Microbiol Immunol 2012; 56: 228–237 doi:10.1111/j.1348-0421.2012.00428.x ORIGINAL ARTICLE General stress sigma factor RpoS influences time required to enter the viable but non-culturable state in Salmonella enterica Akiko Kusumoto 1 , Hiroshi Asakura 2 and Keiko Kawamoto 1 1 Research Center for Animal Hygiene and Food Safety, Obihiro University of Agriculture and Veterinary Medicine, 2–11 Inada, Obihiro, Hokkaido 080–8555, JAPAN, and 2 Division of Biomedical Food Research, National Institute of Health Sciences, Kami-yoga 1–18-1, Setagaya-ku, Tokyo 158–8501, Japan ABSTRACT In stressful conditions, bacteria enter into the viable but non-culturable (VBNC) state; in this state, they are alive but fail to grow on conventional media on which they normally grow and develop into colonies. The molecular basis underlying this state is unknown. We investigated the role of the alternative sigma factor RpoS (σ 38 ) in the VBNC induction using Salmonella Dublin, Salmonella Oranienburg and Salmonella Typhimurium LT2. VBNC was induced by osmotic stress in LT2 and Oranienburg. Dublin also entered the VBNC state, but more slowly than LT2 and Oranienburg did. The LT2 rpoS gene was initiated from an alternative initiation codon, TTG; therefore, LT2 had smaller amounts of RpoS than Dublin and Oranienburg. Oranienburg had a single amino acid substitution (D118N) in RpoS (RpoS SO ). Disruption of rpoS caused rapid VBNC induction. VBNC induction was significantly delayed by Dublin-type RpoS (RpoS SD ), but only slightly by RpoS SO . These results indicate that RpoS delays VBNC induction and that the rapid induction of VBNC in LT2 and Oranienburg may be due to lower levels of RpoS and to the D118N amino acid substitution, respectively. Reduced RpoS intracellular level was observed during VBNC induction. During the VBNC induction, Salmonella might regulate RpoS which is important for maintenance of culturablity under stresses. Key words osmotic stress, RpoS, stress response, VBNC. When exposed to harsh environmental stresses, bacteria struggle to adapt for their survival. One of the strategies, there is a mounting evidence that bacteria enter the viable but non-culturable (VBNC) state (1). In this state, bac- teria lose their culturability in conventional media, while retaining metabolic activity (2, 3), membrane integrity (4), respiration (2) and slow gene transcription (5, 6). A series of bacteria could be recovered from the VBNC state to restart cell division by favorable conditions (2, 3, 7–10). Correspondence Keiko Kawamoto, Research Center for Animal Hygiene and Food Safety, Obihiro University of Agriculture and Veterinary Medicine, 2–11 In- ada, Obihiro, Hokkaido 080–8555, Japan. Tel: 81 155 49 5890; Fax: 81 155 49 5890; email: [email protected]Received 5 October 2011; revised 26 December 2011; accepted 11 January 2012. List of Abbreviations: CFU, colony forming unit; E., Escherichia; NB, Nutrient Broth; S., Salmonella; SD, Salmonella Dublin; SO, Salmonella Oranienburg; VBNC, viable but non-culturable. The alternative sigma factor RpoS (or σ 38 ) is con- served within γ , β, and δ-proteobacteria (11), and func- tions as a master regulator of the general stress response (12). The strain-dependent DNA polymorphism of rpoS gene was reported in E. coli (13, 14). The RpoS binds to core RNA polymerase, which consists of 2 α,1 β, 1 β , and 1 ω subunit, to form the RNA polymerase holoenzyme complex (15). By specifically recognizing the RpoS-dependent promoter sequence, RpoS directs 228 c 2012 The Societies and Blackwell Publishing Asia Pty Ltd
General stress sigma factor RpoS influences time requiredto enter the viable but non-culturable state in SalmonellaentericaAkiko Kusumoto1, Hiroshi Asakura2 and Keiko Kawamoto1
1Research Center for Animal Hygiene and Food Safety, Obihiro University of Agriculture and Veterinary Medicine, 2–11Inada, Obihiro, Hokkaido 080–8555, JAPAN, and 2Division of Biomedical Food Research, National Institute of Health Sciences, Kami-yoga 1–18-1,Setagaya-ku, Tokyo 158–8501, Japan
ABSTRACTIn stressful conditions, bacteria enter into the viable but non-culturable (VBNC) state; in this state,they are alive but fail to grow on conventional media on which they normally grow and develop intocolonies. The molecular basis underlying this state is unknown. We investigated the role of the alternativesigma factor RpoS (σ38) in the VBNC induction using Salmonella Dublin, Salmonella Oranienburg andSalmonella Typhimurium LT2. VBNC was induced by osmotic stress in LT2 and Oranienburg. Dublin alsoentered the VBNC state, but more slowly than LT2 and Oranienburg did. The LT2 rpoS gene was initiatedfrom an alternative initiation codon, TTG; therefore, LT2 had smaller amounts of RpoS than Dublin andOranienburg. Oranienburg had a single amino acid substitution (D118N) in RpoS (RpoSSO). Disruptionof rpoS caused rapid VBNC induction. VBNC induction was significantly delayed by Dublin-type RpoS(RpoSSD), but only slightly by RpoSSO. These results indicate that RpoS delays VBNC induction andthat the rapid induction of VBNC in LT2 and Oranienburg may be due to lower levels of RpoS and tothe D118N amino acid substitution, respectively. Reduced RpoS intracellular level was observed duringVBNC induction. During the VBNC induction, Salmonella might regulate RpoS which is important formaintenance of culturablity under stresses.
Key words osmotic stress, RpoS, stress response, VBNC.
