Salmonella can survive for long periods under extreme desiccation conditions. This stress tolerance poses a risk for food safety, but relatively little is known about the molecular and cellular regulation of this adaptation mechanism. To determine the genetic components involved in Salmonella’s cellular response to desiccation, we performed a global transcriptomic analysis comparing S. enterica serovar Typhimurium cells equilibrated to low water activity (aw 0.11) and cells equilibrated to high water activity (aw 1.0). The analysis revealed that 719 genes were differentially regulated between the two conditions, of which 290 genes were up-regulated at aw 0.11. Most of these genes were involved in metabolic pathways, transporter regulation, DNA replication/repair, transcription and translation, and, more importantly, virulence genes. Among these, we decided to focus on the role of sopD and sseD. Deletion mutants were created and their ability to survive desiccation and exposure to aw 0.11 was compared to the wild-type strain and to an E. coli O157:H7 strain. The sopD and sseD mutants exhibited significant cell viability reductions of 2.5 and 1.3 Log (CFU/g), respectively, compared to the wild-type after desiccation for 4 days on glass beads. Additional viability differences of the mutants were observed after exposure to aw 0.11 for 7 days. E. coli O157:H7 lost viability similarly to the mutants. Scanning electron microscopy showed that both mutants displayed a different morphology compared to the wild-type and differences in production of the extracellular matrix under the same conditions. These findings suggested that sopD and sseD are required for Salmonella’s survival during desiccation.
Citation: Maserati A, Fink RC, Lourenco A, Julius ML, Diez-Gonzalez F (2017) General response of Salmonella enterica serovar Typhimurium to desiccation: A new role for the virulence factors sopD and sseD in survival. PLoS ONE 12(11): e0187692. https://doi.org/10.1371/journal.pone.0187692
Editor: William C. Nierman, J Craig Venter Institute, UNITED STATES
Received: May 25, 2016; Accepted: October 10, 2017; Published: November 8, 2017
Copyright: © 2017 Maserati et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The RNA-Seq results are available in the supporting information files. The RNA-Seq raw data files have been deposited in NCBI's Gene Expression Omnibus and are accessible through GEO Series accession number GSE86580 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE86580).
Funding: This work was supported by grant #2012-67005-19613, provided by the United States Department of Agriculture, National Institute of Food and Agriculture, Agriculture and Food Research Initiative, received by RCF and FDG (https://nifa.usda.gov/program/agriculture-and-food-research-initiative-afri). AM salary was partially supported by the University of Minnesota MnDRIVE Global Food Supply Initiative scholarship (https://mndrive.umn.edu/food). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Salmonella enterica, a gram-negative bacterium belonging to the Enterobacteriaceae family, is a foodborne human pathogen that can cycle from the environment to animals and humans through their fecal matter [1–3]. Because of the wide variety of environments Salmonella can be exposed to, it can adapt to very diverse physical or chemical conditions. Generally, one of the most important factors impacting the ability of an organism to survive in a certain environment is the presence and, more importantly, the availability of water for chemical and biological reactions, a concept defined as water activity (aw). aw is expressed as the ratio between the vapor pressure of water with a solute and the vapor pressure of pure water.
A relatively high aw is essential for microbial growth, since at low aw enzymatic reactions are inhibited and metabolism is reduced [4, 5]. Gram-negative bacteria such as E. coli, Salmonella and Vibrio require an aw greater than 0.95 to grow . As a result, one of the main strategies to control bacterial proliferation in food matrices is the reduction of aw to create a low moisture environment [5, 7]. Most Enterobateriaceae are well adapted to persist in dry environmental conditions and Salmonella is no exception. To survive under harsh conditions, such as those found in a dry environment, bacteria need to activate a variety of cellular stress responses.
One of the first protection mechanisms activated by the shift from humid to dry environments is the response to osmotic stress induced by the decrease of water due to evaporation and the relative increase of the solute. In particular, this process makes the environment increasingly hypertonic, thus triggering specific molecular mechanisms that allow the cell to regulate the internal solute concentration to maintain the appropriate turgor [8–10]. Indeed, when exposed to low aw, Salmonella prevents and minimizes the loss of intracellular water and avoids membrane and protein damage due to the progressively hypertonic environment by increasing the influx of osmoprotectants. A recent study reported that up-regulation of osmoprotectant genes and operons such as proP, proVWX, and osmU can be observed after short-term desiccation on stainless steel coupons . Up-regulation of some of these same genes was also observed after 2 h exposure to aw 0.11 .
As aforementioned, osmotic protection is only one of the first mechanisms deployed by the cell in response to desiccation. A recent study from Gruzdev et al. suggested that it is likely to be part of a network of stress responses, such as oxidative and thermal stress, that act in a concerted fashion and modulate each other . In fact, genes involved in the oxidative stress response through the formation and/or protection of iron-sulfur (Fe-S) clusters, such as nifU, nifS, iscA, part of the nitrogen fixation system (NIF), and sufD [14–18], have been found up-regulated after desiccation of Salmonella on Petri dishes . In the same study, fnr, encoding for the fumarate and nitrate reduction protein and one of the master regulators for the metabolic shift from aerobic to anaerobic conditions, was also induced following desiccation. The knockout mutant Δfnr showed impaired ability to survive dehydration and long-term storage at room temperature .
Interestingly, pre-exposure to desiccation has also been shown to induce protection against heat treatment [20–22]. Most of the theories concerning this cross-protection are based on physicochemical properties of the cell and focus on the stabilization of proteins during thermal exposure when less water is present [23–25]. However, some reports indicated that Salmonella thermal tolerance persists for a short period of time after rehydration . Additionally, non-typhoidal Salmonella spp. and STEC E. coli showed higher tolerance to desiccation than non-pathogenic E. coli , although these species share almost identical physicochemical properties. These observations suggest that the thermal tolerance is not limited to chemical and physical phenomena, but could be in part a consequence of the complex network of overlapping stress responses induced by desiccation.
In addition to heat, exposure to desiccation has been linked with cross-protection for a multitude of other stressors in Salmonella, such as sodium hypochlorite, sodium chloride, bile salts, and hydrogen peroxide . The ability of Salmonella to overcome different stresses is crucial for its virulence, since it is an essential capability during the infection process. To colonize the host, Salmonella has to survive extra- and intra-cellularly in the stomach and the intestine, where it is exposed to acid, osmotic, and oxidative stress, as well as starvation. Induction of virulence genes hilA, invA, and spiC was found after drying and storage in dry milk for short periods in Salmonella cells, both in planktonic and biofilm state .
The production of curli, thin aggregative fimbriae, cellulose, and lipopolysaccharides has also been reported to be important for survival under dry conditions [12, 27–29]. One of the hypothesis concerning the production of these cellular structures is that due to their high water retention quality, exopolysaccharides work as a water deposit . Mutants in lipolysaccharides have been shown to be more sensitive to desiccation than parental Salmonella strains . Salmonella’s ability to produce biofilm has also been reported to be important for desiccation survival on polypropylene discs .
In this work, we identify genetic components that are involved in the ability of Salmonella’s cells to survive reduction in moisture and exposure to very low water activity. In particular, two virulence genes, sopD and sseD, that are important for its survival to desiccation.
Materials and methods
Bacterial strains and culture preparation
The strains used in this study included Salmonella enterica serovar Typhimurium strain ATCC 14028 (S. enterica serovar Typhimurium), Escherichia coli O157:H7 strain ATCC 43895 (E. coli O157:H7), and two deletion mutants of the strain S. enterica serovar Typhimurium, ΔsopD and ΔsseD, obtained as described below by camR and aphA-2 insertions, respectively. The stock cultures were stored at -55°C in a 5:1 solution of Luria-Bertani broth (LB; BBL, Detroit, MI) and glycerol. Working cultures were prepared in tryptic soy broth (TSB, Neogen, Inc., Lansing, MI) or 0.01 M glucose-supplemented LB broth (LBglc) from frozen stock cultures and grown overnight at 37°C shaking at 250 rpm. For mutant strain working cultures, chloramphenicol and kanamycin were added to LBglc to a final concentration of 50 and 100 μg/mL respectively.
