Prevalence and transmission characteristics of Listeria species from ruminants in farm and slaughtering environments in China

ABSTRACT Listeria monocytogenes is an important foodborne pathogen, and is ubiquitously distributed in the natural environment. Cattle and sheep, as natural hosts, can transmit L. monocytogenes to related meat and dairy products. In this study, the prevalence, distribution, and transmission characteristics of Listeria were analysed by investigating 5214 samples of cattle and sheep in farm and slaughtering environments in China. A low contamination incidence of L. monocytogenes (0.5%, 20/4430) was observed in farm environment, but there was a high contamination incidence in slaughtering environment (9.4%, 74/784). The incidence of L. innocua in cattle and sheep farm and slaughtering environments is more common and significantly higher (9.7%, 508/5214) than that of L. monocytogenes (1.8%, 94/5214). The distinct molecular and genetic characteristics of Listeria by PFGE and MLST indicated that L. monocytogenes and L. innocua were gradually transmitted from the farm and slaughtering environments to end products, such as beef and mutton along the slaughtering chain. The ST7, ST9, ST91, and ST155 found in our study were associated with the human listeriosis cases in China. In addition, the findings of virulence markers (inlC, inlJ, LIPI-3, LIPI-4, and ECIII) concerned with the pathogenesis of human listeriosis and antibiotics resistance of L. monocytogenes in this study implies a potential public health risk. This study fills the gap in the epidemiology of beef cattle and sheep that carry Listeria in farm and slaughtering environments in major cattle and sheep producing areas in China.


Introduction
animal-derived products. The hides and intestines are thought to be the most important sources for microbial contamination [14]. Therefore, the investigation of livestock animals in farm environment is very important to elucidate the contaminant source of L. monocytogenes, which will provide crucial data for control of L. monocytogenes at the farm level [15].
The output of beef and mutton in China ranks first in the world, and the demand and proportion of meat are increasing year by year. However, the prevalence and transmission background of L. monocytogenes in cattle and sheep farm environments, and meat and dairy products in major cattle and sheep producing areas in China remain unclear. At present, the surveillance data of Listeria distribution in farm and slaughtering environments of cattle and sheep remain limited in China. Thus, it has become difficult for policymakers to formulate effective policies to control the Listeria contamination in food of cattle and sheep origin. This study was conducted to reveal the prevalence, transmission, and pheno-and genotypic characteristics of Listeria species in farm and slaughtering environments in major cattle and sheep producing areas in China.

Sample collection
Faeces, silage, drinking water, hide swabs, and raw milk samples directly from the nipple of dairy cattle in milk halls in farm environment, and faeces, hide swabs, knife swabs, rinsed water used to clean the knife, chopping board swabs, instrument swabs, and beef for sale after slaughter in the slaughtering environment were gathered in major cattle and sheep producing areas in China from 1 September 2018 to 1 October 2019. A total of 5214 samples were randomly collected from farms and abattoirs located in Northeast China, Northwest China, and Inner Mongolia (Figure 1), covering 12 farms and 6 abattoirs in six provinces in China.
Isolation, PCR and serological identification of Listeria spp.
Listeria were isolated according to the GB 4789.30-2010 method in China, with moderate modifications [16]. Ten millilitres Listeria Enrichment Broth Base LB 1 (Haibo Biology, Qingdao, China) were prepared before the sampling was carried out each time. Approximately 2 g, 2 mL, or 2 cm 2 of samples were collected and directly added into 10 mL of LB 1 once samples were gathered in farms or abattoirs. Then, 25 g of beef or mutton samples and 225 mL of LB 1 were used for immediate homogenizing when these were transported to the laboratory. After the samples were incubated at 30°C for 24 h, 0.1 mL of preliminary enrichment broth was added into 10 mL of LB 2 for the second-step enrichment at 30°C for 24 h. Then, two loopful of secondary enrichment broth were inoculated on PALCAM Agar (Haibo Biology, Qingdao, China) at 30°C for 28 h. Afterwards, 3-5 presumptive colonies were randomly chosen for purification and subculture on the PALCAM agar at 37°C for 24 h. Then, the purified bacterial cultures were chosen for Gram-staining. After Gram-staining, cultures suspected of being Listeria spp. were further checked using the multiplex PCR method that targets specific genes of different Listeria spp. [17]. Primers targeting the putative transcriptional regulator gene were used to check the L. innocua strains [18]. In addition, a multiplex PCR method to differentiate L. monocytogenes, L. ivanovii, typical L. innocua, and atypical L. innocua was used to check the above results again [11]. The Listeria strains were preserved in LB liquid medium that contained 20% glycerol at −80°C.
The bacterial wall-breaking treatment and multiplex PCR method to differentiate the major serogroups of L. monocytogenes were applied, according to the description reported by Doumith et al. [19]. When the serogroups were identified, traditional slide agglutination with the Listeria Antisera (Denka Seiken, Japan) and polyclonal crossed absorbed factor (L. innocua: 6a, 6b) antiserum, in accordance with the methods described by Seeliger and Hönne [20], which were preserved by our laboratory, was operated according to the instructions of the manufacturer, in order to identify the specific serotype of Listeria isolates.