When exposed to harsh environmental stresses, bacteriastruggle to adapt for their survival. One of the strategies,there is a mounting evidence that bacteria enter the viablebut non-culturable (VBNC) state (1). In this state, bac-teria lose their culturability in conventional media, whileretaining metabolic activity (2, 3), membrane integrity(4), respiration (2) and slow gene transcription (5, 6). Aseries of bacteria could be recovered from the VBNC stateto restart cell division by favorable conditions (2, 3, 7–10).
CorrespondenceKeiko Kawamoto, Research Center for Animal Hygiene and Food Safety, Obihiro University of Agriculture and Veterinary Medicine, 2–11 In-ada, Obihiro, Hokkaido 080–8555, Japan.Tel: 81 155 49 5890; Fax: 81 155 49 5890; email: [email protected]
Received 5 October 2011; revised 26 December 2011; accepted 11 January 2012.
List of Abbreviations: CFU, colony forming unit; E., Escherichia; NB, Nutrient Broth; S., Salmonella; SD, Salmonella Dublin; SO, SalmonellaOranienburg; VBNC, viable but non-culturable.
The alternative sigma factor RpoS (or σ38) is con-served within γ, β, and δ-proteobacteria (11), and func-tions as a master regulator of the general stress response(12). The strain-dependent DNA polymorphism of rpoSgene was reported in E. coli (13, 14). The RpoS bindsto core RNA polymerase, which consists of 2 α, 1 β,1 β′, and 1 ω subunit, to form the RNA polymeraseholoenzyme complex (15). By specifically recognizingthe RpoS-dependent promoter sequence, RpoS directs
RNA polymerase to initiate transcription from promoterscontaining this sequence, thereby, governing the expres-sion of numerous genes, most of which work undervarious stresses, such as nutrient starvation (16), os-motic stress (17), high hydrostatic pressure (18), oxida-tive stress (19), DNA damage (20), and acid stress (21).Thus, it seems likely that RpoS plays an important role inadaptation to the harsh foreign stresses (22).
Four conserved regions (numbered 1 to 4 from the N-terminal end), have been identified in the σ70 family towhich RpoS belongs; some of these are further dividedinto subregions (23). Region 2 and 4 are the most highlyconserved among the various σ factors, and bind to the−10 and −35 promoter sequence motifs, respectively (24,25). Region 1 is further divided into two subregions, 1.1and 1.2. Subregion 1.1 in Escherichia coli RpoD has beensuggested to function as preventing region 2 and 4 frombinding to DNA by covering them in free RpoD (26, 27),and mediating the binding of RpoD with core enzyme(28, 29), thereby influencing promoter binding and thetranscription initiation by holoenzyme (25, 30–34).
Salmonella is one of the leading enteric pathogen, andits infection in human is most often associated with con-sumption of contaminated foodstuffs. We previously re-ported that Salmonella entered into the VBNC state underosmotic stress with its ability to return to the prolifera-tion stage (35). The RpoS function has been attributed tothe osmotolerance mainly in E. coli, a model microor-ganism (12), nevertheless it remained veiled how theRpoS associates with the entry into the VBNC state inSalmonella.
In this study, we show the nucleotide polymorphismin the rpoS gene among Salmonella strains. Through-out the gene disruption and overexpression researches,we could demonstrate that the RpoS expression deter-mines the VBNC entry of Salmonella cells under osmoticstress.
MATERIALS AND METHODS
Bacterial strains and growth conditions
Salmonella and E. coli strains used in this study are listedin Table 1. Salmonella was cultured at 30 or 37 ◦C in Lmedium [1% (w/v) tryptone, 0.5% (w/v) yeast extract, and0.5% (w/v) NaCl], LB medium [1% (w/v) tryptone, 0.5%(w/v) yeast extract, and 1% (w/v) NaCl] or Nutrient Brothmedium (Becton Dickinson, Franklin Lakes, NJ, USA). E.coli was cultured in LB medium. When necessary, thefollowing antibiotics were used: kanamycin (50 μg ml−1)and carbenicillin (100 μg ml−1).
Table 1. Bacterial strains and plasmids used in this study
Strain or Genotype or Reference orplasmid description∗ source
Salmonella strainsLT2 S. Typhimurium, wild type (51)KDX1 S. Dublin, wild type Laboratory stockSa99004 S. Oranienburg, wild type (42)KLX1 LT2 rpoS::kan This studyKDX2 KDX1 rpoS::kan This study
PlasmidspKD46 λRed (γ, β, exo) (36)pKD4 Kn (36)pBAD18 PBAD araC Ampr (52)pAKO1 pBAD18 rpoSLT2 This studypAKO2 pBAD18 rpoSSD This studypAKO3 pBAD18 rpoSSO This study
The chromosomal rpoS gene in S. Typhimurium strainLT2 and S. Dublin strain KDX1 was deleted as previouslydescribed (36), using the following primers: delta rpoSF (5′-GAAATCCGTAAACCCGCTGCGTTATTTACCGCAGCGATAAGTGTAGGCTGGAGCTGCTTC-3′) anddelta rpoS R (5′-TTACTCGCGGAACAGCGCTTCGATATTCAGCCCCTGCGTCCATATGAATATCCTCCTTAG-3′). The 64-bp region upstream of the initiation codon ofthe rpoS gene and the 973-bp initial region of the 993-bpopen reading frame of the rpoS gene were deleted. Thesuccessful disruptants were subjected to western blot toconfirm inactivation of RpoS protein, and named KLX1(LT2 rpoS) and KDX2 (KDX1 rpoS).
Plasmid construction
Routine DNA manipulations were carried out accordingto standard procedures (37). Restriction endonucleasesand other enzymes for DNA manipulations were pur-chased from TaKaRa Shuzo (Japan), Toyobo (Japan), andNew England Biolabs (Ipswich, MA, USA). Nucleotidesequences were determined using the BigDye Terminatorv3.1 Cycle Sequencing Kit (Applied Biosystems, Carlsbad,CA, USA) and an ABI PRISM 310 Genetic Analyzer (Ap-plied Biosystems).
