Prevalence, Genetic Heterogeneity, and Antibiotic Resistance Profile of Listeria spp. and Listeria monocytogenes at Farm Level: A Highlight of ERIC- and BOX-PCR to Reveal Genetic Diversity

This study aimed to identify Listeria spp. and L. monocytogenes, characterize the isolates, and determine the antibiotic resistance profiles of the isolates Listeria spp. and L. monocytogenes in fresh produce, fertilizer, and environmental samples from vegetable farms (organic and conventional farms). A total of 386 samples (vegetables, soil, water, and fertilizer with manure) were examined. The identification of bacterial isolates was performed using PCR and characterized using ERIC-PCR and BOX-PCR. The discriminating power of the typing method was analyzed using Simpson's Index of Diversity. Thirty-four (n=34) Listeria isolates were subjected to antimicrobial susceptibility test using the disc-diffusion technique. The PCR analysis revealed that Listeria spp. were present in 7.51% (29/386) of all the samples (vegetable, soil, fertilizer, and water). None of the samples examined were positive for the presence of L. monocytogenes. Percentages of 100% (15/15) and 73.30% (11/15) of the Listeria spp. isolated from vegetables, fertilizer, and soil from organic farm B had indistinguishable DNA fingerprints by using ERIC-PCR and BOX-PCR, respectively. Listeria spp. isolated from 86 samples of vegetable, fertilizer, and environment of organic farm A and conventional farm C had distinct DNA fingerprints. Simpson's Index of Diversity, D, of ERIC-PCR and BOX-PCR is 0.604 and 0.888, respectively. Antibiotic susceptibility test revealed that most of the Listeria spp. in this study were found to be resistant to ampicillin, rifampin, penicillin G, tetracycline, clindamycin, cephalothin, and ceftriaxone. The isolates had MAR index ranging between 0.31 and 0.85. In conclusion, hygienic measures at farm level are crucial to the reduction of Listeria transmission along the food chain.


Introduction
Listeria is a gram-positive, rod-shaped, and non-sporeforming bacterium [1]. Genus of Listeria is classified into 17 species including Listeria monocytogenes that is a common causative agent of human foodborne infection, listeriosis [2]. Listeriosis is usually treated with antibiotic therapy involving the use of penicillin, ampicillin, rifampin, gentamicin, tetracycline, erythromycin, chloramphenicol, or trimethoprim with sulfamethoxazole alone or in combination [3,4]. Previous researches have shown that Listeria spp. may be resistant to several antibiotics such as clindamycin, daptomycin, oxacillin, tetracycline, and nalidixic acid [5,6]. Therefore, it is important to monitor the antibiotic susceptibility of Listeria spp. and L. monocytogenes to ensure the effectiveness of listeriosis treatment.
Repetitive sequence-based PCR (Rep-PCR) is a DNA amplification technique for bacterial genomic fingerprinting by using repetitive DNA elements present within bacterial genome. There are four main types of repetitive sequences used for molecular typing which include enterobacterial repetitive intergenic consensus (ERIC) sequence, BOX elements, repetitive extragenic palindromic sequences (REP elements), and (GTG) 5 [7]. Utilization interspersed repetitive 2 BioMed Research International sequence-based tools can be used in bacterial fingerprinting since the distance between each of the sequences varies among strains [8] and have been used to type wide range of gram-negative and several gram-positive bacteria [7]. ERIC-PCR has been used for intraspecies fingerprinting of Bacillus anthracis and Bacillus cereus [9], Enterobacter sakazakii [10], Lactobacillus [11], Listeria monocytogenes [12], and Salmonella Enteritidis [13,14]. Meanwhile, BOX-PCR has been well used in typing of Escherichia coli [15][16][17], Bifidobacterium [7], Streptomyces [18], Aeromonas spp. [19], Burkholderia pseudomallei [20], and Bacillus anthracis [21]. Nonetheless, ERIC sequence-based PCR (ERIC-PCR) and BOX-PCR were used in this study as they are rapid subtyping methods and have high discrimination power [2,22,23]. According to Strawn et al. [24], agricultural practices (irrigation with contaminated water, fertilization with contaminated manure and contaminated soil) could increase the risk of bacterial contamination of vegetables. Therefore, this study was carried out to assess the contamination levels of Listeria spp. and L. monocytogenes in vegetables, fertilizer, and environmental samples (soil and water) at farm level practicing organic and conventional farming in Sarawak, to obtain information on the genetic diversity of the Listeria spp. isolates using Repetitive Intergenic Consensus Polymerase Chain Reaction (ERIC-PCR) and BOX-PCR. Further, the study aimed to compare the effectiveness of ERIC-PCR and BOX-PCR for genetic diversity of Listeria spp. and determine antibiotic resistance profiles of the Listeria spp. isolates.

