Virulence Characterization of Listeria monocytogenes, Listeria innocua, and Listeria welshimeri Isolated from Fish and Shrimp Using In Vivo Early Zebrafish Larvae Models and Molecular Study

Listeriosis is one of the most notable foodborne diseases and is characterized by high rates of mortality. L. monocytogenes is the main cause of human listeriosis outbreaks, however, there are isolated cases of disease caused by other species of the genus Listeria. The aim of this study was to evaluate strains of L. monocytogenes (n = 7), L. innocua (n = 6), and L. welshimeri (n = 2) isolated from fish and shrimps for their virulence based on the presence of virulence genes and the in vivo Danio rerio (zebrafish) larvae models. A total of 15 strains were analyzed. The zebrafish larvae model showed that the larvae injected with L. monocytogenes strains were characterized by the lowest survival rate (46.5%), followed by L. innocua strains (64.2%) and L. welshimeri (83.0%) strains. Multiplex PCRs were used for detection of selected virulence genes (luxS, actA2, prfA, inlB, rrn, iap, sigB, plcB, actA, hlyA), the majority of which were present in L. monocytogenes. Only a few virulence-related genes were found in L. welshimeri, however, no correlation between the occurrence of these genes and larval survival was confirmed. This research highlights the importance of the potential impact that Listeria spp. strains isolated from fish and shrimps may have on consumers.


Introduction
Species L. monocytogenes, L. ivanovii, and L. seeligeri are considered to be the main causes of listeriosis in animals and humans, however, due to the occurrence of single cases of infection caused by other species, including L. innocua, it seems necessary to study the virulent potential among others species of the genus Listeria [1].
The genus Listeria currently includes seventeen species [2]. For humans, L. monocytogenes is most often involved in foodborne outbreaks of listeriosis [3]. The other two species mentioned above are rarely reported as causes of human listeriosis [4][5][6], though there are reports of L. welshimeri being isolated from human fecal samples [7].
Listeriosis can develop as gastroenteritis with fever or present a more severe course with meningitis and sepsis. Parenteral symptoms are primarily found in people at risk, including pregnant women, newborns, the elderly, and people with reduced immunity (transplant recipients, patients Listeriosis can develop as gastroenteritis with fever or present a more severe course with meningitis and sepsis. Parenteral symptoms are primarily found in people at risk, including pregnant women, newborns, the elderly, and people with reduced immunity (transplant recipients, patients receiving anti-cancer drugs or undergoing immunosuppressive therapy, or Acquired immunodeficiency syndrome (AIDS) patients). In this group, listeriosis has a mortality rate of 20-30% [3].
L. monocytogenes has a sophisticated intracellular lifecycle and pathogenic strategy. Research shows that the pathogenicity island 1 is responsible for this strategy. Pathogenicity island 1 (LIPI-1) contains the prfA-cluster. The prfA gene is a transcription activator that regulates the expression of various L. monocytogenes virulence genes, and plcA is a phospholipase C with phosphatidylinositol activity. LIPI-1 also contains a lecithinase operon, which contains genes whose presence was investigated in the study, including actA and plcB. actA gene coding for actin polymerizing protein, which recruits and polymerizes actin filaments to Listeria when it is in the cell to promote its intracellular motility. plcB gene encodes phospholipase C with lecithinase activity and helps Listeria to escape from phagosomes, thus promoting the spread of Listeria to other cells. LIPI-1 also contains hly, which encodes listeriolysin O (LLO), which lyses erythrocytes and other cells, but also lyses vacuoles of eukaryotic cells, allowing Listeria to spread through the cytoplasm. Removal of the hly gene leads to loss of virulence, which proves the importance of LLO in the pathogenesis of L. monocytogenes [8].
The aim of the study was to evaluate strains of L. monocytogenes, L. innocua, and L. welshimeri isolated from fish and shrimp for their virulence based on the presence of virulence genes and the in vivo zebrafish larvae models.