The plasmids used in this study are listed in Table 1.The 993-bp open reading frame of rpoS gene and its 32-bp
upstream sequence containing natural ribosomal bindingsite was PCR-amplified from purified DNA of the LT2strain and cloned into pBAD18 vector plasmid to obtainpAKO1 (pBAD18 rpoSLT2). The alternative TTG initiationcodon of rpoSLT2 in pAKO1 was replaced by ATG to ob-tain pAKO2 (pBAD18 rpoSSD). The D118N substitutionwas introduced into rpoSSD of pAKO2 to obtain pAKO3(pBAD18 rpoSSO).
Induction of the VBNC state
Induction of the VBNC state was performed as previouslydescribed (35). Briefly, cells grown in NB medium at 37 ◦Cfor 20 h were washed 3 times with 0.85% NaCl, followed byincubation in 7% NaCl (5× volume of the original culture)at 37 ◦C. For plasmid-bearing cells, arabinose was addedto the 16h-incubated cell cultures to a final concentrationof 2 mM, followed by a 4 h incubation. Cells were thenwashed and subjected to incubation in 7% NaCl solution,as described above. At the designed time points, 0.1 ml ofthe suspension and its serial dilutions were plated onto LBagar plates to determine CFUs. Viability was determinedby the LIVE/DEAD BacLight Bacterial Viability Kit (In-vitrogen, Carlsbad, CA, USA). BacLight-stained cells wereobserved under fluorescence microscopy (Olympus BX51,Olympus, Melville, NY, USA), and images of more than5 random fields were taken by Olympus DP70 system(Olympus). Viability of at least 200 cells from a represen-tative image was analyzed.
In this study, we defined VBNC state as bacterial cellswhich form no colonies on LB plates (<10 CFU/mL) yetmaintain high viability (>70%).
Detection of RpoS protein by westernblotting
Bacterial cells were harvested by centrifugation, suspendedin distilled water at an OD660 of 1.0, and then preparedwith Lammli sample buffer (Bio-Rad, CA, USA). Pro-teins were resolved by SDS-PAGE, and transferred to aPVDF membrane (Millipore, Billerica, MA, USA) using asemi-dry blotting apparatus (BIO CRAFT, Japan). West-ern blotting was performed with an anti-RpoS antibody1RS1 (Abcam, Cambridge, UK).
RESULTS
Genetic variation of the rpoS gene amongSalmonella enterica subspecies enterica
To know the existence of rpoS alleles among Salmonella,we searched the nucleotide sequences of rpoS gene from15 Salmonella strains mainly through the NCBI database,except for S. Oranienburg Sa99004 (Fig. 1). Since the ge-
nomic sequence of S. Oranienburg was not available, wesequenced the rpoS gene of S. Oranienburg strain Sa99004,and deposited it in DDBJ under accession No. AB619833.S. Typhimurium LT2 rpoS (rpoSLT2) is known to startwith the alternative initiation codon TTG instead of ATG,as previously reported (38). The amino acid sequence ofRpoS of 12 Salmonella strains, S. Typhimurium SL1344, S.Typhimurium UK-1, S. Typhimurium 14028S, S. Dublin,S. Choleraesuis, S. Schwarzengrund, S. Enteritidis, S.Heidelberg, S. Newport, S. Paratyphi A, S. Paratyphi C,and S. Typhi CT18, was identical to LT2 RpoS; however,their rpoS started with ATG unlike rpoSLT2. S. Gallinarumand S. Oranienburg had a substitution at D14E or D118N,respectively. S. Paratyphi B had a nonsense mutation atthe 243rd amino acid, where the GAG codon for gluta-mate was changed to the TAG stop codon. S. Agona had a1338-bp insertion at the codon corresponding to the 259thamino acid and therefore exhibited extra 11-amino-acidsequence. S. Typhi strain Ty2 had a frameshift mutationcaused by a single nucleotide insertion at the codon cor-responding to the 311th amino acid, providing extra 73amino acids.
We selected three strains, S. Typhimurium strain LT2, S.Oranienburg strain Sa99004, and S. Dublin strain KDX1to investigate the effect of differential rpoS alleles on bacte-rial survival and VBNC induction under osmotic stress.
The rpoS alleles determine RpoS productionin Salmonella
Given the DNA polymorphism of rpoS gene, RpoS proteinexpression was comparatively examined by western blot.In stationary phase, higher levels of RpoS were detected inS. Dublin and S. Oranienburg than in LT2 (2.2- and 2.5-fold, respectively) (Fig. 2a and 2b). We then introducedthe plasmids harbouring different alleles of rpoS gene(rpoSLT2, rpoSSD, and rpoSSO) (Fig. 2c and 2d) into strainLT2 rpoS (designated as KLX1 strain), and their RpoSproduction were comparatively examined under station-ary phase. Plasmid pAKO1 (rpoSLT2) did not alter the RpoSproduction level compared with the wild type LT2 strain,whereas the strains KLX1 harboring pAKO2 (rpoSSD) orpAKO3 (rpoSSO) exhibited greater amounts of RpoS thanthe wild type LT2 strain (5.7- and 8.3-folds, respectively)(Fig. 2c and 2d). Thus, these results indicate that the al-ternative initiation codon TTG caused a reduction in theexpression of RpoS while the D118N substitution did notaffect it.