Sampling Sites and Sample Collection.
A total of 206 vegetable samples, 60 fertilizer samples, 60 soil samples, and 60 water samples were collected from two organic farms (organic farm A and organic farm B) and one conventional farm (conventional farm C) in Kuching, Sarawak. As shown in Table 1, the organic farms practice crop rotations, application of composted chicken waste and plant waste as fertilizer, mechanical methods to control weeds, and restricted use of pesticides. The conventional farm also practices the use of composted animal manures and plant waste as fertilizer; however, it does not practice crop rotation and synthetic chemical pesticides are applied to the produce biweekly.
The vegetable samples and soil samples were collected from the fields while the fertilizer samples were collected from the fertilizer storage places. Water samples were collected from the respective water sources (pond or rainwater). All samples were kept in ice box and transported to the Molecular Microbiology Laboratory at Universiti Malaysia Sarawak.

Sample
Processing and Listeria Enrichment. Sample processing was done as described in Ozbey et al. [25]. First, 25 g (or 25 ml) of each sample (soil, fertilizer, and water) was weighed or measured and then transferred into conical flasks containing 225 ml of Listeria Enrichment Broth (LEB) (Oxoid, United States). After that, the flasks were incubated at 30 ∘ C for 48 hr. Vegetable samples were weighed (25 g) and cut into pieces before being transferred into the conical flasks containing 225 ml of LEB followed by incubation at 30 ∘ C for 48 hr.

Enumeration of Listeria spp.
The enriched bacterial cultures (100 l) from the samples were transferred and spread on PALCAM agar (Merck, Germany). The PALCAM agar plates were incubated at 37 ∘ C for 48 hr. Listeria spp. colonies appeared to be greyish-green or black in color and surrounded by black halo on PALCAM agar [26]. These colonies were observed, counted, and recorded.

DNA Extraction.
Prior to amplification, genomic DNA was extracted using GF-1 Nucleic Acid Extraction Kits (Vivantis, United States) according to manufacturer's guide. DNA template was further subjected to PCR-based analysis.

Identification of Listeria spp. and Listeria monocytogenes
2.5.1. DNA Extraction. Presumptive Listeria spp. colonies were selected from PALCAM agar and subjected to DNA extraction using GF-1 Nucleic Acid Extraction Kits (Vivantis, United States) according to the manufacturer's guide. DNA template was further subjected to PCR-based analysis.

Polymerase Chain Reaction (PCR) for Listeria spp.
PCR detection of Listeria spp. was carried out as described by Jeyaletchumi et al. [27] and Wong et al. [28] with slight modification in the concentration of reagents. The primer pairs used for the detection of Listeria spp. (genus specific) were 5 -CTC CAT AAA GGT GAT CCT-3 and 5 -CAG CAG CCG CGG TAA TAC-3 . These primers were designed to amplify a 938 bp region in the 16S rRNA gene. To prepare 25 l of PCR mixture, 0.2 M of forward primer, 0.2 M of reverse primer, 5.5 l of DNA template, 1.5 l of 10× Taq PCR buffer, 0.2 l dNTP, 1.5 mM MgCl 2 , and 1.5 unit of Taq DNA polymerase were mixed together. Lastly, the PCR products were separated on 1% agarose gel with 100 kb DNA ladder for 75 min, stained with ethidium bromide, and viewed under a UV transilluminator (Maestrogen, Taiwan).