Results
The virulence of L. monocytogenes has been well characterized, however, most previous research has focused on clinical strains, whereas there may be large variations among species and environmental strains [9]. We tested the outcome of experimental infection of early zebrafish larvae with Listeria monocytogenes, Listeria innocua, and Listeria welshimeri. The average survival curves are presented in Figure 1. The lowest survival, after 72 h, among zebrafish larvae can be observed when challenged with L. monocytogenes strains (46.5%). The average survival of L. innocua-infected larvae is higher and reached 64.2%; the highest survival rate was found in L. welshimeri-infected larvae, which was 83.0%. The differences between the species differ significantly (p value < 0.0001) ( Figure 1). The lowest survival, after 72 h, among zebrafish larvae can be observed when challenged with L. monocytogenes strains (46.5%). The average survival of L. innocua-infected larvae is higher and reached 64.2%; the highest survival rate was found in L. welshimeri-infected larvae, which was 83.0%. The differences between the species differ significantly (p value < 0.0001) ( Figure 1).
The survival rate of larvae infected with L. monocytogenes differed statistically significantly (p value < 0.0001) between the strains. After 72 h of the experiment, it was observed that the survival rate varied from 0.0% to 95.5%. In the case of one strain (LM107), none of the infected larvae survived after two days of the experiment (Figure 2). The survival rate of larvae infected with L. monocytogenes differed statistically significantly (p value < 0.0001) between the strains. After 72 hours of the experiment, it was observed that the survival rate varied from 0.0% to 95.5%. In the case of one strain (LM107), none of the infected larvae survived after two days of the experiment (Figure 2). As with L. monocytogenes, survival among larvae infected by L. innocua also varied widely between strains. For the least virulent strain, the survival rate was 95.4%, while for the most virulent strain, it was 19.0%. The differences between the strains were statistically significant (p value < 0.0001). Survival results for individual strains are presented in Figure 3. Of the two L. welshimeri strains, LW104 showed no virulence, and all infected larvae survived the 72-hour experiment. For the second strain (LW105), larvae survival was 72.0%. These differences are statistically significant (p value = 0.0107). Survival results for individual strains are presented in Figure 4. As with L. monocytogenes, survival among larvae infected by L. innocua also varied widely between strains. For the least virulent strain, the survival rate was 95.4%, while for the most virulent strain, it was 19.0%. The differences between the strains were statistically significant (p value < 0.0001). Survival results for individual strains are presented in Figure 3.
value < 0.0001) between the strains. After 72 hours of the experiment, it was observed that the survival rate varied from 0.0% to 95.5%. In the case of one strain (LM107), none of the infected larvae survived after two days of the experiment ( Figure 2). As with L. monocytogenes, survival among larvae infected by L. innocua also varied widely between strains. For the least virulent strain, the survival rate was 95.4%, while for the most virulent strain, it was 19.0%. The differences between the strains were statistically significant (p value < 0.0001). Survival results for individual strains are presented in Figure 3. Of the two L. welshimeri strains, LW104 showed no virulence, and all infected larvae survived the 72-hour experiment. For the second strain (LW105), larvae survival was 72.0%. These differences are statistically significant (p value = 0.0107). Survival results for individual strains are presented in Figure 4. Of the two L. welshimeri strains, LW104 showed no virulence, and all infected larvae survived the 72-h experiment. For the second strain (LW105), larvae survival was 72.0%. These differences are statistically significant (p value = 0.0107). Survival results for individual strains are presented in Figure 4.  Among the tested strains of L. monocytogenes there were two serotypes: 1/2a and 1/2c. The results of average survival for each stereotype are presented in Figure 5. These differences are not large: for serotype 1/2a the average survival was 46.9%, while for serotype 1/2c the average survival was 58.2%; these differences are not statistically significant (p value = 0.1205). The frequencies of the virulence genes identified in Listeria spp. strains are listed in Table 1.   Among the tested strains of L. monocytogenes there were two serotypes: 1/2a and 1/2c. The results of average survival for each stereotype are presented in Figure 5. These differences are not large: for serotype 1/2a the average survival was 46.9%, while for serotype 1/2c the average survival was 58.2%; these differences are not statistically significant (p value = 0.1205).  Among the tested strains of L. monocytogenes there were two serotypes: 1/2a and 1/2c. The results of average survival for each stereotype are presented in Figure 5. These differences are not large: for serotype 1/2a the average survival was 46.9%, while for serotype 1/2c the average survival was 58.2%; these differences are not statistically significant (p value = 0.1205).  Table 1.  The frequencies of the virulence genes identified in Listeria spp. strains are listed in Table 1. Nine virulence profiles were characterized in the study. The most common profile was plcB-actA2-hlyA-sigB-actA-prf A-inlB-rrn-iap-luxS (26.7%), the next most common profiles were sigB-rrn (20.0%) and plcB-hlyA-sigB-actA-prf A-inlB-rrn-iap (13.3%).  The study examined the correlation between the occurrence of genes associated with virulence and the survival of zebrafish larvae. The strongest negative correlation was observed for the actA (−0.3191) and hly (−0.2876) genes. The next genes with negative correlation were the prf A and iap genes, whose values were both −0.1817. A positive correlation was observed for two genes, 0.2648 for the luxS gene and 0.1953 for the rrn gene. These results are not statistically significant (p > 0.005 for each gene).

