Refrigeration of eggs influences the virulence of Salmonella Typhimurium

Salmonella Typhimurium is a human pathogen associated with eggs and egg-derived products. In Australia, it is recommended that eggs should be refrigerated to prevent condensation that can enhance bacterial penetration across the eggshell. Except for the United States, the guidelines on egg refrigeration are not prescriptive. In the current study, in-vitro and in-vivo experiments were conducted to understand the role of egg storage temperatures (refrigerated vs ambient) on bacterial load and the virulence genes expression of Salmonella Typhimurium. The in-vitro egg study showed that the load of Salmonella Typhimurium significantly increased in yolk and albumen stored at 25 °C. The gene expression study showed that ompR, misL, pefA, spvA, shdA, bapA, and csgB were significantly up-regulated in the egg yolk stored at 5 °C and 25 °C for 96 h; however, an in-vivo study revealed that mice infected with egg yolk stored at 25 °C, developed salmonellosis from day 3 post-infection (p.i.). Mice fed with inoculated egg yolk, albumen, or eggshell wash stored at refrigerated temperature did not show signs of salmonellosis during the period of the experiment. Data obtained in this study highlighted the importance of egg refrigeration in terms of improving product safety.

pathogenicity island) 1 and SPI2 for epithelial cell invasion and survival 17 . A previous study has shown that an inoculum of less than 10 colony forming units (CFUs) of Salmonella Typhimurium is sufficient to cause clinical salmonellosis in healthy individuals 18 . Also in field conditions, the load of Salmonella on egg surface and in shed environment has shown to be variable 19 . Therefore, in the current study, the inoculum dose for Salmonella Typhimurium grown in egg components was not adjusted for mice infection.
Various studies have investigated the effects of storage temperatures on the growth and survivability of Salmonella in eggs 10,11,20 ; however, no subsequent investigations on the development of salmonellosis through the consumption of the contaminated egg contents have been conducted. The growth kinetics of Salmonella Typhimurium is regulated by temperature. For example, the survivability of Salmonella Typhimurium on eggshell is better at 22 °C, while bacterial load decreases with storage time at 4 °C 21 . Salmonella Typhimurium grows well in the yolk at either 15 °C or 22 °C, whereas its growth in albumen is temperature dependent 21 . A study has linked the up-regulation of yafD and xthA with the Salmonella Enteritidis survival in albumen 22 .
Salmonella serotypes specific differences in survivability in egg albumen 14,23 suggest that Salmonella Typhimurium may regulate its transcriptional machinery differently. To answer this question, we hypothesized that the virulence of Salmonella Typhimurium is enhanced at 25 °C compared with refrigeration temperature and, therefore, if consumed, contaminated eggs stored at ambient temperature will cause clinical disease in mice. The present study had three main objectives: (1) Understand the growth and survivability of Salmonella Typhimurium in eggs stored at 5 °C and 25 °C. (2) Determine the effects of storage temperatures on Salmonella Typhimurium gene expression in yolk, albumen and on the eggshell surface. (3) Study the virulence of Salmonella Typhimurium cultured in raw egg components in BALB/c mice.

Methods
Salmonella growth and survivability kinetics in eggs at different temperatures. Eggs preparation and inoculation with Salmonella Typhimurium. For both the in-vitro and in-vivo experiments, a pure culture of Salmonella Typhimurium definitive type 9 stored at − 80 °C in 50% glycerol was revived on nutrient agar (NA; ThermoFisher Scientific, Australia) and subcultured in Luria Bertani (LB; ThermoFisher Scientific, Australia) broth until the optical density (OD) at 600 nm reached approximately 1. Fresh eggs directly obtained from a 33-week old cage layer hen flock were sanitized by dipping in 75% ethanol for 90 s and air dried completely in a class II biosafety cabinet. To study the survivability of bacteria on the eggshell surface, individual eggs (n = 6 in each treatment group) were dipped for 90 s in 1 × 10 6 CFU/mL of Salmonella suspension prepared in LB broth. To confirm bacterial deposition on the eggshell surface, an extra set of eggs (n = 6) were dipped in the inoculum and processed immediately for the recovery of Salmonella. For studying Salmonella growth or survivability in the albumen and yolk, individual eggs received a 0.1 mL inoculum of 1 × 10 3 CFU per egg directly injected into the egg albumen or yolk and the injection holes were sealed. This dose was selected to mimic a field scenario where a low load (1.71 log 10 MPN per egg) of Salmonella on an egg has been reported 19 . All the treatment groups were incubated at 5 °C or 25 °C for either 96 h or 28 days. The 5 °C treatment eggs were put in a 55 L plastic container and stored in a cool room and the 25 °C condition eggs were stored in an incubator. Two separate hygrometers were used to monitor the relative humidity (%) levels of the two treatment groups. These two temperature conditions were selected as most likely during storage eggs are exposed to an ambient temperature close to 25 °C. The control eggs in each treatment group (n = 4 per treatment) were treated in sterile LB and processed in the same way as of the Salmonella inoculated eggs.
