The RNA chaperone Hfq regulates expression of fimbrial-related genes and virulence of Salmonella enterica serovar Enteritidis

Authors


Correspondence: Guoqiang Zhu, College of Veterinary Medicine, Yangzhou University, 12 East Wenhui Road, Yangzhou 225009, China. Tel.: +86 514 8797 2590; fax: +86 514 8731 1374; e-mails: yzgqzhu@yzu.edu.cn; yzgqzhu@hotmail.com

Abstract

Salmonella Enteritidis is an intracellular pathogen that causes enteritis and systemic disease in humans and other animals. The RNA chaperone protein Hfq mediates the binding of small noncoding RNAs to target mRNA and assists in post-transcriptional gene regulation in bacteria. In this study, we constructed an hfq deletion mutant in S. Enteritidis SE50336 and analyzed the expression of major fimbrial subunits sefA, bcfA, fimA, safA, stbA, sthA, csgA, csgD, and pegA using quantitative real-time PCR. The gene expression of sefA increased about 14-fold in the hfq mutant, as compared with its expression in the wild-type strain. The expression of fimA and pegA did not change significantly, while the expression of the other fimbrial genes was significantly down-regulated in the hfq mutant. The ability of SE50336Δhfq adhering to Caco-2 cells was also reduced as compared with wild-type adherence. The virulence of the hfq mutant was significantly reduced in a 1-day-old chicken model of S. Enteritidis disease, as determined by quantifying the lethal dose 50% of the bacterial strains. We conclude that Hfq critically contributes to S. Enteritidis virulence, likely partially affected by regulating fimbrial gene expression.

Introduction

Salmonella enterica serovar Enteritidis is a Gram-negative facultative intracellular pathogen with a broad range of hosts (Saeed et al., 1999). Salmonella Enteritidis infects young chickens and produces a spectrum of symptoms from enteritis to systemic infection and death (Jeurissen & Janse, 1996). Although S. Enteritidis rarely causes clinical disease in adult chickens, latent infections and chronic carriers contaminate poultry products, resulting in foodborne infections in humans (Centers for Disease Control & Prevention, 2000; De Buck et al., 2004a). Salmonella Enteritidis enters the host by invading and colonizing the gastrointestinal tract mucosa, using fimbrial proteins to adhere to host tissues (Hohmann et al., 1978; Rajashekara et al., 2000; Clayton et al., 2008). Thirteen fimbrial loci have been defined based on the complete genome sequence of S. Enteritidis strain P125109 (Clayton et al., 2008). Although the contribution of all of these fimbriae to adhesion or virulence has not yet been established, it is clear that the type I fimbriae and curli contribute to reproductive tract infections in laying hens and in egg contamination (Cogan et al., 2004; De Buck et al., 2004b). SEF14 fimbriae are only expressed in S. Enteritidis and in closely related serovars (Turcotte & Woodward, 1993). SEF14 fimbriae are essential for efficient uptake and survival of S. Enteritidis in macrophages in intraperitoneal infections (Edwards et al., 2000).

The Hfq protein is a conserved RNA-binding protein, which forms homohexamers of c. 12-kDa subunits (Franze de Fernandez et al., 1972). Hfq was first identified as a host factor (HF-1) for bacteriophage Qβ RNA replication and subsequently shown to be a global post-transcriptional regulator (Franze de Fernandez et al., 1968; Chao & Vogel, 2010). Hfq maintains small noncoding RNAs (sRNAs) stability and/or interacts with the target mRNAs by binding AU-rich sequences and facilitating base pairing between sRNAs and mRNAs (Møller et al., 2002; Valentin-Hansen et al., 2004). The role of Hfq in bacterial physiology becomes apparent through the pleiotropic phenotypes of hfq mutants in many eubacterial species such as altered growth rates, sensitivity to environmental stress, virulence attenuation, and altered drug resistance (Tsui et al., 1994; Chao & Vogel, 2010; Hayashi-Nishino et al., 2012). For example, Hfq deficiency decreased growth rates and affected both osmosensitivity and the oxidation of carbon sources in Escherichia coli K12 (Tsui et al., 1994). A hfq mutant of Azorhizobium caulinodans exhibits a defect in nitrogen fixation (Kaminski et al., 1994). In recent years, deletion of Hfq was also found to impact virulence in many bacterial pathogens, including Salmonella Typhimurium, Vibrio cholerae, and Pseudomonas aeruginosa (Sonnleitner et al., 2003; Ding et al., 2004; Sittka et al., 2007). A S. Typhimurium hfq deletion was highly attenuated in mice and showed a severe defects in invading epithelial cells, secreting virulence factors, and growing within epithelial cells and macrophages (Sittka et al., 2007). A V. cholerae strain lacking hfq failed to colonize the suckling mouse intestine (Ding et al., 2004). The hfq mutation also reduced the virulence of P. aeruginosa by impairing twitching and swarming behaviors mediated by type IV pili (Sonnleitner et al., 2003).