Discussion
The latest European Food Safety Authority (EFSA) data shows that almost half (41.7%) of the Rapid Alert System for Food and Feed notifications for L. monocytogenes in 2008-2016 concerned fish and their products [10]. It is L. monocytogenes that is considered the main threat to humans among species of the genus Listeria, while there have been rare cases of infection with other species of this genus [11]. Pathogenicity is closely related to the prf A-virulence gene cluster (pVGC), containing the prf A, plcA, hly, mpl, actA, and plcB genes, although there are other genes that are indirectly treated as virulence factors [12].
In the present study, the intraspecies variations in the virulence profiles were investigated. Research suggests that the hly gene is crucial for the virulence of L. monocytogenes strains [12,13]. In the present study, the hly gene was present in each L. monocytogenes strain for both serotype 1/2a and 1/2c. Additionally, it was found in one atypical strain of L. innocua. The atypical L. innocua strain was characterized by very high survival among larvae (80.9%), while the most virulent strain 103 had only the sigB and rrn genes. The most likely explanation seems to be the inactive form of the genes from prfA-cluster, while the high mortality rate of larvae infected with strain 103 may be due to genes other than those tested in this experiment. Cases of such strains have already been reported [14,15]. There is also a study indicating other species of the genus Listeria possessing the hly gene, including L. seeligeri [16]. In the study of the frequency of genes from pVGC cluster, they were found in each of the L. monocytogenes strains. Similar results were obtained by other researchers in the study of food products [17][18][19]. According to Orsi et al. (2016), L. welshimeri possessing a pVGC cluster have not been confirmed [2]. Due to possible polymorphism in the actA gene, two sets of primers, actA1 and actA2, were used [20]. For one strain, the actA gene was detected by the sequence for actA2 primer and was not detected for the second set of primers. In the case of genes unrelated to the pVGC cluster, they most often occurred in L. monocytogenes and L. innocua. Only two of ten genes (sigB and rrn) occurred in L. welshimeri.
In this work, a model for infection was established in zebrafish larvae in order to study the pathogenesis of Listeria sp. In vivo research on zebrafish larvae has also been used to study the virulence of various pathogens including Escherichia coli [21], Staphylococcus aureus [22], and Aeromonas hydrophila [23] but also fungal or viral agents. As for the use of zebrafish larvae to study the pathogenicity of Listeria, only L. monocytogenes has been used in previous research [24]. Until now, in vivo models using Danio rerio to study the virulence of L. monocytogenes focused on the model itself and not on testing strains from environmental samples. One study from 1996 investigated Listeria spp. strains from food samples, however, experiments were carried out at 22 • C, a temperature that differs from the optimal temperature for the expression of Listeria virulence genes, therefore this study will not be considered in the discussion [25].
In this study, the pathogenicity of L. welshimeri and L. innocua strains isolated from raw fish and shrimp was investigated. L. monocytogenes strains were characterized by their significant virulence potential. Mortality among zebrafish larvae was 65.6%. The study did not prove statistically significant correlation between the occurrence of known virulence genes and virulence in an in vivo model.
Comparing the virulence of L. innocua and L. monocytogenes, it can be concluded that L. innocua has been observed to be less virulent, however, mortality among larvae varied between 4.6% and 80.9%. Although this species is considered to be avirulent, there are individual data reporting patients with diseases caused by L. innocua. In 2014, L. innocua was the cause of meningitis. Researchers showed the presence of the prf A, inlA, inlB, inlC, hly, plcA, plcB, mpl, iap, clpE, daaA, and actA genes in the investigated strain [26]. The first fatal case caused by L. innocua occurred in 2003. The cause of death was bacteremia. In the cited study, the authors did not specify a virulence profile [27].
There is little data on L. welshimeri virulence. In our study, only two L. welshimeri strains were tested, none of which showed high mortality among Danio rerio larvae. A similar study on a live model, in this case mice, was carried out in 1986. The virulence of 16 L. welshimeri strains were investigated, each of which proved to be avirulent [28].

Ethics Statement
All fish were housed in the fish facility of Laboratory of Genomics and Transcriptomics in the University of Warmia and Mazury, Olsztyn, Poland, which was built according to local animal welfare standards. All animal procedures were performed in accordance with Polish and European Union animal welfare guidelines. According to the European Directive 2010/63/EU and Polish law regulations Dz.U. z 2015 r. poz. 266, all procedures performed in the present study including the use of early life-stages zebrafish and the euthanasia for the purpose of organ dissection did not require ethics committee permission.

Bacteria
The different Listeria species and serotypes isolated from fish (Salmo salar and Oncorhynchus mykiss) and shrimp (Penaeus monodon) sources used in this study were obtained from the microbiological collection of the Department of Industrial and Food Microbiology University of Warmia and Mazury in Olsztyn and are listed in Table 2. Strains were grown from frozen stocks in 5 mL of Brain Heart Infusionbroth (Merck, Darmstadt, Germany) overnight at 37 • C. Subsequently, the strains were transferred to 5 mL of BHI broth and incubated for 24 h at 37 • C. After incubation, the strains were centrifuged and washed three times with sterile saline. Eventually, phenol red (Merck, Darmstadt, Germany) was added to obtain a 0.05% solution in order to be able to observe the inoculation process. Table 3. List of primers used in research [30].

Conclusions
The data presented in this study show that the diversity according to the virulence level of Listeria sp. strains isolated from fish and shrimp is complex and based on different mechanisms that seem to differ according to the species of the strains. In addition, as demonstrated by the experiment carried out in vivo, although L. innocua is not considered dangerous to human life, it may pose a potential threat.

Conflicts of Interest:
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.