The bacterial strains and plasmids used in this study are listed in Table 1. Bacteriological media and components were from Difco (Franklin Lakes, NJ). Antibiotics and reagents were from Sigma (St. Louis, MO). Salmonella strains were grown at 37°C in either buffered magnesium minimal medium pH 5.5 (MgM) [32] or LB medium [33]. For plates, media was solidified with 1.5% (wt/vol) agar. When required, medium was supplemented with tetracycline (tet; 12.5 μg/ml), L-cysteine-HCl (cys; 22 μg/ml), DL-tryptophan (trp; 20 μg/ml) and L-histidine-HCl (his; 22 μg/ml). Buffered saline with gelatin (BSG) [34] was used as a diluent and to suspend bacteria prior to inoculation of mice.

Expression of spvR-lacZ and spvA-lacZ fusions in RpoS⁺ and RpoS⁻S. Typhi was determined by β-galactosidase activity assay [35]. Strains were transformed with pGTR72 (spvR-lacZ, operon fusion), pGTR90 (spvA-lacZ, operon fusion) or pGTR75 (tet-lacZ operon fusion control) [36] and grown in MgM medium to stationary phase. The spv operon is up-regulated under conditions that mimic the Salmonella containing vacuole (SCV) [33]. MgM media was used for this experiment because it mimics the environment of SCV [37].

S Typhi and S. Paratyphi A strains harboring the S. Typhimurium virulence plasmid were constructed by conjugation on minimal media [38]. Plasmid pStSR101 is a Tnmini-tet-labeled virulence plasmid derivative, which can restore the full virulence of pSTV-cured strains of S. Typhimurium [25]. Like the wild-type virulence plasmid, it is self-transmissible [38]. S. Typhimurium χ3351 SL1344 hisG carrying pStSR101 was used as the donor. Conjugation was performed in M9 minimal media agar [39], supplemented with cys, trp, and his. Selection was performed on M9 media supplemented with cys, trp and tet to select for S. Typhi and S. Paratyphi A transconjugants and against the histidine-requiring S. Typhimurium donor. Transconjugants were characterized for LPS, Vi antigen (S. Typhi), biochemical properties, and nutritional requirements as described [39] and were found to exhibit the expected phenotypes (Table 1). The stability of pStSR101 in S. Typhi and S. Paratyphi A was determined essentially as described by Konjufca et al. [40], except that strains were grown in the absence of tetracycline for fifty generations, at which point cells were plated onto LB agar and individual colonies were screened for tetracycline resistance.

Expression of spvR and spvA in S. Typhi and S. Paratyphi A from pStSR101 was evaluated by RT-PCR. Total RNA extraction was performed by RNeasy QIAgene kit (Hilden, Germany) from strains grown in MgM media at 37°C to stationary phase [33,37]. Reverse transcription and PCR was performed using the one-step RT-PCR QIAgene kit. Specific primers were used for the spvR (5′-GGAAACAGGTTCCTTCAGTATCGC-3′ and 5′-TATTTGGCTGTTAACGGCTCTCCC-3′) (size of the spvR amplified fragment: 349 bp) and spvA (5′-TTGTCCGTCAGACCCGTAAACAGT-3′ and 5′-TCTTCCAGCGACACATCGGTATT CAG-3′) (size of the spvA amplified fragment: 358 bp) genes. 16S rRNA primers were used as control of expression (5′-ACTGGCAGGCTTGAGCTTGTAGA-3′ and 5′-AAGGGCACAACCTCCAAGTA GACA-3′) (size of the 16S rRNA amplified fragment: 158 bp).

BALB/c newborn and infant mice (Charles River Laboratories, Wilmington, MA) were bred and maintained at 22°C to 23°C with 12 hours of illumination daily. Mice older than two weeks were separated from their mothers 4 hours before infection and fed with regular food. Bacterial strains were grown overnight in standing cultures that were diluted 1:100 in prewarmed LB broth and grown with mild aeration to an OD₆₀₀ of 0.8 to 0.9. Bacteria were sedimented by centrifugation at room temperature and resuspended in BSG to densities appropriate for the inoculation route and dose. Newborn mice (3 to 24 hours) and infant mice (48 hours to 3 weeks old) were challenged with ~10⁹ CFU of S. Typhi, S. Paratyphi A, or S. Typhimurium. Ten microliters of ~10⁹ CFU of the bacterial strain suspended in BSG were orally administered. Intranasal inoculations consisted in 5 μl of ~10⁹ CFU of the bacterial strain suspended in BSG, administered without anesthesia. Mice were euthanized via asphyxiation with CO₂ and necropsied at various times. The bacterial titers in newborn mice inoculated orally were determined at 3, 7, 14 and 21 days post oral infection. Spleen, liver, and the intestines were collected and washed with BSG until homogenization. The homogenizer (Brinkman, Westburg, NY) was washed with 5% Amphyl, followed by a wash with 70% ethanol, followed by two washes with dH2O. Homogenized tissues were plated onto MacConkey agar plates supplemented with 1% lactose to determine the number of viable bacteria. Salmonella colonies were white on the MacConkey plates. Isolated colonies were further identified by agglutination with Salmonella specific antiserum (Table 1) and antibiotic resistance marker.