Inoculation on filters and RNA extraction
Working cultures of S. enterica serovar Typhimurium were freshly inoculated in TSB and grown for 3 h at 37°C shaking at 250 rpm. The cultures were collected through centrifugation (10 min at 4,696×g) and washed twice with distilled sterile water (DSW) to eliminate nutrient residues. Approximately 109 CFU were spotted on 0.2 μm polycarbonate filters (Merck Millipore Ltd., Billerica, MA) and allowed to air-dry for 24 h at room temperature in a biosafety cabinet. Filters were placed in desiccators containing water or a saturated solution of lithium chloride 99% (Acros Organics, Thermo Fisher Scientific, Waltham, MA) to allow equilibration to aw 1.0 and 0.11 respectively. Separate samples of corn starch were included as controls to monitor the aw of the desiccators. After 4 days in the desiccators at 25°C, the aw of corn starch samples were measured and the total RNA was extracted from Salmonella cells using the RNAprotect Bacteria Reagent and RNeasy Mini Kit (Qiagen, Hilden, Germany). The experiments were repeated three times in different days. Each time, three technical replicates were performed. The total RNA was extracted individually for each replicate and then the RNA was pooled together from all the replicates for the same conditions.
RNA-Seq and global transcriptional analysis
Total RNA was processed at the University of Minnesota Genomic Center. Briefly, samples were quantified using a fluorimetric RiboGreen assay. Total RNA integrity was assessed using capillary electrophoresis with 2100 BioAnalyzer (Agilent Technologies, Santa Clara, CA), generating a RNA Integrity Number (RIN). Samples were converted to Illumina sequencing libraries using the TruSeq RNA Sample Preparation and Library preparation Kit (Illumina Inc., San Diego, CA). In summary, 1 μg of total RNA was reverse transcribed into cDNA. The cDNA was fragmented, blunt-ended, ligated to barcoded adaptors, and amplified using 15 cycles of PCR. Final library size distribution was validated using capillary electrophoresis and quantified using fluorimetry (PicoGreen) and via Q-PCR. Indexed TruSeq libraries were then normalized, pooled, hybridized to a paired end flow cell, and individual fragments were clonally amplified by bridge amplification on the Illumina cBot (Illumina Inc., San Diego, CA). Once clustering was completed, the flow cell was loaded on the HiSeq 2000 (Illumina Inc., San Diego, CA). Sequencing was performed on both strands.
Base call files for each cycle of sequencing were generated by Real-Time Analysis software (Illumina Inc., San Diego, CA). Primary analysis and de-multiplexing were performed using CASAVA software v1.8.2 (Illumina Inc., San Diego, CA). The end results of the CASAVA workflow, de-multiplexed FASTQ files, were analyzed using the DNASTAR SeqManNgen and ArrayStar softwares (DNASTAR, Inc., Madison, WI). The reads were assembled and mapped to the genome using SeqManNgen, while the levels of expression were estimated using ArrayStar.
The data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus  and are accessible through GEO Series accession number GSE86580 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE86580).
Construction of deletion mutants
Two mutant strains were generated using the λ Red-mediated homologous recombination . The plasmid pKD46 was introduced into S. enterica serovar Typhimurium using a MicroPulser electroporator (Bio-Rad Labs., Hercules, CA) at 1.7 kV. Primers used for knockout of sopD and sseD genes are reported in Table 1. All knockout primers were 60 nucleotides long, with 39 nucleotides homologous to the upstream or downstream regions flanking the targeted gene, and 21 nucleotides homologous to a universal cap designed in the drug resistance cassette (kindly provided by Dr. Roth, University of California at Davis, Davis, CA). The procedure used to make donor DNA fragments followed the protocol previously described . Briefly, the PCR amplification protocol was: 95°C for 5 min; 95°C for 1 min, 55°C for 1 min, 72°C for 1:40 min × 30 cycles, and 72°C for 5 min. The 55°C melting temperature of our fragments was calculated on the 21-nucleotide caps.
The sopD and sseD coding sequences were disrupted with the chloramphenicol (camR) and the kanamycin (aphA-2) resistance cassette, respectively, leaving the ATG region intact. The resistance cassettes were inserted in 3’-5’ to avoid polar effects of the universal cassette promoter on the downstream genes (S1 Fig).
The deletion of the sopD and sseD genes by substitution and insertion of the antibiotic cassette was verified by PCR amplification (S1 Fig) and Sanger sequencing. Primers used are listed in S1 Table. The sequenced reads were matched against NCBI database using the BLAST function  for the antibiotic cassette portion, while the gene’s upstream and downstream regions were identified using Artemis platform (Wellcome Trust Sanger Institute) . The total genome of the two mutants was also extracted using the GenElute Bacterial Genomic DNA Kit (Sigma-Aldrich, St. Louis, MO) and sequenced on a HiSeq 2500 sequencer (Illumina, San Diego, CA). The assembling and mapping results were obtained using the DNASTAR SeqManNgen software (DNASTAR, Inc., Madison, WI).
The growth rates and generation times were determined using the optical density (OD) measured at 600 nm with the Epoch 2 microplate reader (Bio Tek Instruments, Inc., Winooski, VT). Liquid cultures in LBglc were incubated overnight at 37°C with shaking at 250 rpm. The OD of each strain was then adjusted to 0.02 in LBglc and 200 μL aliquots were pipetted to 96-well plates. The plates were transferred to the plate reader with an incubation temperature of 37°C and orbital shaking. The ODs were recorded every 10 minutes for 24 h. Growth rates during the exponential phase were calculated by a regression of ln(OD) vs. time, where the slope was the growth rate based on the Monod equation (dN/dt = μN, where N is cell concentration expressed as OD, t is time, and μ is growth rate).
Viability experiments on micro glass beads
Bacterial cultures were grown overnight at 37°C with shaking at 250 rpm in LBglc. The cultures were collected through centrifugation (10 min at 4,696×g) and washed twice with DSW to eliminate nutrient residues. For these experiments we decided to use glass beads (150–250 μm) (Corpuscular Inc., Cold Spring, NY). Our decision was motivated by the need to increase the number of cells, while still ensuring that the cells were evenly distributed and exposed to low aw. The greater total surface area offered by the beads compared to the filters allowed for a larger number of cells and for the formation of a thinner layer of adhering cells (as confirmed by the SEM micrographs). The washed pellets were re-suspended in 10 mL DSW and used to inoculate 10 g of sterile glass beads. Inoculated glass beads were spread on a sterile Petri dish and dried for 4 days at 38.5±0.5°C. For viable cell enumeration, 100 μL of each re-suspension were serially diluted in sterile saline (NaCl 0.9%), and 100 μL were spread plated on differential tryptic soy agar (dTSA) [TSA (Neogen, Inc.) supplemented with ammonium iron (III) citrate 16% (Fluka Analytical, Sigma-Aldrich, St. Louis, MO) (0.8 g/L) and sodium thiosulfate 98.5% (Acros Organics, Thermo Fisher Scientific, Waltham, MA) (6.8 g/L)].
After drying, the beads were distributed into sterile 200 μL PCR plastic tubes (Thermo Fisher Scientific). For exposure to aw 0.11 and 1.0, the samples were equilibrated for 7 days at 25°C in desiccators containing a saturated lithium chloride (Acros Organics, Thermo Fisher Scientific) solution or sterile distilled water (SDW). After 7 days, the water activity of samples was measured (cutoff value: aw reference ±0.02), and the samples were sealed. To determine the survival rate, beads for every sample were serially diluted in saline and spread plated on dTSA for cell enumeration. The recovery after every treatment was measured as cell viability in Log (CFU/g), and the survival rate was calculated as viability change in Log (CFU/g).
Scanning electron microscopy
Samples were collected after inoculation on glass beads, dried for 4 days, and equilibrated 7 days to aw 0.11 and 1.0. Immediately after collection, samples were fixed with a solution of 1% paraformaldehyde, 1% glutaraldehyde, and 0.05 M sodium cacodylate as previously described . The samples were fixed overnight and then dehydrated through an ethanol series (10, 25, 70, 90, 95, and 100% ×2 for 24 h each) and HMDS series (30, 60 100%, 20 min each). Samples were transferred into 100% HDMS, air-dried for 48 h at room temperature, placed on 9.5 mm aluminum stubs with adhesive carbon tape, and coated with 20 nm of gold-palladium using the SC7620 Mini Sputter Coater (Quorum Technologies, Inc, Guelph, Canada). Scanning electron microscopy (SEM) imaging was performed with the JSM 6060LV scanning electron microscope (JEOL USA, Inc., Peabody, MA) using a 15 kV accelerating voltage.
