linked to a jump from humans to animals (Voss et al., 2005; Price et al., 2012). The prevalence of livestock-associated (LA)-MRSA in farm animals is increasing and the resulting food products may become contaminated. This may be an important ecological niche, favoring the evolution of LA-MRSA (Yan et al., 2014; Verhegghe et al., 2016), and the apparently increasing prevalence can lead to human outbreaks (Grøntvedt et al., 2016).

Lineages other than LA-MRSA are found in food intended for human consumption. Contaminated food of animal origin may contribute to the prevalence of CA-MRSA (Ogata et al., 2012; Rodríguez-Lázaro et al., 2015) or hospital-acquired (HA)-MRSA (Pu et al., 2009; Weese et al., 2010). We demonstrated the presence of CA-MRSA in foods confiscated from passengers on international flights, mostly from Central and South America (Rodríguez-Lázaro et al., 2015). This has food safety implications: contamination with the CA-MRSA lineages isolated may have been due to incorrect food handling, and not zoonotic transmission.

Staphylococcal enterotoxins (SEs, se genes) have been detected in many foodstuff causing staphylococcal food poisoning, toxic shock, and allergic and autoimmune reactions (Gonano et al., 2009). Staphylococcal enterotoxins are classified into 23 different SEs and SE-like toxins (SELs) including the five major serological types SEA, SEB, SEC, SED, and SEE (Hennekinne et al., 2012; Carfora et al., 2015; Ono et al., 2015). However, little is known about the prevalence of MRSA in foods involved in international trade, or about their enterotoxin production. The role, if any, of foods illegally transported between different parts of the globe in MRSA epidemiology is unknown. We investigated whether the uncontrolled entry of foods into Europe through international flight passengers or from open markets at European Union (EU) borders is a potential route of transmission of enterotoxigenic antibiotic-resistant strains, particularly MRSA. Our study reveals the entry of MRSA by this route, and that the lineages involved showed enterotoxigenicity. Uncontrolled entry of foodstuffs into the EU can be a relevant neglected route of MRSA dissemination.

We document S. aureus and MRSA in food confiscated to international flight travelers in UE airports or illegally sold foods in open markets at EU borders. Approximately one in six of the foods tested (136 of 868; 15.7%) were positive for S. aureus, and 3.0 % were positive for MRSA (26 of 868). The confiscated foods in which S. aureus counts exceeded international microbiological recommendations and EU regulations (Anonymous, 2005) were of very diverse geographical origins, including South and Central America, Europe, Africa, and Asia (Figures 1, 2). This is a serious public health concern because S. aureus can produce enterotoxins that can be present in food products. Our results agree with previous reports of the prevalence of S. aureus contamination (13–60%) of food products of animal origin (Crago et al., 2012; Rodríguez-Lázaro et al., 2015; Sun et al., 2015; Ciolacu et al., 2016). Various confiscated cheeses, including soft, semi hard, or hard had counts of up to 3.17 log₁₀ above the limit established by Anonymous (2005). The high S. aureus counts may be in part the consequence of improper storage and poor hygiene according to the nature of surfaces and number of people handling the food (Chen et al., 2001). There may also be two way bacterial transfer between clothes and food preparation surfaces (Bloomfield et al., 2011). Twenty-six of the S. aureus strains isolated were MRSA (3.0%) of which 19 were enterotoxigenic. Under inappropriate food transportation conditions, heat-stable enterotoxins could emerge leading to gastroenteritis outbreaks. Many of the MRSA strains were positive for one or more se genes, consistent with previous studies (Pereira et al., 2009; Carfora et al., 2015). Contaminated milk and dairy products are major vectors of enterotoxins but, the link between the source of food contamination and transfer of antibiotic resistance determinants remains unclear; there are few reports describing the presence and possible origin of MRSA in foods (Ortega et al., 2010). However, there has been one case of community-acquired foodborne illness caused by SEC-producing MRSA (Jones et al., 2002) in the United States, and production of SE types SEB, SEC, SED, and SEE in two MRSA strains of milk origin from Minnesota farms (Haran et al., 2012).

Little is known about the prevalence of MRSA in food (Oniciuc et al., 2017), and particularly on in foods entering the EU. Our research group has performed several studies in airports and at borders (Oniciuc et al., 2015; Rodríguez-Lázaro et al., 2015) and found S. aureus and MRSA prevalence values of 8–33.9% and 0.5–3.1%, respectively. The values we report here are similar: 15.7 and 3.0% for S. aureus and MRSA, respectively. The number, the type of food sample, the geographical origin and the sampling place (Vienna International Airport was included in this study) are larger than in previous studies, but no significant differences in the overall percentages were observed. Nevertheless, 19 of 263 food items (7.2%) were MRSA positive at Bilbao Airport but only 7 of 595 (0.8%) at Vienna Airport. This significant difference could be a consequence of the geographical origins of the flights: most of the MRSA-positive foods at the Bilbao Airport were from Central and South America (14 of 19; 73.7%).

