Enterotoxin and Emetic Toxin Genes Profiles and Genetic Diversity of Bacillus cereus Isolated from Food, Environmental and Clinical Samples in Serbia

Abstract Bacillus cereus, usually ingested by food, can cause two types of disease due to the presence of toxins: vomiting and diarrhea syndrome. Systemic infections can also occur. The aim was to detect genes for enterotoxins (hblA, entFM) and emetic toxin (cer) and to investigate the genetic heterogeneity of B. cereus isolates from food, environment and human stool. Identification of B. cereus was performed by means of selective medium, classical biochemical test and polymerase chain reaction (PCR). Toxin genes were detected by PCR. Typing was performed by random amplified polymorphic DNA (RAPD). EntFM gene was present in all stool and food samples and in 28/30 environmental isolates. HblA gene was present in 29/30 stool, 23/30 food and 24/30 environmental isolates. Cer gene was present in 30/30 stool, 28/30 food and 25/30 environmental isolates. The RAPD results show high heterogeneity among the isolates from each group. In the cumulative dendrogram, representative isolates from all three groups formed two clusters with a difference of 53%. The detection of toxin genes in all B. cereus isolates indicated these bacteria as potentially pathogenic and a serious threat for human health. The presence of isolates from all three groups in the same cluster suggests the existence of similar strains in the environment, food and patients, which is in line with the circulation of strains in nature through the food chain.


INTRODUCTION
Bacillus cereus (B.cereus) is a Gram-positive spore forming food pathogen commonly found in a wide variety of different foods and environments (soil, water, air), as well as in the feces of humans and animals [1]. The vegetative cell or spores of B. cereus into the environment enter the food chain, from soil to plants, animals and humans, leading to a major problem in the agricultural and food industry, as well as in veterinary and human medicine [2].
In recent decades, the incidence of infections after consuming microbiologically contaminated food has increased, despite microbiological food control. The reasons are associated with the latest trends in global food production in industrialized countries, expansion in production of semi-processed, ready to eat food and storage of food at an inadequate temperature. Unfortunately, not all countries investigated stool samples for the presence of B. cereus, neither report cases nor outbreaks.
Although the presence of pathogenic strains of B. cereus is most often associated with intoxication and toxicoinfection of the gastrointestinal system, recent studies have linked this bacterium with infections in neonatology, surgical and traumatic wounds, intravenous drug delivery, catheter placement, and invasive bacteremia of the central nervous system (meningitis and brain abscess), endophthalmitis, pneumonia and gas gangrene [3]. The registration of fatalities [3][4][5] caused by strains of B. cereus, due to delayed diagnosis or inadequate response to antibiotic therapy, has highlighted the great importance of introducing methods for the rapid detection of B. cereus.
At appropriate physical -chemical conditions, B. cereus can cause two types of intoxications: vomiting, due to the presence of heat and acid stable emetic toxin and diarrheal syndrome, due to the presence of heat-labile enterotoxins. Thereby, gene detection gives us proper information about the possible production of enterotoxins and emetic toxins [6,7].
Due to the wide distribution of B. cereus strains in the soil, water, air, food, as well as in the feces of humans and animals [1], there is a need to investigate the genetic heterogeneity of isolates from different sources [8,9].
In this study, we detected the presence of genes responsible for diarrhegenic and emetic toxin and fortifi ed the genetic diversity of B. cereus isolates from investigated groups (food, environment and patients' stool) by RAPD analysis (Random Amplifi ed Polymorphic) DNA.

Samples
The samples were classifi ed into three groups, each of them comprising 30 isolates: different types of food (F1-F30), isolates from patients (P1-P30), and from the environment (E1-E30). B. cereus ATCC 11778 was used as the positive control.

Identifi cation of B. cereus isolates
All types of samples were cultured on selective mannitol egg yolk polymyxin agar (MYP) (HiMedia, India), and incubated for 18-24 h at 37 0 C. Pink colonies with a lecithinase reaction were transferred on 5% sheep blood agar (HiMedia, India), and thereon incubated. Presence of β-hemolysis was screened on 5% sheep blood agar following the procedure of Collins et al. [10]. In Gram-staining preparations, we confi rmed the presence of terminal to subterminal spores, and existence of parasporal crystalline inclusions [11]. Morphological properties such as β-hemolysis, spores and inclusions were used to identify Bacillus spp. Finally, B. cereus was identifi ed with an interactive database by using BBL Crystal GP ID Biochemical profi les (Becton, Dickinson and Company, USA). B. cereus group was confi rmed by PCR, yielded 533 bp amplifi ed fragments with primer pair BalF/BalR.

Preparation of DNA template and polymerase chain reaction (PCR)
A single colony of each isolate of B. cereus was incubated in the Brain-heart infusion broth (HiMedia, India) at 37 ºC for 18-24 h. A pellet of 1 ml of culture was rinsed in saline solutions overnight, resuspended in 500 µl of distilled water, and boiled for 10 min. The prepared DNA was used directly for PCR or stored at -20 ºC until use.
PCR was used for the identifi cation of B. cereus (balFR gene), as well as for the detection of enterotoxin genes (hbla, entFM and bceT), and emetic toxin gene (cer) using specifi c primers (Invitrogen, Vivogen D.O.O.) ( Table 1). The PCR mixture was prepared in a volume of 25 µl, with DreamTaqGreen Master Mix (ThermoScientifi c, Lithuania), 0.2 µmol l -1 fi nal concentration of each primer, and 2.5 µl of prepared DNA template. The primer sequences and PCR conditions were the same as described earlier [12,13]. PCR was performed on thermocycler Eppeddorf MasterCycler (Eppendorf, Germany). The fi nal polymerization lasted 5 minutes The fi nal polymerization lasted 5 minutes The PCR products were separated on 1.5% agarose gel (ICN Biomedicals) using electrophoresis (Pharmacia LKB), stained with ethidium bromide, visualized on a UV transilluminator (Shimadzu 160UV-Vis) and photographed by gel documentation system.