Generation times for the wild-type and mutant strains were calculated averaging three independent biological replicates (n = 3). The generation time for each replicate was determined using the average OD of four technical replicates. All the viability experiments on glass beads were repeated at least four times (n ≥ 4) in different days (biological replicates). Each biological replicate consisted of three technical replicates. Technical replicate results were averaged to obtain the Log (CFU/g) for each biological replicate.
Significance, expressed as p-value, was calculated using a two-tailed Student’s t-test assuming equal variance for all experiments. Threshold for significance was set at p ≤ 0.05. Standard error of the mean (SE) was used to calculate the variation among samples. Averages, p-values, and SE were performed on the results of the biological replicates for each strain at each condition.
Global transcriptional analysis
The transcriptional profile of S. enterica serovar Typhimurium cells air-dried and equilibrated to aw 0.11 was compared to that of cells dried and equilibrated to aw 1.0. Out of 4,489 genes  719 (16%) were differentially expressed between the two conditions. Among these, 290 genes (40.3%) were up-regulated (2-fold cutoff) (S2 Table). We decided to focus on the up-regulated genes because are those most likely necessary for the adaptation to low water activity. These genes were categorized based on the KEGG Orthology (KO) database . We found 5 functional classes: 1) metabolism (52 genes); 2) genetic information processing (24 genes); 3) environmental information processing (25 genes); 4) cellular processes (3 genes), and 5) infectious diseases (2 genes, both classified as virulence factors). The remaining 184 genes did not belong to any orthology group and were, therefore, unclassified. Table 2 lists a selected group of up-regulated genes and their functions by KO categories and sub-categories.
More specifically, multiple amino acid transporters where found up-regulated, including the alanine/glycine transporter yaaJ, the histidine/lysine/arginine/ornithine transporter hisQ, and the phenylalanine transporter pheP. Other genes involved with ion transportation were also found induced by low water activity: for example trkH for potassium, znuB for zinc, nhaB for sodium, and also the transcriptional regulator mntR for manganese. Several tRNA species were up-regulated: the tRNA for leucine (leuP and leuT), glutamate (gltU), histidine (hisR), arginine (argW), aspartate (aspU) and serine (serX) were the most abundant in low water activity. DNA replication genes, such as holA, rnhA, dinG, and DNA repair genes (ruvC, priC), were also induced in low water activity conditions.
Other up-regulated gene groups included transcriptional and translational regulators (i.e. gutM, yneJ, yciH) and ribosomal genes (i.e. rluA, rpsJ, rrmA). Genes involved in fimbriae and flagella biosynthesis were also up-regulated (i.e. flhA, sthE, stiA). Although we detected only two virulence genes by KO classification, sopD and sseD, four other known virulence factors were up-regulated in low water activity conditions. These virulence factors included sscA, a chaperon for the SseC translocon ; sseA, a class II chaperone specific for translocon proteins SseB and SseD [40–42]; mviN, involved in peptidoglycan biosynthesis and required for virulence in mice [43, 44]; and mgtC, part of the mgtBC operon in SPI-3, involved in the regulation of the membrane potential , and transcriptionally controlled by PhoP/PhoQ system .
ΔsopD and ΔsseD mutant verification and sequencing
Based on the transcriptomic analysis results, we generated mutants in two virulence factors, sopD and sseD. Disruption of sopD by insertion of the chloramphenicol resistance cassette (camR) and of sseD by the kanamycin resistance cassette (aphA-2) was confirmed by PCR and sequencing. Genome-wide sequencing of the two mutants confirmed that the only variants present were localized in the targeted genes. When grown in LBglc at 37°C with aeration, both the mutants had the same generation time (G) compared to the wild-type (WT) strain: 22 min (SE 0.4 min) for WT, 21 min (SE 0.3 min) for ΔsopD (p = 0.33) and 21 min (SE 0.1 min) for ΔsseD (p = 0.11).
Survival on glass beads
To test the effect of sopD and sseD genes in response to desiccation, S. enterica serovar Typhimurium wild-type (WT), mutant strains, and E. coli O157:H7 were inoculated on micro glass beads, dried, and equilibrated to aw 0.11 and 1.0 (Figs 1 and 2). As mentioned in the Materials and Methods, the use of glass beads was chosen to allow for a larger starting inoculum, while ensuring a homogenous surface distribution of the cells.
Changes in cell viability [Log (CFU/g)] of S. enterica serovar Typhimurium wild-type (WT; dark grey), ΔsopD (light grey), ΔsseD (white), and E. coli O157:H7 (purple) after 4 days drying (A) and after 7 days of equilibration to aw 0.11 and 1.0 (B). Bars indicate standard error of the mean (SE). Stars indicate p-values < 0.05 (*), < 0.01 (**), and < 0.005 (***) in comparison with Salmonella WT under the same condition.
Total changes in cell viability [Log (CFU/g)] of S. enterica serovar Typhimurium wild-type (WT, dark grey), ΔsopD (light grey), ΔsseD (white), and E. coli O157:H7 (purple) after 11 days of treatment (4 days drying and 7 days of equilibration to aw 1.0 or 0.11). Bars indicate standard error of the mean (SE). Stars indicate p-values < 0.05 (*), < 0.01 (**) and < 0.005 (***) in comparison with Salmonella WT under the same condition.
Cell counts decreased for all the strains after drying (Fig 1A) [Salmonella WT, -3.0 Log (CFU/g); ΔsopD, -5.5 Log (CFU/g); ΔsseD, -4.3 Log (CFU/g); and E. coli O157:H7, -4.5 Log (CFU/g)]. The differences between Salmonella WT and the mutants were significant (p-values 0.0003 for ΔsopD and 0.02 for ΔsseD). Although the E. coli O157:H7 count decreased, it was not significantly different from Salmonella WT (p = 0.06).
After equilibration to aw 1.0 (Fig 1B), the cell counts for all the strains except ΔsseD increased compared to drying [Salmonella WT, 1.3 Log (CFU/g); ΔsopD, 2.3 Log (CFU/g); and E. coli O157:H7, 0.9 Log (CFU/g)]. Conversely, ΔsseD decreased by 2.4 Log (CFU/g). This difference was statistically significant when compared to the WT and ΔsopD, but not to E. coli O157:H7 (p-values 0.007, 0.014, and 0.06, respectively).
After equilibration to aw 0.11 (Fig 1B), the cell counts for all the strains decreased compared to drying [Salmonella WT, -1.3 Log (CFU/g); ΔsopD, -1.7 Log (CFU/g); ΔsseD, -2.7 Log (CFU/g); and E. coli O157:H7, -2.2 Log (CFU/g)], although only ΔsseD had a significantly lower recovery than Salmonella WT (p = 0.012).
The large decrease in cell count of ΔsopD during drying lowered the cell count close to the detection limit of the experiment [2.7 Log (CFU/g)]. Therefore, it was difficult to estimate the additional decrease in viability and its significance after exposure to aw 0.11. For this reason, we also calculated the total change in cell count for both aw over the entire treatment period (11 days, Fig 2). At aw 1.0, the total change in cell counts after the 11 day treatment was not significantly different between Salmonella WT, ΔsopD, and E. coli O157:H7 [Salmonella WT, -1.6 Log (CFU/g); ΔsopD, -3.1 Log (CFU/g); and E. coli O157:H7, -3.7 Log (CFU/g)]. On the contrary, ΔsseD had an overall reduction of 6.7 Log (CFU/g), significantly larger than Salmonella WT and ΔsopD, but not E. coli O157:H7 (p-values 0.0004, 0.03, and 0.14, respectively). At aw 0.11 (Fig 2), we observed large total reductions in cell counts [Salmonella WT, -4.5 Log (CFU/g); ΔsopD, -7.4 Log (CFU/g); ΔsseD, -7.1 Log (CFU/g); and E. coli O157:H7, -6.8 Log (CFU/g)]. The mutants and E. coli O157:H7 had significantly lower cell viability compared to Salmonella WT (p-values 0.0001 for ΔsopD, 0.003 for ΔsseD, and 0.01 for E. coli O157:H7).