The MRSA lineages found included ST5, ST8, ST1649, ST1 and other lineages represented in a lesser extent (ST7, ST22, ST72, ST97, and ST398) (Figure 5). The ST most frequently isolated was ST5 (30.8%). This ST is believed to be the result of a species jump followed by adaption to the new host (Lowder et al., 2009). ST5 has mostly been found among poultry isolates (Lowder et al., 2009), but in our study was mostly isolated from dairy products. The ST5 lineage is a major component of hospital- and community-associated MRSA and MSSA worldwide (Miko et al., 2013). Other MRSA strains identified in our study were ST8-MRSA-IV/V and ST1649-MRSA-IV, which are successful of CA-MRSA clones. The presence of PVL genes and different antimicrobial susceptibility patterns in ST8-MRSA are causes of concern; it is not clear whether human handlers played any role in the preliminary post slaughter process.

The LA-MRSA ST398 is widespread in Europe (Oniciuc et al., 2017). It emerged rapidly in the Netherlands and now accounts for 20% of all human MRSA cases and for 42% of newly detected MRSA, indicating that animals are important reservoirs for human MRSA infection (Kadariya et al., 2014). There have already been outbreaks due to LA-MRSA ST398 (Wulf et al., 2008; Verkade et al., 2012). However, LA-MRSA ST398 was found only in one dairy product from Egypt in our study; this isolate harbors the luk-PVL genes but not enterotoxin-producing genes, in accordance with previous studies (Argudín et al., 2011). We isolated one OS-MRSA strain from a cheese from Turkey at Vienna airport: this MRSA variant is important because it may be misidentified as MSSA (Ariza-Miguel et al., 2015), such that β-lactam antibiotics may be used for treatment resulting in the potential development of highly resistant MRSA.

Many of our MRSA isolates were resistant to three or more antimicrobial agents (Table 1). This relatively high prevalence of these strains and their antimicrobial resistance profiles reveals the potential public health problems associated with illegal import of food in passenger luggage. Multiresistant strains are being distributed worldwide by air travel, and indeed undoubtedly by the various other forms travel across borders. United States custom officers ask all passengers from outside the United States to fill a form about contact with animals (Category A referring to zoonotic disease transmissible through animal contact) and any animal-origin food product potentially infected with zoonotic agents (Category B) (Noordhuizen et al., 2013). The EU forbids the importation of personal foodstuff of animal origin including meat, milk, and derived foodstuffs, from non-EU-countries (other than the Faeroe Islands, Greenland, and Iceland) (Anonymous, 2009). The EU has issued several regulations for the import of animals and food products of animal origin, but these regulations generally refer to commercial trade and bulk food products (Anonymous, 2007, 2009). This disregard of the risk associated with illegally transported foodstuffs combined with insufficient border controls could lead to foodborne outbreaks (Noordhuizen et al., 2013). The prevalence of MRSA in food depends on, among other factors, origin and country of provenance (Rodríguez-Lázaro et al., 2015; Oniciuc et al., 2015). Foodborne outbreaks of MRSA infection have been reported (Jones et al., 2002), confirming that food serves as a transmission pathway for MRSA (Oniciuc et al., 2017).

This study shows the presence of enterotoxigenic HA-, CA-, and LA-MRSA in foods confiscated from passengers on flights entering the EU and in food illegally introduced and sold at EU borders. There is therefore a risk of foodborne transmission of MRSA, and illegal entry of food is a possible route for enterotoxigenic MRSA transmission and spread. Efficient control measures (such as the increase of the in situ examination and testing for the presence of uncontrolled entry of foods at EU borders) are required to avoid transmission of antimicrobial-resistant strains to humans by the consumption of such foods. Travelers should be made aware of the dangers, to promote acceptation of prohibition of food transport and thereby reduce the risk of spreading of foodborne pathogens.

DR-L designed, supervised the experiments, analyzed the results, revised the first draft, and prepare the last draft of the manuscript. E-AO performed part of the experiments, analyst the data and prepare the first draft of the manuscript. PG, IF-N, DG, and MD-G performed part of the experiments. JE-B performed part of the experiments and revised the different version of the manuscript. MW and AIN collaborated in the design of the experiment and revised the different version of the manuscript. MH collaborated in the design of the experiment, in the supervision of the experiments, and in the analysis of the results and revised the different versions of the manuscript.

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.