Assessment of Listeria monocytogenes virulence in the Galleria mellonella insect larvae model

Several animal models have been used to understand the molecular basis of the pathogenicity, infectious dose and strain to strain variation of Listeria monocytogenes. The greater wax worm Galleria mellonella, as an alternative model, provides some useful advantages not available with other models and has already been described as suitable for the virulence assessment of various pathogens including L. monocytogenes. The objectives of this study are: 1) confirming the usefulness of this model with a wide panel of Listeria spp. including non-pathogenic L. innocua, L. seeligeri, L. welshimeri and animal pathogen L. ivanovii; 2) assessment of virulence of several isogenic in-frame deletion mutants in virulence and stress related genes of L. monocytogenes and 3) virulence assessment of paired food and clinical isolates of L. monocytogenes from 14 major listeriosis outbreaks occurred worldwide between 1980 and 2015. Larvae injected with different concentrations of Listeria were incubated at 37°C and monitored over seven days for time needed to kill 50% of larvae (LT50) and to determine change of bacterial population in G. mellonella, 2 and 24 hours post-inoculation. Non-pathogenic members of Listeria and L. ivanovii showed significantly (P < 0.05) higher LT50 (lower virulence) than the wild type L. monocytogenes strains. Isogenic mutants of L. monocytogenes with the deletions in prfA, plcA, hly, actA and virR genes, also showed significantly (P < 0.05) higher LT50 than the wild type strain at the inoculum of 106CFU/larva. Food isolates had significantly (P < 0.05) lower virulence than the paired clinical isolates, at all three inoculum concentrations. L. monocytogenes strains related to non-invasive (gastroenteritis) outbreaks of listeriosis showed significantly (P < 0.05) lower virulence than isolates of the same serotype obtained from outbreaks with invasive symptoms. The difference, however, was dose and strain- dependent. No significant differences in virulence were observed among the serotype tested in this study.

Introduction Human listeriosis, caused by the pathogen Listeria monocytogenes, accounts for about 1600 cases and 250 deaths per year in the USA [1]. Outbreaks of listeriosis are commonly associated with severe invasive symptoms including septicemia, meningitis and spontaneous abortion among pregnant women. In general, people with compromised immune systems, neonates and the elderly are most often affected [2]. A few outbreaks with milder symptoms of febrile gastroenteritis have also been reported [3]. The non-invasive febrile gastroenteritis however has shown to have significantly higher occurrence rates and is not particularly associated with any underlying illnesses [4]. In the past, outbreaks of human listeriosis were often linked to deli meats and dairy products and most frequently involved susceptible groups including the elderly, infants, pregnant women and individuals with suppressed immune system [5]. Emergence of newer food vehicles including cantaloupe, stone fruit, caramel-coated apples and most recently, frozen foods [6][7][8][9] along with the several unusual cases of invasive listeriosis (meningitis) found among healthy children aged 5-15 years [8] emphasize the importance of a better understanding how L. monocytogenes survives and grows in fruits and vegetables and weather the different environment change virulence potential of the pathogen.
Historically, L. monocytogenes virulence assessment has been performed using laboratory animals including mice, guinea pigs and monkeys [10]. However, use of such models is frequently associated with both ethical and financial burdens. Furthermore, the number of animals needed and length of the reproduction cycle of mammals became challenging factors in research involving animals. Various invertebrate models such as Caenorhabditis elegans (C. elegans), Drosophila melanogaster (D. melanogaster) and Zebra fish (Danio rerio) have also been used in the assessment of the host-pathogen interactions with L. monocytogenes [11][12][13]. Although some of these models have shown promise, the evaluation of the role of temperature-dependent virulence factors of L. monocytogenes is limited as many of these invertebrate models are incubated at or near room temperature. In recent years, use of larvae of the greater wax moth Galleria mellonella has emerged as a promising model for the assessment of virulence of numerous human pathogens including L. monocytogenes [14]. Among the various advantages of this model are low cost, easy manipulation, ethical acceptability, and the ability to incubate larvae at 37˚C, the temperature of the human body and one required for the optimal expression of many key virulence factors in L. monocytogenes [15]. Most importantly, innate immune system of G. mellonella resembles to one of mammal's, with enzymes, reactive oxygen species and antimicrobial peptides necessary to protect from bacterial infection [16].