Although the role of Hfq in virulence has been studied in many pathogens, its role in regulating fimbrial gene expression is not well understood. In this study, we constructed an S. Enteritidis hfq deletion mutant and analyzed fimbrial gene expression and virulence in a chicken model of disease.

Materials and methods

Bacterial strains, plasmids, and cell line culture conditions

The bacteria and plasmids used in this study are listed in Table 1. Bacteria were grown in Luria–Bertani broth (LB) or on LB plates at 37°C, except for strains containing the temperature-sensitive plasmids pCP20 or pKD46, which were grown at 30 °C. Bacteria harboring antibiotic resistance genes were cultured in LB containing 100 μg mL−1 of ampicillin (Amp) and 34 μg mL−1 of chloramphenicol (Cm) when appropriate. To determine the growth rates, bacteria were grown at 37 °C with agitation (180 r.p.m./min) in LB broth and the OD600 nm was measured once every hour. Human colorectal adenocarcinoma epithelial cells (Caco-2) were cultivated in Dulbecco's minimal Eagle's medium (DMEM) containing glutamine (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (Gibco). Cells were maintained in an atmosphere of 5% CO2 at 37 °C.

Table 1. Bacterial strains and plasmids used in this study
Strains/PlasmidsCharacteristicsReferences
  1. pBR322 carrying the entire.

Strains
CMCC(B)50336Salmonella enterica serovar Enteritidis wild typeNICPBP, China
50336△hfqhfq-deficient mutantThis study
50336△hfq/phfq50336△hfq carrying pBR-hfq (Ampr)This study
Plasmids
pKD3Cmr; Cm cassette templateDatsenko & Wanner (2000)
pKD46Ampr, λRed recombinase expressionDatsenko & Wanner (2000)
pCP20Ampr, Cmr; Flp recombinase expressionDatsenko & Wanner (2000)
pBR-hfqpBR322 carrying the entire hfq gene (Ampr)This study
pGEM-T EasyCloning vector, AmprTakara
pMD19 T-simpleCloning vector, AmprTakara

Construction of hfq deletion mutant and the complemented strain

The hfq gene of S. Enteritidis strain 50336 was PCR amplified by designing primers that flank the hfq gene in S. Enteritidis strain P125109. The hfq deletion mutant was constructed by λ-Red-mediated recombination system, as described previously (Datsenko & Wanner, 2000; Duan et al., 2013). The primers used are listed in Table 2. Briefly, primers hfq-F and hfq-R were designed to amplify the chloramphenicol cassette from plasmid pKD3, including 48-bp homology extensions from the 5′ and 3′ of the hfq gene. The PCR products were purified and introduced into plasmid pKD46-containing S. Enteritidis 50336 by electroporation. Recombinant bacteria 50336Δhfq::cat was screened and selected on both Cm and Amp resistance LB agar plates. Allelic replacement of hfq by the Cm cassette was verified by PCR screening using primers (vhfq-F and vhfq-R) and DNA sequencing. Then, Cm cassette gene of 50336Δhfq::cat was excised by introducing the Flp recombinase-expressing vector pCP20. Then, the hfq complete deletion mutant was obtained, confirmed by PCR and DNA sequencing, and used for later function analysis. The full-length hfq gene was cloned by primers pBR-hfq-F and pBR-hfq-F, and ligated to plasmid pBR322. The recombined plasmid pBR322-hfq was transferred to the hfq mutant and then obtained complemented strain.