VBNC induction in S. Dublin, S. Oranienburg,and LT2
The three strains were incubated in 7% NaCl to mea-sure the culturability and membrane integrity (used as
Fig. 1. Alignment of the amino acid sequences of RpoS proteins from Salmonella enterica subsp. enterica strains: S. Typhimuriumstrain LT2, S. Typhimurium strain SL1344 (SL), S. Typhimurium strain UK-1 (UK1), S. Typhimurium strain 14028S (140), S. Dublin strainCT_02021853 (SD), S. Choleraesuis strain SC-B67 (SC), S. Schwarzengrund strain CVM19633 (SS), S. Enteritidis strain P125109 (SE), S.Heidelberg strain SL476 (SH), S. Newport strain SL254 (SN), S. Paratyphi A strain AKU_12601 (SPA), S. Paratyphi C strain RKS4594 (SPC),S. Typhi strain CT18 (STC), S. Oranienburg strain Sa99004 (SO), S. Gallinarum strain 287/91 (SG), S. Agona strain SL483 (SA), S. ParatyphiB strain SPB7 (SPB), and S. Typhi strain Ty2 (STT). The amino acid sequences of Salmonella strains other than S. Oranienburg were retrieved fromthe protein database of NCBI. Amino acid sequence of LT2 RpoS is shown at the top of the alignment. Dots indicate amino acid residues identical toLT2 RpoS. Boxes indicate positions not in agreement with LT2 sequence, and the actual amino acid residue for that position is shown.
Fig. 2. Detection of RpoS protein bywestern blotting using an anti-RpoSantibody. (a) RpoS protein in S. Typhimuriumstrain LT2, KLX1 (LT2 rpoS), S. Dublin strainKDX1 (SD), KDX2 (SD rpoS), and S.Oranienburg strain Sa99004 (SO) during thestationary phase. We analyzed the signalintensity of full-length RpoS protein comparedwith that in LT2 (b). (c) RpoS protein in KLX1 (LT2rpoS) cells containing the plasmids; pBAD18(vector control), pAKO1 (rpoSLT2), pAKO2(rpoSSD) or pAKO3 (rpoSSO). In plasmid-bearingcells grown to the stationary phase, RpoSexpression was induced by 2 mM arabinose for4 h. The signal intensity of full-length RpoSprotein compared with LT2 harboring pBAD18was analyzed (d). The closed and open barsindicate the presence or absence of 2 mMarabinose, respectively. The arrowheads indicatethe full-length RpoS protein.
viability marker). At 2 days post incubation, LT2 did notform colonies (< 10 CFU/mL), however, the viability re-mained at 72% even at 10 days post incubation (Fig. 3),indicating that LT2 entered the VBNC state after 2 days
0
10
20
30
40
50
60
70
80
90
100
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10
Via
bilit
y (%
)
log 1
0C
FU
/mL
Incubation time (day)
LT2, log10CFU/mL
SD, log10CFU/mL
SO, log10CFU/mL
LT2, viability
SD, viability
SO, viability
Fig. 3. Induction of the VBNC state in S. Typhimurium strain LT2, S.Dublin strain KDX1, and S. Oranienburg strain Sa99004. The VBNCstate was induced in LT2 (circles), S. Dublin strain KDX1 (triangles), andS. Oranienburg strain Sa99004 (squares) by osmotic stress as describedin the Experimental Procedures. The culturability (log10CFU/mL) and vi-ability of Salmonella cells are indicated by the solid and dashed lines,respectively. Data are representative of more than three independentexperiments.
of incubation. Similarly, S. Oranienburg Sa99004 enteredthe VBNC state after 3 days of incubation (Fig. 3). On theother hand, S. Dublin KDX1 still retained culturability at1.1 × 102 CFU/mL at 3 days post incubation, and it took5 days to enter the VBNC (Fig. 3). Together, these dataindicated that those Salmonella strains displayed differ-ent time-course for the VBNC entry under the osmoticstimuli.
The rpoS alleles determine the salt-inducibleVBNC state in Salmonella enterica
Osmotic stress reduced the amount of RpoS proteinin the three strains (S. Dublin KDX1, S. OranienburgSa99004, and LT2) during the induction of the VBNC state(Fig. 4). The amount of RpoS was rapidly decreased in LT2and S. Oranienburg. The reduction rate at day 1 post os-motic stress in LT2 and S. Oranienburg was 50% and 40%,respectively. In contrast, RpoS protein was decreased grad-ually in S. Dublin, which took longest period to becomenon-culturable state. These results provide us to assumethe potent involvement of the DNA polymorphism of rpoSgene in the salt-induced VBNC in Salmonella.
In order to explore this idea, we disrupted the rpoS genefrom the strains LT2 and KDX1 (designated as strainsKLX1 and KDX2) and their survival and viability weremeasured under the osmotic stress (Fig. 5). Microbialcolony count assays showed that the strain KLX1 (LT2rpoS) lost culturability earlier than the parental strainLT2 (Fig. 5a), although both of them retained high viability
Fig. 4. Intracellular levels of RpoS protein during the induction ofVBNC. RpoS protein in S. Typhimurium strain LT2, S. Dublin strain KDX1(SD), and S. Oranienburg strain Sa99004 (SO) during the induction ofVBNC was detected by western blotting using an anti-RpoS antibody(a). The signal intensity of full-length RpoS protein compared with thatat day 0 was analyzed (b). Circles, triangles, and squares indicate S. Ty-phimurium strain LT2, S. Dublin strain KDX1, and S. Oranienburg strainSa99004, respectively. The arrowheads indicate the full-length RpoSprotein.
even after 7 days of incubation (85.4% and 84.8%, respec-tively, Fig. 5a). Similarly, S. Dublin strain KDX2 (KDX1rpoS) also exhibited faster course of VBNC entry underthe stress than the parental strain KDX1 whereas they didnot show any significance in the viability shift (Fig. 5b).Together, these results indicated that disruption of rpoSgene accelerated course of VBNC induction in Salmonellaby osmotic stress.
