Conditions of In Vitro Biofilm Formation by Serogroups of Listeria monocytogenes Isolated from Hass Avocados Sold at Markets in Mexico

Listeria monocytogenes is an important pathogen that has been implicated in foodborne illnesses and the recall of products such as fruit and vegetables. This study determines the prevalence of virulence-associated genes and serogroups and evaluates the effects of different growth media and environmental conditions on biofilm formation by L. monocytogenes. Eighteen L. monocytogenes isolates from Hass avocados sold at markets in Guadalajara, Mexico, were characterized by virulence-associated genes and serogroup detection with PCR. All isolates harbored 88.8% actA, 88.8% plcA, 83.3% mpl, 77.7% inlB, 77.7% hly, 66.6% prfA, 55.5% plcB, and 33.3% inlA. The results showed that 38.8% of isolates harbored virulence genes belonging to Listeria pathogenicity island 1 (LIPI-1). PCR revealed that the most prevalent serogroup was serogroup III (1/2b, 3b, and 7 (n = 18, 66.65%)), followed by serogroup IV (4b, 4d–4e (n = 5, 27.7%)) and serogroup I (1/2a–3a (n = 1, 5.5%)). The assessment of the ability to develop biofilms using a crystal violet staining method revealed that L. monocytogenes responded to supplement medium TSBA, 1/10 diluted TSBA, and TSB in comparison with 1/10 diluted TSB (p < 0.05) on polystyrene at 240 h (p < 0.05). In particular, the biofilm formation by L. monocytogenes (7.78 ± 0.03–8.82 ± 0.03 log10 CFU/cm2) was significantly different in terms of TSBA on polypropylene type B (PP) (p < 0.05). In addition, visualization by epifluorescence microscopy, scanning electron microscopy (SEM), and treatment (DNase I and proteinase K) revealed the metabolically active cells and extracellular polymeric substances of biofilms on PP. L. monocytogenes has the ability to develop biofilms that harbor virulence-associated genes, which represent a serious threat to human health and food safety.


Introduction
Listeria monocytogenes is a bacterium ubiquitous in the environment and is the causative agent of listeriosis, leading to septicemia, encephalitis, endocarditis, meningitis, abortions, and stillbirths [1,2]. The severity of the pathology is associated with several at-risk groups, such as those with a weak immune system, adults >65 years, pregnant women, and newborn babies [3,4]. Listeria, Salmonella, and Shiga toxin-producing Escherichia coli (STEC) are

L. monocytogenes Isolates
In total, 18 L. monocytogenes isolates were evaluated in this study, which were previously obtained from Hass avocados sold at retail markets in Guadalajara, Mexico [12]. Stocks of L. monocytogenes were stored using the protocol described by Avila-Novoa et al. [24]. Working cultures of L. monocytogenes were maintained in tryptic soy broth (TSB; Becton Dickinson Bioxon, Le Pont de Claix, France) with 0.6% yeast extract (TSBYE) (Sigma-Aldrich, St. Louis, MO, USA) and confirmed by PCR.

PCR-Serogroup Analysis and Virulence Genes
Bacterial DNA was extracted from 24 h/30 • C cultures in TBSYE using the protocol described by Avila-Novoa et al. [24]. All L. monocytogenes strains were investigated for the detection of prfA, plcA, hly, mpl, actA, plcB, inlA, and inlB genes by PCR using the protocol by Montero et al. [25]. The amplification conditions used were as follows: 5 min at 94 • C; 35 cycles of 40 s at 94 • C, 75 s at different temperatures for different genes, and 75 s at 72 • C; followed by a final extension of 10 min at 72 • C (Table 1). Alternatively, the serogroups of L. monocytogenes isolates were determined by Doumith et al. [14] for the detection of the four PCR serogroups of L. monocytogenes, as described: I (1/2a and 3a), II (1/2c and 3c), III (1/2b, 3b, and 7), and IV (4b, 4d, and 4e). L. monocytogenes ATCC 19111 was used as the positive control. Table 1. Primers used in this study.

