Reduced invasion to human epithelial cell lines of Salmonella enterica serovar Typhi carrying S. Typhimurium sopD2


  • Present address: Annette N. Trombert, Centro de Genomica y Bioinformatica, Instituto de Biotecnologia, Universidad Mayor, Santiago, Chile.

  • Editor: Ian Henderson

Correspondence: Guido C. Mora, Laboratorio de Microbiología, Facultad de Medicina, Universidad Andrés Bello, Avda. República 217, Santiago, Chile. Tel.: +56 02 661 8373; fax: +56 02 661 8069; e-mail:


Salmonella enterica serovar Typhi and Typhimurium are closely related serovars. However, S. Typhi, a human-specific pathogen, has 5% of genes as pseudogenes, far more than S. Typhimurium, which only has 1%. One of these pseudogenes corresponds to sopD2, which in S. Typhimurium encodes an effector protein involved in Salmonella-containing vacuole biogenesis in human epithelial cell lines, which is needed for full virulence of the pathogen. We investigated whether S. Typhi trans-complemented with the functional sopD2 gene from S. Typhimurium (sopD2STM) would reduce the invasion of human epithelial cell lines. Our results showed that the presence of sopD2STM in S. Typhi significantly modified the bacterial ability to alter cellular permeability and decrease the CFUs recovered after cell invasion of human epithelial cell line. These results add to mounting evidence that pseudogenes contribute to S. Typhi adaptation to humans.


Salmonella enterica serovar Typhi is a strictly human-specific pathogen causing the systemic disease typhoid fever (Parry et al., 2002). In contrast, Salmonella enterica serovar Typhimurium is a pathogen with a broad host range that causes gastroenteritis and septicemia, including enteric fever in mice (Tsolis et al., 1999; Parry et al., 2002; Zhang et al., 2003). Although these are closely related serovars sharing over 96% of DNA sequence identity, S. Typhimurium does not cause enteric fever in humans (Parry et al., 2002). This suggests that genetic differences between the serovars are crucial for disease development. These differences could be produced during Salmonella evolution due to horizontal transfer mechanisms and/or loss of genetic information by deletion or pseudogenization (Andersson & Andersson, 1999; Moran & Plague, 2004). The process by which a microorganism becomes adapted to its host by the generation of pseudogenes is termed ‘reductive evolution.’ This process has been observed in several human pathogens such as Shigella flexneri, Mycobacterium leprae and S. Typhi (Arber, 2000; Dagan et al., 2006). Salmonella enterica serovar Typhi contains approximately 200 pseudogenes, several of them associated with processes related to pathogenicity. In this context, some Salmonella pathogenicity island-2 (SPI-2)-dependent effector proteins (sseJ, sopD2) are annotated as pseudogenes (Parkhill et al., 2001; McClelland et al., 2004). We recently reported that the trans-complementing S. Typhi sseJ pseudogene with the functional gene from S. Typhimurium decreases cytotoxicity in human-derived epithelial cell lines (HT-29) (Trombert et al., 2010). Thus, our results suggest that the loss of sseJ in S. Typhi contributes to the adaptation to the systemic infection in humans by increasing the bacterially induced cytotoxicity and decreasing the retention/proliferation within epithelial cells (Trombert et al., 2010).

Upon entry into host cells, S. Typhimurium resides in a membrane-bound compartment termed the Salmonella-containing vacuole (SCV) (Knodler & Steele-Mortimer, 2003). Effector proteins that are translocated by type III secretion system (T3SS) encoded and/or regulated by Salmonella pathogenicity island-1 (SPI-1) and SPI-2 are essential for the modulation of the SCV endocytic traffic (Brumell & Grinstein, 2004). One of these effectors corresponds to SifA, a protein required for the formation of lysosomal glycoprotein (lgp)-containing structures (Sifs) in epithelial cells, which emerge from the vacuole (Stein et al., 1996; Boucrot et al., 2003). SifA binds to SseJ, host proteins SKIP (SifA kinesin-interacting protein), and RhoA family GTPases to cooperatively regulate the dynamics of SCV membrane in infected cells (Ohlson et al., 2008; Dumont et al., 2010). In addition, SopD2 corresponds to an SPI-2-regulated protein that promotes Sif formation, which contributes to virulence in mice (Ruiz-Albert et al., 2002; Brown et al., 2006; Schroeder et al., 2010).