Concerning the association of rpoS alleles with the RpoSproduction, we next used the strains KLX1 (LT2 rpoS)harbouring plasmids pAKO1 (rpoSLT2), pAKO2 (rpoSSD),or pAKO3 (rpoSSO) to examine their physiological traitsunder the high osmolarity. These strains were osmoticallyinduced to the VBNC state (Fig. 6); strain KLX1 harbor-ing empty vector entered the VBNC state at 1 day afterexposure to the osmotic stress, while rpoSSD-expressingcells took longer to enter the VBNC state (6 days), in-dicating that the expression of rpoS delayed the entry ofSalmonella into the VBNC state. The rpoSLT2- or rpoSSO-expressing strains entered into the VBNC state at 2 dayspost incubation (Fig. 6). Together, we could show that
0102030405060708090100
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12
Via
bilit
y (%
)
log 1
0CF
U/m
L
Incubation time (day)
(a) LT2
0102030405060708090100
0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7
Via
bilit
y (%
)
log 1
0CF
U/m
L
Incubation time (day)
rpoS+, log10CFU/mL
ΔrpoS, log10CFU/mL
rpoS+, viability
ΔrpoS, viability
(b) SD
Fig. 5. Induction of the VBNC state in the rpoS gene disruptants.The VBNC state was induced in S. Typhimurium strain LT2 (rpoS+, cir-cles) and strain KLX1 (LT2 rpoS, triangles) (a), or S. Dublin strain KDX1(rpoS+, circles) and strain KDX2 (S. Dublin rpoS, triangles) (b) by os-motic stress, as described in the Experimental Procedures. The cultura-bility (log10CFU/mL) and viability of Salmonella cells are indicated by thesolid and dashed lines, respectively. Data are representative of more thanthree independent experiments.
the overexpression of RpoS, delays the VBNC entry inSalmonella.
DISCUSSION
The VBNC state is a strategy of non-spore-forming bacte-ria to survive under various stresses that are unfavorablefor proliferation. In this state, bacterial cells are unable toproliferate to form colonies, but retain high viability; thusconventional microbiological assays are unable to detectVBNC bacteria (39). When the environmental conditionsare once again suitable for proliferation, such as in a host’sintestine, the bacteria exit the VBNC state to a cultur-able state (3,7–10). It is reported that numerous bacteria,including pathogens, are able to enter the VBNC stateand retain their virulence genes or factors in this state;
Fig. 6. The effect of the expression of rpoS alleles on the inductionof the VBNC state. The VBNC state was induced in KLX1 (LT2 rpoS)cells containing pBAD18 (vector control, circles), pAKO1 (rpoSLT2, trian-gles), pAKO2 (rpoSSD, squares) or pAKO3 (rpoSSO, rhombi). In plasmid-bearing cells grown to the stationary phase, RpoS expression was in-duced by 2 mM arabinose for 4 h. After RpoS expression, the cells wereexposed to osmotic stress to induce the VBNC state. The culturability(log10CFU/mL) and viability are indicated by the solid and dashed lines,respectively. Data are representative of more than three independentexperiments.
therefore, contamination of food products by VBNC bac-teria is of public health concern (1).
The loss of culturability is one of the significant featuresof VBNC cells. The possible involvement of RpoS in thebacterial culturability was suggested by previous studies(40, 41). However, little is known about its role in VBNC.In this study, we investigated effect of rpoS disruptionand expression on VBNC state and intracellular levels ofRpoS during induction of VBNC state, using S. Oranien-burg, S. Dublin, and S. Typhimurium strain LT2. The rpoSgene disruption caused rapid induction of VBNC state(Fig. 5). Expression of RpoS delayed induction of VBNC(Fig. 6). These results indicate that RpoS was not an es-sential protein to induce VBNC and that the presence ofRpoS delayed VBNC entry. Rapid entry of rpoS-deficientstrains into non-culturable state was also observed previ-ously in E.coli (40, 41). The E. coli parental strains reachedthe non-culturable state in 33 days under laboratory mi-crocosms consisting of an artificial oligotrophic mediumincubated at 4 ◦C, whereas rpoS mutants lost their cul-turability in 21 days. However, the rpoS-deficient strainsfailed to retain high viability, and rapidly died after enter-ing VBNC. Thus, they concluded the involvement of the
rpoS gene in E. coli persistence in the VBNC state. Theirobservation that rpoS-deficient strains lost culturabilitymore rapidly than parental strains was consistent with ourfinding. However, in our study, rpoS-deficient Salmonellastill retained the high viability as approximately 70−90%during the observation period, which was comparable toparental strains. RpoS may not be essential to enter VBNCstate in Salmonella spp., however the presence of RpoS af-fects the length of time required to enter this state.
It has been shown that considerable heterogeneity existsin the rpoS gene in E. coli (13, 14). In this study, we foundsuch sequence variability at the rpoS locus in Salmonellaby analyzing the gene sequence of different serovars of S.enterica (Fig. 1). Amino acid sequence of RpoS from 10of 15 Salmonella strains, including S. Dublin strain KDX1and S. Typhimurim strain LT2, was identical. As previouslyreported, the rpoS gene in LT2 contains the rare initiationcodon TTG (38). S. Oranienburg strain Sa99004 was iso-lated from a patient in nation-wide food-borne outbreakin Japan caused by dried processed squid (42), and foundto have the substitution D118N in rpoS gene. We foundthat VBNC induction was rapidly induced in LT2 and S.Oranienburg compared to S. Dublin (Fig. 3). The varia-tion of the gene may associates with functional variationsthat are involved in phenotypic variation. Thus, we in-vestigated the influence of the alternative initiation codonTTG and the D118N mutation on VBNC phenotype ofLT2 and S. Oranienburg, respectively. Since the presenceof RpoS delayed VBNC induction, we examined the timerequired to enter VBNC and the intracellular levels ofRpoS in LT2, S. Dublin, and S. Oranienburg. The amountof RpoS was decreased in all strains during the course ofosmotic incubation (Fig. 4).