Genetic Diversity Analysis Using ERIC and BOX-PCR.
The ERIC-PCR condition for this method was in accordance with Indrawattana et al. [30] and Laciar et al. [31]. Meanwhile, the BOX-PCR condition followed Jamali and Thong [32] and Versalovic et al. [8] with slight modifications on the reagent concentration and reaction condition. In ERIC-PCR, the primer pairs used were 5 -ATGTAAGCTCCTGGGGATTCAC-3 and 5 -AAG-TAAGTGACTGGGGTGAGCG-3 . To prepare 25 l of PCR mixture, 1.0 M of forward primer, 1.0 M of reverse primer, 3.0 l of DNA template, 5.0 l of 5×Taq PCR buffer, 0.2 mM dNTP, 2.0 mM MgCl 2 , and 1.0 unit of Taq DNA polymerase were mixed together. The PCR reaction was carried out according to the condition in Table 2. In BOX-PCR, the primer used was BOX A1R (5 -CTACGG CAA GGC GAC GCT GAC G-3 ). To prepare 25 l of PCR mixture, 400 M of each dNTPs, 1×PCR buffer, 3 mM MgCl 2 , 4 M of primer, and 2.5 U Taq DNA polymerase (Promega) were mixed together. The PCR reaction was carried out according to the condition in Table 3. The PCR products from ERIC-and BOX-PCR were separated in 2% agarose gel with 100 kb DNA ladder for 90 min. Then, the gel was stained with ethidium bromide and viewed under a UV transilluminator (Maestrogen, Taiwan). The DNA band patterns were analyzed and a dendrogram was generated for the Listeria isolates by using BioNumerics 7.5 software program (Applied Maths, Sint-Martens-Latem, Belgium) using Dice coefficient and the unweighted pair group method (UPGMA) [33]. Simpson's Index of Diversity, D, was also calculated.
The discriminating power of this typing method was calculated by using Simpson's Index of Diversity, D [34]. The higher the discriminatory index, the greater the effectiveness of a particular fingerprinting method to discriminate different strains [35]. This index was given by the following equation: "N" denotes the total number of strains in the sample population, "s" denotes the total number of types described, "n j " denotes the number of strains belonging to the jth type.

Antibiotic Susceptibility Test.
Antibiotic susceptibility of isolated Listeria spp. was carried out with the disc-diffusion method by Chen et al. [5] and Morobe et al. [36] with a slight modification. The antibiotic discs (Oxoid, the United States) used were ampicillin (10 g), cephalothin (30 g), chloramphenicol (30 g), clindamycin (2 g), erythromycin (15 g), gentamycin (10 g), penicillin G (10 g), rifampin (5 g), streptomycin (10 g), tetracycline (30 g), sulfamethoxazole/trimethoprim (23.75 g/1.25 g), novobiocin (30 g), nitrofurantoin (10 g), and ceftriaxone. First, the overnight culture grown in Mueller-Hinton broth was spread uniformly onto the Mueller-Hinton agar plate. Antibiotic discs were then placed onto the surface of each plate (4 antibiotics/Petri dish) using antibiotic-disc dispenser (Oxoid, United States). After incubation at 37 ∘ C for 24 hr, the diameter of growth inhibition zone surrounding each disc was measured and interpreted according to the CLSI (Clinical and Laboratory Interpretive Criteria (nearest whole mm) of a particular antibiotic of CLSI. The CLSI criteria for staphylococci were referred to in this study because interpretative criteria for Listeria are not available from CLSI with the exception of susceptibility breakpoints for ampicillin and penicillin. Multiple antibiotic resistance (MAR) index of an isolate was calculated as defined by Krumperman [37]: "a" denotes number of antibiotics to which the particular isolate was resistant and "b" denotes number of antibiotics to which the particular isolate was exposed.

Prevalence of Listeria spp. Based on PCR Analysis.
Analysis using PCR assay revealed that Listeria spp. were present in 7.51% (29/386) of all the samples (vegetable, soil, fertilizer, and water) collected. It was present in 9.10% (6/66), 8.13% (13/160), and 6.25% (10/160) of the samples collected from organic farm A, organic farm B, and conventional farm C, respectively. The prevalence of Listeria spp. from all the samples was shown in Table 4. The gel picture for PCR amplification of 16S rRNA gene of Listeria spp. was shown in Figure 1.