In S. Typhimurium, sopD2 encodes a protein sharing 42% identity with SopD, a known SPI-1-dependent effector that plays a major role in gastroenteritis in animal models of Salmonella infection (Jones et al., 1998). Mutation of sopD2 in S. Typhimurium led to a prolonged survival in infected mice compared with survival in mice infected with the otherwise isogenic wild-type strain. Furthermore, in a competition index assay, the sopD2 mutant was recovered at a significantly lower level compared with the wild type after the two strains coinfected the same mouse, indicating a significant role of this effector in Salmonella pathogenesis (Brumell et al., 2003).

Salmonella enterica serovar Typhi lacks several effector proteins that in S. Typhimurium are crucial for the pathogenicity of the serovar (Raffatellu et al., 2005), such as sopD2, which in S. Typhi is described as a pseudogene (Parkhill et al., 2001; McClelland et al., 2004). We suggest that sopD2 inactivation is involved in human host adaptation of S. Typhi. To evaluate this, in this study we examined the effect of trans-complementation of S. Typhi with sopD2 from S. Typhimurium (sopD2STM) and its effect on reducing invasion of the epithelial cell line.

Materials and methods

Bacterial strains, media and growth conditions

Salmonella enterica serovar Typhi and S. Typhimurium strains used in this study are described in Table 1. Strains were routinely grown in Luria–Bertani (LB) at 37 °C with vigorous shaking, or anaerobically grown by adding 500 μL of sterile mineral oil as a barrier to oxygen before invasion assays in cultured human cells. When required, the medium was supplemented with chloramphenicol (20 μg mL−1).

Table 1.   Bacterial strains and plasmids used in this study
 Relevant characteristicReferences or sources
 Serovar Typhimurium
  ATCC 14028sWild-type strain, virulentATCC
  LT2Wild-type strainS. Maloy
  NT060ATCC 14028s ΔsopD2∷FRTThis work
 Serovar Typhi
  STH2370Clinical strain, virulentHospital Dr Lucio Córdova
  STH001Clinical strain, virulentHospital Dr Lucio Córdova
  STH004Clinical strain, virulentHospital Dr Lucio Córdova
  STH005Clinical strain, virulentHospital Dr Lucio Córdova
  STH006Clinical strain, virulentHospital Dr Lucio Córdova
  STH007Clinical strain, virulentHospital Dr Lucio Córdova
  STH008Clinical strain, virulentHospital Dr Lucio Córdova
  STH009Clinical strain, virulentHospital Dr Lucio Córdova
 pCC1Monocopy vector, F plasmid derivedEpicentre
 pNT007pCC1 carrying S. Typhimurium sopD2 geneThis work

Bioinformatic analyses

Comparative sequence analyses were made with the complete genome sequences of S. Typhi strains CT18 (AL513382) and Ty2 (AE014613), and the serovar Typhimurium (AE006468.1). The sequences were analyzed using blast, alignment and phylogeny tools available at and vector nt suite v.8 software (Invitrogen).

PCR amplifications

PCR amplifications of S. Typhimurium 14028s sopD2 gene were performed using an Eppendorf thermal cycler and Taq DNA polymerase (Invitrogen). The reaction mixture contained 1 × PCR buffer, 1.5 mM MgCl2, each dNTP (200 mM), primers (1 mM), 100 ng of template DNA and 2 U of polymerase. Amplifications included 30 cycles (94 °C for 30 s, 58 °C for 1 min and 72 °C for 2 min 30 s), followed by a final extension step at 72 °C for 10 min. Template S. Typhi and S. Typhimurium chromosomal DNA was prepared as described previously (Santiviago et al., 2001). Primers sopD21 (GTGTGGCTGTTCCAGAATGTGCTG) and sopD22 (CCGTTGCTAAACTGCCGTTTGCTTA) were used to amplify a fragment of 1800 bp. For S. Typhimurium 14028s mutagenesis, primers sopD23W (ATGCCAGTTACGTTAAGTTTTGGTAATCGTCATAACTATGTGTAGGCTGGAGCTGCTTCG) and sopD24W (TATATAAGCATATTGCGACAACTCGACTTTTCACTTATACATATGAATATCCTCCTTAG) were used to amplify the aph- and cat-resistance cassette from pKD3 and pKD4, respectively (Datsenko & Wanner, 2000). Letters in italics highlight primer sequences that annealed with both resistance cassettes.