The main objective of this study was the risk assessment of L. monocytogenes paired clinical and food isolates of L. monocytogenes associated with the major foodborne outbreaks using G. mellonella model. We also assessed the virulence potential of set of L. monocytogenes strains associated with a few self-limiting febrile gastroenteritis outbreaks. Furthermore, we aimed to demonstrate the importance of L. monocytogenes virulence factors by employing an extended panel of isogenic mutants. An understanding of the virulence potential of different L. monocytogenes strains would improve Listeria risk assessments and help in developing better food safety guidance as well.

Bacterial strains and growth conditions
The different Listeria species, serotypes, mutations and isolates from different sources used in this study are listed in the tables. Bacteria were grown overnight at 37˚C in brain heart infusion broth (BHI) (Becton Dickinson and Co., Sparks, MD) and on BHI agar plates. Bacterial cultures were washed twice and serially diluted with phosphate buffered saline (PBS). Appropriate dilutions were plated on BHI agar and incubated for 24h at 37˚C for bacterial CFU count. Colony counts were used to calculated bacterial inoculum.

G. mellonella injection and death assay
Last-instar larvae purchased from a commercial vendor (Vanderhorst, Inc., St. Marys, Ohio) were injected with bacterial inoculum in groups of 30 for each strain and for each dilution. Inoculum was administered directly to the larval hemocoel through the last left pro-leg as previously described [17]. Every trial included a group of 10 un-injected larvae as an environmental control and 10 larvae injected with PBS as a method control. Experiments were performed in at least three independent trials. Injected insects were monitored over seven days at 37˚C, and the time necessary to kill 50% of larvae (LT 50 ), by each inoculum was recorded. By the day seven, pupa formation was recorded in survived larvae.

Changes in populations of Listeria spp. in inoculated G. mellonella larvae
To assess bacterial population change, Galleria larvae were infected with selected panel of different Listeria spp (10 6 CFU/larva). Panel included selected non-L. monocytogenes, wild type L. monocytogenes LS1209 and isogenic mutants with the deletion in virulence and stress related genes (Tables 1 and 2). Clinical and food isolate related to the Jalisco cheese listeriosis outbreak [5] were also included. At the time points of 2 and 24 hours post-infection, respectively, five surviving larvae (approximately 1g) in each test group were randomly selected, surface sterilized with 70% ethanol, added to 1ml of PBS and crushed by vortexing. Appropriate dilutions were plated on RAPID'L.mono Medium (BIO RAD, USA) agar and incubated at 37˚C for 24-36h before enumeration of the typical L.monocytotgenes colonies.

Statistical analysis
The life-table method was used to produce non-parametric estimates of the survivor functions and the homogeneity of the survival curves across strains was tested using the Wilcoxon test [18]. The analyses were stratified by dose and p-values were Bonferroni adjusted. Data was analyzed using SAS/STAT software, VERSION 9.4, (c) 2002-2012, SAS Institute Inc., Cary, NC.

