Genomic Diversity and Virulence Genes Characterization of Bacillus cereus sensu lato Isolated from Processing Equipment of an Algerian Dairy Plant

to toxins produced by the B. cereus group in the food safety and dairy industries in Algeria. Tis study emphasized the potential of whole-genome sequencing for genotyping and predicting the virulence-associated genes and antimicrobial resistance. Te present study will help to better assess the health and spoilage risk associated with B. cereus in dairy processing plants and to incorporate adequate preventive measures.


Introduction
Te Bacillus (B.) cereus group is composed of several species that are able to put the public health at risk due to the production of toxins.Tese bacteria also cause economic losses to the food and dairy industries [1].Consuming contaminated food mainly with pathogens and microbial toxins may lead to severe diseases; hence, food safety is seen as a major concern in any society.For this reason, cleaning and disinfection are important practices for the elimination of both food residues and the reduction of microbial contamination in the food industries [2].Te bioflm-forming bacteria settling in dairy industry materials might infuence both the safety and quality of milk and its products [3].In fact, such bacteria is found on the surfaces of milk transport tools such as pipes and containers and is described as a major cause of contamination of the dairy products.Tis leads to the corroding of metal pipes, reducing heat transfer, and increasing fuid frictional resistance [4,5].Unfortunately, there is no defnite method that can radically resolve the bioflm issue in the dairy industry [6].
Bacteria from the Bacillus cereus group have been frequently isolated from various food environments including food contact surfaces in the dairy industry, which constitutes a hygienic as well as a technological problem [7,8].Bacillus cereus is able to form bioflms on many types of substrata, including stainless steel surfaces.Te major microbiological challenge for food production lines is the presence of B. cereus bioflms and their substantial physiological versatility.Te capacity of Bacillus cereus to endure in the diferent parts of the dairy manufacturing plants allows this microorganism to be a potential source of the product's contamination and recontamination.Moreover, the low-temperature and the short period of sterilization are not efective in completing the elimination of B. cereus in the milk.Tis could result in food spoilage which subsequently leads to rejection of the products by the customers and create a fnancial defcit for the dairy industry [9][10][11].Furthermore, the task of removing these bacteria is made challenging because of bioflm and heatresistant spore formation [10,11].
Te B. cereus group, also known as B. cereus sensu lato, is a subdivision of the Bacillus genus that consists of eight closely related species, i.e., B. cereus sensustricto, Bacillus anthracis, Bacillus thuringiensis, Bacillus mycoides, Bacillus pseudomycoides, Bacillus weihenstephanensis, Bacillus toyonensis and Bacillus cytotoxicus [12,13].Tese microorganisms are clustered into seven phylogenetic subgroups according to the panC gene sequence [14][15][16].Tey also include phenotypic features of the strains in each group such as the ranges of temperature growth [15].Bacillus cereus is able to secrete an array of enzymes and virulence factors, such as toxins and enterotoxins.Moreover, the B. cereus poisoning efect on food shows two diferent syndromes: the diarrheal and/or the emetic syndromes [17].Te latter is triggered by a small cyclic heat-stable peptide (cereulide) which leads to vomiting 0.5-6 after ingestion [18].Te diarrheal type is due to enterotoxins, a group of heat-labile proteins.Tese enterotoxins include the tripartite hemolysin BL (Hbl) constituted by B, L 2 and L 1 components encoded by hblC, hblD, and hblA genes, respectively [19], the nonhaemolytic enterotoxin (Nhe) encoded by nheA, nheB and nheC genes [20] and the single protein, cytotoxin K encoded by the cytK [21,22].Products from other genes, such as entFM, entS and plcA are equally included in the pathogenesis of B. cereus [23][24][25][26].Another signifcant virulence factor of B. cereus is its capacity to adhere to various surfaces [27].Te ability of spores to adhere to stainless steel is linked to various factors such as their hydrophobicity potential, the number of appendages on their surface, the exosporium length and the zeta potential [28].Various studies have evaluated the prevalence of pathogenic B. cereus in milk and dairy products [29][30][31].However, data on the presence of such bacteria on the processing equipment, especially on their persistence after the cleaning step, as well as the mechanisms responsible for their ability to withstand the cleaning process are almost inexistent to the best of our knowledge.Terefore, the aim of the current study was to identify potential B. cereus isolates from dairy equipment in a processing plant in Algeria and assess their virulence potential.