Quantitative recovery of Salmonella from the inoculated eggs. For estimation of Salmonella survivability on the eggshell surface, individual eggs from both the treatment groups (5 °C and 25 °C) were washed for 2 min in 10 mL of buffered peptone water (BPW; ThermoFisher Scientific, Australia) in Whirl-Pak sample bags, the rinsate was serially diluted (10 times) in phosphate buffered saline (PBS) and plated onto xylose lysine deoxycholate (XLD; ThermoFisher Scientific, Australia) agar plates for overnight incubation at 37 °C. Salmonella culturability in albumen and yolk of the stored eggs was also determined. Eggs were opened into a sterile 90 mm petri dishes, then poured into sterile bags for thorough mixing, and using a syringe, 0.1 mL of either albumen or yolk was added to 0.9 mL of BPW and mixed. Serial tenfold dilutions were prepared in PBS and plated onto XLD agar plates. The plates were incubated overnight at 37 °C and Salmonella colonies were enumerated to determine the total number of bacteria in each sample. The egg contents of the 5 °C stored samples were enriched for the qualitative assessment of Salmonella following a previously described method 19 . Briefly, 1 mL samples of yolk or albumen of individual eggs were enriched in 9 mL of BPW and incubated overnight at 37 °C. From the incubated samples, 100 μL was added to 10 mL of Rappaport-Vassiliadis soya peptone (ThermoFisher Scientific, Australia) broth and incubated overnight at 42 °C. The samples were streaked onto XLD, incubated overnight at 37 °C and read as positive or negative for characteristic Salmonella colonies. To rule out the chances of external contamination, control groups were processed in the same manner. The growth of Salmonella was recorded as CFU/mL of albumen or yolk or eggshell surface rinsate. The in-vitro egg experiments were repeated for a total of three times.
Role of temperature and storage time in regulation of genes of Salmonella in egg components. To obtain good quality RNA from Salmonella inoculated egg components, albumen and yolk were inoculated with 1 × 10 8 CFU/mL of Salmonella, while for the eggshell surface inoculation, intact individual eggs were dipped for 90 s in BPW (prepared in 1 L sterile beaker) that contained 1 × 10 9 CFU/mL of Salmonella. Based on the protocol of TRIzol Reagent (Invitrogen, Australia), at least 1 × 10 7 cells of bacterial origin are required to extract sufficient RNA. Our pilot experiment showed that 1 × 10 6 CFUs of Salmonella per mL of egg content was not sufficient for quality RNA extraction. www.nature.com/scientificreports/ Preparation of egg components. Intact eggs were sanitized in ethanol and allowed to dry thoroughly as described previously. As the volume of albumen and yolk vary from egg to egg; therefore, for precise measurement of per mL volume for Salmonella inoculation, egg contents were separated in sterile petri dishes. Maximum albumen (both the colloidal and watery parts) or yolk from individual eggs was aspirated using separate syringes and decanted into 50 mL Falcon tubes. The quantity of albumen or yolk in the tubes was adjusted based on Salmonella inoculum (1 × 10 8 CFU/mL of yolk or albumen). There were 3 biological replicates for each of the egg components.
Preparation of Salmonella inoculum. A stock culture of Salmonella Typhimurium was revived on NA media plate and a single colony was subcultured in LB broth until the OD (measured at 600 nm) of the culture reached approximately 1. The inoculum dose for the egg surface was prepared in LB, while for the egg internal contents, Salmonella culture was pelleted by centrifugation (4000 × g for 15 min). The pellet was resuspended in 1 mL of PBS and its OD at 600 nm was read as described earlier. The culture was diluted to obtain 1 × 10 9 CFU/mL of Salmonella and a 0.1 mL of this inoculum was added into either 0.9 mL of yolk or albumen. Thus, the final dose of Salmonella was 1 × 10 8 CFU/mL of egg content. For the positive control group, Salmonella Typhimurium was incubated (with 3 biological replicates) in a shaking incubator at 37 °C in LB broth for 12 h and 1 × 10 8 CFU/mL was processed for RNA extraction.
