Listeria monocytogenes is a ubiquitous foodborne pathogen that causes listeriosis and is frequently linked to consumption of ready-to-eat (RTE) foods, such as fresh produce, soft cheese, and especially RTE meat products (Fagerlund et al., 2020). RTE meat products have caused several outbreaks of listeriosis, including the largest outbreak that occurred in South Africa (Smith et al., 2019; Gilmour et al., 2020). Furthermore, listeriosis remains a significant public health concern because of the high case fatality and mortality (Thomas et al., 2015).

RTE foods are intended to be consumed without further cooking or other processing, thus requiring the elimination or reduction of microorganisms of concern to acceptable levels. The microbiological safety of RTE foods relies on the adoption by manufacturers and processors of a number of measures aimed at preventing food contamination (FAO., and WHO., 2019). Contamination by L. monocytogenes is a major characteristic at food processing plants because bacterium can exist in harsh environments, such as low pH and high salt concentrations, and can grow and persist at low temperatures (Ferreira et al., 2014). The epidemiological investigation of meat-related outbreaks has revealed substantial deficiencies in hygiene in food processing environments (Smith et al., 2019). In a previous study, the contamination of L. monocytogenes in an RTE food processing plant suggested that clone transmission may occur in the food processing environments in Shanghai (Zhang et al., 2021) and thus pose a potential hazard to food safety. Therefore, the constant risk of L. monocytogenes transmission to consumers remains a central challenge to the food processing plants (Almeida et al., 2013; Muhterem-Uyar et al., 2015).

The multilocus sequence typing (MLST) method has been used to type L. monocytogenes into many different sequence types (STs). Epidemiological investigations have identified that several STs are predominantly found in clinical patients, whereas other STs have been shown to be particularly abundant in food or food processing environments (Maury et al., 2016). In China, ST87 and ST8 were the most STs isolated from patients (Li et al., 2019), whereas ST9 was the most common ST isolated from foods (Wang et al., 2012; Zhang et al., 2020). ST5 and ST121 isolates have been reported to be predominantly found in food processing environments (Muhterem-Uyar et al., 2015; Naditz et al., 2019; Zhang et al., 2021). Therefore, MLST of L. monocytogenes can help establish links between isolates from different sources and provide information regarding the source of contamination (Chen et al., 2013).

Pulsed-field gel electrophoresis (PFGE) has been used to characterize clusters of L. monocytogenes isolates and can even identify persistence (Buchanan et al., 2017; Li et al., 2019). Persistent strains are defined as those with an indistinguishable PFGE signature (Buchanan et al., 2017). Whole-genome sequencing (WGS) is a powerful tool for analyzing molecular characteristics of L. monocytogenes together with the development of core-genome MLST (cgMLST) and single nucleotide polymorphisms (SNPs) that have been used to investigate listeriosis outbreaks and offer various advantages over PFGE (Chen et al., 2016; Jackson et al., 2016). WGS has recently been used to determine the contamination and/or colonization route of pathogens within food processing environments (Zhen et al., 2017). Tracking the primary contamination route of L. monocytogenes in RTE meat processing plants is essential as this could help guide food business operators to remove or reduce resident L. monocytogenes. In this study, we investigated the contamination of L. monocytogenes in two RTE plants in Shanghai. The study aimed to (1) describe the occurrence of L. monocytogenes in different areas, such as thawing, trimming, boiling, cooled, packaging, and sterilizing rooms, at two RTE meat plants in Shanghai; (2) trace the relevant sources of contamination for RTE meat products using PFGE and WGS and clarify the transmission of L. monocytogenes at RTE meat plants; and (3) identify the molecular characteristics of persistent isolates in RTE meat plants in relation to their ST, serogroup, and virulence profile.

The control of L. monocytogenes in RTE meat processing plants is an ongoing and important challenge. The introduction of L. monocytogenes to processing plants continues to occur, and persistent L. monocytogenes isolates can exist in food processing plants because of their strong ability to survive in different types of conditions (Conficoni et al., 2016). Therefore, samples from different areas in the two RTE meat processing plants were collected to investigate the distribution of L. monocytogenes and track the sources of contamination. A total of 239 samples were collected in the two RTE meat processing plants during 2019–2020 in Shanghai (Table 1); 58 of these samples were from raw and accessory materials, and 12 samples were from process water, by which L. monocytogenes isolates could enter food processing environments (Marriott et al., 2018). Overall, 99 of the 239 samples were from processing environments and facilities, such as floor, wall, transport units, pipes, trash can, and weighing tools, as well as hands, clothes, and shoes of the handlers. These niches can harbor sites where L. monocytogenes can survive and even grow (Buchanan et al., 2017). Overall, 70 of the 239 samples were intermediate and end products, collected and also identified to be contaminated by L. monocytogenes.