Scanning electron microscopy of S. enterica serovar Typhimurium wild-type, ΔsopD, ΔsseD, and E. coli O157:H7
Observations by SEM of S. enterica serovar Typhimurium wild-type (WT), ΔsopD, ΔsseD, and E. coli O157:H7 cells on glass beads showed WT and E. coli O157:H7 cells as rod-shaped cells with an average length of 1.5 μm and an average width of 0.5 μm. In contrast, both mutants displayed different cell morphology, more coccobacillary, with markedly smaller, rounder, and shorter cells, 1 μm in length or less (Fig 3).
Cells were collected from the overnight cell re-suspension used to inoculate the glass beads. The images show the change in cell morphology for the two mutants. Magnification and scale bar are embedded in the images.
After drying, WT cells still appeared rod-shaped, although some cells displayed cell surface corrugations indicating a loss of turgidity (Fig 4). Cells were also embedded in a thick extracellular matrix. After drying, both mutants maintained a smaller size than the WT with a spheroidal shape and an evident indentation in the middle of the cell. Additionally, the cell density on the glass bead surface was lower than what was observed in the parental strain. Although ΔsopD also produced an extracellular matrix, it lacked the three dimensional structure observed for the WT. The matrix did not embed the cells but was attached to the bead surface. The matrix also had characteristic cell-shaped discontinuities where cells detached. In the same conditions, ΔsseD did not produce a homogeneous matrix, but presented an intricate network of ropy filaments. E. coli O157:H7 cells had morphology similar to Salmonella WT, but they produced a matrix similar to ΔsopD.
Cells of S. enterica serovar Typhimurium WT, ΔsopD, ΔsseD, and E. coli O157:H7 inoculated and dried on glass beads for 4 days at 38.5°C. Note the differences in cell morphology and matrix structure among the four strains. Arrows and letters indicate specific elements present in each strain: matrix (m), fenestrations (f), cell concavity (c), and ropy filaments (r). The images are representative of the sample population. Magnification and scale bar are embedded in the images.
After 7 days of equilibration to aw 0.11 (Fig 5), Salmonella WT cells were still characterized by the presence of an extracellular matrix, although the matrix was partially disrupted and detached. The cells maintained the rod morphology, but membrane corrugations and distortion, indicating loss of turgidity and cellular damage, were more evident than in the control. Both mutants, as well as E. coli O157:H7, lacked an extracellular matrix. Cells showed surface corrugation indicating membrane damage, and cell debris was present on the bead surface.
Images of S. enterica serovar Typhimurium WT, ΔsopD, ΔsseD, and E. coli O157:H7 cells inoculated, dried and equilibrated for 7 days to aw 0.11 on glass beads. Arrows and letters indicate specific elements present in each strain: matrix (m), cell concavity (c), damaged cells (d), and debris (s). The images are representative of the sample population. Magnification and scale bar are embedded in the images.
As observed at aw 0.11, after equilibration for 7 days to aw 1.0 (Fig 6) the two mutants maintained the characteristic morphology and presented loss of turgidity and wrinkling of the membrane, suggesting cell damage. Cell debris was also observed on the beads for both mutants as well as for the WT and E. coli O157:H7. Differently than at aw 0.11, the two mutants had an extracellular matrix similar to the one formed by the WT and E. coli O157:H7.
Images of S. enterica serovar Typhimurium WT, ΔsopD, ΔsseD, and E. coli O157:H7 cells inoculated, dried and equilibrated for 7 days to aw 1.0 on glass beads. Arrows and letters indicate specific elements present in each strain: matrix (m), cell concavity (c), damaged cells (d), and debris (s). The images are representative of the sample population. Magnification and scale bar are embedded in the images.
The response of S. enterica serovar Typhimurium to desiccation and exposure to low aw is extremely important because it can trigger other, seemingly unrelated, stress tolerance responses [13, 20, 21, 47, 48], but its underlying molecular mechanisms are still largely unknown. Several groups have performed different global transcription analysis on desiccated Salmonella. Deng et al. studied the transcriptome of this microorganism by RNA sequencing in low aw peanut oil  and there are a good number of reports on microarray-based transcriptomic analyses on Salmonella survival or adaptation to abiotic surfaces such as filter paper, stainless steel, and plastic [11, 12, 19]. However, all of these studies have focused either on matrix-based low aw or desiccation. To focus solely on the low aw effect without the confounding factor of variations in the chemical composition of food matrices (e.g., batch to batch, sourcing of raw materials, and aging of the product), we decided to perform a global transcriptomic analysis on cells exposed to extreme low aw using abiotic surfaces.
Our transcriptomic analysis showed that exposure to low aw has a broad impact on the expression of many genes involved in anabolic and catabolic pathways. This was expected because the rate of enzymatic reactions slows when occurring in low aw, thus causing a decrease in the metabolic rate. This phenomenon has already been suggested in other studies . However, we also found that low aw induced the expression of many genes involved in DNA replication and repair. DNA damage is known to be an effect of desiccation, in particular covalent modifications and double-stranded breaks . DNA repair genes were up-regulated under desiccation conditions in many different microorganisms, including Deinococcus radiodurans  and Bradyrhizobium japonicum . We can speculate that during the shift to low aw, a portion of the replicating cells were unable to complete their replication due to either lack of energy and building blocks or water available for chemical reactions. This halt in the process would most likely be sensed as replication errors by the DNA replication checkpoints, thus inducing DNA repair mechanisms.
We also observed a stark increase in many tRNAs in cells exposed to low aw. tRNAs are usually difficult to observe by RNA sequencing, mainly because of their strong secondary and tertiary structures and post-transcriptional modifications . Therefore, the observed increase in readable tRNA sequences for some tRNAs in desiccated cells might indicate that the post-transcriptional processing of those tRNAs is less efficient, possibly due to the same processes that cause DNA replication errors. However, an intriguing possibility is that the differential processing of tRNAs is a long-term adaptation strategy for the cell to coordinate amino acid transport and control translation. This would fit well with the high number of amino acid transporters and the transcriptional/translational regulators we found induced at low aw. Similar observations—up-regulation of amino acid transport and metabolism, and transcription and translation-associated genes—led Gruzdev et al.  to the conclusion that de novo protein synthesis is a requirement for the cell adapting to desiccation.
Osmolarity homeostasis is important in the desiccation process and, in our study, we observed the induction of osmV (2.91 fold), one of the genes involved in the transport of osmoprotectants during osmotic stress, in cells exposed to aw 0.11 for four days. However, this was the only osmolarity-related gene induced in our treatment conditions. This is interesting, because one of the main long term responses the cell deploys to counteract osmotic stress during desiccation is the intracellular accumulation of osmoprotectants . Osmoprotectants can be accumulated intracellularly from the medium or by de novo synthesis [55–57]. During desiccation, the ABC transporters ProU (ProVWX) and OsmU (OsmVWXY), in conjunction with the permease ProP, transport in the cytoplasm glycine/betaine and proline, the main osmoprotectants used by the cell [58–61]. Li et al. and Finn et al. found the ProU system up-regulated in S. enterica serovar Typhimurium desiccated on filter paper and stainless steel, respectively [11, 12].
The difference we observed in the induction of the ProU system may be explained by differences in methodology, including desiccation conditions and length of exposure to low aw. In particular, we washed and re-suspended the cells in pure water prior to desiccation, whereas Li et al.  used cells in PBS and Finn et al.  used cells in LB. As suggested elsewhere , the use of isotonic solutions in this kind of experiments can be problematic, as it would induce osmotic response during desiccation through the increase of solute concentration caused by water evaporation rather than through a decrease in aw. Moreover, those two studies focused on very short-term effects of desiccation (2 hours  and 4 hours ), whereas our analysis was performed after an extended period of time (4 days). Therefore, in our experiment, cells were already adapted to the initial osmotic shock and no longer needed many osmotic stress genes. These two non-exclusive explanations are corroborated by previous research that showed that 22 hour air-drying of water-resuspended Salmonella cells on plastic Petri dishes did not induce the Pro/Osm genes . It is important to mention that production of the ProU transport system is usually induced by the presence of environmental glycine or betaine, and in our case, as well as in Gruzdev’s group, glycine was not present in the environment . A similar explanation can be applied to the absence of up-regulation of the trehalose biosynthesis genes (otsA and otsB), which is another important osmoprotectant synthesized by the cell [63–65].