Epifluorescence Microscopy and Scanning Electron Microscopy (SEM)
After incubation at 30 • C for 240 h, the PP coupons were removed and processed using the protocol described by Avila-Novoa et al. [24]. Alternatively, SEM was performed using the methodology described by Borucki et al. [22] and Fratesi et al. [28]. Biofilms were observed using a TESCAN Mira3 LMU scanning electron microscope (Brno-Kohoutovice; Czech Republic). L. monocytogenes ATCC 19111 was used as the positive control and a PP coupon without inoculum was included in all assays.
2.6. Determination of Biofilm Components 2.6.1. Matrix Characterization Biofilm detachment assays were carried out as described by Fredheim et al. [29] and Avila-Novoa et al. [30]. Previously mature biofilms cultivated in TSB at 30 • C for 240 h in polystyrene microtiter plates were washed with 0.9% NaCl. The major components of L. monocytogenes biofilms were treated with (i) proteinase K (Promega, Madison, WI, USA) and (ii) 0.5 mg/mL DNase I (Roche, Mannheim, Germany) using the protocol described by Avila-Novoa et al. [24].

Phenotype Analysis of Biofilm Production
All L. monocytogenes strains were characterized phenotypically by culture on Congo red agar (CRA) plates according to the protocol described by Arciola et al. [31], with some modifications. Briefly, CRA was prepared with TSBYE, 30 g/L of glucose (Sigma-Aldrich, St. Louis, MO, USA), 15 g/L of bacteriological agar (Becton Dickinson Bioxon, Le Pont de Claix, France), and 40 mg/L of Congo red (Sigma-Aldrich, Steinheim, Germany). According to the macroscopic characteristics developed by L. monocytogenes in the CRA, they were interpreted as (a) biofilm producers: black colonies with a dry filamentous, or (b) nonproducers: smooth pink colonies.

Statistical Analysis
All the experiments were performed in triplicate, and the data were evaluated using analysis of variance (ANOVA), followed by a Fisher's least significant difference (LDS) test, using the Statgraphics Centurion XVI software program (StatPoint Technologies, Inc., Warrenton, VA, USA).
Furthermore, L. monocytogenes biofilms on PP were observed as cells irreversibly attached and microcolonies of metabolically active cells examined by an epifluorescence microscope (Figure 1). Mono-species L. monocytogenes showed that cells were linked and embedded in dense EPS (Figure 2).

Quantification and Components of the Matrix Biofilm Formation
DNase I and proteinase K were used to determine the components that make up the matrix of the L. monocytogenes biofilm; we found eDNA (9.44-47.69%) and protein (16.90-46.82%) (p < 0.05). In turn, there was a difference between each of the L. monocytogenes strains that made up the groups (p < 0.05) ( Table 5). It was determined that 100% of L. monocytogenes strains were biofilm producers on CRA (Figure 3).

Discussion
In recent years, the consumption of fruit and vegetables has increased; consequently, there have been L. monocytogenes outbreaks associated with foods such as enoki mushrooms, cantaloupe, frozen vegetables, and packaged salads [7]. In addition, L. monocyto-