Our study confirmed the different contamination models of L. monocytogenes observed in the two RTE meat plants. Indeed, a higher level of L. monocytogenes contamination occurred in the packaging room than in other areas (p < 0.05), indicating this may be the high-risk area. At the same plant, the positive rate of L. monocytogenes from samples in the packaging room was significantly higher than that in other rooms (p < 0.05). These results suggest that contamination of L. monocytogenes in the packaging room was more likely to occur. In packaging rooms, more of the tasks required human handling, such as weighing, cutting, and packaging. Furthermore, L. monocytogenes isolates in CL5 were found in the clothes and shoes of the workers, as well as facilities and environments in packaging rooms; this suggests the possibility of cross contamination. Accordingly, on the one hand, it is essential to pay attention to the hygiene of the workers, i.e., their clothes needed to be cleaned and disinfected regularly. On the other hand, it is vital to clean and disinfect the facilities and environments regularly. Moreover, L. monocytogenes isolates have been detected in 2019 as well as in 2020. Therefore, regular cleaning and disinfection are essential.

Six L. monocytogenes isolates were detected in raw materials at plant B. However, L. monocytogenes was not detected in processing environments and facilities at plant B. These results suggested that raw materials contaminated with L. monocytogenes were the means whereby bacterium could enter food processing environments. However, this contamination could be eradicated by some measure. Raw materials should be cleaned and disinfected, and good hygiene of the plants should be essential.

The PFGE profiles of 19 L. monocytogenes isolates in PT7 were indistinguishable (> 95.7%), suggesting that they were from the same ancestor. The epidemiological data showed that these 19 L. monocytogenes isolates detected at plant A during 2019–2020 were persistent isolates that might have been presented since 2 years at the plant. A similar contamination model reported that when L. monocytogenes isolates entered food processing plants, recontamination and persistence frequently occurred (Chambel et al., 2007), and such samples represented 65.5% (19/29) of the overall L. monocytogenes isolates. The presence of L. monocytogenes in PT7 in the cooling, pickling, boiling, and packaging rooms and their circulation at plant A through the transmission between food products and the facilities, probably represented the mechanism whereby L. monocytogenes could persist in the food processing plants. Indeed, a significant difference (p < 0.05) in the occurrence of contamination between plants A and B was observed in the packaging room, and a significant difference (p < 0.05) was also observed between the packaging room and other areas at plant A. This contamination model suggested the existence of cross contamination between different areas within the same plant and reflected the operative features of a plant regarding sanitization procedures and behavior of workers. Based on the findings observed in the packaging room at plant A, the hands, clothes, and shoes of the handlers were contaminated with L. monocytogenes isolates belonging to PT7 and could be considered carriers for processing environments, facilities, and even end products. In contrast, processing facilities, such as conveyors, and the inside and outside surface of facilities were significant niches that supported the tendency of L. monocytogenes to persist in processing facilities because of the difficulty in completely eradicating the bacterium from these areas. It has been reported that similar hard-to-clean surfaces allow the bacteria to survive or even proliferate, thereby making it difficult to completely eradicate from these surfaces (Ferreira et al., 2014). Therefore, ensuring correct sanitization practices is essential to avoid cross contamination to products. Where necessary, facilities should be autoclaved to completely eradicate L. monocytogenes.

There was a high diversity of PFGE patterns of L. monocytogenes from plant B. Eight isolates were typed into five PTs (PT1–PT2 and PT4–PT6). ST121-IIa was predominant with three L. monocytogenes isolates identified from raw materials in plant B in 2020. However, one ST121-IIa was detected in the processing facility at plant A in 2020, which belonged to PT3. The similarity of the isolate in PT3 to ST121 isolates in PT2 was low. Furthermore, only one ST121 isolate was detected in plant A. It was difficult to judge how ST121 isolates entered plant A. Although there was no relationship between plant A and plant B, these results could suggest that ST121 L. monocytogenes could enter plants via raw materials. Raw meat has rarely been implicated in foodborne disease, although this may be considered a potential source of domestic cross contamination of other foods (Thévenot et al., 2006). ST121 L. monocytogenes has been reported to persist in food processing facilities (Naditz et al., 2019). Therefore, strengthening surveillance of L. monocytogenes in raw materials as well as food processing environments is essential to prevent contamination of L. monocytogenes.