Our class of orthology gene classification of the differentially expressed genes identified two genes encoding for virulence factors, sseD and sopD. While sseD and sopD have been characterized for their role in virulence and infection mechanisms, to our knowledge, they have never been reported to be involved in desiccation adaptation and tolerance, and for this reason we targeted them for further analysis. SseD is a translocon component of the Salmonella Pathogenicity Island 2 (SPI-2) Type III Secretion System (T3SS) , and it is part of the sseABCD operon, encoding a chaperon protein and the other translocon components . We also observed the induction of sseA, encoding the chaperonin for sseB and sseD [40–42], and sscA, encoding the chaperonin of sseC , in cells exposed to aw 0.11. This indicates that the entire SPI-2 T3SS translocon is likely involved during adaptation and survival to desiccation and low aw. Interestingly, although SopD was initially identified as an effector translocated by the T3SS of SPI-1 , it is now suggested that the SPI-2 encoded injectisome can also be involved in its deployment . Even though SopD release is activated under SPI-1 inducing conditions , sopD expression remains elevated during later stages of infection and is involved in survival and replication inside the macrophage [69, 71].
To better understand the impact of these two genes on the ability of S. enterica serovar Typhimurium to survive desiccation, we compared the survival rates and morphological characteristics of the WT, ΔsopD and ΔsseD mutants, and a virulent strain of E. coli (O157:H7). We chose to include E. coli in our comparisons because of its genetic and physiological relatedness to Salmonella and lack of the same kind of virulence mechanisms encoded by the SPIs.
The mutants were first compared to the WT under optimal growth conditions (LBglc, 37°C, shaking), and no differences were detected in generation time and in the ability to reach stationary phase. Additionally, the cell viability of the mutants in the initial cell-resuspension was checked, and the mutants, as well as E. coli O157:H7, had the same cell viability of Salmonella WT before being inoculated on glass beads (data not shown). On the contrary, interesting differences were detected when observing the different cell-resuspensions with SEM. Both mutants exhibited a shorter and coccoidal shape compared to the WT. Both ΔsopD and ΔsseD mutants had decreased tolerance to desiccation compared to the WT, clearly indicating that SopD and SseD play an important role during desiccation survival. To the best of our knowledge, this is the first time that different cell shape and size have been observed for mutants in these two genes.
In a previous study, field emission scanning microscopy (FESEM) was performed on a large number of SPI-2 effectors mutants and no changes in cell morphology were reported for ΔsseD in that case . The use of a different S. enterica serovar Typhimurium strain—ATCC14028 instead of NCTC 12023—and different growth conditions might justify the different morphologies detected in our study. Furthermore, Chakravortty et al. used specific SPI-2 inducing conditions. As a result, the induction of other SPI-2 T3SS genes could have masked defects in cell morphology caused by the mutation in sseD.
A similar round shape was observed for mreC mutant in Salmonella  and E. coli . This gene belongs to the mreBCD operon, which is responsible for the cell shape and correct formation of the cytoskeletons [75–79]. For E. coli, deletion in mreBCD caused the formation of spheroid-like cells defective in adjusting the rate of phospholipid synthesis . Interestingly, we observed up-regulation of the entire mreBCD operon in WT cells of S. enterica serovar Typhimurium equilibrated to aw 0.11, suggesting a role of the operon in the cell response to low aw, probably through the maintenance of a correct membrane composition and cell-shape. Although we cannot directly prove a connection between sopD/sseD and the mre operon, the fact that the two mutants, defective in cell-shape, are less tolerant to desiccation, suggests that the desiccation response mechanisms may include the sopD/sseD-dependent induction of the mre operon.
The effects of the mutations persisted after the additional week at aw 0.11. The cell viability of the two mutants decreased once more. This indicates that sopD and sseD are not only involved in Salmonella adaptation to desiccation but are also essential for long period survival. It is interesting to note that drying and exposure to aw 0.11 had a dramatic effect also on E. coli O157:H7, which lacks SPI-1 and 2 genes.
An additional possibility is that the mutants are more prone than the WT to enter a viable but not culturable state (VBNC). The VBNC state is described as a dormant state in which cells are metabolically inactive and are not culturable using standard methods . This phenomenon has already been observed for Salmonella cells under different stress conditions, including drying and desiccation [81–83].
Structurally, after drying and prolonged exposure to aw 0.11, Salmonella WT displayed a thick layer of solid extracellular matrix. In general, production of EPS has been associated to a higher desiccation tolerance in a variety of bacterial species, such as E. coli, Acinetobacter calcoaceticus, Erwinia stewartii, and Rizhobium sullae [84, 85]. It is believed that EPS works as a water reservoir and is protective from desiccation . Moreover, genes encoding for fimbrial subunits were found induced in cells equilibrated to aw 0.11 on filters—i.e. sthA, sthB, and sthE, part of the fimbrial operon sthABCDE, important for colonization in mice  as well as chickens . The matrix observed for Salmonella WT after the 7-day equilibration showed some signs of damage, possibly due to the desiccation of the hydrogel.
After drying, the extracellular solid matrices of both E. coli O157:H7 and ΔsopD presented signs of damage as well as cell detachment, while after exposure to aw 0.11, neither ΔsopD nor E. coli O157:H7 had any kind of extra-cellular material production. Interestingly, ΔsseD did not produce any solid matrix after drying, although cells were encased in a network of ropy filaments, which appeared to connect all cells. ΔsseD did not present these filaments after exposure to aw 0.11 and the cell viability decreased compared to after drying.
It is possible that the mutant cells can temporarily produce extracellular structures with the role of protecting the cell from desiccation, but this response might be ineffective for long periods of exposure to low aw. Based on the different role and characteristics of SopD and SseD, we hypothesize that the decrease in desiccation and low aw tolerance that we observed in our mutants is mainly due to lack of secretion of the effector protein SopD. As previously mentioned, SopD is an effector secreted by both T3SSs, while SseD is part of the injectosome of the SPI-2 T3SS. SPI-1 and SPI-2 T3SS are expressed during the infection process at different stages . While SPI-1 expression is activated for the invasion of the host cell and the initial formation of the Salmonella containing vacuole (SCV), SPI-2 is necessary at a later stage, for Salmonella induced filaments (SIF) formation and bacterial replication inside the SCV.
In the case of the ΔsopD mutant, SopD cannot be produced nor secreted, and cell viability decreases dramatically from the first treatment (desiccation). ΔsseD is less susceptible than ΔsopD to the effect of initial desiccation, but after prolonged exposure to aw 0.11, the effect of sseD depletion becomes as dramatic as sopD. These data suggest that during the initial desiccation, the SopD effector could still be secreted in the ΔsseD mutant, probably by SPI-1 T3SS, but during long-term low aw exposure, the role of SPI-2 T3SS may become fundamental, and due to malfunctioning of the injectsome in ΔsseD, SopD may no longer be secreted. Additionally, the induction in the WT of sseD, as well as sseA and sscA, in cells equilibrated to aw 0.11 strongly indicates that the correct assembly and functioning of the SPI-2 T3SS is required for survival at extreme low aw conditions.
When cells were exposed to high aw (aw 1.0) after drying, cell viability for the WT, ΔsopD, and E. coli O157:H7 increased. This may be due to the ability of the cells to utilize proteins and nutrients released by the dead cells still on the beads, similarly to what hypothesized by Gruzdev et al. . The presence of dead cells and cellular debris was observed at SEM after equilibration to aw 1.0. Additionally, it has also been suggested that residues of extracellular polymeric substances (EPS) formed during desiccation can serve as a source of nutrients [90, 91]. The ΔsseD mutant, instead, did not increase in cell viability after equilibration to aw 1.0. The differences in recovery from desiccation between the two mutants could be linked to the different role of the two effectors. Possibly, the lack of a completely developed needle structure in ΔsseD results in defective secretion of several other effectors required for the recovery process. At aw 1.0 no differences were detected in EPS production, confirming that the differences in the exopolymeric matrix formation between the WT and the mutants/E. coli O157:H7 are due to differences mainly in response to desiccation.