Prevalence of Listeria monocytogenes Based on PCR
Analysis. PCR detection of L. monocytogenes in the samples was carried out by using primer pairs which amplified 730 bp region of the listeriolysin (hlyA) gene. However, L. monocytogenes was absent in all the samples analyzed.

Genetic Diversity of Listeria Isolates Using ERIC-PCR.
The electrophoretic profile of DNA fragments obtained after ERIC-PCR amplification yielded 1-5 bands with size approximately 120 bp to 1450 bp. A common band with molecular size of approximately 520 bp was observed in the electrophoretic profile from most of the isolates. Based on the ERIC-PCR dendrogram shown in Figure 2, the Listeria spp. isolated from organic farm A, organic farm B, and conventional farm C were genetically diverse and heterogeneous as they were not classified into specific cluster by either sampling area or the type of samples. ERIC-PCR analysis produced 11 different DNA fingerprint profiles. Simpson's Index of Diversity, D, was calculated for ERIC-PCR based on Hunter and Gaston [34]. The D value of this technique was calculated to be 0.604.

Genetic Diversity of Listeria Isolates Using BOX-PCR.
The electrophoretic band pattern of BOX-PCR amplification yielded 2-13 bands with size approximately 120 bp to 1550 bp. Based on the BOX-PCR dendrogram shown in Figure 3, Listeria spp. isolated from all the three farms were not classified according to the types of sample or sampling area. Therefore, these Listeria spp. isolates were genetically diverse and heterogeneous. BOX-PCR analysis produced 14 different fingerprint profiles. Simpson's Index of Diversity, D, was calculated for BOX-PCR based on Hunter and Gaston [34]. The D value of this technique was calculated to be 0.888.
Listeria spp. were isolated from 11 vegetable samples, 2 (Chinese mustard and cucumber) from organic farm A and 9 (Chinese cabbage, romaine/cos lettuce, and Chinese white cabbage) from organic farm B. Antibiotic resistance graph of vegetable samples from the three farms is shown in Figure 4.
All Listeria spp. isolated from vegetable samples from organic farms A and B were resistant to penicillin G, tetracycline, and clindamycin. For soil samples, Listeria spp. were isolated from 6 soil samples, 4 from organic farm A and 2 from organic farm B. Antibiotic resistance graph of soil samples from the three farms is shown in Figure 5.
In this study, all Listeria spp. isolated from soil samples from organic farms A and B were resistant to clindamycin, cephalothin, and ceftriaxone. Two Listeria spp., 4 Listeria spp., and 2 Listeria spp. were isolated from the fertilizer        denotes Listeria spp. isolated from vegetables, soil, fertilizer, and water from organic farm A, organic farm B, and conventional farm C, "(b)" denotes Listeria spp. isolated from soil and water from organic farm A and conventional farm C, "(c)" denotes Listeria spp. isolated from vegetable and water from organic farm B and conventional farm C, and "(d)" denotes Listeria spp. isolated from soil from organic farm A. samples collected from organic farm A, organic farm B, and conventional farm C, respectively. Antibiotic resistance graph of fertilizer samples from the three farms is shown in Figure 6.
The results revealed that all Listeria spp. isolated from fertilizer samples were resistant to clindamycin and ceftriaxone. Conventional farm C was the only farm where the water samples were detected with Listeria spp., with 9 Listeria spp. isolated. Antibiotic resistance graph of water samples from the 3 farms is shown in Figure 5. All (100%) (9/9) of the Listeria spp. were resistant to penicillin G. MAR index defined by Krumperman [37] was evaluated for all the isolates. In this study, Listeria spp. isolates demonstrated MAR; they were resistant to at least four of the thirteen antibiotics tested. The MAR indexes for all the isolates are recorded in Table 5 Table 4, a total of five vegetables from organic farms A and B had high concentration of Listeria spp. in the vegetables (ranging from 9.50 × 10 2 to 2.10 × 10 5 CFU/g). However, none of vegetables from conventional farm C was positive. According to the Public Health England [38], samples consisting of more than 100 CFU/g of Listeria spp. are considered unsatisfactory and investigation is required. Therefore, the vegetable samples collected from organic farms A and B were considered unsatisfactory and this represents the risk of contracting listeriosis associated with fresh produce consumption. Listeria spp. were present in 16.70% (2/12) and 8.30% (2/24) Figure 3: Dendrogram constructed for BOX-PCR of Listeria spp. in vegetable, fertilizer, and environmental samples collected from organic farm A, organic farm B, and conventional farm C. "A" denotes organic farm A, "B" denotes organic farm B, "C" denotes organic farm C, "(a)" denotes Listeria spp. isolated from vegetables, soil, and fertilizer from organic farm A and organic farm B, "(b)" denotes Listeria spp. isolated from vegetable, soil, and water from organic farm A and conventional farm C, "(c)" denotes Listeria spp. isolated from vegetables, fertilizer, and water from organic farm A, organic farm B, and conventional farm C, and "(d)" denotes Listeria spp. isolated from water from conventional farm C. of soil from organic farms A and B. Listeria spp. are widely distributed in the environment including soil, vegetation, surface water, sewage, animal feeds, farm environments, and food-processing environments [39]. According to Vackachan et al. [40], contaminated fertilizer and humidity of the soil may increase the risk of soil contamination. Therefore, measures should be taken for the use of contaminated soil to reduce the presence of the bacteria. The fertilizer used by the three farms in the present study was animal waste compost (chicken litter) and plant waste. Such fertilizers are usually used as they are of low cost, organic, and containing notable amount of nutrients. Normally, composting of animal waste can inactivate large populations of human pathogens but improper composting or cross-contamination results in the high survival rate of these pathogens. Improper composting may also result in the regrowth of the pathogens in the finished compost products under a range of favorable conditions [41]. Listeria spp. were absent in the water samples from organic farms A and B. However, Listeria spp. were detected in 33.30% (8/24) of the water from conventional farm C. According to Galvez et al. [42] and Chitarra et al. [43], pathogenic bacteria such as Salmonella, pathogenic E. coli, and L. monocytogenes can be found in irrigation water for fresh produce. These pathogenic bacteria can internalize crops through the roots and survive in them. This study also revealed no presence of L. monocytogenes in all the samples (vegetable, soil, water, and fertilizer) from all farms which could indicate lower potential of disease burden as the species is commonly causing human infections [2].