Origin of B. cereus Strains.
Seventeen Bacillus strains were isolated from 160 samples collected by aseptically swabbing an area of 1 cm 2 of the inner surfaces of milk tanks that had contained pasteurized, unpasteurized local raw cow and recombined milk, as well as from packaging lines of dairy plant processing located in Tlemcen (north-west of Algeria) in 2010-2012.Swab premoistened with sterile distilled water containing 3% (v/v) Tween 80 was carefully used to ensure maximum recovery of bacteria from the equipment surface by overlapping horizontal and vertical strokes.Ten, the swabs were placed in test tubes containing 10 ml of maximum recovery diluent (tryptone salt broth).Te diluent tubes containing the swabs were vortexed for at least one minute to release the attached bacteria.Samples were then heat treated at 80 °C for 10 minutes before being incubated at 30 °C for 24 to 48 hours.
Recombined milk is obtained by merging water with skim milk powder and carefully adding milk fat in such a way that the target fat content is reached.Tese strains were collected through diferent samplings that were performed after the cleaning and sanitizing procedures.
Tese strains belong to the collection of the Laboratory of Food and Environmental Microbiology (University of Tlemcen-Algeria) and were previously characterized using biochemical tests and physicochemical tests for the surface properties of their spores (hydrophobicity and zeta potential) [32].

Determination of the Afliation of the Presumptive B. cereus Isolates.
Te genotypic group of the presumptive B. cereus was determined as defned previously by Guinebretière et al. [15].Amplifcation and partial sequencing of the panC gene were performed using Forward (5′-GAG GCG AGA GAA TAC GGA ATA CG-3′) and reverse (5′-GCC CAT TTG ACT CGG ATC CAC T-3′) primers [15].Te following PCR conditions were applied: 94 °C for 5 min as an initial denaturation step, followed by 30 cycles of 15 s at 94 °C, 30 s at 55 °C, 30 s at 72 °C, and 7 min at 72 °C as a fnal extension [36].Amplicons were purifed and sequenced as mentioned above.Te sequences were analyzed using the Sym'Previus online software (https://www.tools.symprevius.org/bcereus/),a tool that allows a rapid classifcation of B. cereus strains in four groups (from I to IV) by comparing the panC gene sequences of new isolates with reference strains published by Guinebretière et al. [15] and stored in the database [15].

Pulse Field Gel Electrophoresis (PFGE).
To screen potential intra and interspecies diversity, the isolates were further investigated using PFGE [37].Te fngerprints obtained were analyzed, compared and grouped based on the number and position of the fragments observed using the Bionumerics system (Dice's Coefcient of similarity, UPGMA: Unweighted Pair Group Method with Arithmetic Average; Applied Maths, Saint-Martens-Latem, Belgium).

Hemolysis Assay.
For the evaluation of hemolytic activity, the isolates were streaked on Columbia agar (Oxoid, Basingstoke, UK) containing 5% horse-blood according to the technique described by Mulligan et al. [38].Te plates were incubated at 30 °C for 24 to 48 h and the presence of hemolytic haloes around the colonies was examined visually.Depending on the ability to degrade red blood cells, three categories can be identifed: no hemolysis (c-hemolysis), incomplete hemolysis (α-hemolysis) and complete hemolysis (β-hemolysis) [39].