All primers were designed on the basis of the reported sequence of S. Typhimurium LT2 sopD2 (AE006468.1).

Cloning of sopD2STM

The sopD2 PCR product was cloned directly in the pCC1 vector according to the manufacturer's instructions (CopyControl PCR Cloning Kit, Epicentre) to yield the plasmid pNT007. The presence of the gene and its promoter region in the plasmid was confirmed by PCR amplification and restriction endonuclease analyses. The cloned PCR product was sequenced to ensure that it did not harbor any mutation (data not show).


sopD2 pseudogene sequencing was performed by Macrogen Corp. (Rockville, MD) using S. Typhi chromosomal DNA prepared as described (Santiviago et al., 2001) and pNT007 previously isolated using the Wizard miniprep kit (Promega).

Construction of mutant strains

To generate the chromosomal deletion of sopD2, a ‘one-step inactivation’ protocol was performed (Datsenko & Wanner, 2000). Following mutagenesis, aph- and cat-resistance cassettes were removed by FLP-mediated recombination.

Gentamicin protection assay

To measure bacterial invasion, the method described by Lissner et al. (1983) and modified by Contreras et al. (1997) was used. Briefly, HEp-2 monolayers were grown at 37 °C in a 5% CO2/95% air mixture in RPMIFS (RPMI medium supplemented with 10% fetal bovine serum pretreated for 30 min at 60 °C). Bacterial strains were grown anaerobically to mid-exponential phase and then harvested by centrifugation before infection of the confluent HEp-2 monolayers in 96-well microtiter plates at a multiplicity of infection of 100 : 1. After incubation for 1 h to allow bacterial entry into cells, monolayers were washed twice with phosphate-buffered saline (PBS), and 100 μL of RPMI containing gentamicin (200 μg mL−1) was added to each well. The plates were then incubated for 2 h to kill any remaining extracellular bacteria. For strains carrying vectors, the medium was supplemented with chloramphenicol throughout the assay. The medium was removed and cells were washed twice with PBS. The cells were then lysed with sodium deoxycholate (0.5% w/v in PBS). The number of intracellular bacteria (CFUs, counted 3-h postinfection, t3) was determined by plating onto LB agar plates with chloramphenicol. Quantitative invasion assay values were calculated as follows:


Cell permeability assay

We used an in vitro assay according to Trombert et al. (2010), modified from the method described by McCormick et al. (1993). Briefly, the colon carcinoma HT-29 cell line was grown to confluence (18–21 days) on 3.0-μm pore-size filters (or transwells, Millicell®, Millipore) with glucose-free RPMI (Gibco). Each transwell was inoculated individually to the apical surface with 400 μL of approximately 1 × 107 CFU mL−1 of bacterial cultures and immediately incubated for 60 min at 37 °C. After extensive washing with sterile PBS, the extracellular bacteria were killed by treatment of monolayers with gentamicin (50 μg mL−1). Immediately after gentamicin treatment, the medium from basal compartment of the epithelial cell monolayer was collected and plated for CFU to assess the number of bacteria that passed through the cell monolayer. The polarization of cells was confirmed by transepithelial electrical resistance and transmission electron microscopy (data not shown).


All results are expressed as means±SD of an individual experiment performed in triplicate. P-values were calculated according to Student's t-test, and P<0.05 or P<0.01 values were considered statistically significant.


Bioinformatic analyses of S. Typhi sopD2 pseudogene

To assess whether the sopD2 locus is a pseudogene in the serovar Typhi, we compared the available sequences of S. Typhi CT18 and S. Typhi Ty2 (McClelland et al., 2001; Parkhill et al., 2001; Deng et al., 2003). We observed that the sequence of sopD2 in S. Typhi contains a frameshift at codon 48 resulting in a premature stop codon that disrupts the expected ORF. Accordingly, the online databases report sopD2 as a pseudogene (Fig. 1). To confirm the presence of this frameshift in our S. Typhi clinical strain collection, PCR assays were carried out using SopD21 and SopD22 primer pairs. The primers yield an 1800-bp amplicon in both S. Typhi and S. Typhimurium and were used to test the clinical strains obtained from Chilean typhoid patients (STH collection, Hospital Dr Lucio Córdova, Chile). The PCR products were sequenced and in silico comparison was performed with the reference strains (S. Typhi CT18, S. Typhimurium LT2 and Salmonella enterica serovar Paratyphi A ATCC 9150) using the bioinformatic available program (clustalw; Our results indicated that the deletion described in S. Typhi CT18 is conserved among S. Typhi clinical strains (Fig. 2). Therefore, the sopD2 frameshift mutation seems to be a feature in the genome of serovar Typhi.