Mann-Whitney U Test (version 5.0; GraphPad Software, Inc.) was used for comparing the expression of spv-lacZ fusions.

Newborn and infant mice infected orally with S. Typhi and S. Paratyphi A survived without any symptoms of disease, while the mice inoculated with a S. Typhimurium strain cured of plasmid pSTV succumbed to the infection (Table 2). Infant mice, one to three weeks of age, and newborn mice infected intranasally with S. Typhi and S. Paratyphi A strains also survived without any symptoms of disease (Table 2). The bacterial titers in newborn mice inoculated orally were determined at 3, 7, 14 and 21 days post infection. Spleen, liver, and the intestines were collected and the number of viable bacteria was determined. S. Typhi RpoS⁺ and S. Paratyphi A RpoS⁺ were able to colonize the intestines for three weeks (Figure 1A). S. Typhi Ty2 RpoS⁻ was less able to persist in the intestines and cleared after one week. S. Typhi and S. Paratyphi A were able to colonize the spleen and liver of infected mice and the RpoS- strain Ty2 was more effectively cleared than the RpoS⁺ strains (Figures 1B and 1C). These results are consistent with a report that rpoS S. Typhimurium mutants are less persistent in mice than their wild-type RpoS⁺ parent strains [41, 45]. Taken together, these results show that although wild-type S. Typhi and S. Paratyphi A can transiently colonize young mice, they are not capable of establishing a disseminating infection.

The S. Typhimurium virulence plasmid is required for colonization of mouse tissues [24]. An investigation into whether the addition of the virulence plasmid to S. Typhi and S. Paratyphi A could enhance their ability to colonize young mice was conducted. One virulence plasmid operon that is critical for host invasion is encoded in the spv region [24,33,46]. The spv region consists of five genes, spvRABCD which are all transcribed in the same direction [46]. The spvR gene encodes SpvR, which activates the transcription of both the spvR and spvABCD transcriptional units [12,47,48]. In addition, the sigma factor RpoS is also required for maximum expression of the operon [49,50,51,52].

Because of the central role of spv in tissue colonization, expression of spvR-lacZ and spvA-lacZ fusions in RpoS⁺ and RpoS⁻ S. Typhi by β-galactosidase activity were compared to establish that these genes were able to be transcribed in S. Typhi and that transcription is dependent upon RpoS. [35]. The RpoS⁺ S. Typhi strains produced β-galactosidase levels comparable to the S. Typhimurium RpoS⁺ control for both fusions (Figures 2A and 2B). In contrast, reduced expression was observed for S. Typhi Ty2, which has a defective rpoS allele due to a frame-shift mutation at nucleotide 993 [53]. These results indicate that S. Typhi is able to transcribe the spv genes to the same levels as S. Typhimurium and confirm that maximum spv expression requires RpoS.

A tetracycline-marked derivative of the S. Typhimurium virulence plasmid, pStSR101 [24], was moved into the S. Typhi and S. Paratyphi A strains. Because these host-restricted strains do not normally carry this plasmid, it is possible that the plasmid may not be maintained for enough generations to colonize a mouse. Therefore, the stability of pStSR101 in S. Typhi and S. Paratyphi A was examined. The pStSR101 virulence plasmid was stably maintained for 50 or more generations in all S. Typhi and S. Paratyphi A strains. In addition, whether the spv genes were transcribed from this plasmid was evaluated. RT-PCR products were detected in all strains harboring pStSR101 (Figure 3), indicating that the spv genes are transcribed in S. Typhi and S. Paratyphi A.

The effect of the S. Typhimurium virulence plasmid on the ability of S. Typhi and S. Paratyphi A to cause disease in newborn and infant mice was evaluated. When newborn and infant mice were orally inoculated with S. Typhi and S. Paratyphi A harboring pStSR101, they survived without any disease symptoms, whereas all the mice inoculated with S. Typhimurium died (Table 3). Infant mice and newborn mice infected intranasally with S. Typhi or S. Paratyphi A carrying pStSR101 also survived without any symptoms of disease (Table 3).

The presence of pStSR101 did not enhance the ability of S. Typhi and S. Paratyphi A to colonize infant mice (Figure 1). Unexpectedly, the presence of pStSR101 in S. Typhi Ty2 RpoS⁻ resulted in a slight increase in persistence in the intestines compared to the plasmid-free strain (Figures 1A and 1D), indicating that there may be a virulence plasmid gene(s) that can complement the defect in intestinal colonization imparted by the RpoS⁻ phenotype. Recently, it has been reported that pP_ST98, a promiscuous R plasmid found in a multi-drug resistant isolate of S. Typhi from Asia, carries the spv region. This plasmid confers antibiotic resistance and increases virulence in mice when transferred to pSTV⁻ S. Typhimurium [54]. While it is possible that spv genes may confer an enhancement of virulence in human hosts, our results indicate that they have no effect on infant mouse colonization in wild-type S. Typhi.