Table 2. Primers used in this study
PrimerSequence (5′-3′)Product size (bp)
hfq-FTTCAGAATCGAAAGGTTCAAAGTACAAATAAGCATATAAGGAAAAGAGTGTGTAGGCTGGAGCTGCTTCG1114
hfq-RCGACGCCCCCGACATGGATAAACAGCGCGTGAACTTATTCAGTCTCTTCCATATGAATATCCTCCTTAG
vhfq-FAAGGCATTGCTGGAAGGGT1097/837
vhfq-RGGCATCATAACGGTCAAACAA
pBR-hfq-FCGCGGATCCATGGCTAAGGGGCAATCTT327
pBR-hfq-RACATGCATGCTTATTCAGTCTCTTCGCTG
bcfA-FTGACGCTGCCTGTTCTGTTT136
bcfA-RGCAGTCTTCCAGTTTGATGGTG
fimA-FGACTGCGATCCGAAAGTGG91
fimA-RCAGAGGAGACAGCCAGCAAA
sefA-FTGCTGCTGGTCAGAAAGTTG167
sefA-RTATTGGCTCCCTGAATACGC
safA-FGGTTGCTAACACGACACTGG152
safA-RCAAAGGTGAACCAGCTCCTC
stbA-FATCACAGGCTCGCTTCTTGT211
stbA-RCGGACAGCCCAATATCAAAC
sthA-FAAGAAATCCGTGTTGGTCGT165
sthA-RGGCAACCGATAACCATGACT
csgA-FATGCCCGTAAATCTGAAACG172
csgA-RCCGTATTGGCCGACAGTAAT
pegA-FGAAGTGGTGGGAACATCCTG249
pegA-RGAAGCCCGCACCATTATTAG
csgD-FGCCTCATATTAACGGCGTGT157
csgD-RGGACTCGGTGCTGTTGTAGC
gyrAFGCATGACTTCGTCAGAACCA278
gyrARGGTCTATCAGTTGCCGGAAG

Adhesion assay

The adhesion assay was performed as previously described (Jouve et al., 1997). In brief, bacteria were grown to an OD600 nm of 2.0 in LB broth at 37 °C. Caco-2 cells were seeded in 96-well tissue culture plates at a concentration of 1 × 105 cells per well and grown for 20 h. Cell monolayers were washed three times with sterile phosphate-buffered saline (PBS, pH 7.2). Then, the Caco-2 cells were incubated with 100 μL of bacteria suspension in DMEM medium for 2 h with a multiplicity of infection (moi) of 100 at 37 °C in each well of a 96-well plate. Infections were carried out in triplicate. After 2 h of incubation, the infected cell monolayer was gently washed three times with PBS to remove loosely adherent bacteria. Cells were lysed with 0.5% Triton X-100 for 30 min. The lysates were serially diluted and plated on LB agar plates for bacterial enumeration.

Macrophage survival assay

Infection of activated mouse peritoneal macrophages was performed as described previously (Ehrt et al., 1997). The peritoneal macrophages were harvested from 6- to 8-week-old female Balb/c mice (Comparative Medicine Center, Yangzhou University, Jiangsu, China) according to the method by Ehrt et al. (1997). About 3 × 104 activated mouse peritoneal macrophages were inoculated to each well in a 96-well plate and incubated for 24–48 h for survival assay. Bacteria were grown to an OD600 nm of 2.0 in LB broth at 37 °C and then harvested for infection. The peritoneal macrophages were infected with a moi of 100, and infections were carried out in triplicates. After 30-min incubation, the cells were washed three times and incubated with 150 μL RPMI 1640 medium (containing 50 μg mL−1 gentamicin) to kill noninvasive bacteria. After 2 or 6 h, the infected cells were washed three times and lysed in PBS containing 1% Triton X-100 for 10 min. The cells were lysed to release the intracellular bacteria, and the lysates were diluted and plated onto LB agar plates for bacterial enumeration.

RNA isolation and quantitative real-time PCR

Wild-type Salmonella strain 50336, 50336Δhfq, and 50336Δhfq/phfq were grown to an OD600 nm of 2.0 and collected by centrifugation. Total RNA was extracted using TRIzol reagent (Invitrogen). cDNA was synthesized using the PrimeScript RT reagent Kit with gDNA Eraser (Takara Bio, Shiga, Japan) for reverse transcription PCR. Transcript abundance was quantified using SYBR Premix Ex Taq II (Takara Bio) in an ABI7500 instrument (Applied Biosystems). Assays were performed in triplicates, and all data were normalized to the endogenous reference gene gyrA. Melting curve analysis demonstrated that the accumulation of SYBR green-bound DNA was gene-specific and not due to primer dimers. The 2−ΔΔCT method was used for data quantification (Livak & Schmittgen, 2001).

Animal infections

Bacteria were grown in LB to early stationary phase, harvested by centrifugation, and diluted in PBS prior to inoculation into chickens. Groups (n = 19) of 1-day-old chickens (National Chickens Genetic Resources, Yangzhou, China) were inoculated by subcutaneous injection. Control groups received 200 μL PBS. Chickens were monitored daily for signs of illness and death. The lethal dose 50% (LD50) was calculated 14 days postinfection using the method described previously (van der Velden et al., 1998). All procedures were approved by the Beijing Animal Welfare Committee.