Pulsed-field gel electrophoresis
The pulsed-field gel electrophoresis (PFGE) of the strains on 57 representative Listeria strains based on different regions, types of samples, and serotypes was performed using the primary restriction enzyme ApaI (Takara, Dalian, China), according to the standard operating procedure provided by the PulseNet of Centers for Disease Control and Prevention [21]. Similarities among the fingerprint profiles of strains were analysed by the unweighted pair group method with arithmetic mean using the BioNumerics software (Version 5.10, Applied Maths, Belgium). The Salmonella strain H9812 restricted with ApaI was used for molecular weight standards in all the PFGE gels.

Multi-locus sequence typing
Multi-locus sequence typing (MLST) on 50 L. monocytogenes and 50 L. innocua isolates from different regions based on seven house-keeping genes (abcZ, bglA, cat, dapE, dat, ldh, and lhkA) was performed, according to the primers and the methods provided in http://bigsdb.web.pasteur.fr/Listeria/. The sequences types (STs) were determined by comparing with the allelic profiles for Listeria in the MLST database. The minimum spanning tree construction based on the seven house-keeping genes' sequences was analysed using the BioNumerics software.

Detection of virulence markers of L. monocytogenes
Virulence genes inlC and inlJ were detected simultaneously using PCR methods described by Liu et al. [22]. The presence of LIPI-3 encoding listeriolysin S and LIPI-4 encoding a cellobiose-family phosphotransferase system were confirmed by using PCR targeting llsX and ptsA genes, respectively, to screen for the potential hypervirulent L. monocytogenes [8,23]. ECIII associated with outbreaks in the United States was also determined for 1/2a isolates by using PCR [24]. In addition, mutations leading to a premature stop codon (PMSC) in inlA that significantly reduce the virulence of L. monocytogenes were analysed [8].

Antimicrobial susceptibility test
Antimicrobial susceptibility test of all L. monocytogenes strains was done using the disk diffusion method according to the Clinical and Laboratory Standard Institute (CLSI) guidelines [25]. A total of 18 antibiotic agents were tested at specific concentrations per disk (Table 4). Staphylococcus aureus ATCC 25923 and Escherichia coli ATCC 25922 were used as quality control strains. Strains with three or more antibiotic resistance were defined as multidrug-resistant strains [26].

Statistical analysis
The chi-square test was used to determine statistical differences of the prevalence of Listeria spp. among the farms, abattoirs, and sample categories. P < .01 was considered as having a highly significant difference, while P < .05 was regarded as having a significant difference. All statistical analyses were performed using the SPSS v25.0 software.

Results
Occurrence of Listeria spp. in farm and slaughtering environments of ruminants The incidences of Listeria spp. in various samples or different farms and abattoirs are summarized in Tables  1 and 2, and Tables S1-S3. The prevalence rate of Listeria spp. in farm environment varied from 0% to 18.4% among different farms. Among the samples of different categories in farm environment, the prevalence rates of Listeria spp. varied within 1.9-7.8%. Furthermore, the prevalence rates of Listeria spp. varied from 22.9% to 69.5% among different abattoirs. Among different categories of samples in slaughtering environment, the prevalence rates of Listeria spp. varied within 17.4-60.0%. The incidence of Listeria spp. in farm F in Daqing or farm H in Shenyang was highly significant, when compared with that in the other farms (P <.01). However, the difference of these between farm F and farm H was not statistically significant (P > .05). The incidence of Listeria spp. in abattoir S in Changchun exhibited highly significant difference, when compared with that in other abattoirs (P <.01). The incidence of Listeria spp. in faeces or silage was highly significant than in the other sample categories in farm environment. There were no statistically significant differences between faeces and silage (P > .05). The incidence of Listeria spp. in beef was highly significant than in the other sample categories expect rinsed water and chopping board in slaughtering environment (P <.01). There were no statistically significant differences among beef, rinsed water, and chopping board (P > .05).
A total of 1025 Listeria strains were isolated from 5214 samples, in which the amount of isolated L. monocytogenes strains was 117 (11.41%) and the amount of isolated L. innocua strains was 907 (88.49%). Merely one atypical L. innocua strain was isolated from farm H, while the remaining L. innocua strains were all typical ones. All L. monocytogenes strains were divided into 1/2a (70.09%) and 1/2c (29.91%) serotypes by multiplex PCR and Listeria antiserum.

Pulsed-field gel electrophoresis
The PFGE analysis with the ApaI enzyme divided the 34 L. monocytogenes strains into 27 pulsotypes (PTs) (Figure 2), and 23 L. innocua strains into 17 PTs ( Figure S1), with diverse clusters. Identical L. monocytogenes strains with same serotypes, PTs and STs from different samples were observed only in the same region. The 11 L. monocytogenes strains (32.4%) exhibited six kinds of identical PTs and STs. Identical L. monocytogenes isolated from cattle and sheep farm and slaughtering environments had regional correlation. The strains which were identical among faeces, silage, and drinking water in farm environment, and the identical strains among slaughtering environment, intestinal tract, and beef were found. The Listeria strains from the same abattoir had genetic correlation in time and space. PFGE showed that there were transmission correlation and genetic association of Listeria in cattle and sheep farm, slaughtering environment and meat chain, which had a typical continuous transmission chain.