The levels of intracellular RpoS protein are regulatedat the levels of transcription, translation, and proteolysis(reviewed in 22). We previously demonstrated that os-motic stress induced expression of Histone-like proteinH-NS and caused DNA topological change in S. Oranien-burg (43). H-NS preferentially binds to intrinsically bentDNA. Although H-NS reduces the transcription of morethan 100 genes, it also is shown to be a negative regulatorfor rpoS translation (44,45). H-NS binds directly to rpoSmRNA to enhance cleavage of the mRNA (44). Small reg-ulatory RNA (sRNA) DsrA, which promotes rpoS transla-tion by forming a complex with rpoS mRNA, is suppressedby H-NS in the same way. In addition to the regulatoryeffect of H-NS at the translational levels of RpoS produc-tion, H-NS is thought to positively regulate RpoS prote-olysis since RpoS protein is more stable in hns-deficientbackground (45). Smaller fragments indicating degradedRpoS proteins were detected in stationary phase of LT2, S.Dublin, and especially in S. Oranineburg before osmoticstress. But there were no significant increase in amounts of
smaller RpoS fragments during the course of high osmoticincubation, implying that down-regulation at the transla-tional levels was presumably involved in rapid decrease ofRpoS. We previously showed that H-NS was up-regulatedin response to osmotic stress in Salmonella (43). Fromthese findings, it is suggested that H-NS is possibly in-volved in decrease in RpoS during induction of VBNCstate. However, further work is needed to investigate thispossibility.
We introduced different types of rpoS alleles (rpoSLT2,rpoSSD, and rpoSSO) from a plasmid in LT2 rpoS, andcompared the VBNC phenotypes and intracellular levelsof RpoS. The LT2 rpoS expressing rpoSLT2 showed thelowest levels of RpoS protein compared to other trans-formants (Fig. 2c and 2d). As it is reported that usageof alternative initiation codons affects translational effi-ciency (46), the alternative initiation codon of rpoS genemight influence the intracellular levels of RpoS protein attranslational level, resulting in lower levels of intracellularRpoS. In fact, LT2 exhibited the lowest levels of intracel-lular RpoS protein compared to other S. enterica serovarsOranienburg and Dublin (Fig. 2a and 2b). It seems thatLT2 entered VBNC state rapidly due to lower levels ofintracellular RpoS caused by the alternative initiationcodon in rpoS gene. In LT2 rpoS background, RpoSSO
showed similar amounts of RpoS protein to RpoSSD (Fig.2c and 2d). The time required to enter VBNC in RpoSSO-expressing cells is shorter than RpoSSD-expressing cells(Fig. 6). It is likely that LT2 rpoS with rpoSSO mightcause rapid RpoS reduction resulting in rapid inductionof VBNC entry, as well as S. Oranineburg did. This singleamino acid substitution appears to be sufficient to generaterapid decrease of intracellular RpoS in response to osmoticstress. As a result, VBNC state might be induced rapidly inS. Oranienburg.
In addition to the possibility that the D118N substitu-tion might cause rapid decrease of RpoS, there is anotherpossibility that the D118N substitution might affect RpoSfunctions. This possibility is brought by crystal structureand mutational analyses of RpoD (σ70). The D118 residueof RpoS is located in region 2.2 which is the most highlyconserved region among the σ70 family to which RpoSbelongs (23). Crystal structure analysis of RpoD (σ70) re-vealed that this region made a helix and formed a coiled-coil structure with the helices of the other conserved re-gions, and this coiled-coil structure contained the coreRNA polymerase binding region and promoter recogni-tion site (47). The residue corresponding to D118 of RpoSis conserved in RpoD (23). A mutation in this residueof RpoD causes a severe defect in core RNA polymerasebinding; therefore, activity of the mutant RpoD to initiatethe transcription is severely lowered (48). From crystalstructure and mutational analyses of RpoD, there is a pos-
sibility that RpoS D118N may be unable to successfullyinteract with core RNA polymerase and thus may fail toinitiate the transcription of RpoS-dependent genes.
As we described above, the substitution D118N seemsto affect reduction rate of RpoS during VBNC inductionor RpoS function to induce expression of RpoS regulon. Asa result, S. Oranienburg might rapidly enter VBNC state.To clarify this point, further experiments are needed.
In this study, under incubation in 7% NaCl, Salmonellalost culturability (107-fold decrease in CFU) but retainedhigh viability (>70%) (Fig. 3), suggesting that cell divisionmight slow down severely or be blocked in VBNC state.Although RpoS-regulating genes during VBNC inductionis not understood yet, it is of interest that ftsQAZ genes,whose products are essential for cell division, are underregulation of RpoS (49). Further studies about RpoS reg-ulon during induction of VBNC are required to reveal themolecular mechanism of loss of culturability in VBNCstate.
The detection of a bacterial contamination has beentraditionally determined by its ability to grow and makecolonies on agar plates. The outbreaks of salmonellosishave frequently been reported as Salmonella was isolatedfrom patients. But it is often the case that the bacteriacould not be isolated from suspected food or environ-mental samples. There are many reasons for failure toidentify the source of contamination, however it could bepartly explained by the probable contamination of VBNCbacteria. S. Oranienburg strain Sa99004 used in this studywas the clinical isolate of food-borne outbreak caused byingestion of dried, salted squid, and can shift smoothlyto non-culturable state under high osmotic environment.In food processing, the bacteria that face various kindsof environmental stresses including high osmolarity mayenter a VBNC state, making their detection on culturemedia difficult during quality control tests in food in-dustries. Understanding of the molecular basis of VBNCstate would provide insight into the physiology of bacte-rial stress responses and VBNC state, and may give a clueto develop novel detection method for VBNC bacteria.