Random amplifi ed polymorphic DNA (RAPD) analysis
RAPD analysis was used for the molecular comparison of selected isolates. The four primers were used and the amplifi cation conditions for them were described in the papers of the above authors: DJ16 [14], SPH1 [15], AG15 and AX16 [16] ( Table 2). PCR was carried out in a 25 µl volume with DreamTaqGreen Master Mix (ThermoScientifi c, Lithuania), 2.5 µl of the prepared bacterial DNA as template, and 0.2 µmol l -1 fi nal concentration of an appropriate primer.
The amplifi ed DNA fragments were analyzed as in the PCR reaction described above. The sizes of fragments were determined by comparison with DNA molecular weight markers GeneRuler DNA Ladder mix SM0331 (Thermo Scientifi c, Lithuania) and HyperLadder 50bp, Bioline, Germany.
The cluster analysis of the RAPD patterns was performed using STATISTICA 8 program.

Detection of enterotoxin and emetic toxin genes
The presence of enterotoxigenic genes hblA, and entFM was detected by PCR. The hblA gene specifi c PCR yielded an amplifi ed product of 834 bp in 29/30 B. cereus isolates from stools, 23/30 isolates from food and in 24/30 isolates from the environment (Figure 1 A, B, C).   The isolates were formed by two clusters, with a difference of 53%. Cluster I contained two subclasses, with a difference of 34%, and cluster II contained two subclasses with a 47% difference. Five isolates, i.e. two isolates from patients and three environmental isolates grouped in one cluster. All other isolates were grouped in another cluster. Between them, two isolates from food (F6 milk powder and F18 dietetic product) showed the highest degree of similarity: 97.5%. Another seven isolates from food (F13 beet salad, F3 oregano F10 cinnamon, F15 cheese, F29, F17 and F24 dietetic products) grouped in a branch with 83% similarity. All other subclusters and branches grouped isolates from different sources. (Figure 5).

DISCUSSION
The aim of our study was to detect genes for enterotoxins (hblA, entFM) and emetic toxin (cer) and to investigate the genetic heterogeneity/similarity of B. cereus isolates from food, environment and human stool. The high prevalence of gene for diarrheal and emetic toxins in samples of different origin [17][18][19] which indicates a potential risk to human health has been confi rmed by our results from all three groups of isolates tested. Similar results were obtained in several studies [20][21][22] confi rming the presence of both genes in all tested food samples, especially in fast food [23,24]. On the other hand, there are studies [7,21] which confi rmed only the presence of entFM gene in all food samples tested, or only signifi cant presence of hblA gene in milk and milk products samples [22]. Also there are reports [24] which confi rmed a high percentage of these genes in samples from patients and food. Similar to our results a high percentage (68%) of the hblA gene detection in B. cereus strains from the soil was confi rmed, too [25]. It is of most importance to point out that all three tested genes were detected in most of our B. cereus isolates. Moreover, we noted that in the isolates in which the hblA gene was not detected, entFM or entFM was revealed. Thus, the obtained results are consistent with the authors whose results confi rm the presence of the entFM gene in emetic strains. We also obtained higher percentage of cer gene in food and clinical samples then other authors [26]. Based on the data we recorded, strains we investigated possess a substantial virulence potential, or due to hblA and entFM genes, or due to cer gene. It cannot be excluded that in the same time when enterotoxins are produced, cereulin is synthesized, too. We could notice that there are more toxinogenic strains in Serbia in food, environmental and clinical samples than in other countries [12,27].
Cumulative RAPD dendrogram of representative isolates of all three groups, the most food isolates grouped together at the level of 78%, indicating high genetic similarity among food isolates. The other food isolates were present in both clusters and subclasters together with patient and environmental isolates indicating the widespread distribution of genetically similar B. cereus strains nevertheless their different geographic distribution. Analyzing cumulative RAPD dendrograms within each group of B. cereus isolates, a large heterogeneity was determined. This data may indicate a greater similarity of RAPD profi le between isolates within the group of patients compared to isolates from the food and environment. Similarly, Prasad et al. [8] confi rmed high genetic heterogeneity among strains of B. cereus by RAPD in fi sh from tropical seas. According to Ghelardi et al. [28] who characterized B. cereus isolated from diseased and food samples, in two related epidemics, their fi ndings indicated clonality between isolates from patients and from food, thus confi rming the source of contamination. RAPD method with PCR for detection of toxin genes could be successfully used in the characterization of epidemic and non-epidemic strains and monitoring both clinical and environmental B. cereus isolates.
Generally, there are not enough data on epidemic and individual cases [29]. Recording clinical cases and monitoring of B. cereus infections and outbreaks in Serbia could give us relevant data on enteric and emetic diseases. Many conditions [3] impose the need to improve the diagnostics of these bacteria. As B. cereus is increasingly associated with non-gastrointestinal infections, controlling the presence of genes for toxin production should take a signifi cant role in the control of intrahospital infections. Additionally, characterization of strains from food, environment and patients by RAPD analysis could provide more valuable information on the sources of infection.
In conclusion, the detection of toxins genes in all B. cereus isolates indicated these bacteria as potentially pathogenic and a serious threat for human health. To the best of our knowledge, this is the fi rst study to on the presence of entero and emetic toxin genes and genetic diversity in B. cereus isolates from different sources in Serbia. The presence of isolates from all three groups in the same cluster suggests the existence of similar strains in the environment, food and stools of patients, which is in line with the circulation of strains in nature through the food chain.