Figure 1.

 Genomic organization of sopD2 in Salmonella enterica serovar Typhi. Genomic context of sopD2 in S. Typhi CT18, location of the primers designed (small arrows) and the expected amplicon of 1800 bp (thick black line).

Figure 2.

 Alignment of a genetic section of sopD2. This section was amplified from chromosomal DNA of a Salmonella enterica serovar Typhi clinical collection (STH2370 and numbers, see Table 1) sequenced and compared with the genomic sequences of S. Typhi CT18 (CT18), Salmonella enterica serovar Paratyphi A ATCC 9150 (ParatA) and Salmonella enterica serovar Typhimurium LT2 (LT2) available in online databases ( Black long squares show the punctual deletion and the TGA STOP codon.

HT-29 cellular permeability altered by S. Typhi carrying sopD2STM gene

Recently, we reported that S. Typhi harboring the sseJ gene from S. Typhimurium significantly disrupted HT-29 monolayer compared with the wild-type strain (Trombert et al., 2010). In the same way, we infected polarized HT-29 monolayers with the S. Typhi wild-type and S. Typhi sopD2STM strains. If the integrity of the monolayer is disrupted by the bacteria, gentamicin will leak through the lower chamber, decreasing the recovered CFU mL−1. When the monolayer is not disrupted, the recovered CFU mL−1 should remain essentially constant over the same time course. The S. Typhimurium 14028s (black diamonds) and S. Typhimurium 14028s ΔsopD2∷FRT (NT060) (white circles) showed a slight decline over the time course of the assay, suggesting that the monolayer integrity was not significantly affected by these strains (Fig. 3). In contrast, CFU mL−1 of S. Typhi STH2370 abruptly decreased until they became undetectable, strongly suggesting that gentamicin leaked due to a monolayer disruption (white squares). When S. Typhi was complemented with sopD2STM gene (in the pNT007 plasmid, see Materials and methods) and used to infect the monolayer, we observed that the corresponding CFU mL−1 showed a sharp difference with the otherwise isogenic wild-type strain resembling the S. Typhimurium phenotype (black triangles). The CFU mL−1 numbers from infected cells with S. Typhi carrying the empty plasmid (pCC1) showed no differences with respect to the wild-type strain (data not shown).

Figure 3.

 Cell permeability assay of Salmonella enterica serovar Typhi and Salmonella enterica serovar Typhimurium through HT-29 human cell monolayers. Black diamonds, S. Typhimurium 14028s; white squares, S. Typhi STH2370; black triangles, S. Typhi STH2370/pNT007; white circles, S. Typhimurium 14028s ΔsopD2∷FRT. The arrow indicates the time at which gentamicin was added. The results represent the average of three independent experiments. Each experiment was performed in duplicate. Values are expressed as means±SD of three independent experiments.

Decrease of S. Typhi sopD2STM intracellular proliferation within HEp-2 cells

It has been reported that sopD2 contributes to the synthesis of Sifs, lipid filaments essential for S. Typhimurium intracellular proliferation (Brumell et al., 2003; Jiang et al., 2004; Birmingham et al., 2005). When we performed a gentamicin protection assay, we observed that S. Typhi sopD2STM showed a significant decrease of CFU recovered from HEp-2-infected monolayers compared with the wild-type strain (Fig. 4). In contrast, S. Typhi sopD2STM showed similar invasion levels compared with S. Typhimurium 14028s ΔsopD2∷FRT (NT060) (P=0.13749). The results suggest that loss of SopD2 function in the serovar Typhi contributes to the bacterial intracellular proliferation in human epithelial cells.

Figure 4.

 Gentamicin protection assay of wild-type Salmonella enterica serovar Typhi and its complemented strain. HEp-2 cells were grown and infected with Salmonella enterica serovar Typhimurium 14028s, S. Typhi STH2370, S. Typhi STH2370/sopD2STM (pNT007) and S. Typhimurium 14028s ΔsopD2∷FRT (NT060). The recovered CFUs were counted 3-h postinfection. Values correspond to means±SD of three different experiments, each performed in triplicate.