Listeria spp. in G. mellonella model
In order to establish a range of Listeria concentrations for the virulence studies in G. mellonella model, we compared a few representative species of Listeria, including L. monocytogenes LS1209 ( Table 1). The results (Fig 1) revealed similar LT 50 (time to kill 50% of the population) values of 18-24 hours at the highest doses of 10 7 CFU/larva. The LT 50 was significantly (P < 0.05) higher for L. seeligeri LS6 and L. welshimeri LS166, 24 hours post inoculation, while the rest of non-L.monocytogenes strains did not significantly differ from the reference L. monocytogenes strain (Fig 1A). At the inoculum of 10 6 CFU/larva the LT 50 for the reference L. monocytogenes LS1209, was 2 fold lesser than the L. ivanovii and 4-6 fold smaller than the rest of the non-L. monocytogenes strains tested. Significantly (P < 0.05) lower mortality was observed at 24, 72 and 168 hours ( Fig 1B) for L. seeligeri LS6 and L. welshimeri LS166 compared to L.monocytogenes. Similar results were also obtained when larvae were infected with the doses of 10 5 and 10 4 CFU, separately (Fig 1C and 1D). Among the non-L. monocytogenes strains employed in this study, L. ivanovii expressed the highest virulence potential, most significantly (P < 0.05) at the doses of 10 5 CFU/larva after 72 h of incubation ( Fig 1C). The lowest doses of 10 4 CFU/larva produced LT 50 of < 168 hours for L. monocytogenes, while the mortality of larvae infected with the rest of the tested strains remained well below 50% (Fig 1D). Throughout the experiment no death among environmental and method control groups have been detected.
Change in populations of Listeria spp. in inoculated G. mellonella larvae In order to monitor the fate of the injected Listeria, we enumerated viable Listeria (CFU) in the selected Listeria spp., 2 and 24 hours post inoculation ( Table 2). Bacterial counts indicated rapid decreas of all tested strains 2 hours post inoculation followed by the increase recorded at the 24h time point. Increase, however, was strain related. Populations of LS411 and LS412, isolates from Jalisco cheese listeriosis outbreak were significantly (P<0.05) higher 24 hours post inoculation than the other tested Listeria strains. L. monocytogenes hly mutant at 2h was reduced by a factor of 10 3 following the inoculation and stayed at that level even after 24h.
L. monocytogenes isogenic mutants in G. mellonella model A panel of 16 strains (Table 3) with the deletion in relevant virulence and oxidative stress genes along with the parental strains LS1209 and LS1223 was employed for virulence assessment in the G. mellonella model. Although deletion of the hly and prfA genes, separately, did not completely attenuate the virulence in larvae, mortality of larvae over the 168 hours of incubation was significantly (P < 0.05) lower compared to the parental strain, as well as the other isogenic mutants tested in all three doses (Fig 2A-2C). Interestingly, deletion of the plcA gene resulted in a lower mortality of larvae compared to the virulence of ΔplcB, Δhly, ΔprfA and, ΔactA mutants and was dose-dependent (Fig 2A-2C). Compared to the parental strain, the virulence of ΔplcA was significantly (P < 0.05) lower (higher LT 50 ) at each of the tested doses. Deletion of actA gene resulted with the significantly (P < 0.05) lower mortality of larvae compared to the wild type strain. Mutants with deletions of the inlB, and inlAB genes expressed lower killing potential than the parental strain. The difference was statistically significant (P < 0.05) only at the doses of 10 6 and 10 5 CFU/larva (Fig 3B and 3C). The mutant strain with the deletion of inlA demonstrated similar virulence to the parental strain ( Fig 3A-3C).

L. monocytogenes response regulator mutants in G. mellonella model
One of the important virulence factors in L. monocytogenes is a two-component response-regulator system termed VirR/VirS [19]. The system, encoded by virR and virS genes, regulates modifications of bacterial surface components. Isogenic deletion mutants of virR and virS in the EGD strain showed that the virR mutant is severely deficient in virulence while the virS mutation had very little effect [20]. Testing of the L. monocytogenes H7858 isogenic mutants with the deletion in virR and virS genes indicates that virS gene is not as essential for L. monocytogenes virulence as the presence of the virR gene (Fig 4). At each of the tested concentrations ( Fig 4A-4C) mortality of larvae caused by the virR mutant was significantly (P < 0.05) lower than the larvae infected with the parental strain or of the ΔvirS mutant.