Discussion
In recent years, the consumption of fruit and vegetables has increased; consequently, there have been L. monocytogenes outbreaks associated with foods such as enoki mushrooms, cantaloupe, frozen vegetables, and packaged salads [7]. In addition, L. monocytogenes has the ability to develop biofilms in a food environment [21]. Sixty percent of outbreaks are caused by biofilm-associated infections by pathogens and antimicrobial resistance, which have a significant impact on the food industry [32,33].
In our study, the analysis of serogroups by PCR revealed that 66.6% of the L. monocytogenes isolates from Hass avocado were identified as serogroup III (1/2b, 3b, and 7), 27.7% as IV (4b, 4d, and 4e), and 5.5% as serogroup I (1/2a and 3a). Other investigators have demonstrated a higher prevalence of some serotypes (1/2b, 3b, 1/2a, 1/2c, and 4b) in food or processing environments [20,25,34,35]. According to the CDC (2011), serotypes 1/2a and 1/2b were associated with 147 cantaloupe-associated listeriosis cases in 28 U.S. states in 2011 [36]. This could be relevant for food safety and public health, as it shows the diversity of serogroups of L. monocytogenes in Hass avocados. It is essential to confirm the serotype with the incorporation of techniques used for subtyping L. monocytogenes such as random amplification of polymorphic DNA-polymerase chain reaction (RAPD-PCR), repetitive extragenic palindromes-PCR (REP-PCR), and pulsed field gel electrophoresis (PFGE), which allow for the sources and mechanisms of contamination during food processing to be determined and fresh food produce to be commercialized, at the same time identifying serotypes related to disease outbreaks. Likewise, Roche et al. [37] argued that the virulence of L. monocytogenes is related to its serotype and lineage (I-IV).
Additionally, 38.8% of L. monocytogenes isolates had virulence-associated genes (LIPI-1) ( Table 2). Montero et al. [25] and Vilchis-Rangel et al. [38] reported similar percentages (20-40%) for LIPI-1 in L. monocytogenes isolates from orange juice, vegetables, and frozen vegetables. Moreover, several studies have reported LIPI-1 in 100% of L. monocytogenes isolates from duck, beef, pork, chicken, vegetables, fried rice, fish, goat meat, pasteurized milk, and yogurt [13,39,40]. The wide range of virulence-associated L. monocytogenes genes found in this study could be associated with food type, sources, mechanisms of contamination during food processing and marketing, epidemiological factors, countries, geographical differences, the genetic diversity of strains, and methodologies for the serotyping of L. monocytogenes in food. Overall, 55.5% of L. monocytogenes isolates carried genes corresponding to LIPI-1. However, 5.5% of L. monocytogenes did not carry LIPI-1 in this study (Table 2). These results agree with those of Montero et al. [25] and Vilchis-Rangel et al. [38], who reported that virulence-associated genes (LIPI-1) were not detected in L. monocytogenes isolates collected from frozen vegetables, orange juice, and fresh vegetables. Montero et al. [25] argued that the absence of one of these genes did not imply that a strain was not virulent. L. monocytogenes strains are known to differ in virulence [20]. For example, multiple distinct genetic mechanisms (mutations in the inlA and prfA genes) could be responsible for natural virulence attenuation in L. monocytogenes [41].
In this study, L. monocytogenes isolates from Hass avocados had a high capacity for biofilm formation in supplement medium (TSBA, 1/10 diluted TSBA) and TSB in comparison with 1/10 diluted TSB (p < 0.05) at 48-240 h. Similarly, other studies have reported a high capacity for biofilm formation on polystyrene by L. monocytogenes isolated from cheese-processing plants, cheeses, and milk samples [26,27,44]. In contrast to our results, Kadam et al. [45] demonstrated that L. monocytogenes biofilm production was higher in a minimal medium compared to a nutrient-rich medium. This could be due to the organic matter or food residues that favored preconditioning and irreversible adhesion for the formation of biofilm. In fact, several studies have reported that the adherence and formation of biofilm by L. monocytogenes on surfaces is affected by factors such as temperature, pH, medium, incubation time, strain, serotype, and certain fatty acids (iso-C 14:0 , anteiso-C 15:0 , and iso-C 16:0 ) [46,47].
Accordingly, the development of the biofilm of L. monocytogenes was favored at 240 h (p < 0.05) in this study. The difference was associated with several factors. First, consider that in 48 h at 30 • C, it is possible that L. monocytogenes isolates are developing an irreversible adhesion phase or the formation of microcolonies is occurring, compared to at 240 h, where the cell density has increased and bacteria would be in the stage of maturation or dispersal. Researchers argue that the biofilm formation capacity of L. monocytogenes increases after 72 h and the biofilm maturation stage occurs at 240 h [48,49]. Several studies have argued for a possible correlation between the serotype or phylogenetic division and biofilm-forming ability [22,50,51]. In addition, a CV staining assay is used for biofilm biomass quantification, but does not reveal the number of viable cells within the biofilm matrix [52,53].