In the process of adaptation to humans, bacterial genes no longer compatible with the lifestyle of facultative pathogens within the host are selectively inactivated. These inactivated genes are called ‘antivirulence genes’ and their loss of function results in the adaptation to a given host (Maurelli, 2007). Salmonella enterica serovar Typhi is a facultative bacterial pathogen that has accumulated a large number of pseudogenes (approximately 5% of the genome), over 75% of which have completely lost their function (McClelland et al., 2004; Dagan et al., 2006). Compared with free-living organism genomes, facultative pathogens harbor several pseudogenes and a gene population structure that promotes the maintenance of specific mutations. In contrast to free-living bacteria (large genomes, a great diversity of functional genes and low percentage of laterally transferred genes) and obligate parasites (extremely reduced genomes), S. Typhi represents an intermediate step exhibiting some genome erosion directed to inactivation and loss of detrimental or nonessential functions for its environment, i.e. the host (Ochman & Moran, 2001). Therefore, we hypothesized that loss of some genes contributed to the adaptation of S. Typhi to produce a systemic infection in humans.

sopD2 gene corresponds to an SPI-2-regulated effector protein (Brumell et al., 2003). In S. Typhi this gene carries a nucleotide deletion that produced a nonsense mutation and probably the loss of an essential protein translocation motif [WEK(I/M)xxFF; data not shown] (Brumell et al., 2003; Brown et al., 2006). Our results confirm that sopD2 in S. Typhi is a pseudogene supporting previous studies of Salmonella spp. comparative genomics (Parkhill et al., 2001; McClelland et al., 2004).

The SPI-1 encodes T3SS effectors that mediate the invasion by Salmonella of nonphagocytic cells via cellular host cytoskeleton manipulation (Bueno et al., 2010). SPI-2 is induced after bacterial internalization and is essential for bacterial survival and proliferation within SCV. In addition, intracellular Salmonella interferes with the actin cytoskeleton in an SPI-1 T3SS-independent manner and in SPI-1-dependent effectors contributing to the SCV establishment (Abrahams & Hensel, 2006; Bakowski et al., 2008). This has been described as a regulatory and functional SPI cross-talk (Knodler et al., 2002; Steele-Mortimer et al., 2002). It is therefore not surprising that although sopD2 in S. Typhimurium corresponds to a SPI-2-regulated effector, it is also expressed under SPI-1 conditions and participates in epithelial cell invasion (Brumell et al., 2003). The previous observation is supported by S. Typhimurium ΔsopD2∷FRT null mutant strain (NT060), which showed a decreased invasion level in HEp-2 human epithelial cell line.

In S. Typhimurium, SCV synthesis depends on SopD2 and the sopD2 gene mutation decreases intracellular proliferation and bacterial virulence in mice (Jiang et al., 2004; Birmingham et al., 2005). However, the presence of a fully functional sopD2 gene interferes with S. Typhi proliferation within human epithelial cells and the bacterial capacity to alter cellular permeability. Because S. Typhimurium SopD2 participates in the endosome (SCV) synthesis and concomitantly in the generation of Sif structures (Brumell et al., 2003; Jiang et al., 2004; Birmingham et al., 2005), we suggest that SopD2STM in S. Typhi interferes directly in the intracellular traffic of this pathogen in epithelial cells.

Our previous studies showed that this permeability alteration is directly related to cytotoxicity (Trombert et al., 2010). In this work, we observed that the functional transfer of sopD2 from S. Typhimurium to S. Typhi decreased cellular permeability and more likely decreased cytotoxicity. It has been observed that S. Typhi increases cytotoxicity within the host by inactivating or acquiring new genes (Oscarsson et al., 2002; Fuentes et al., 2008; Faucher et al., 2009; Trombert et al., 2010). Thus, the inactivation of some genes (i.e. sopD2) and the acquisition of others could contribute to cytotoxicity toward the epithelial barrier and might ensure the development of systemic infection in the human host.

Our results together suggest that pseudogenization of S. Typhi sopD2 caused modification in the bacterial pathogenicity within eukaryotic cells in vitro, plausibly contributing to the S. Typhi adaptation to the human host.


UNAB Grant DI-05/I (A.N.T.), CONICYT Grant D-21060491 and AT-24080052 (G.P.R.) and FONDECYT Grant 1110120 (G.C.M.).