Genetic Heterogeneity of Listeria spp. Based on ERICand BOX-PCR
Analysis. The findings from both ERIC-PCR and BOX-PCR analysis in the present study showed that the Listeria spp. isolates were not grouped together based on the types of samples and the source of isolation. They were not classified into specific cluster by either sampling area or the type of samples. In the present study, the Listeria spp. isolated from organic farm A, organic farm B, and conventional farm C were genetically diverse and heterogeneous. The heterogeneity was expected as the isolates were collected from different types of sample (vegetable, soil, fertilizer, and water) and sampling locations (organic farm A, organic farm B, and conventional farm C). In this study, Simpson's Index of Diversity, D, value for ERIC-and BOX-PCR was 0.604 and 0.888, respectively. According to Kqueen et al. [35], the higher the discriminatory index, the greater the effectiveness of a particular fingerprinting method to discriminate different strains. Thus, it was shown that BOX-PCR had greater discriminatory power than ERIC-PCR for fingerprinting Listeria spp. isolates of this study. In comparison, the discriminatory power for both BOX-PCR and ERIC-PCR analysis was lower as compared to the finding by Jersek et al. [44] which revealed ERIC-PCR was suitable for the typing of L. monocytogenes isolates as the index of discrimination was high (0.98). Another study conducted by Jamali and Thong [32] reported that the discrimination indexes for REP-PCR, BOX-PCR, RAPD, and PFGE were 0.992, 0.998, 1, and 0.916, respectively. They suggested that different subtyping methods often give different discriminatory powers. Therefore, it is necessary to use more than one subtyping approach to provide a more accurate description of the genetic diversity of microorganisms in the study. On the other hand, other fingerprinting tools such as REP-PCR and (GTG) 5 are well employed in bacterial typing [8] which can be tested in further study.