Statistical analysis

Data were analyzed by Student's t-test for independent samples. Differences were considered significant if  0.05.

Results and discussion

Construction of the mutant 50336Δhfq and complemented strain 50336Δhfq/phfq

In this study, we investigated the virulence of an S. Enteritidis hfq mutant and the regulation of fimbrial gene expression by Hfq. Salmonella Enteritidis strain 50336 contains an hfq gene (309 bp) with 100% and 99.35% identity to the S. Enteritidis P125109 and S. Typhimurium strain LT2 hfq genes, respectively (McClelland et al., 2001). We constructed an hfq mutant in S. Enteritidis 50336 using λ Red recombination (Supporting information, Fig. S1a; 50336Δhfq) and also complemented this mutant by expressing Hfq from pBR322 (Fig. S1b; 50336Δhfq/phfq). The growth characteristics of the hfq mutant and control strains were determined in LB liquid medium. The mutant 50336Δhfq exhibited slightly slower growth in the log phase and reached stationary phase at a lower optical density as compared with the wild-type strain, although the difference was not significant (Fig. 1). The growth of hfq mutant in other organisms such as S. Typhimurium SL1344 (Sittka et al., 2007), S. Enteritidis strain 147 (Karasova et al., 2009), V. cholerae (Ding et al., 2004), and Yersinia pestis (Geng et al., 2009) were also not significantly different from that in the wild-type strain. It indicated that deletion of Hfq caused slight change in growth in many eubacteria.

Figure 1.

Growth curves of wild-type Salmonella Enteritidis 50336, the mutant 50336Δhfq and complemented strain 50336Δhfq/phfq. Bacteria were grown in liquid LB medium at 37 °C for 12 h with agitation, and the OD600 nm values of triplicate cultures in LB medium were determined in 1-h intervals. Data are the means of three independent experiments.

The hfq mutation influences adherence to Caco-2 cells

Successful adhesion to epithelial cells is important for pathogenesis. To study whether deleting hfq affected bacterial adhesion, we performed bacterial binding to Caco-2 cells. Salmonella Enteritidis 50336Δhfq was significantly attenuated in adhering to Caco-2 cells, as compared with the wild-type strain (Fig. 2, P = 0.01). Complementing the deletion strain with hfq expressed from pBR322 partially restored bacterial adherence.

Figure 2.

Adherence to Caco-2 cells by wild-type Salmonella Enteritidis 50336, the mutant 50336Δhfq, and complemented strain 50336Δhfq/phfq. The Caco-2 cells were incubated with bacteria suspension for 2 h with a moi of 100 at 37 °C. The bacteria number adhered to Caco-2 cells were enumerated by dilution plate count. Data are expressed as mean ± standard deviation of triplicate experiments. *Indicates statistically significant difference compared with the wild-type strain (P < 0.05).

The hfq mutation impairs survival in macrophages

Invasion and survival of S. Enteritidis into the host cells, especially macrophages, are an important prerequisite for successful pathogenesis in the host. We also assayed for S. Enteritidis survival in mice peritoneal macrophages in vitro. At 2 and 6 h postinfection, we observed 7- and 12-fold reductions, respectively, in the number of intracellular 50336Δhfq as compared with the number of WT bacteria (Fig. 3, P < 0.05). The complemented strain 50336Δhfq/phfq restored survival ability partially.

Figure 3.

Survival and replication of wild-type Salmonella Enteritidis 50336, the mutant 50336Δhfq, and complemented strain 50336Δhfq/phfq in the activated mice peritoneal macrophages. After infecting 2 or 6 h, the cells were washed and lysed to release the intracellular bacteria, and the lysates were diluted and plated onto LB agar plates for bacterial enumeration. Data are expressed as mean ± standard deviation of triplicate experiments. *Statistically significant difference compared with the wild-type strain (P < 0.05).

Hfq regulates fimbrial-related genes expression

Hfq regulates the expression of almost a fifth of all Salmonella genes, including several pathogenicity islands (SPI-1, -2, -4, -5) and several clusters of fimbrial genes (Sittka et al., 2008). Sitta's research revealed that about 20% in 24 fimbrial-related genes such as fimA, fimC, fimD, and bcfA were down-regulated at SPI-induced conditions (Sittka et al., 2008). The major biofilm regulator csgD expression was also down-regulated in an hfq mutant (Monteiro, et al., 2012). We hypothesized that the reduced adhesion and intracellular survival of 50336Δhfq were due to changes in the expression of fimbrial genes regulated by Hfq. We quantified the expression of several fimbrial subunits and determined that sefA expression was up-regulated about 14-fold in the hfq mutant as compared with WT (Fig. 4). While the expression of fimA and pegA were almost unchanged, the expression of all of the other fimbrial-related genes we assayed was significantly down-regulated in the hfq mutant (Fig. 4).