Salmonella RNA extraction from egg components and pure culture of bacteria. To recover Salmonella from the egg surface for RNA extraction, individual eggs (n = 3) at each incubation time-point (12,72, and 96 h) from each treatment group (5 °C and 25 °C) were rinsed for 2 min in 5 mL BPW. The rinsate was centrifuged at 5000 × g for 10 min. The bacterial pellet was resuspended in 0.25 mL of PBS into which 0.75 mL of TRIzol Reagent was added and maintained on ice. To recover Salmonella from egg contents, 1 mL of albumen or yolk (n = 3) at each incubation time-point (12, 72 and 96 h) and treatment group (5 °C and 25 °C) was solubilized in 1 mL of sodium citrate (0.2 M, pH unadjusted), vortexed and centrifuged at 18,000 × g for 5 min at 4 °C. The supernatant was discarded, and the bacterial pellet was resuspended in 0.25 mL of PBS into which 0.75 mL of TRIZol Reagent was added and the samples were maintained on ice. For the extraction of bacterial RNA from pure culture, 1 mL of Salmonella inoculated LB broth (1 × 10 8 CFU/mL) was centrifuged at 18,000 × g for 5 min at 4 °C, the pellet was resuspended in 0.25 mL of PBS into which 0.75 mL of TRIZol Reagent was added and the samples were maintained on ice until processed for RNA extraction. The TRIZol Reagent added samples were briefly homogenised using a hand-held IKA T10 Basic Ultra Turrax Homogenizer (Wilmington, NC, USA) while maintaining the tubes on ice in 20 mL container. The homogenised samples were incubated for 7 min on ice and 0.1 mL of 1-Bromo-3-chloropropane (BCP; Sigma, Australia) was added into each sample. The samples were incubated for 3 min at room temperature and centrifuged at 12,000 × g for 15 min at 4 °C to separate RNA from DNA and protein. Maintaining the samples on ice, from the top aqueous layer, 0.5 mL was transferred into new tubes and the process for RNA extraction was completed as per the manufacturer's instructions. The quality and purity were tested in Nanodrop 1000 and the RNA was stored at − 80 °C until used for cDNA synthesis and Fluidigm PCR.
Primer, cDNA synthesis and Fluidigm PCR. A total of 95 primers for targeted genes of Salmonella Typhimurium involved in stress response, virulence, survivability, metabolism, and host colonization were designed using NCBI as described in our previous publications 24,25 and were synthesized by Merck, Australia. The main aim behind selecting these genes was to understand their regulations in egg components about storage temperature and their subsequent role in causing clinical salmonellosis in mice. Using PCR (see Supplementary material Text S1) and 2% agarose gel electrophoresis, all the primers were tested for target specificity in cDNA synthesized from Salmonella Typhimurium extracted RNA. To cross check the issue of genomic DNA contamination, RNA samples were also included in the PCR run. cDNA was synthesized from the extracted RNA as per the protocol of the QuantiTect Reverse Transcription Kit (Qiagen, Australia). Briefly, 1 μg of RNA per sample was subjected to genomic DNA removal in a 14 μL total reaction volume for 3 min at 42 °C as per the protocol of the kit. For the reverse transcription step, the samples (in 20 μL reaction volume) were incubated for 25 min at 42 °C and to inactivate the Quantiscript reverse transcriptase, a final incubation step for 3 min at 95 °C was executed. The cDNA samples were stored at − 80 °C until used for downstream analysis. A Fluidigm PCR was performed on all the cDNA samples using 95 genes in a 96.96 DYNAMIC ARRAY IFCS at the Australian Centre for Cancer Biology, Adelaide (Supplemental material Text S2). The quantification cycle (Cq) values obtained were analyzed for relative fold change expression (2 −ΔΔCq ) using the data of Salmonella Typhimurium grown in LB broth for 12 h as a reference control and rrsG (16S rRNA) as a reference gene for data normalization. The rrsG was selected as a reference gene based on its least standard deviation value in the current experimental conditions. For a clear interpretation of up-or down-regulated genes, the relative expression data were presented as log 2 fold change. Mice inoculation with egg components containing Salmonella. Mice were divided into 16 treatment groups with 7 mice per group (Table 1). Yolk, albumen, and eggshell wash used as inoculums were prepared as described for the in-vitro studies above. Briefly, the inoculated eggs were stored at either 5 °C or 25 °C for 96 h and individual mice received through oral gavage 0.1 mL of