Outbreak related clinical and food isolates of L. monocytogenes in the G. mellonella model Paired clinical and food isolates from 13 out of 14 major historical outbreaks of listeriosis (Table 4) were tested for possible differences in virulence potential. Parallel testing revealed higher virulence potential of clinical isolates of L. monocytogenes compared to their pairs recovered from food. These differences were, however, strain-and dose-dependent. The clinical isolate of L. monocytogenes strain LS620 related to celery-chicken salad outbreak [21] demonstrated significantly (P<0.05) higher virulence potential at each of the tested doses compared to the isolates obtained from the salad and celery associated with the same outbreak ( Fig 5A). On the other hand, strain LS414 a clinical isolate related to the coleslaw outbreak [5] expressed significantly higher virulence potential at the lower doses of 10 5 and 10 4 CFU/larva, whereas doses of 10 6 CFU/larvae were equally lethal for both the clinical and food isolate ( Fig  5B). Lastly, strain LS412, a clinical isolate from the Jalisco cheese outbreak [5] resulted in significantly (P<0.05) higher mortality of larvae at the doses of 10 6 and 10 4 CFU/larva, while difference was not significant at the doses of 10 5 CFU/larva ( Fig 5C). Overall, 11 out of 13 tested clinical isolates demonstrated significantly higher virulence potential compared to their pairs isolated from the foods related to the same listeriosis outbreaks at the inoculum of 10 6 CFU/larva. At the lower concentrations of 10 5 and 10 4 CFU/larva all the clinical isolates  Listeria monocytogenes virulence assessment in Galleria model demonstrated significantly (P<0.05) higher virulence than their pairs isolated from food. We did not observe any correlation between serotypes and severity of the effect since strains LS660 (serotype 1/2a), LS411 (serotype 4b) and LS740 related to the 2011cantaloupe listeriosis outbreak [6] (serotype 1/2b); all the strains showed similar LT 50 of 24 hours at the inoculum concentration of 10 6 CFU/larva and no significant difference in mortality of larvae.

L. monocytogenes strains related to outbreaks of listeriosis with invasive and non-invasive (gastroenteritis) symptoms
In order to compare the virulence of the strains related to febrile gastroenteritis and invasive outbreaks we compared isolates from three different outbreaks belonging to serotype 1/2b and 4b (Table 4). Comparison has been conducted between clinical (Fig 6A, 6B and 6C) and Listeria monocytogenes virulence assessment in Galleria model food isolates (Fig 6A1, 6B1 and 6C1) of same serotype related to non-invasive and invasive outbreaks at the inoculum concentrations of 10 6 CFU/larva (Fig 6A and 6A1)), 10 5 CFU/ larva (Fig 6B and 6B1) and 10 4 CFU/larva (Fig 6C and 6C1). Both clinical and food isolates related to invasive outbreaks showed significantly (P<0.05) higher virulence reflected as a percentage of larval mortality over the monitoring period of 168 hours. Statistical analysis were conducted relative to illness type (invasive vs non-invasive) for clinical and food isolates separately. Results from other strains listed in the Table 4 relative to illness type, not presented in Fig 6, also revealed significantly (P<0.05) higher virulence of clinical and food isolates related to invasive listeriosis outbreaks compared to the non-invasive febrile gastroenteritis outbreaks.