Antibiotic Susceptibility of the Listeria Isolates.
This present study revealed that 97.06% (33/34) of Listeria spp. isolated from vegetables, soil, fertilizer, and water from organic farm A, organic farm B, and conventional farm C were resistant to clindamycin. Chen et al. [5] found that all Listeria spp. isolates in catfish fillets and processing environment were resistant to clindamycin. Gamboa-Marin et al. [45] also revealed that L. monocytogenes, Listeria spp., and L. ivanovii from swine processing facilities in Colombia had major resistance and intermediate susceptibility to clindamycin. In the present study, Listeria spp. from the three farms showed the lowest resistance against gentamicin. This is comparable to a study by Li et al. [46] which revealed gentamicin exhibited good activity against Listeria spp. from processed bison in the USA. This study showed that all the Listeria spp. isolates were resistant to more than one antibiotic and therefore demonstrated MAR. According to Krumperman [37], MAR index "Amp" denotes ampicillin, "Cf" denotes cephalothin, "Cro" denotes chloramphenicol, "Sli" denotes clindamycin, "Ery" denotes erythromycin, "Gen" denotes gentamycin, "PenG" denotes penicillin G, "Rif" denotes rifampin, "Str" denotes streptomycin, "Tet" denotes tetracycline, "Sxt" denotes sulfamethoxazole/trimethoprim, "Nor" denotes nitrofurantoin, and Cf denotes ceftriaxone.
value lower than 0.20 indicates that the organism originated from a lower risk source in which the antibiotics are seldom or never used. MAR index value higher than 0.20 indicates that they are originated from a higher risk source which is greatly exposed to antibiotics. A study conducted by Jamali et al. [47] reported that 8.40% of Listeria spp. isolated from raw milk in Iran showed multiple antibiotic resistances. In the present study, all Listeria spp. had MAR index higher than 0.20, suggesting that the Listeria spp. isolates from the three farms were originated from a higher risk source in which they had been constantly exposed to antibiotics. MAR of Listeria spp. in vegetables, soil, and irrigation water could be a result of the usage of animal waste as fertilizer which might contain antibiotics used to prevent or treat animal diseases and promote animal growth. Hu et al. [48] conducted a study on the migration of antibiotics from manure to soil and from soil to vegetables and groundwater. In the study, they applied manure containing antibiotics to organic vegetable bases and revealed that the soil, vegetables, and water were detected with antibiotic residues. Some antibiotics are still biologically active despite being in environment. This eventually can initiate development of antibiotics resistance genes in microorganisms. The findings in the present study showed that Listeria spp. isolated from the samples from all the three farms were multiresistant to the antibiotics tested. The presence of Listeria spp. that were resistant to antibiotics commonly used to treat human listeriosis (including ampicillin, chloramphenicol, erythromycin, gentamicin, penicillin G, rifampin, tetracycline, and sulfamethoxazole/trimethoprim) raises the possibility of future acquisition of resistance by L. monocytogenes and Listeria spp.

Conclusion
The findings of this study show current occurrence of Listeria spp. at farm level of selected organic and conventional vegetable farms in Sarawak, Malaysia. Personal hygiene and good manufacturing practice at farm level are essential for prevention of the transmission of the organism along the food chain. Based on the genotyping analysis, all Listeria spp. isolates were heterogenous. Nonetheless, BOX-PCR was shown to be better in discriminatory power than ERIC-PCR and can be utilized in Listeria typing. This study also presented high occurrence of multiple antibiotic resistant strains in the fresh produce and farm environment which could be an indicator of the excessive use of antibiotics in the agriculture field.

Data Availability
Data generated in this study are included in this article.

Conflicts of Interest
The authors declare that they have no conflicts of interest.

Authors' Contributions
Lesley Maurice Bilung conceived the study and was involved in the design and coordination of the study and manuscript drafting. Lai Sin Chai and Ahmad Syatir Tahar were involved in the manuscript drafting and editing. Lai Sin Chai was involved in the sampling collection, processing, and data analysis. Kasing Apun and Lesley Maurice Bilung were involved in the final editing of the manuscript. All authors read and approved the final manuscript.