shell wash, yolk, or albumen that had either been inoculated or non-inoculated with Salmonella Typhimurium. The positive control groups received through oral gavage 1 × 10 3 CFU/ 0.1 mL of Salmonella Typhimurium in LB broth stored at either 5 °C or 25 °C (inoculum adjusted) for 12 h. Further detail of the inoculum that individual mice in each treatment group received is outlined in Table 1. The leftover inoculums from the treatment groups were maintained on ice, serially diluted and plated on XLD agar to confirm the actual dose of Salmonella Typhimurium that mice received. The experiment was conducted over 21 days. During the post-infection period (day 0 to 21), the mice were routinely observed for clinical signs of non-typhoidal salmonellosis (e.g. lethargy, hunching, ruffled fur) and mortality. Adhering to the Animal Ethics Committee Guidelines, mice suffering from the clinical disease with a score of five or above were humanely euthanized using carbon dioxide and the collected organs were processed for the quantification of Salmonella.

Effect of
Salmonella detection in mice feces. From all the treatment groups, fecal samples were collected on day 3, 6, 9, 12, 15, and 18 post-infection and processed for Salmonella isolation. The cage and bedding materials were changed at every fecal sampling to minimize the risk of carryover. For the qualitative assessment of Salmonella, 1 g of fecal samples was mixed into 9 mL of BPW, vortexed and incubated overnight at 37 °C. The samples were processed following the RVS enrichment method previously described 19 . The plates were then examined to determine if the samples were positive (scored as 1) or negative (scored as 0) for Salmonella Typhimurium.
Salmonella load in organs. At the point of cull, segments of liver, spleen, ileum and cecum were collected and homogenized in tubes containing 0.5 mL of 0.9% saline and stainless-steel beads (2-8 mm). Homogenates were serially (tenfold) diluted and processed for the quantitative and qualitative assessment of Salmonella. For the quantitative assessment, 0.1 mL of the homogenate was directly plated onto XLD, incubated overnight at 37 °C and colonies were counted. The CFU data were expressed as log 10 load of Salmonella per gram of the organ. Statistical analysis. The Salmonella load data were analyzed in GraphPad Prism version 8.0 using one-and two-way ANOVA with Tukey's multiple comparison test for determining the level of significance (P < 0.05). The relative gene expression fold change was calculated by 2^− ΔΔCq method and LB 12 h was used as a reference control. The log 2 fold change data between the 12, 72 and 96 h were analyzed in StatView version 5.0.1 by repeated measure analysis and the level of significance was determined by PLSD (P < 0.05).

Results
Growth kinetics and survivability of Salmonella in egg components. Post-inoculation with Salmonella, the relative humidity at both 5 °C and 25 °C varied from 76 to 82% over the 96 h egg storage experiment. The in-vitro experiment showed that Salmonella Typhimurium inoculated into egg yolk grew significantly to log 10 7.88 CFU/mL at day 4 of storage at 25 °C, whereas its growth plateaued from day 4 until day 28 of storage (Fig. 1A). This showed that after 96 h of storage, the load of Salmonella Typhimurium increased by log 10 www.nature.com/scientificreports/ 3.88 CFU/mL in the yolk. The data showed that when inoculated in egg albumen, by day 4 post-inoculation, Salmonella Typhimurium migrated from albumen and grew to log 10 5.70 CFU/mL in the yolk (Fig. 1A). Load of Salmonella Typhimurium in the albumen of albumen-inoculated eggs remained constant from day 0 to day 4 of storage and then increased to log 10 6.84 CFU/mL (Fig. 1B). Overall, these data showed that the growth of Salmonella Typhimurium significantly increased in the yolk within 96 h of storage at 25 °C of the inoculated eggs. In the qualitative assessment, 75.6% of the yolk and 15.6% of the albumen samples of the eggs stored at 5 °C were positive for Salmonella after the enrichment process ( Supplementary Fig. S1A), demonstrating that Salmonella Typhimurium survived better in yolk than albumen over the 28 days of storage ( Supplementary Fig. S1B). As the samples stored at 5 °C did not show growth of Salmonella after the direct plating; therefore, in the subsequent in-vitro experiments, inoculated eggs were stored only for 96 h. In contrast to 25 °C, Salmonella Typhimurium counts remained constant for yolk and albumen samples stored at 5 °C for 96 h (Fig. 2A). However, the load of Salmonella Typhimurium was significantly higher in the yolk compared with the albumen. Salmonella survived numerically higher on the eggshell surface stored at 5 °C compared with the 25 °C (Fig. 2B). For the eggshell treatment groups, on average, each egg had 9.7 × 10 5 (or 5.99 on log 10 scale) CFU of Salmonella, which was just under 1 mL of inoculum that contained 1 × 10 6 CFU. Therefore, the survivability of Salmonella on the eggshell surface stored at 5 °C for 96 h was 3.53 CFU/mL (on a log 10 scale) that showed a 0.46 log reduction (Fig. 2B). The survivability of Salmonella on the eggshell surface stored at 25 °C for 96 h was 2.22 CFU/mL that showed a 1.76 log reduction (Fig. 2B). Overall, the eggshell data showed that Salmonella Typhimurium survived numerically better at 5 °C compared with the 25 °C storage temperature.
Salmonella transcription profile is affected by egg storage temperature. To understand the effects of storage temperatures on the gene expression of Salmonella Typhimurium in the yolk, albumen and on eggshell, Fluidigm PCR was performed on Salmonella RNA extracted from inoculated egg contents whereas  Stress response, host colonization, virulence and invasion. Among the genes involved in host colonization, aggregation and infection, the expression levels of bapA, cadB, cadC, misL, ompR, shdA, spvA and spvC in at least one egg content and storage time-points were up-regulated (Fig. 3). Among them, bapA is involved in biofilm formation and host colonization, and the deletion of bapA has shown reduced colonization potential in the gut of mice 26 . bapA was significantly up-regulated in the yolk and its mean expression was significantly higher in the 25 °C stored yolk compared with the 5 °C (Fig. 3A). However, in albumen and on the egg surface, the expression of bapA was not consistent, where it was down-regulated on the egg surface of the samples stored at 25 °C. cadB, cadC and misL were up-regulated in the yolk stored both at 5 °C and 25 °C, while in albumen, these genes were up-regulated at 96 h of incubation in the samples stored at 5 °C (Fig. 3B-D). Among the genes involved in stress response, ompR was the only up-regulated gene across all the samples at the stored conditions (Fig. 4A). For albumen and egg surface samples, the mean fold change of ompR was significantly higher at 5 °C compared with the 25 °C stored samples. shdA was not consistently regulated in albumen and on egg surface, while in yolk, it was consistently upregulated both at 5 °C and 25 °C at the storage time-points (Fig. 4B). The expression of sdhA is involved in colonization as shown in mice challenged with Salmonella Typhimurium 27 . spvA and spvC were up-regulated in yolk but were down-regulated in albumen and on egg surface stored at 5 °C and 25 °C for 96 h (Fig. 4C,D).
Flagella, fimbrae and biofilm formation. In biofilm formation, both curli and cellulose synthesis are co-regulated by a complex regulatory network in which csgD (agfD) acts as a global regulator 28 in the expression of csgB and csgA 29 . In the yolk stored both at 5 °C and 25 °C, csgB, fimH, pefA and pefB were up-regulated across all the time-points, while their up-regulation was consistent in albumen and on the surface of eggs across the storage time-points (Fig. 5A-D). In the albumen stored at 5 °C, csgB, csgD, fimH, pefA were significantly up-regulated at only 96 h. The expression of bapA is up-regulated in the formation of curli and cellulose through the involvement of csgD 26 . This indicates that Salmonella on the egg surface at 5 °C resisted well to the low temperature.  www.nature.com/scientificreports/ Glycolysis, purine and pyrimidine metabolism. Multiple genes involved in cell metabolism were up-regulated in the yolk stored both at 5 °C and 25 °C for 96 h (Fig. 6A,B). Among the investigated genes, carA, carB, purD, pyrB, pyrD and pyrE were up-regulated in the yolk stored both at 5 °C and 25 °C across all the storage time-points, while the up-regulations of purE, purG and pyrC were inconsistent (Fig. 6A,B). Unlike samples stored at 5 °C, there were not many Salmonella genes significantly up-regulated in albumen stored at 25 °C. In the albumen stored at 5 °C, adK and yaiC (adrA) were significantly up-regulated at 96 h, while aroC, fbp, pgm, ptsG, purA, purD, purE, purF, pyrB, pyrC, pyrD and pyrE were up-regulated both at 72 and 96 h. At the surface of eggs stored

Two-component system and Salmonella pathogenicity islands.