In this study, WGS was used to obtain more detailed information on the genetic similarity between isolates and has clear advantages over PFGE (Zhen et al., 2017). The indistinguishable L. monocytogenes isolates that have the same PFGE pattern could be differentiated by cgMLST (Figure 2). cgMLST analysis of L. monocytogenes isolates in PT7 exhibited up to nine allelic differences, and these isolates could be identified as being the same clone. The closest pairwise differences between these ST5 isolates from plant A ranged from 0 to 16 SNPs. These numbers were well below the threshold of 25 SNPs commonly employed during outbreak investigations as an indication that two isolates had originated in the same facility (Wang et al., 2018; Allard et al., 2019). This indicated that persistence of L. monocytogenes at plant A was the most common scenario. These isolates were from different areas, such as the packaging, cooling, and pickling rooms, which suggested that clone transmission had occurred at plant A during 2019–2020 by cross-transmission. This cross contamination has been reported as the primary route of contamination by L. monocytogenes in RTE food products in food processing environments (Møretrø and Langsrud, 2004; Ferreira et al., 2014); clones of L. monocytogenes are known to have persisted for years in processing environments, including those in meat industries (Cherifi et al., 2018; Melero et al., 2019).

The primary origin of ST5 from plant A is unknown. In-depth epidemiological studies have identified that ST5 isolates can be highly abundant in foods and in food processing environments (Muhterem-Uyar et al., 2018; Naditz et al., 2019). Our previous study indicated that the predominant L. monocytogenes isolates from both food and clinical patients in Shanghai were ST5 (Zhang et al., 2020). Why do ST5 isolates tend to be persistent in food processing environments? At the genome level, the ST5 isolates contain plasmids harboring an efflux pump system (bcrABC cassette) and heavy metal resistance genes, which may be important for the persistence of ST5 isolates in food processing environments (Muhterem-Uyar et al., 2018). In contrast, our previous study proved that these ST5 isolates at plant A could form biofilms, which provided a protective environment for bacterial survival and thus increased the risk of subsequent contamination (Colagiorgi et al., 2017; Zhang et al., 2021). Although ST5 isolates with LIPI-1 were not hypervirulent because they lack LIPI-3 or LIPI-4 (Manso et al., 2019), they still have caused many sporadic patients (Zhang et al., 2020). Several outbreaks caused by L. monocytogenes have been linked to ST5 isolates (Buchanan et al., 2017). Therefore, the potential risk of pathogenicity is worth noting, and further studies are needed to uncover the trait that enables persistence of ST5 in food processing environments.

Raw meat is a common source of introduction of L. monocytogenes into food processing environments (Fagerlund et al., 2020). In this study, raw materials were also positive for L. monocytogenes isolates. Raw meat distribution chains in China are complex, and there are many different raw meat suppliers who sold raw meat to different meat processing plants. This may explain why the positive rate of L. monocytogenes occurrence was different between plants A and B. Furthermore, a total of three STs (ST378, ST121, and ST120) from raw materials were detected in plant B. However, these STs were only detected once, suggesting that these were transient isolates. However, transient isolates could change to persistent isolates under specific conditions (Naditz et al., 2019), and detection of L. monocytogenes from raw materials should be prioritized to ensure food safety. Furthermore, detection of L. monocytogenes isolates from upstream in the raw material chains was required to potentially identify the primary sources of L. monocytogenes isolates.

Our study demonstrated a two-contamination model of L. monocytogenes isolates at different RTE meat processing plants in Shanghai. Using PFGE, ST5 isolates at plant A were confirmed as persistent isolates and may have been persisted here for an extended period (at least 2 years). Clone transmission was confirmed using WGS data from processing environments and facilities by cross contamination. At plant B, L. monocytogenes isolates could enter the food processing environment from raw materials. However, these isolates were considered as transient isolates by PFGE and were not detected in the processing environments, facilities, and RTE products. These results suggested that continuous monitoring, stringent surveillance, and source tracking are crucial to guarantee food safety.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.

HZ and JW collected samples, analyzed samples, performed WGS, and analyzed the data. HZ drafted the manuscript. XZ designed the study and revised this manuscript. ZC and XL tested the samples. YY and JW performed the PFGE. YLY and XW were involved in the collection of isolates. QD and XZ revised this manuscript. All authors contributed to the article and approved the submitted version.

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.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.