Genotypic Determination of Toxin Production Potential.
Genomic DNA was extracted and purifed as previously described by Celandroni et al. [40].Te presence of toxinencoding genes in bacterial genomes was evaluated by PCR as previously described by Taenthanee et al. [41].All strains were tested for the presence of sph (sphingomyelinase), plcA (phosphatidylinositol-specifc phospholipase C; PI-PLC), entFM (enterotoxin FM), entS (enterotoxin S), nheA, nheB and nheC (the three components of NHE), hblC (the L 2 component of HBL) complexes, cytK (cytotoxin K) and ces (emetic toxin) genes.Primers and annealing temperatures used for each PCR are listed in Table 1.Genomic DNA was extracted from B. cereus ATCC 14579 and ATCC 10987 and used as positive controls for each amplifcation.
Journal of Food Quality

Bacillus cereus Identifcation and Diferentiation.
Using 16S rRNA gene sequencing, the isolates were identifed as belonging to the B. cereus sensu lato group.Te microscopic observation of the isolates did not reveal the presence of parasporal crystals characteristic of B. thuringiensis.Sequencing of the panC gene phylogenetic classifed the isolates into 64.7% group III and 35.3% group IV (Table 2).
Te results of PFGE showed that 8 (47.1%), 4 (23.5%), 3 (17.6%)and 2 (11.8%) were diferentiated into four groups A, B, C and D, respectively, based on the number and position of the fragments observed, as shown in Figure 1.It was noticed that each of the two panC phylogenetic groups contain two different types of isolates corresponding to the PFGE group (A and C) and (B and D) for phylogenetic group III and IV, respectively (Table 2).Group A was the dominant group (47.1%) and contains bacteria isolated for all the diferent samples screened except from the sample of the pipelines of pasteurized recombinated milk.Isolates from groups B, C and D were also detected in various samples (Table 2 and Figure 1).

Hemolysis Capacity of the Isolates.
Hemolysin production was evaluated by seeding bacterial cells on blood agar plates.All bacteria investigated exhibited a β-hemolytic characteristic as a clear zone around the colonies due to the complete degradation of red blood cells that was observed.

Putative Pathogenic Potential of B. cereus Strains.
Te presence of sph, nheA/B/C, hblC, plcA, entT, entFM, entS, cytK and ces genes in the bacterial genomes was verifed by PCR.As shown in Table 3, none of the 17 isolates carried the emetic toxin-encoding gene (cse), while the other genes were commonly amplifed.entFM and entS were amplifed in all isolated strains.Nhe components and SMaseencoding genes were detected in most strains (sph 88.2%, nheA 88.2%, nheB 94.1% and nheC 94.1%).For other genes, the hblC, plcA, entT and cytK genes were detected in 29.4%, 94.1%, 35.3% and 41.2%, respectively.Te obtained results indicate a noticeable interstrain diferences in the toxigenic profles of the isolates.One isolate (Bacillus cereus sensu lato100) can be assumed to be the most virulent strain, since it possesses hemolytic activity and 10 out of 11 (90.9%)virulence genes (Table 3).On the other hand, the least virulent potential can be attributed to the isolate Bacillus cereus sensu lato18, with four genes out of 11 (36.4%).Te phylogenetic and genotyping analysis displaying relatedness between the Bacillus cereus sensu lato isolates based on cgSNP distances rooted to the reference genome revealed three diferent groups according to how the isolates were closely related (Figure 2).Te MLST analysis based on WGS revealed three different sequence types (STs) (1431, 2226, and 1018) (Figure 2).Tree isolates belonged to clonal complexes (CC-142) (Figure 2).For eight isolates, so far, an unknown clonal complex was identifed.Hierarchical clustering with a threshold of 500 cgSNPs, grouped eleven strains into three distinct clusters while two strains were less than 20 SNPs diferent (Figure 2).

Genetic Characterization of Bacillus cereus sensu lato
Te genomic analysis of Bacillus cereus sensu lato isolates revealed diferent virulence-associated genes related to motility, chemotaxis, adhesion and invasion.All isolated B. cereus sensu lato harbored inhA, nheA/B/C, entFM, cytK2, plcA, alo, sph and cerA while hblA/B/D, hblC, hlyIIR, hlyII and bpsC were detected in 25% of the isolates.Hemolysin BLbinding components represented by hblA/C/D genes were detected in the strains with the same PFGE profle as sequence type 1431 and clonal complex CC-142.A pore-forming toxin of Bacillus cereus that has been linked to a case of necrotic enteritis (cytK) was detected in all isolated B. cereus.