Our data showed that the phoP and phoQ involved in the two-component system were significantly down-regulated across all the sampling time-points in egg components, except the albumen stored for 72 and 96 h at 5 °C (Fig. 6A). This positively correlated with the expression of other invasion genes such as prgH. All of the investigated genes involved in the regulation of type III secretory systems, such as ssaB, ssaC, ssaD, ssaE, ssaG, ssaH, ssaI, sseA, sseB, sseE, ssrA and ssrB were significantly down-regulated in the egg components stored at 5 °C and 25 °C for 12, 72 and 96 h except the sseA that was up-regulated on the egg surface at 5 °C. hilA, the transcriptional activator of SPI1 and co-activators such as hilC and hilD were all down-regulated in the egg components stored for 96 h both at 5 °C and 25 °C.
Other genes that inconsistently upregulated in the yolk, albumen and on egg surface were proP, narL, narX, rpoS, xthA and yafD. In brief, bapA, csgB, fimH, misL, ompR, pefA, pefB, purD, pyrD, pyrE, shdA, spvA and spvC consistently up-regulated across all the three storage time-points in the yolk are implicated in maintaining the virulence of Salmonella in yolk. Interestingly, these genes were not consistently up-regulated in the albumen and on the egg surface.
Salmonella Typhimurium invasiveness in mice and development of clinical disease. The treatment groups that received Salmonella inoculated yolk stored at 25 °C for 96 h and the positive control group that received Salmonella in LB broth stored at 25 °C for 12 h were culled at day 6 and day 8 post-infection, respectively, due to the development of clinical signs with a score of 5 and above. The percent survival rate of the clinically sick mice was very low compared with all other treatment groups that survived until culled at day 21 p.i. (Supplementary  Fig. S2A). There was no significant difference in the percent survivability in the treatment groups that received yolk and LB stored at 25 °C. Compared with day 0 of infection, treatment groups that developed clinical signs after day 3 p.i. had 100% morbidity (Supplementary Fig. S2B). During the mice dissection, gross pathological lesions including the emptied gastrointestinal tract and enlarged liver and spleen were observed in the Salmonella in yolk and LB inoculated groups.

Fecal shedding profile of Salmonella inoculated mice and Salmonella load in organs.
Prior to the challenge with Salmonella Typhimurium, feces were collected from the mice and tested for the presence of Salmonella. All mice were Salmonella negative prior to commencing the infection challenge. Salmonella was not detected in the feces www.nature.com/scientificreports/ of control mice or mice infected with egg contents stored at 5 °C at any time-point during the experiment. Feces collected from mice inoculated with LB or yolk stored at 25 °C were positive for Salmonella at day 3 and 6 post infection. Mice inoculated with shell wash and egg albumen stored at 25 °C containing Salmonella Typhimurium were culture negative for the bacteria in their feces at day 3, 6, 9, and 12 post-infection. At day 15 and 18 postinfection, however, fecal samples collected from mice in these two treatment groups were culture positive for Salmonella Typhimurium (Fig. 7A). Salmonella was enumerated from all the organs that were sampled from mice culled during the trial. The organs collected from the treatment groups that did not shed Salmonella in the feces were also negative for Salmonella, after the enrichment culture method. Salmonella load was higher (P < 0.05) in cecum, ileum, liver and spleen of the mice fed with Salmonella in LB and yolk compared with the mice received Salmonella in albumen and on shell wash stored at 25 °C (Fig. 7B).