Very recently, the regulatory system of a 97 nt small antisense RNA, STnc3140, encoded on the negative strand of sopD and positioned in its coding region (from position +726 to +822), has been partially characterized in Salmonella [92–95]. This sRNA was first named SLasRNA0334 and was identified in 2012 by Ramachandran et al. using a combination of differential RNA-seq and in silico analysis . A subsequent study published in 2013 by Kröger et al. renamed this antisense RNA STnc3140 and showed that this RNA interacted with the RNA chaperone Hfq specifically during the transition from late exponential to early stationary phase . In 2016, Colgan et al. thoroughly characterized the regulatory networks of 280 sRNAs in 5 different conditions (mid exponential phase, intermediate exponential phase, early stationary phase, late stationary phase, and SPI-2 inducing) and found that STnc3140 is positively regulated by RpoS and Hfq, but is repressed by HilD and Fur . In particular, the authors of that study observed more than 3-fold decrease in STnc3140 expression in RpoS and Hfq mutants in late stationary phase and early stationary phase, respectively, and more than 3-fold increase in HilD and Fur mutants in early stationary phase. In a different study, Smirnov et al. observed a decreased expression of STnc3140 also in a ProQ mutant of Salmonella . In our mutant, this small RNA was removed along with the SopD coding sequence.
The absence of STnc3140 could be at least partially responsible for the low aw-sensitive phenotype observed in the ΔsopD. However, this hypothesis is unlikely based on the fact that i) none of the regulators reported in the above mentioned studies are differentially expressed in our conditions; ii) the induction, under low aw conditions, of sopD and few other genes related to secretion (e.g., sseD, sseA, and sscA) was described in the WT strain first and not limited to STnc3140; and iii) the phenotypes of ΔsopD and ΔsseD, although clearly distinct, they shared several similarities (e.g., coccoidal shape, sensitivity to desiccation and low aw, impacted extracellular matrix production at low aw). Regardless, as more information about small ncRNAs becomes available, it will be interesting to unravel the effects of the SopD effector from the ones of STnc3140, possibly by reintroducing into a ΔsopD background only STnc3140 and its regulatory unit and complementing the ΔsopD mutation with codon bias-modified copy of the gene not encoding STnc3140.
S1 Fig. Schematic representation of the mutants and of the genomic regions amplified for PCR verification.
Schematic drawing of the S. enterica serovar Typhimurium wild-type sopD and sseD gene knock-out mutations with the chloramphenicol resistance cassette and the kanamycin resistance cassette, respectively (A). The sites of λ Red-mediated homologous recombination are indicated with crossing lines, while the primers used for the creation of the cassette are indicated with arrows. UL1 and UR1 are the universal caps part of the drug-cassette kit by Dr. Roth Laboratory (University of California at Davis, Davis, CA). Figure B shows the collocation of the primers used in the PCR reaction for the verification of ΔsopD and ΔsseD mutants of S. enterica serovar Typhimurium.
S1 Table. Primers used for PCR verification and sequencing of the ΔsopD and ΔsseD mutants.
We thank Jonathan Rossbach for his valuable help during the SEM observations and his constructive critique of the manuscript. We also thank Lijun Duan, Megan Andorfer, Hannah Riesner, Joseph Dummann, Doug Snyder, and Samantha Sun for their invaluable help in performing the experiments.
- 1. Crumrine MH, Foltz VD, Harris JO. Transmission of Salmonella Montevideo in wheat by stored-product insects. Appl Microbiol. 1971;22: 578–580. pmid:4943271
- 2. Foster JW, Spector MP. How Salmonella survive against the odds. Annu Rev Microbiol. 1995;49: 145–174. pmid:8561457
- 3. Winfield MD, Groisman EA. Role of nonhost environments in the lifestyles of Salmonella and Escherichia coli. Appl Environ Microbiol. 2003;69: 3687–3694. pmid:12839733
- 4. Troller J, Christian J. Water Activity and Food. New York: Academic Press; 1978.
- 5. Beuchat LR, Komitopoulou E, Beckers H, Betts RP, Bourdichon F, Fanning S, et al. Low-water activity foods: increased concern as vehicles of foodborne pathogens. J Food Prot. 2013;76:150–172. pmid:23317872
- 6. Fontana AJ. Appendix D: minimum water activity limits for growth of microorganisms. In: Barbosa-Cánovas GV, Fontana AJ, Schmidt SJ, Labuza TP, editors. Water activity in foods: fundamentals and applications. Ames: Blackwell Publishing Professional; 2007. pp. 405.
- 7. Akdogan H. High moisture food extrusion. Int J Food Sci Technol. 1999;34:195–207.
- 8. Poolman B, Blount P, Folgering JHA, Friesen RHE, Moe PC, van der Heide T. How do membrane proteins sense water stress? Mol Microbiol. 2002;44: 889–902. pmid:12010487
- 9. Wood JM, Bremer E, Csonka LN, Kraemer R, Poolman B, van der Heide T, et al. Osmosensing and osmoregulatory compatible solute accumulation by bacteria. Comp Biochem Physiol A Mol Integr Physiol. 2001;130: 437–460. pmid:11913457
- 10. Wood JM. Bacterial osmosensing transporters. Methods Enzymol. 2007;428: 77–107. pmid:17875413
- 11. Finn S, Handler K, Condell O, Colgan A, Cooney S, McClure P, et al. ProP is required for the survival of desiccated Salmonella enterica serovar Typhimurium cells on a stainless steel surface. Appl Environ Microbiol. 2013;79: 4376–4384. pmid:23666329
- 12. Li H, Bhaskara A, Megalis C, Tortorello ML. Transcriptomic analysis of Salmonella desiccation resistance. Foodborne Pathog Dis. 2012;9: 1143–1151. pmid:23237410
- 13. Gruzdev N, Pinto R, Sela S. Effect of desiccation on tolerance of Salmonella enterica to multiple stresses. Appl Environ Microbiol. 2011;77: 1667–1673. pmid:21216905
- 14. Zheng L, White RH, Cash VL, Jack RF, Dean DR. Cysteine desulfurase activity indicates a role for NIFS in metallocluster biosynthesis. Proc Natl Acad Sci U S A. 1993;90: 2754–2758. pmid:8464885
- 15. Yuvaniyama P, Agar JN, Cash VL, Johnson MK, Dean DR. NifS-directed assembly of a transient [2Fe-2S] cluster within the NifU protein. Proc Natl Acad Sci U S A. 2000;97: 599–604. pmid:10639125
- 16. Zheng L, White RH, Cash VL, Dean DR. Mechanism for the desulfurization of L-cysteine catalyzed by the nifS gene product. Biochem. 1994;33: 4714–4720.
- 17. Krebs C, Agar JN, Smith AD, Frazzon J, Dean DR, Huynh BH, et al. IscA, an alternate scaffold for Fe-S cluster biosynthesis. Biochem. 2001;40: 14069–14080.
- 18. Saini A, Mapolelo DT, Chahal HK, Johnson MK, Outten FW. SufD and SufC ATPase activity are required for iron acquisition during in vivo Fe-S cluster formation on SufB. Biochem. 2010;49: 9402–9412.