Discussion
Te ecological adaptations play a major role in the emerging of B. cereus.Te temperature tolerance is considered a factor that leads the strains of various groups of B. cereus to adjust to new environments [15].Microorganisms belonging to diferent groups possess diferent physiological properties and safety potentials.For example, microorganism of groups     II and VI are psychrotrophic, microorganism of groups III, IV and VII are mesophilic and those of group V have an intermediate behavior [15].Group III seems to be involved in the highest risk of food-borne diseases, while groups VII, IV and II correspond to a moderate risk, and fnally group VI represents a very low risk.Te strains belonging to group III are mostly cytotoxic, and this group particularly includes the emetic strains [55,56].
Most isolates screened in the current study belong to Group III, thus indicating that their presence in the milk environments constitutes a serious safety issue.A smaller amount of group IV isolates as compared to Group III was noticed which could be explained by the fact that isolates of group IV are rather sensitive to heat treatments in comparison to those in group III [57,58].Te strains that are highly toxic and are mostly involved in food-borne disease belong to mesophilic groups III and IV, that is, consistent with the results of the present study [59].
Te existence of B. cereus in dairy plants can be attributed to diferent sources such as the cow udders, collecting containers, transportation systems, and milk powder in the case of recombined milk [60].Teir presence in the processing equipment demonstrates that the cleaning and disinfection procedures have not been performed carefully on the production equipment [61] or that the methods and products used are not efcient enough.In a previous study, the spores of the isolates herein investigated were shown to have diferent lengths of appendages, the surface of the exosporium and spore hydrophobicity which enable these spores to adhere to stainless steel surfaces [32].
In order to trace the sources and comprehend the epidemiology of food-borne pathogens, a number of molecular characterization methods are used such as random amplifcation of polymorphic DNA (RAPD) [62], multilocus sequence typing (MLST) [63] and amplifed fragment length polymorphism (AFLP) [64].Among these methods, pulsefeld gel electrophoresis (PFGE) is an efective way to clearly analyze the genetic variation among strains and the relations between phenotypic and genotypic features [65].Te fngerprints gained by PFGE allow us to observe the presence of similar isolates in diferent samples, thus suggesting a cross-contamination with similar strains across the production chain.
Te fndings of this study revealed that pathogens can create a recontamination risk on stainless steel surfaces for a long time because of their ability to survive there.Besides, the outstanding food on these surfaces may well encourage the bacterial survival.Hence, two major aspects are responsible for the risk of food-borne illnesses: the degree of contamination on the surfaces and the possibility of its transmission to food [66].
Furthermore, since milk packaging is commonly performed using very clean, but not strictly aseptic fllers, the danger of postpasteurization recontamination of processed milk is considered [67].Te factors that determine the shelf life of pasteurized milk include the existence and evolution of aerobic psychrotrophic endospore formers where members of B. cereus are key spoilers beside other species of the genus Bacillus.Te resistance of endospores to the pasteurization process reduces the shelf life of the pasteurized milk through the underlying germination and the production of spoilage enzymes [68], such as lipases (SMase, PI-PLC, PC-PLC) and proteases, can lead to food degradation and spoilage, thus economically impacting the food produced [69,70].In a previous study, all strains were shown to possess proteolytic and lipolytic activities [32].
Contamination of dairy processing plants by B. cereus constitutes a serious health hazard for the consumers, since this microorganism is mainly responsible for food-borne diseases [13].Te pathogenicity of B. cereus is primarily due to its ability to secrete a plethora of toxins [71,72].
Te high occurrence of the nhe genes observed is not unusual, however, the occurrence of the hblC gene is quite low compared to that reported (45%-65%) in previous studies on B. cereus isolates from other origins [73,74].Consequently, based on the various published literature on this matter [18,73], the amount of strains holding only cytK is assumed to be fairly limited, yet the latter (cytK) might well be toxic [22].Bacillus cereus 14 did not exhibit hblC or sph that are linked to hemolysis potential, suggesting that other factors are associated to such a characteristic.In fact, several hemolytic factors are expressed in the B. cereus group Journal of Food Quality of bacteria such as cereolysin [75], sphingomyelinase [76], cereolysin AB [77] and cereolysin-like hemolysin [78].Although few emetic toxin encoding genes are found in dairy food [79,80].In this study, B. cereus revealed no inclusion of the emetic toxin encoding gene.In controversy, these genes were detected in B. cereus in 10.2%; [81]; 9%, [82]; 1.0-3.8%,[80]; 2%, [83]; and 1.1% [84].Te isolates used in the present work revealed no inclusion of the emetic toxin encoding gene.
Te fndings of the present research show that the B. cereus investigated in this study are likely to represent a signifcant health risk because these isolates contain the enterotoxin gene and have the ability to adhere to stainless steel surfaces.In addition, the presence of proteolytic and lipolytic enzymes in these isolates increases the chances of cross-contamination and hinders the decontamination of food industry surfaces.
In this study, the detection of genetic factors for AMR was performed using diferent tools and databases (AMR-Finder, ResFinder and NCBI) in order not to miss any loci, as there is no single method that might be sufcient for the purpose alone (Table 4) and we found that AMRFinder appears to be a highly accurate AMR gene detection system, as previously mentioned [88].
Terefore, it is crucial to include B. cereus group of isolates in routine food quality control in laboratories in this region of Algeria.Also, evaluating and upgrading the cleaning and disinfecting procedures of the production materials in the milk processing environment are helpful ways to avoid economic losses and detrimental health consequences for consumers.Finally, eliminating B. cereus in the food processing industries requires taking the appropriate hygiene measures.Tese measures include both the availability of sufcient cleaning equipment and the reassurance of the employees' personal hygiene, and importantly, increasing the points of control with a regular check of the sanitization products and the temperature of the pasteurization process [24].