In this study, L. monocytogenes in mono-species biofilms (Lm-303, Lm-320, Lm-352, and Lm-356) had a higher cellular density in TSBA (7.78 ± 0.03-8.82 ± 0.03 log 10 CFU cm −2 ; p < 0.05) in comparison with TSB, 1/10 diluted TSB, and 1/10 diluted TSBA at 240 h onto polypropylene type B (Table 4). In particular, Lm-133 had a lower cellular density in TSB, 1/10 diluted TSB, and TSBA in comparison with Lm-320 (p < 0.05). This may be associated with the biofilm formation capacity of this particular serotype, even though it is serogroup III (1/2b, 3b, and 7), and with intrinsic factors like the nutrient level in the culture medium. In addition, this agrees with previous studies demonstrating the biofilm-forming ability of L. monocytogenes on materials encountered in the food-processing industry, such as stainless steel, aluminum, polycarbonate, polypropylene, polyurethane, polyvinylchloride, silicone rubber, natural white rubber, PETG, PTFE, Lexan, Nitryl rubber, and glass [17,54].
However, Pan et al. [55] and Nilsson et al. [56] revealed that the serotype or strain origin can have an effect on the biofilm-forming behavior of L. monocytogenes. Hence, a biofilm is a complex system where the different stages of biofilm formation alternate, depending on many factors such the composition of the medium or different levels of nutrients, which can influence the cell-cell communication; this drives the physiological and metabolic processes within the biofilm and affects biofilm development [50,57]. In addition, the biofilm matrix contains one or more extracellular polymeric substances (EPS) such as polysaccharides, proteins, extracellular DNA (eDNA), lipids, polyglutamate, teichoic acids, humic substances, etc. [33,58,59].
Our epifluorescence microscopy results revealed the metabolically active cells in the microcolonies formed (Figure 1), and SEM was used to assess the biofilm architecture (EPS and embedded bacterial cells) (Figure 2) onto polypropylene type B. In particular, treatment with DNase I and proteinase K revealed the presence of proteins (16.90-46.82%) and eDNA (9.44-47.69%). In addition, on CRA, 100% of the L. monocytogenes isolates were biofilm producers, because the extracellular polysaccharides combine with the Congo red dye, demonstrating exopolysaccharides of L. monocytogenes (Figure 3).
These results agree with those of Jiao et al. [60] and Muthukrishnan et al. [61], who showed that extracellular proteins are a major EPS components of the biofilm dry mass. Likewise, Kadam et al. [45] showed the presence of eDNA after DNase I was added to the microtiter plates during biofilm formation by L. monocytogenes (strains 18 and 55). Recently, studies have reported the structural components of L. monocytogenes within the EPS, such as eDNA, proteins (InlA, BapL, PlcA, FlaA, PBP, and ActA), polysaccharides (poly-β-(1-4)-N-acetylmannosamine (poly-NAM), and teichoic acids (WTA and LTA)); moreover, their roles in bacterial adhesion and aggregation, and as a structural component within the biofilm that provides stability to the entire structure and horizontal gene transfer, have been revealed [18,[62][63][64][65].
Additionally, EPS strengthen the survival of microorganisms embedded on a substratum; moreover, they decrease the effectiveness of disinfectants, having a similar effect as organic matter or food residues present on the surface, resulting in less disinfectant coming into contact with the microorganism and reducing the effectiveness of the disinfectant or antimicrobial agent [66][67][68][69][70]. Our results emphasize the importance of incorporating other types of materials into the food industry so that they do not allow for the adherence and formation of biofilms in fresh produce. However, the design of strategies for the prevention and/or removal of biofilms of L. monocytogenes will also be necessary. In addition, techniques such as confocal laser scanning microscopy (CLSM) to reveal the structural components that make up the biofilm matrix and serotyping of L. monocytogenes can be used to determine the sources and mechanisms of contamination during the processing and marketing of fresh food products.

Conclusions
L. monocytogenes has the ability to develop a biofilm that harbors virulence-associated genes (LIPI-1) of L. monocytogenes isolates from Hass avocados. This could be a relevant food safety and public health consideration. However, serotyping is required to determine the prevalence, severity, and association of serotypes of L. monocytogenes with the ability to develop biofilm. Our study showed that the development of biofilms of L. monocytogenes is affected by 1/10 diluted TSB at 48-240 h using a CV staining assay. However, the development of biofilms of L. monocytogenes had a higher cellular density in TSBA onto PP. Additionally, future research should consider the detection of genes involved in other islands of pathogenicity (LIPI-2, LIPI-3, and LIPI-4) of L. monocytogenes isolates from Hass avocados.

Data Availability Statement:
The data used to support the findings of this study are available from the corresponding author upon request.