- 19. Gruzdev N, McClelland M, Porwollik S, Ofaim S, Pinto R, Saldinger-Sela S. Global transcriptional analysis of dehydrated Salmonella enterica serovar Typhimurium. Appl Environ Microbiol. 2012;78: 7866–7875. pmid:22941081
- 20. Kirby RM, Davies R. Survival of dehydrated cells of Salmonella Typhimurium LT2 at high temperatures. J Appl Bacteriol. 1990;68: 241–246. pmid:2187845
- 21. Archer J, Jervis ET, Bird J, Gaze JE. Heat resistance of Salmonella Weltevreden in low-moisture environments. J Food Prot. 1998;61: 969–973. pmid:9713756
- 22. Baird-Parker AC, Boothroyd M, Jones E. The effect of water activity on the heat resistance of heat sensitive and heat resistant strains of salmonellae. J Appl Bacteriol. 1970;33: 515–522. pmid:5491626
- 23. Hiramatsu R, Matsumoto M, Sakae K, Miyazaki Y. Ability of Shiga toxin-producing Escherichia coli and Salmonella spp. to survive in a desiccation model system and in dry foods. Appl Environ Microbiol. 2005;71: 6657–6663. pmid:16269694
- 24. Taneja S, Ahmad F. Increased thermal stability of proteins in the presence of amino acids. Biochem J. 1994;303: 147–153. pmid:7945233
- 25. Earnshaw RG, Appleyard J, Hurst RM. Understanding physical inactivation processes: combined preservation opportunities using heat, ultrasound and pressure. Int J Food Microbiol. 1995;28: 197–219. pmid:8750667
- 26. Aviles B, Klotz C, Eifert J, Williams R, Ponder M. Biofilms promote survival and virulence of Salmonella enterica sv. Tennessee during prolonged dry storage and after passage through an in vitro digestion system. Int J Food Microbiol. 2013;162: 252–259. pmid:23454816
- 27. Mattick KL, Jorgensen F, Legan JD, Cole MB, Porter J, Lappin-Scott HM, et al. Survival and filamentation of Salmonella enterica serovar Enteritidis PT4 and Salmonella enterica serovar Typhimurium DT104 at low water activity. Appl Environ Microbiol. 2000;66: 1274–1279. pmid:10742199
- 28. White AP, Gibson DL, Kim W, Kay WW, Surette MG. Thin aggregative fimbriae and cellulose enhance long-term survival and persistence of Salmonella. J Bacteriol. 2006;188: 3219–3227. pmid:16621814
- 29. Garmiri P, Coles KE, Humphrey TJ, Cogan TA. Role of outer membrane lipopolysaccharides in the protection of Salmonella enterica serovar Typhimurium from desiccation damage. FEMS Microbiol Lett. 2008;281: 155–159. pmid:18312578
- 30. Iibuchi R, Hara-Kudo Y, Hasegawa A, Kumagai S. Survival of Salmonella on a polypropylene surface under dry conditions in relation to biofilm-formation capability. J Food Prot. 2010;73: 1506–1510. pmid:20819362
- 31. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002;30: 207–210. pmid:11752295
- 32. Datsenko KA, Wanner BL. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A. 2000;97: 6640–6645. pmid:10829079
- 33. Karlinsey JE. lambda-Red genetic engineering in Salmonella enterica serovar Typhimurium. Methods Enzymol. 2007;421: 199–209. pmid:17352924
- 34. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215: 403–410. pmid:2231712
- 35. Carver T, Harris SR, Berriman M, Parkhill J, McQuillan JA. Artemis: an integrated platform for visualization and analysis of high-throughput sequence-based experimental data. Bioinformatics. 2012;28: 464–469. pmid:22199388
- 36. Lazinsky D, Sicko-Goad L. Paraformaldehyde-glutaraldehyde as a routine phytoplankton fixative. Micron. 1979;10: 49–50.
- 37. McClelland M, Sanderson KE, Spieth J, Clifton SW, Latreille P, Courtney L, et al. Complete genome sequence of Salmonella enterica serovar Typhimurium LT2. Nature. 2001;413: 852–856. pmid:11677609
- 38. Kanehisa M, Goto S. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 2000;28: 27–30. pmid:10592173
- 39. Cooper CA, Mulder DT, Allison SE, Pilar AVC, Coombes BK. The SseC translocon component in Salmonella enterica serovar Typhimurium is chaperoned by SscA. BMC Microbiol. 2013;13: 221. pmid:24090070
- 40. Coombes BK, Brown NF, Kujat-Choy S, Vallance BA, Finlay BB. SseA is required for translocation of Salmonella pathogenicity island-2 effectors into host cells. Microbes Infect. 2003;5: 561–570. pmid:12787732
- 41. Ruiz-Albert J, Mundy R, Yu XJ, Beuzon CR, Holden DW. SseA is a chaperone for the SseB and SseD translocon components of the Salmonella pathogenicity-island-2-encoded type III secretion system. Microbiol. 2003;149: 1103–1111.
- 42. Zurawski DV, Stein MA. SseA acts as the chaperone for the SseB component of the Salmonella Pathogenicity Island 2 translocon. Mol Microbiol. 2003;47: 1341–1351. pmid:12603739
- 43. Carsiotis M, Stocker BA, Weinstein DL, O'Brien AD. A Salmonella Typhimurium virulence gene linked to flg. Infect Immun. 1989;57: 3276–3280. pmid:2680969
- 44. Inoue A, Murata Y, Takahashi H, Tsuji N, Fujisaki S, Kato J. Involvement of an essential gene, mviN, in murein synthesis in Escherichia coli. J Bacteriol. 2008;190: 7298–7301. pmid:18708495
- 45. Gunzel D, Kucharski LM, Kehres DG, Romero MF, Maguire ME. The MgtC virulence factor of Salmonella enterica serovar Typhimurium activates Na+,K+-ATPase. J Bacteriol. 2006;188: 5586–5594. pmid:16855249
- 46. Blanc-Potard AB, Solomon F, Kayser J, Groisman EA. The SPI-3 pathogenicity island of Salmonella enterica. J Bacteriol. 1999;181: 998–1004. pmid:9922266
- 47. Kieboom J, Kusumaningrum HD, Tempelaars MH, Hazeleger WC, Abee T, Beumer RR. Survival, elongation, and elevated tolerance of Salmonella enterica serovar Enteritidis at reduced water activity. J Food Prot. 2006;69: 2681–2686. pmid:17133811
- 48. McDonough FE, Hargrove RE. Heat resistance of Salmonella in dried milk. J Dairy Sci. 1968;51: 1587–1591. pmid:5693159
- 49. Deng X, Li Z, Zhang W. Transcriptome sequencing of Salmonella enterica serovar Enteritidis under desiccation and starvation stress in peanut oil. Food Microbiol. 2012;30: 311–315. pmid:22265317
- 50. Humann JL, Kahn ML. Genes involved in desiccation resistance of Rhizobia and other bacteria. In: de Bruign FJ, editor. Biological nitrogen fixation. Hoboken: Wiley-Blackwell; 2015. pp. 397–404.
- 51. Tanaka M, Earl AM, Howell HA, Park MJ, Eisen JA, Peterson SN, et al. Analysis of Deinococcus radiodurans's transcriptional response to ionizing radiation and desiccation reveals novel proteins that contribute to extreme radioresistance. Genetics. 2004;168: 21–33. pmid:15454524
- 52. Cytryn EJ, Sangurdekar DP, Streeter JG, Franck WL, Chang WS, Stacey G, et al. Transcriptional and physiological responses of Bradyrhizobium japonicum to desiccation-induced stress. J Bacteriol. 2007;189: 6751–6762. pmid:17660288
- 53. Su AA, Tripp V, Randau L. RNA-Seq analyses reveal the order of tRNA processing events and the maturation of C/D box and CRISPR RNAs in the hyperthermophile Methanopyrus kandleri. Nucleic Acids Res. 2013;41: 6250–6258. pmid:23620296
- 54. Csonka LN. Physiological and genetic responses of bacteria to osmotic stress. Microbiol Rev. 1989;53: 121–147. pmid:2651863
- 55. da Costa MS, Santos H, Galinski EA. An overview of the role and diversity of compatible solutes in bacteria and archaea. Adv Biochem Eng Biotechnol. 1998;61: 117–153. pmid:9670799
- 56. Roessler M, Muller V. Osmoadaptation in bacteria and archaea: common principles and differences. Environ Microbiol. 2001;3: 743–754. pmid:11846768
- 57. Empadinhas N, da Costa MS. Osmoadaptation mechanisms in prokaryotes: distribution of compatible solutes. Int Microbiol. 2008;11: 151–161. pmid:18843593
- 58. Cairney J, Booth IR, Higgins CF. Osmoregulation of gene expression in Salmonella Typhimurium: proU encodes an osmotically induced betaine transport system. J Bacteriol. 1985;164: 1224–1232. pmid:3905768