Conclusion
While the health risks resulting from the recontamination of pasteurized milk by B. cereus adhered to pipelines are well known, little attention has been given to molecular analysis of genotypes and virulence in the Algerian diary industry.In the present work, we highlighted the high virulence factors of B. cereus in the pipeline of dairy plant in Algeria.Te panC and PFGE analyses clearly showed the genetic diversity of B. cereus isolated in this study.To the best of our knowledge, this is the frst study concerned with the analysis of virulence factors and defnition of genotypes by PanC and PFGE of diferent B. cereus isolates from the pipeline of the dairy industry in Algeria that revealed a possible high risk of B. cereus to the dairy industry and public health.From food safety perspectives, the presence and pathogenicity of B. cereus in the dairy industry play a key role and should not be disregarded.

Figure 1 :
Figure 1: Dendrogram of cluster analysis of PFGE fngerprints of B. cereus sensu lato isolated from diferent processing equipment of an Algerian dairy plant.Te dendrogram based on Dice's coefcient of similarity with the unweighted pair method with arithmetic averages clustering algorithm (UPGMA).
milk storage tank Pasteurized recombined milk storage tank Pipeline of pasteurized milk Pipeline of pasteurized recombined milk Pipeline of raw milk Raw milk storage tank Raw recombined milk storage tank

Figure 2 :
Figure 2: Phylogenetic relation and genotypes of 12 Bacillus cereus sensu lato isolated in this study rooted to the reference genome with regard to the sequence types and clonal complex.

Table 1 :
Genes, primers sequences and PCR conditions used in this study.

Table 2 :
Phylogenetic group and isolate origin of Bacillus cereus sensu lato isolated in this study.

Table 3 :
Toxin gene profles of the screened Bacillus cereus sensu lato isolated in this study using PCR.

Table 4 :
Genetic characterization of antimicrobial resistance of the tested Bacillus cereus sensu lato.