ACKNOWLEDGMENTS
We express thanks to Professor Takayuki Ezaki (Gifu Uni-versity) for kindly providing Salmonella Typhimuriumstrain LT2, and to Professor Hirofumi Aiba (Nagoya Uni-versity) for the pKD46 and pKD4 plasmids. We also thankthe National BioResource Project (NIG, Japan) for pro-viding the pBAD18 plasmid. This work was supported inpart by grants-in-aid for scientific research from JapanSociety For the Promotion of Science (JSPS) (Project No.21880005, to A. K.) and the Ministry of Education, Sci-ence, and Culture of Japan (MEXT) (to K. K.).
1. Oliver J.D. (2005) The viable but nonculturable state in bacteria. JMicrobiol 43 Spec No: 93–100.
2. Rahman I., Shahamat M., Kirchman P.A., Russek-Cohen E., ColwellR.R. (1994) Methionine uptake and cytopathogenicity of viable butnonculturable Shigella dysenteriae type 1. Appl Environ Microbiol60: 3573–8.
3. Lleo M.M., Tafi M.C., Canepari P. (1998) NonculturableEnterococcus faecalis cells are metabolically active and capable ofresuming active growth. Syst Appl Microbiol 21: 333–9.
4. Lloyd D., Hayes A.J. (1995) Vigor, Vitality and Viability ofMicroorganisms. FEMS Microbiol Lett , 133: 1–7.
5. Lleo M.M., Pierobon S., Tafi M.C., Signoretto C., Canepari P.(2000) mRNA detection by reverse transcription-PCR formonitoring viability over time in an Enterococcus faecalis viable butnonculturable population maintained in a laboratory microcosm.Appl Environ Microbiol 66: 4564–7.
6. Heim S., Lleo M.M., Bonato B., Guzman C.A., Canepari P. (2002)The viable but nonculturable state and starvation are differentstress responses of Enterococcus faecalis, as determined by proteomeanalysis. J Bacteriol 184: 6739–45.
7. Oliver J.D., Bockian R. (1995) In vivo resuscitation, and virulencetowards mice, of viable but nonculturable cells of Vibrio vulnificus.Appl Environ Microbiol 61: 2620–3.
8. Wai S.N., Mizunoe Y., Takade A., Yoshida S. (2000) A comparisonof solid and liquid media for resuscitation of starvation- andlow-temperature-induced nonculturable cells of Aeromonashydrophila. Arch Microbiol 173: 307–10.
9. Whitesides M.D., Oliver J.D. (1997) Resuscitation of Vibriovulnificus from the Viable but Nonculturable State. Appl EnvironMicrobiol 63: 1002–5.
10. Lleo M.M., Bonato B., Tafi M.C., Signoretto C., Boaretti M.,Canepari P. (2001) Resuscitation rate in different enterococcalspecies in the viable but non-culturable state. J Appl Microbiol 91:1095–102.
11. Chiang S.M., Schellhorn H.E. Evolution of the RpoS regulon: originof RpoS and the conservation of RpoS-dependent regulation inbacteria. J Mol Evol 70: 557–71.
12. Hengge-Aronis R. (2000) The general stress response in Escherichiacoli. In:G. Storz A.R.H.-A., ed. Bacterial Stress Responses.Washington, DC: ASM press.
13. Ivanova A., Renshaw M., Guntaka R.V., Eisenstark A. (1992) DNAbase sequence variability in katF (putative sigma factor) gene ofEscherichia coli. Nucleic Acids Res 20: 5479–80.
14. Tanaka K., Takayanagi Y., Fujita N., Ishihama A., Takahashi H.(1993) Heterogeneity of the principal sigma factor in Escherichiacoli: the rpoS gene product, sigma 38, is a second principal sigmafactor of RNA polymerase in stationary-phase Escherichia coli. ProcNatl Acad Sci U S A 90: 3511–5.
15. Ishihama A. (2000) Functional modulation of Escherichia coli RNApolymerase. Annu Rev Microbiol 54: 499–518.
16. Notley L., Ferenci T. (1996) Induction of RpoS-dependentfunctions in glucose-limited continuous culture: what level ofnutrient limitation induces the stationary phase of Escherichia coli?J Bacteriol 178: 1465–8.
17. Cheung K.J., Badarinarayana V., Selinger D.W., Janse D., ChurchG.M. (2003) A microarray-based antibiotic screen identifies a
regulatory role for supercoiling in the osmotic stress response ofEscherichia coli. Genome Res 13: 206–15.
18. Robey M., Benito A., Hutson R.H., Pascual C., Park S.F., MackeyB.M. (2001) Variation in resistance to high hydrostatic pressure andrpoS heterogeneity in natural isolates of Escherichia coli O157:H7.Appl Environ Microbiol 67: 4901–7.
19. Schellhorn H.E., Hassan H.M. (1988) Response of hydroperoxidaseand superoxide dismutase deficient mutants of Escherichia coli K-12to oxidative stress. Can J Microbiol 34: 1171–6.
20. Khil P.P., Camerini-Otero R.D. (2002) Over 1000 genes are involvedin the DNA damage response of Escherichia coli. Mol Microbiol 44:89–105.
21. Lin J., Smith M.P., Chapin K.C., Baik H.S., Bennett G.N., FosterJ.W. (1996) Mechanisms of acid resistance in enterohemorrhagicEscherichia coli. Appl Environ Microbiol 62: 3094–100.
22. Hengge R. (2009) Proteolysis of sigmaS (RpoS) and the generalstress response in Escherichia coli. Res Microbiol 160: 667–76.
23. Lonetto M., Gribskov M., Gross C.A. (1992) The sigma 70 family:sequence conservation and evolutionary relationships. J Bacteriol174: 3843–9.