Discussion
Due to the ubiquitous presence and wide distribution of Listeria spp. in the environment, assessment of their pathogenicity and virulence potential is of high importance. As the only food-borne human pathogen among Listeria spp., L. monocytogenes is of particular interest. Extensive research into human infectious diseases often requires the use of various animal species as experimental models. The most common mammalian models used are rodents, specifically mice [22]. Mice are used because of their easy availability, relatively easy manipulation, shorter life cycle and high reproductive rates. Furthermore, employment of animal models allows control of conditions and variables associated with the experiment as well as use of large number of animals necessary for valid statistical analysis. However, in recent years, use of mammalian animal models has become increasingly less acceptable. This led to the introduction of various alternative insect and nematode models [11,[23][24][25]. These organisms Listeria monocytogenes virulence assessment in Galleria model developed a very effective immune system whose function relies on humoral and cellular innate mechanisms-mimicking human host. Both, the nematode C. elegans [11] and the fruit fly, D. melanogaster [26] offer several advantages, but the inability of C. elegans worms to incubate at 37˚C and lack of colonization and dissemination [27] and self-fertilization and narrow choice of infection routes in D. melanogaster [28] make them less suitable models for the study of Listeria spp. Invertebrate infection model of greater wax worm G. mellonella has been employed in virulence and pathogenicity studies of various microbes including L. monocytogenes [16,17]. Table 4. LT 50 of G. mellonella larvae after inoculation with L. monocytogenes isolates* related to the major listeriosis outbreaks**. Comparative analysis of various Listeria spp. and serotypes are also available [16]. However, the virulence assessment of different L. monocytogenes clinical and food /environmental isolates has not been previously reported. In this study we demonstrate different pathogenic potentials of clinical and food isolates related to the major historic outbreaks of listeriosis. Additionally, we compare isolates from the listeriosis outbreaks with systematic (invasive) symptoms with the isolates related to the listeriosis outbreaks with only febrile-gastroenteritis symptoms. Finally, we report the difference in pathogenicity of extended panel of L. monocytogenes isolates with single or multiple mutations in the virulence and stress related genes. Several previous studies indicated variability in virulence among L. monocytogenes strains [29,30]. The advantage of the employment of G. mellonella as a model for the assessment of L. monocytogenes virulence is that a large number of strains can be tested at the different infectious doses in a relatively short time and at low cost. These tests are necessary in order to better understand mechanisms of virulence and the pathogen-host immune response. Monitoring of Listeria monocytogenes virulence assessment in Galleria model the populations of Listeria spp., in the insect host, 2 and 24 hours post inoculation revealed different growth potentials among tested strains. Fast decrease in population, recorded by CFU count 2 hours post inoculation, indicates effective immune response of the insect as previously reported by Mukherjee et al. [16]. However, most of the tested strains recovered and grew thereafter (Table 2), with the exception of L. monocytogenes hly mutant. Our results demonstrated that the level of growth of pathogen inside the host does not necessarily correlate with the virulence (LT 50 ).

Strain (origin/serotype) Source/Illness type LT 50 (hours) at10 6 CFU/larva LT 50 (hours) at10 5 CFU/larva
The evaluation of L. monocytogenes strains isolated from food (n = 15), patients diagnosed with listeriosis (n = 13) and the environment (n = 6) demonstrated strong strain and dose dependent virulence potential of the pathogen. These results are also in agreement with Comparison of the virulence potential of clinical and food isolates of L. monocytogenes strains related to listeriosis outbreaks with invasive and non-invasive (gastroenteritis) symptoms. Parallel testing of the clinical isolates of the same serotype related to non-invasive (LS405designated with black square and LS403-designated with dark gray square) and invasive (LS740-designated with white square and LS414-designated with light gray square) outbreaks at the inoculum concentrations of 10 6 CFU/larva (6A), 10 5 CFU/larva (6B) and 10 4 CFU/larva (6C); Paired food isolates related to non-invasive (LS404-designated with black square and LS402-designated with dark gray square) and invasive (LS670-designated with white square and LS413-designated with light gray square) outbreaks have also been compared at the inoculum concentrations of 10 6 CFU/larva (6A1), 10 5 CFU/larva (6B1) and 10 4 CFU/larva (6C1); a < b (P<0.05), c < d (P<0.05).
https://doi.org/10.1371/journal.pone.0184557.g006 available reports [16,17,31]. The novelty of this study, however, is that our data indicates higher virulence capacity of the clinical L. monocytogenes isolates when compared with their pairs isolated from foods related to the same outbreaks of listeriosis. Such difference could be attributable to epigenetic differences in the isolates originating from different sources as previously suggested [32,33]. These authors found SNPs identified in human isolates diverge less than those from food/environmental isolates. They speculate that regulated in-host environment provides less survival pressure for pathogens compared to the environmental conditions. In our study, differences in virulence capacity were confirmed by higher larval death and a shorter LT 50 caused by clinical isolates compared to the food isolates. The mortality and morbidity outcome of listeriosis outbreaks are complicated and depend on a host of factors including amount of pathogen ingested, food type and immune function of the consumers. Another important factor could be genotype and phenotype of the organism. A simple screening of virulence potential of L. monocytogenes strains involved in different outbreaks may provide some useful data for correlation between the severity of the disease and their virulence potential. If this is true, such information could be useful in refining Listeria risk assessment and may provide additional means of controlling human listeriosis.