- 59. Cairney J, Booth IR, Higgins CF. Salmonella Typhimurium proP gene encodes a transport system for the osmoprotectant betaine. J Bacteriol. 1985;164: 1218–1223. pmid:3905767
- 60. Stirling DA, Hulton CS, Waddell L, Park SF, Stewart GS, Booth IR, et al. Molecular characterization of the proU loci of Salmonella Typhimurium and Escherichia coli encoding osmoregulated glycine betaine transport systems. Mol Microbiol. 1989;3: 1025–1038. pmid:2691838
- 61. Frossard SM, Khan AA, Warrick EC, Gately JM, Hanson AD, Oldham ML, et al. Identification of a third osmoprotectant transport system, the osmU system, in Salmonella enterica. J Bacteriol. 2012;194: 3861–3871. pmid:22609924
- 62. Sleator RD, Hill C. Bacterial osmoadaptation: the role of osmolytes in bacterial stress and virulence. FEMS Microbiol Rev. 2002;26: 49–71. pmid:12007642
- 63. Csonka LN, Hanson AD. Prokaryotic osmoregulation: genetics and physiology. Annu Rev Microbiol. 1991;45: 569–606. pmid:1741624
- 64. Kempf B, Bremer E. Uptake and synthesis of compatible solutes as microbial stress responses to high-osmolality environments. Arch Microbiol. 1998;170:319–330. pmid:9818351
- 65. Strom AR, Kaasen I. Trehalose metabolism in Escherichia coli: stress protection and stress regulation of gene expression. Mol Microbiol. 1993;8: 205–210. pmid:8391102
- 66. Nikolaus T, Deiwick J, Rappl C, Freeman JA, Schroder W, Miller SI, et al. SseBCD proteins are secreted by the type III secretion system of Salmonella pathogenicity island 2 and function as a translocon. J Bacteriol. 2001;183: 6036–6045. pmid:11567004
- 67. Cirillo DM, Valdivia RH, Monack DM, Falkow S. Macrophage-dependent induction of the Salmonella pathogenicity island 2 type III secretion system and its role in intracellular survival. Mol Microbiol. 1998;30: 175–188. pmid:9786194
- 68. Jones MA, Wood MW, Mullan PB, Watson PR, Wallis TS, Galyov EE. Secreted effector proteins of Salmonella Dublin act in concert to induce enteritis. Infect Immun. 1998;66: 5799–5804. pmid:9826357
- 69. Brumell JH, Kujat-Choy S, Brown NF, Vallance BA, Knodler LA, Finlay BB. SopD2 is a novel type III secreted effector of Salmonella Typhimurium that targets late endocytic compartments upon delivery into host cells. Traffic. 2003;4: 36–48. pmid:12535274
- 70. Haraga A, Ohlson MB, Miller SI. Salmonellae interplay with host cells. Nat Rev Microbiol. 2008;6: 53–66. pmid:18026123
- 71. Jiang XJ, Rossanese OW, Brown NF, Kujat-Choy S, Galan JE, Finlay BB, et al. The related effector proteins SopD and SopD2 from Salmonella enterica serovar Typhimurium contribute to virulence during systemic infection of mice. Mol Microbiol. 2004;54: 1186–1198. pmid:15554961
- 72. Chakravortty D, Rohde M, Jager L, Deiwick J, Hensel M. Formation of a novel surface structure encoded by Salmonella Pathogenicity Island 2. EMBO J. 2005;24: 2043–2052. pmid:15889142
- 73. Bulmer DM, Kharraz L, Grant AJ, Dean P, Morgan FJ, Karavolos MH, et al. The bacterial cytoskeleton modulates motility, type 3 secretion, and colonization in Salmonella. PLoS Pathog. 2012;8: e1002500. pmid:22291596
- 74. Bendezu FO, de Boer PA. Conditional lethality, division defects, membrane involution, and endocytosis in mre and mrd shape mutants of Escherichia coli. J Bacteriol. 2008;190: 1792–1811. pmid:17993535
- 75. Wachi M, Doi M, Tamaki S, Park W, Nakajima-Iijima S, Matsuhashi M. Mutant isolation and molecular cloning of mre genes, which determine cell shape, sensitivity to mecillinam, and amount of penicillin-binding proteins in Escherichia coli. J Bacteriol. 1987;169: 4935–4940. pmid:2822655
- 76. Doi M, Wachi M, Ishino F, Tomioka S, Ito M, Sakagami Y, et al. Determinations of the DNA sequence of the mreB gene and of the gene products of the mre region that function in formation of the rod shape of Escherichia coli cells. J Bacteriol. 1988;170: 4619–4624. pmid:3049542
- 77. Wachi M, Doi M, Okada Y, Matsuhashi M. New mre genes mreC and mreD, responsible for formation of the rod shape of Escherichia coli cells. J Bacteriol. 1989;171:6511–6516. pmid:2687239
- 78. Jones LJ, Carballido-Lopez R, Errington J. Control of cell shape in bacteria: helical, actin-like filaments in Bacillus subtilis. Cell. 2001;104: 913–922. pmid:11290328
- 79. van den Ent F, Amos LA, Löwe J. Prokaryotic origin of the actin cytoskeleton. Nature. 2001;413: 39–44. pmid:11544518
- 80. Oliver JD. Recent findings on the viable but nonculturable state in pathogenic bacteria. FEMS Microbiol Rev. 2010;34: 415–425. pmid:20059548
- 81. Eriksson de Rezende CL, Mallinson ET, Gupte A, Joseph SW. Salmonella spp. are affected by different levels of water activity in closed microcosms. J Ind Microbiol Biotechnol. 2001;26: 222–225. pmid:11464270
- 82. Podolak R, Enache E, Stone W, Black DG, Elliott PH. Sources and risk factors for contamination, survival, persistence, and heat resistance of Salmonella in low-moisture foods. J Food Prot. 2010;73: 1919–1936. pmid:21067682
- 83. Gruzdev N, Pinto R, Sela Saldinger S. Persistence of Salmonella enterica during dehydration and subsequent cold storage. Food Microbiol. 2012;32: 415–422. pmid:22986208
- 84. Ophir T, Gutnick DL. A role for exopolysaccharides in the protection of microorganisms from desiccation. Appl Environ Microbiol. 1994;60: 740–745. pmid:16349202
- 85. Gharzouli R, Carpene MA, Couderc F, Benguedouar A, Poinsot V. Relevance of fucose-rich extracellular polysaccharides produced by Rhizobium sullae strains nodulating Hedysarum coronarium L. legumes. Appl Environ Microbiol. 2013;79: 1764–1776. pmid:23183977
- 86. Roberson EB, Firestone MK. Relationship between desiccation and exopolysaccharide production in a soil Pseudomonas sp. Appl Environ Microbiol. 1992;58: 1284–1291. pmid:16348695
- 87. Weening EH, Barker JD, Laarakker MC, Humphries AD, Tsolis RM, Bäumler AJ. The Salmonella enterica serotype Typhimurium lpf, bcf, stb, stc, std, and sth fimbrial operons are required for intestinal persistence in mice. Infect Immun. 2005;73: 3358–3366. pmid:15908362
- 88. Morgan E, Campbell JD, Rowe SC, Bispham J, Stevens MP, Bowen AJ, et al. Identification of host-specific colonization factors of Salmonella enterica serovar Typhimurium. Mol Microbiol. 2004;54: 994–1010. pmid:15522082
- 89. Szeto J, Namolovan A, Osborne SE, Coombes BK, Brumell JH. Salmonella-containing vacuoles display centrifugal movement associated with cell-to-cell transfer in epithelial cells. Infect Immun. 2009;77: 996–1007. pmid:19103768
- 90. Billi D, Potts M. Life and death of dried prokaryotes. Res Microbiol. 2002;153: 7–12. pmid:11881900
- 91. Potts M. Desiccation tolerance of prokaryotes. Microbiol Rev. 1994;58: 755–805. pmid:7854254
- 92. Colgan AM, Kroger C, Diard M, Hardt WD, Puente JL, Sivasankaran SK, et al. The Impact of 18 Ancestral and Horizontally-Acquired Regulatory Proteins upon the Transcriptome and sRNA Landscape of Salmonella enterica serovar Typhimurium. PLoS Genet. 2016;12: e1006258. pmid:27564394
- 93. Smirnov A, Forstner KU, Holmqvist E, Otto A, Gunster R, Becher D, et al. Grad-seq guides the discovery of ProQ as a major small RNA-binding protein. Proc Natl Acad Sci U S A. 2016;113: 11591–11596. pmid:27671629
- 94. Ramachandran VK, Shearer N, Jacob JJ, Sharma CM, Thompson A. The architecture and ppGpp-dependent expression of the primary transcriptome of Salmonella Typhimurium during invasion gene expression. BMC Genomics. 2012; 13–25.
- 95. Kröger C, Colgan A, Srikumar S, Handler K, Sivasankaran SK, Hammarlof DL, et al. An infection-relevant transcriptomic compendium for Salmonella enterica serovar Typhimurium. Cell Host Microbe. 2013;14: 683–695. pmid:24331466