28. Gruber T.M., Markov D., Sharp M.M., Young B.A., Lu C.Z., ZhongH.J., Artsimovitch I., Geszvain K.M., Arthur T.M., Burgess R.R.,Landick R., Severinov K., Gross C.A. (2001) Binding of theinitiation factor sigma(70) to core RNA polymerase is a multistepprocess. Mol Cell 8: 21–31.
29. Nagai H., Shimamoto N. (1997) Regions of the Escherichia coliprimary sigma factor sigma70 that are involved in interaction withRNA polymerase core enzyme. Genes Cells 2: 725–34.
30. Bowers C.W., Dombroski A.J. (1999) A mutation in region 1.1 ofsigma70 affects promoter DNA binding by Escherichia coli RNApolymerase holoenzyme. EMBO J 18: 709–16.
31. Bowers C.W., Mccracken A., Dombroski A.J. (2000) Effectsof amino acid substitutions at conserved and acidic residueswithin region 1.1 of Escherichia coli sigma(70). J Bacteriol 182:221–4.
32. Gopal V., Chatterji D. (1997) Mutations in the 1.1 subdomain ofEscherichia coli sigma factor sigma70 and disruption of its overallstructure. Eur J Biochem 244: 613–8.
33. Vuthoori S., Bowers C.W., Mccracken A., Dombroski A.J.,Hinton D.M. (2001) Domain 1.1 of the sigma(70) subunit ofEscherichia coli RNA polymerase modulates the formation of stablepolymerase/promoter complexes. J Mol Biol 309: 561–72.
34. Wilson C., Dombroski A.J. (1997) Region 1 of sigma70 is requiredfor efficient isomerization and initiation of transcription byEscherichia coli RNA polymerase. J Mol Biol 267: 60–74.
35. Panutdaporn N., Kawamoto K., Asakura H., Makino S.I. (2006)Resuscitation of the viable but non-culturable state of Salmonellaenterica serovar Oranienburg by recombinant
resuscitation-promoting factor derived from SalmonellaTyphimurium strain LT2. Int J Food Microbiol 106: 241–7.
36. Datsenko K.A., Wanner B.L. (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR products.Proc Natl Acad Sci U S A 97: 6640–5.
38. Mcclelland M., Sanderson K.E., Spieth J., Clifton S.W., Latreille P.,Courtney L., Porwollik S., Ali J., Dante M., Du F., Hou S., LaymanD., Leonard S., Nguyen C., Scott K., Holmes A., Grewal N.,Mulvaney E., Ryan E., Sun H., Florea L., Miller W., Stoneking T.,Nhan M., Waterston R., Wilson R.K. (2001) Complete genomesequence of Salmonella enterica serovar Typhimurium LT2. Nature413: 852–6.
39. Oliver J.D. (2000) The public health significance of viable butnonculturable bacteria. In:Colwell R.R., Grimes D.J., eds.Nonculturable Microorganisms in the Environment . Washington, D.C.: ASM Press.
40. Boaretti M., Lleo M.M., Bonato B., Signoretto C., Canepari P.(2003) Involvement of rpoS in the survival of Escherichia coli in theviable but non-culturable state. Environ Microbiol 5: 986–96.
41. Munro P.M., Flatau G.N., Clement R.L., Gauthier M.J. (1995)Influence of the RpoS (KatF) sigma factor on maintenance ofviability and culturability of Escherichia coli and Salmonellatyphimurium in seawater. Appl Environ Microbiol 61: 1853–8.
42. Asakura H., Makino S., Takagi T., Kuri A., Kurazono T., Watarai M.,Shirahata T. (2002) Passage in mice causes a change in the ability ofSalmonella enterica serovar Oranienburg to survive NaCl osmoticstress: resuscitation from the viable but non-culturable state. FEMSMicrobiol Lett 212: 87–93.
43. Asakura H., Kawamoto K., Shirahata T., Makino S. (2004) Changesin Salmonella enterica serovar Oranienburg viability caused by
NaCl-induced osmotic stress is related to DNA relaxationby the H-NS protein during host infection. Microb Pathog 36:147–51.
44. Brescia C.C., Kaw M.K., Sledjeski D.D. (2004) The DNA bindingprotein H-NS binds to and alters the stability of RNA in vitro and invivo. J Mol Biol 339: 505–14.
45. Yamashino T., Ueguchi C., Mizuno T. (1995) Quantitative controlof the stationary phase-specific sigma factor, sigma S, in Escherichiacoli: involvement of the nucleoid protein H-NS. EMBO J 14:594–602.
46. Stenstrom C.M., Holmgren E., Isaksson L.A. (2001) Cooperativeeffects by the initiation codon and its flanking regions ontranslation initiation. Gene 273: 259–65.
47. Malhotra A., Severinova E., Darst S.A. (1996) Crystal structure of asigma 70 subunit fragment from E. coli RNA polymerase. Cell 87:127–36.
48. Sharp M.M., Chan C.L., Lu C.Z., Marr M.T., Nechaev S., MerrittE.W., Severinov K., Roberts J.W., Gross C.A. (1999) The interface ofsigma with core RNA polymerase is extensive, conserved, andfunctionally specialized. Genes Dev 13: 3015–26.
49. Ballesteros M., Kusano S., Ishihama A., Vicente M. (1998) TheftsQ1p gearbox promoter of Escherichia coli is a major sigmaS-dependent promoter in the ddlB-ftsA region. Mol Microbiol 30:419–30.
50. Yanisch-Perron C., Vieira J., Messing J. (1985) Improved M13phage cloning vectors and host strains: nucleotide sequences of theM13mp18 and pUC19 vectors. Gene 33: 103–19.
52. Guzman L.M., Belin D., Carson M.J., Beckwith J. (1995) Tightregulation, modulation, and high-level expression by vectorscontaining the arabinose PBAD promoter. J Bacteriol 177:4121–30.