Figure 4.

Fold changes of major fimbrial subunit genes mRNA level were determined in the mutant 50336Δhfq and complemented strain 50336Δhfq/phfq by quantitative RT-PCR compared with the wild-type strain. Assays were performed in triplicate. The 2−ΔΔCT method was used for data quantification.

SefA is the major fimbrial subunit of SEF14 fimbriae, which are expressed only by S. Enteritidis and its most closely related serovars. But the deletion of sefA did not affect adhesion to Caco-2 cells (Fig. S2). It was supposed that a dramatic up-regulation of sefA gene expression did not affect adhesion ability of 50336Δhfq. In S. Typhimurium, the major fimbrial subunits as bcfA, fimA, safA, stbA, sthA, and csgA were related to virulence and essential for colonization in the murine gut (van der Velden et al., 1998; Edwards et al., 2000; Weening et al., 2005), whereas the csgD genes was implicated in colonization of the avian gut (Morgan et al., 2004). In S. Enteritidis strain P125109, single deletion of stbA, sthA, pegA, or bcfA affected colonization in the cecum of chickens in a certain degree (Clayton et al., 2008). Thus, fimbriae play important roles in pathogenicity in Salmonella. In this study, the expression level of bcfA, safA, stbA, sthA, and csgA declined evidently, especially csgD decreased about 20-fold in 50336Δhfq compared with the wild-type strain. It was supposed that decreased expression of bcfA, safA, stbA, sthA, csgA, and csgD induced by hfq deletion affected fimbriae formation, and further affected adhesion and invasion in vitro.

The hfq mutation attenuates virulence in chickens

An LD50 assay was conducted to compare the effect of deleting hfq on S. Enteritidis virulence in chickens. After infecting 1-day-old chicks with 109 CFU of each of the three strains, we observed that 68% (13/19) to 84% (16/19) of chickens displayed intestinal hyperemia and diarrhea 10 h postinfection. When infected with 108 CFU, the chickens infected with 50336Δhfq showed lower mortality, as compared with wild-type S. Enteritidis 50336 and the complemented strain 50336Δhfq/phfq. The LD50s were calculated 14 days postinfection. The results showed that the LD50 of the WT strain 50336, 50336Δhfq, and 50336Δhfq/phfq were 9.472 × 108, 2.442 × 109, and 2.974 × 108 CFU, respectively. It indicated that the mutant 50336Δhfq was attenuated 2.6-fold as compared with WT (Table S1). The complementation strain 50336Δhfq/phfq recovered virulence completely. All three strains of S. Enteritidis were recovered from the liver, spleen, and cecum of the infected chickens. As a global regulator, Hfq regulates several pathogenicity islands (SPI-1, -2, -4, -5) genes and fimbrial genes of S. Typhimurium (Sittka et al., 2008). It was supposed that the virulence decrease of S. Enteritidis hfq mutant on Salmonella-infected chickens caused by many pathogenicity-related factors. The changes of fimbrial genes expression might be one of the reasons for virulence decline.

In summary, we have constructed an S. Enteritidis hfq mutant, reported virulence of the strain in vitro and in vivo and regulation function of Hfq in fimbrial-related genes expression. The deletion of hfq significantly reduced adhesion ability to Caco-2 cells and survival in mice peritoneal macrophages, meanwhile, reduced virulence of S. Enteritidis in chickens. The qRT-PCR results demonstrated that the expression of several fimbrial genes such as bcfA, safA, stbA, sthA, csgA, and csgD declined evidently. Thus, Hfq is an important virulence-regulated factor in S. Enteritidis 50336.

Acknowledgements

This study was supported by grants from the Chinese National Science Foundation (Nos. 31101826, 31270171, and 31101833), National Science and Technology Support Plan Grant (No. 2012BAK17B10), China 863 High Technology Plan Grant (No. SS2012AA101001), Program for Changjiang Scholars and Innovative Research Team in University (No. IRT0978). A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD) and National Science Foundation Grants of Jiangsu Province (Nos. BK2012265 and BK2011430). The authors or their institution do not have any relationships that may influence or bias the results and data presented in this manuscript.

Authors’ contribution

X.M. and X.M. contributed equally to this study.

Ancillary