Multi-locus sequence typing
A total of 24 different STs were classified among all 100 Listeria spp. isolates, which were further assigned to 23 clonal complexes (CCs): 50 L. monocytogenes strains were divided into 7 STs (6 CCs) and 50 L. innocua were divided into 17 STs (17 CCs). The MLST data revealed that the most prevalent STs of L. monocytogenes were ST9 (9 strains, 36%) and ST91 (8 strains, 32%). The minimum spanning tree of Listeria isolates is shown in Figure 3 and Figure  S2, presenting the genetic correlation among strains of different regions or categories of samples.

Antimicrobial susceptibility test
The antimicrobial resistance analysis of 18 antibiotic agents against the 117 L. monocytogenes isolates is shown in Table 4. All strains were susceptible to two antibiotics, penicillin and imipenem. The most frequent antibiotic resistance was to clindamycin  (71.8%), streptomycin (71.8%), amikacin (65.8%), and cefuroxime (61.5%). Notably, three strains belonged to ECIII were resistant to 10 antibiotics.

Discussion
There have been a few reports on the prevalence and distribution of Listeria spp. in food processing environments and retail food markets in China [27][28][29][30][31]. However, the prevalence and distribution of Listeria spp. from ruminants in farm and slaughtering environments have been rarely reported in China at present. In this study, L. monocytogenes mainly presented in faeces samples (0.2%) in the farm environment, and the incidence of L. monocytogenes was 0.5% (20/4430). A higher prevalence rate of L. monocytogenes in the farm environment was reported in 1738 faeces samples of black beef cattle (6%) in Japan [15], and in 734 faeces samples of dairy cattle (43%) in New York State [32]. The prevalence of L. monocytogenes in raw milk samples produced by dairy cattle was 1.8% (9/441), which was lower than that reported by Kalorey et al. [33], which was 5.1% of the 2060 raw milk samples in Central India. The occurrence of Listeria spp. in raw milk samples may be associated with faeces, silage, and milking hygiene [34].
L. monocytogenes was found in 3.4% of bovine carcass in Turkey [35] and in 2.5% of bovine carcass in Poland [36]. The prevalence of which was lower than that in the present results for bovine carcass (19.6%) in the slaughtering environment. The incidence of L. monocytogenes and Listeria spp. in the farm environment (0.5% and 5.3%, respectively) was much lower than that in the slaughtering environment (9.4% and 46.0%, respectively). The incidence of L. innocua in cattle and sheep farm and slaughtering environments is more common and significantly higher (9.7%, 508/5214) than that of L. monocytogenes (1.8%, 94/5214).
The 1/2a, 1/2b, 1/2c, and 4b are the dominant serotypes for food strains in China, in which serotype 1/ 2a, 1/2b, and 4b accounted for most of the human clinical cases [3,37]. In addition, 1/2c was also found in human listeriosis in China [38]. In our study, merely strains of 1/2a and 1/2c serotypes were isolated in farm and slaughtering environments of cattle and sheep. This is consistent with the results, in which only the 1/2a and 1/2c serotypes were isolated by Zhu et al., Takashi et al. and Wieczorek et al. [36,39,40].
It is noteworthy that 1-3 Listeria strains of different serotypes in the same sample were isolated during the isolation and identification process of samples. In general, strains with one serotype were encountered in most samples. However, strains with 2-3 serotypes of Listeria in some samples were isolated in this study. Different species of the Listeria genus or different serotypes of one species in one sample should be considered as discrepant strains. It is suggested that at least five suspicious colonies should be selected from each sample in the process of isolating strains, when possible, in order to avoid the possibility of missing detection, and provide more comprehensive and objective data for large-scale laboratory and epidemiological investigation, instead of selecting only one viable colony for further identification.
The simple data for the incidence of Listeria spp. in the diagrams limitedly revealed the transmission characteristics. The objective transmission characteristics must be in conjunction with the epidemiologic, laboratory, and environmental investigations. Listeria has an obvious regional epidemic in the cattle and sheep farm environment, in which the prevalence of Listeria in cattle and sheep farm environments is very low and L. monocytogenes was isolated only from Shenyang H farm and Dunhua L farm. It was surprising to note that the samples collected in farm B, D, E, J, K, and N in the breeding environment were negative for L. monocytogenes and other Listeria spp. In Inner Mongolia (Farm E) and Zhenlai (Farm D and farm N), the land was full of desertified soil, and   (10) 117 (100) 0 (0) 0 (0) the local annual rainfall was pretty less than the east slaughtering environments in China, and provides a scientific basis for the prevention and control of listeriosis in humans and animals.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This study was funded by the National Key Research and Development Program of China (No. 2018YFD0500504).