Discussion
The main objective of this study was to understand the temperature driven changes in Salmonella Typhimurium inoculated into table eggs. The in-vitro study showed that Salmonella Typhimurium grew significantly in yolk by day 4 of storage at 25 °C and from day 4 to day 28, the growth plateaued. A significantly higher load in the albumen inoculated eggs stored at 25 °C on day 7 compared with day 4 showed that Salmonella Typhimurium grew in albumen at a slower rate. The slower growth rate of Salmonella Typhimurium in albumen compared with yolk could be due to its antimicrobial properties as shown for Salmonella Enteritidis 30 . The increased load of Salmonella Typhimurium in the yolk of albumen inoculated eggs confirmed that yolk favoured its growth. Salmonella Enteritidis has been shown to replicate faster in yolk compared to the albumen of eggs stored at 25 °C 31 . In the albumen inoculated eggs stored at 25 °C, the higher CFUs obtained from yolk further confirmed that Salmonella Typhimurium migrated from the albumen and exhibited significant replication.
The Salmonella Typhimurium load data in the egg contents indicate that if eggs are stored at ambient temperature, a low load of Salmonella can increase in the egg contents significantly. A higher Salmonella load on the egg surface at 5 °C compared with the 25 °C confirms that Salmonella has the potential to survive on the egg surface at refrigerated temperature and, therefore, it is important to follow proper sanitation guidelines in kitchens while handling eggs. No increase in the load of Salmonella in egg albumen and yolk stored at 5 °C supports the notion www.nature.com/scientificreports/ that eggs should be stored at refrigerated temperature to reduce the growth rate of Salmonella. It is important to note that a different inoculum dose was used on egg surface compared to egg internal contents. This was to mimic the field conditions (real-life scenario) where Salmonella infected flock can lay eggs with up to 10 6 CFU/ egg. In order to get an inoculum of 10 3 for mice challenge from the egg surface, egg surface was inoculated with 10 6 CFU, as eggshells have numerous pores and it is possible that Salmonella Typhimurium can penetrate through to pores. Given that Salmonella Typhimurium penetration across the eggshell pores was not tested in this study, further investigation is necessary. Overall, the data in the in-vitro whole egg experiment confirmed that within 4 days of inoculation Salmonella Typhimurium grew at a faster rate in the yolk of both the yolk and albumen inoculated eggs stored at 25 °C. Therefore, in the subsequent in-vitro and in-vivo experiments, eggs were stored until day 4 post Salmonella Typhimurium inoculation. The Fluidigm PCR data showed that Salmonella Typhimurium up-regulated genes involved in cell metabolism, fimbriae formation, stress response, virulence and survival in egg. However, the expression pattern varied with temperature and egg component. An in-vitro study with egg albumen showed that when incubated below 30 °C for a maximum of 24 h, egg albumen exhibited bacteriostatic activity against Salmonella Enteritidis 32 . Differences in the survival ability of two genetically similar strains of Salmonella Enteritidis in an egg 33 show that Salmonella adopts a wide range of strategies that include nucleotide deletion or insertion. Differences in the gene expression pattern observed in the current study could be due to the variable behaviour of the bacteria, as Salmonella Typhimurium expresses its pathogenicity islands differently 34 .
The csg region in Salmonella spp., encodes protein polymers known as curli fimbriae, which promote community behaviour and host colonization. In low temperature, curli are important for cell aggregation, adhesion to surfaces and biofilm formation. The up-regulation of csgB both at 5 °C and 25 °C in the yolk, albumen and at 5 °C at the egg surface shows that Salmonella expressed the curli protein to maintain the physiology of survival and aggregation. Interestingly, in this study, the non-consistent regulation of csgD with csgB showed their independent regulations in the experimental conditions. yafD in Salmonella Enteritidis provides resistance to albumen 22 , and in the current study, a condition dependent up-regulation of expression in the albumen stored at 5 °C for 12 and 72 h was observed. The consistent up-regulation of ompR across all the sampling time-points in the egg components stored both at 5 °C and 25 °C shows its role in the survivability of Salmonella in an egg. The data showed that ompR played a greater role in Salmonella survival at 5 °C, as it acts as a central regulator in reprogramming the Salmonella transcriptome in a stressful environment 35 . Among the genes involved in osmotic stress response, the up-regulation of rpoS both in albumen and yolk at 5 °C shows that Salmonella diverted the egg component and temperature driven stress mainly through this gene. However, it is important to note that not all the genes involved in osmotic stress response were included in the current study.