Human listeriosis in healthy populations may results in self-limited gastroenteritis or can be asymptomatic; therefore such cases are rarely recorded. Several outbreaks of febrile gastroenteritis caused by L. monocytogenes have been reported [3]. The non-invasive outbreaks are often associated with high attack rates (50-80%) without any reported hospitalization and/or death. Among the tested L. monocytogenes strains three food-clinical pairs were from noninvasive febrile gastroenteritis listeriosis outbreaks [3]. Although such cases are not reported these three outbreaks had significant number of affected individuals and were linked to single food sources. Parallel testing of isolates related to the non-invasive and invasive listeriosis outbreaks revealed higher virulence potential of isolates belonging to the same serotypes and from the latter outbreak. The difference is consistent regardless of source of the tested isolates (food or clinical). Strains related to the non-invasive cases of listeriosis (with the symptoms of gastroenteritis) showed significantly (p< 0.05) lower LT 50 when compared with isolates related to the invasive cases with severe listeriosis symptoms. The extent of difference was, however, dose and strain dependent. Laksanalamai et al. [34] compared the strains with a pan-genomic microarray and showed significance difference in genotype and transcriptome profiles between invasive and non-invasive strains. Our results with Galleria model showed that the non-invasive strains are also less virulent than the invasive strains. Similar results have also been reported by Franciosa et al., 2001 [35], suggesting the differences in DNA sequences as a possible cause of such occurrence. Previous studies with a custom designed microarray also indicated that there are differences in gene contents between the isolates from non-invasive and invasive listeriosis outbreak strains [34]. Further study is needed to understand if any one or more of these genes or difference in transcription or both are responsible for the difference in virulence observed in the present study. Although majority of the known listeriosis outbreaks had been associated with the serotype 4b strains, we did not observe any significant serotype related differences.
The genetic basis of L. monocytogenes pathogenesis is well characterized [36]. Among the genes identified, prfA plays a significant role in controlling L. monocytogenes pathogenicity [37]. The prfA gene not only controls transcription of a set of virulence related genes it also controls stress related genes and genes may not be directly associated with the pathogenicity [38]. Along with the emergence of L. monocytogenes as a major food-borne pathogen, molecular mechanisms associated with pathogen's virulence have been extensively investigated. Creation of the isogenic mutants with deletion in the major virulence or environmental stress genes proved to be a useful tool in the assessment of L. monocytogenes pathogenicity. We used a panel of 16 wild type and mutant strains which included in-frame deletions in prfA, hly, actA, plcA, plcB, inlA, inlB, inlAB, virR, virS, and sigB genes. Although all of these mutants have been studied in some animal models and in-vitro tissue culture assays, to our knowledge, this is the most complete isogenic mutant panel tested in the G.mellonella model reported so far. A significantly (p< 0.05) higher virulence potential of the parent L. monocytogenes strain compared to isogenic mutants with the deletions in prfA, plcA, plcB, hly, actA genes, was observed in this study. Deletion of the actA gene in L. monocytogenes resulted with the significant decrease of larval mortality. This does not come as a surprise knowing the importance of actin based motility for the infection in vertebrate cells. Mukherjee et al. [16] previously detected intracellular rapid movement of bacteria or intracellular presence of actin "cloud", which indicate similarity of the infection development in Galleria hemocytes and vertebrate cells. The significant decrease in the larval mortality caused by the absence of these five key well characterized virulence factors in L. monocytogenes confirmed G. mellonella as a useful animal model for the assessment of pathogenicity and virulence of L. monocytogenes. Thus the Galleria model may be useful in identifying newer virulence related genes of this foodborne pathogen.