The two-component signal transduction system is involved to modify the cellular output in response to environmental signals both in-vitro and in-vivo conditions 36,37 . The down-regulation of phoP and phoQ in the egg components show that Salmonella was unable to up-regulate its two-component system in the conditions applied in this study. The current study confirms that Salmonella's persistence in yolk, albumen and on the egg surface leads to the downregulation of type III secretory system. Overall, Salmonella Typhimurium regulated its transcriptional machinery differently in eggs stored at 5 °C and 25 °C, whereas the temperature driven changes affected the in-vivo virulence capacity of Salmonella in the murine model.
The findings in the in-vitro egg studies correlated with the in-vivo mice trial, where the data demonstrated that Salmonella Typhimurium in the yolk stored at 25 °C showed the invasive potential that resulted in the development of clinical salmonellosis. Salmonella load in the 25 °C stored albumen slightly increased by day 4 of storage; however, the infected mice treatment group that received 10 2 CFU in albumen did not start shedding Salmonella in feces until day 15 p.i. The yolk contains high levels of iron, which is an important requirement for bacterial growth 38 . Growth of Salmonella Typhimurium has been shown to increase in response to the presence of iron, and adhesion of Salmonella Typhimurium to epithelial cells increases when the bacteria are pre-incubated in higher iron environments 38 . Therefore, this study confirmed that egg yolk can become a higher risk food product if eggs are not stored at refrigeration temperature. Unlike yolk, albumen contains a considerably lower level of free iron 39 . The antimicrobial functions of the main components of fresh albumen, such as lysozyme, ovalbumin, ovostatin, avidin and ovotransferrin depend upon storage temperature and time, and the functions decline faster in eggs stored at 37 °C compared with 20 °C and 4 °C as shown in Salmonella inoculated eggs 40 . In the current study, no significant increase in Salmonella growth in albumen until day 4 post-inoculation could be due to the antimicrobial function of some of these proteins.
In the current study, the significantly lower percent survival rate, 100% morbidity and higher Salmonella load in cecum, ileum, liver and spleen of the 25 °C stored yolk and LB infected mice proved that yolk increased the invasive capacity of Salmonella that resulted in the development of salmonellosis. Additionally, the quantitative differences in Salmonella load in egg components stored at 5 °C and 25 °C might be contributing factors to the development of salmonellosis in mice. The culling of the sick mice on days 6 and 8 p.i. indicated that although the mice received Salmonella through yolk and albumen, Salmonella in the yolk resulted in fecal shedding from day 3, while the latter group did not shed the bacteria until day 15 p.i. Shell wash at 25 °C and albumen at 25 °C inoculated mice began shedding Salmonella from day 15 p.i. with no clinical signs of salmonellosis during the trial period. This delay in shedding could be attributed to the Salmonella being initially stressed due to the storage environment (shell wash and fresh egg albumen). The stress of the treatment group environment (shell wash and albumen) coupled with the exposure to the stressful environment of the digestive tract of the mice 41 could have caused a delay in the development of salmonellosis and fecal shedding. Although Salmonella survived in the egg components at 5 °C storage, the inoculum did not result in the fecal shedding of Salmonella in mice during the trial period. Surprisingly, this was the case for the Salmonella stored LB in the 5 °C treatment group as well. Observing gross pathological lesions during dissection in the Salmonella in yolk and LB (stored at 25 °C) inoculated groups confirmed that these mice developed salmonellosis. However, these findings were not observed www.nature.com/scientificreports/ in the mice from the albumen or shell rinsate inoculated or any other group that did not shed Salmonella in the feces. In Australia, upon collection from the layer sheds, eggs are stored at temperatures ≤ 15 °C on farms. This temperature is then maintained during grading steps at farms. Further studies are necessary to investigate the effect of farm storage temperature on Salmonella virulence.

Conclusions
Overall, the growth and survivability of Salmonella Typhimurium in egg affected by ambient temperature. The temperature influenced the virulence of Salmonella Typhimurium and the storage of inoculated eggs at ambient temperature resulted in salmonellosis. In the in-vitro study, the panel of genes assessed for their functions in maintaining the virulence of Salmonella showed that genes involved in metabolism, stress response, virulence, and colonization were down-regulated in the albumen and on the egg surface. In the in-vivo experiment, mice infected with egg wash and albumen containing Salmonella stored at ambient temperature started shedding Salmonella in feces on day 15 p.i. shows that egg components coupled with storage temperature affected the virulence of the bacteria. The data provide evidence that eggs stored at refrigerated temperature significantly reduces the risk of salmonellosis.