The virRS system, one of the 15 confirmed two-component systems harbored by L. monocytogenes, has important role in the control of the cell envelope stress response [20]. This twocomponent system is comprised of a set of genes controlled by the virR regulator and the putative cognate histidine kinase of VirR known as VirS. In this study we compared isogenic mutants with the deletion of the virR and virS in Galleria. Our results indicate that the virS gene may not be significant for L. monocytogenes virulence since mortality of larvae (LT 50 ) inoculated with ΔvirS did not differ from the parental strain. This finding is in agreement with a previous report by Mandin et al. [20] who found no difference in LD 50 between ΔvirS mutant, inactivated by deleting 200 bp in the middle of the gene, and the parent L. monocytogenes strain EGD, while the LD 50 of ΔvirR mutant was much higher (10 5 CFU) compared to EGD (10 4 CFU). In our study, deletion of the virR gene resulted in a significantly (p< 0.05) lower mortality of Galleria when compared to the wild type L. monocytogenes and virS mutant. The role of virR in antimicrobials and organic acid salts resistance has already been described [19]. These same authors also reported various antimicrobial activities of G. mellonella induced by the antimicrobial hemolymph proteins. We speculate that deletion of virR gene led to the stimulation of immune response of Galleria larvae leading to higher LT 50 (lower virulence). Absence of virS, however, could have been compensated by the presence and activity of other histidine kinases or other activation mechanisms [20].
The alternative sigma factor, sigB, through regulation of host genes, enables survival of L. monocytogenes under challenging environmental conditions and also controls to the transcription of the gene encoding virulence regulator prfA [39,40]. In this study, we investigated the effect of sigB deletion on pathogenic potential of L. monocytogenes. The results demonstrated that pathogenicity of ΔsigB mutant had not significantly differed from that of the parental L. monocytogenes strain. This could be due to the direct inoculation into the hemocoel, hence avoiding stressful conditions during the gastrointestinal passage. Joyce et al. 2010 [17] also reported similar reduction of Galleria haemocytes 24 hours after infection with wild type L. monocytogenes and ΔsigB mutant.
L.monocytogenes harbors a host of internalin genes of which InlA and InlB play a significant role in the invasion of the targeted cells by L. monocytogenes [41]. Testing of the L. monocytogenes mutants with deletions in inlA, inlB and inlAB revealed no significant difference in the virulence potential between these mutants and the wild type parent strain at any of the tested doses. In fact, inlB and inlAB mutants showed higher LT 50 compared to the wild type strain and inlA mutant, with a statistically significant (p< 0.05) difference at the infection doses of 10 5 and 10 4 CFU/larva. In the absence of a genome sequence for Galleria and lack of better understanding of the host immune response, we can only speculate about the possible reasons for these effects. Direct injection of bacterial inoculum into the larval hemocoel may bypass the need for InlA function as previously observed in mice after intravenous infection [16]. Unlike the mouse model, where both inlA and inlB mutants did not show any pathogenic potential, only inB mutant revealed lower pathogenicity in Galleria. We are of the opinion that this could be host and cell-specific property of L. monocytogenes. This could be also due to some other function/s of InlB that is not relevant in mouse model of infection.
Our results confirmed dose dependent killing of G. mellonella larvae with LT 50 of 24 hours or less at the highest dose of 10 7 CFU/larva. These results however applied to both non-L. monocytogenes strains as well as L. monocytogenes and are in agreement with the previous findings [16,17]. These could be due to some non-specific toxicity related to overloading the immune system with such a massive dose. Lower doses, however, demonstrated significant difference in virulence between pathogenic and non-pathogenic Listeria spp. This is expected as only L. monocytogenes and L. ivanovii are known to contain all the known virulence related genes, and both of them are considered to be animal pathogens with the exception that L. ivanovii has rarely caused human infection [42]. Our results thus support the previous findings that the Galleria model can be a suitable alternative model for L. monocytogenes infection assessment. Thus, it appears that this model may be very suitable to screen other Listeria species identified [16] from different environmental sources which would lead to increased understanding about the role of these species in human and animal listeriosis.