Materials and Methods

Animals—The in vivo portion of the study involved the use of 1-to 2-week-old male Holstein calves (n = 36) and 10-day-old mixed-breed pigs of both sexes (72). Animal experiments were approved by the Iowa State University Institutional Animal Care and Use Committee.

Bacterial strains and preparation—Bacterial strains used in the study were summarized^5,7–12 (Appendix 1). Strain LNWI⁹ (designated as Salmonella Typhimurium CYP) served as the cytopathic (ie, CYP) strain, whereas Salmonella Typhimurium DT104 strain 98-795⁸ was used in experiments involving protozoa. Other SGI1-bearing strains included recently acquired isolates of Salmonella Dublin and Salmonella Choleraesuis. Salmonella Saint-paulNPG was chosen as a representative neuropathogenic (ie, NPG) strain, and SARB strains¹² and certain SGI1-free strains served as control strains. Bacteria were stored in cryopreservation tubes containing 50% glycerol and 50% Lennox L broth^a at −80°C and were grown in or on Lennox L broth or culture agar^b prior to use.

Isolation of SGI1-bearing isolates of Salmonella Dublin and Salmonella Choleraesuis—For both serotypes, approximately 150 to 500 isolates were obtained from clinical and nonclinical isolates submitted to the National Veterinary Services Laboratories from 2004 to 2006. Serotype-specific isolates were pooled and grown aerobically for 16 hours at 37°C in 500 ml of Lennox L broth that contained ampicillin^c (100 μg/mL) and chloramphenico^c (32 μg/mL). One milliliter of flocculent culture was then subcultured aerobically for 16 hours at 37°C in 50 ml of Lennox L broth that contained streptomycin^c (64 μg/mL) and sulfamethoxazole^c (512 μm/mL). Individual colonies were isolated by plating 30-μL aliquots on culture agar^b that contained ampicillin, chloramphenciol, streptomycin, and sulfamethoxazole. The presence of SGI1 was assessed in individual colonies via a PCR assay that used the floR-tetR amplicon.⁸

Assessment of other SGI1 genes in isolates of Salmonella Dublin and Salmonella Choleraesuis—To evaluate the fidelity of SGI1 in isolates that had positive results for floR-tetR, 4 other SGI1 genes were examined. These genes were SO13, tnpR, aadA2, and pse-1. The PCR conditions were as described elsewhere⁸ for the floR-tetR amplicon. The oligonucleotide sequences were summarized^6,8,13 (Appendix 2).

Salmonella invasion assays following survival within Acanthamoeba castellanii or HEp-2 cells—Acanthamoeba castellanii^d and HEp-2 cells^e (ie, immortal adherent mammalian epithelial cells that are used for assessing Salmonella virulence¹⁰) were maintained as described elsewhere.⁶ Approximately 10⁹ bacteria were added to approximately 10⁵ A castellanii or HEp-2 cells. The Salmonella-Protozoa mixture was then gently mixed for 16 hours at 37°C in a sealed 5-mL glass tube, whereas the Salmonella-HEp-2 mixture was maintained at 37°C in a 5% CO₂ humidified incubator. At the end of the 16-hour incubation period, extracellular Salmonella organisms were killed by the addition of florfenicol^f (300 μg/mL). Then, HEp-2 cells were trypsinized and resuspended in Coleman buffer.^5,6 Cells were lysed by centrifugation (5,000 × g for 60 seconds) with 2.5mM glass beads and a mini-bead beater.^g The lysate was centrifuged at 1,000 × g for 2 minutes and then re-suspended in 350 μL of Lennox L broth. Of the 350 μL, 25 were used for selective enumeration and 300 were used for an invasion assay performed in triplicate (ie, 100 μL/well), as described elsewhere.^5,6 Percentage invasion was calculated by dividing the number of CFUs recovered by the number of CFUs added.

In vivo infection experiments involving protozoa-laden Salmonella organisms—For all in vivo experiments, animals were assigned to groups by use of a randomization procedure. Acanthamoeba castellanii was chosen for use in these experiments because this protozoan can mediate Salmonella hypervirulence⁶ and is often associated with water and thus is relevant to potable water used in swine operations.

Approximately 10⁹ bacteria were added to approximately 10⁵ A castellanii or HEp-2 cells, and the mixtures were subjected to the same regimen described for the invasion assay. Salmonella-laden HEp-2 cells were trypsinized and removed from tissue culture dishes prior to in vivo challenge exposure. Previous studies^5,6 have revealed that approximately 40% to 60% of Salmonella organisms are recovered from eukaryotic cells after coincubation.

The Salmonella-eukaryotic cell mixture was then placed in a gelatin capsule, and the capsule was orally administered to 1- to 2-week-old male Holstein calves or 10-day-old mixed-breed pigs of both sexes. For each strain of Salmonella spp, 6 pigs and 3 calves were inoculated with A castellana cells that contained Salmonella spp, whereas 3 animals received HEp-2 cells that contained Salmonella spp (the latter was not administered to calves for Salmonella Typhimurium DT104 and TH11). All animals received approximately 4.5 × 10⁶ CFUs/kg. Animals were monitored every 8 to 12 hours to evaluate changes in appetite, fecal consistency, and rectal temperature. Bacteriologic examinations were performed on blood samples collected every 12 hours.

At 36 to 96 hours after inoculation, animals were euthanatized. Calves were premedicated by administration of xylazine hydrochloride^h (0.45 mg/kg, IM), which was followed by administration of pentobarbitalⁱ (1 mg/kg, IV). Pigs were euthanatized by administration of pentobarbital (1 mg/kg, IP). The spleen was then removed from each animal and submitted for bacteriologic examinations.

In vivo infection experiments involving cytopathic strains—Animals were challenge-exposed via oral administration of 4.5 × 10⁶ CFUs of Salmonella Typhimurium CYP/kg; these Salmonella organisms were grown aerobically for 16 hours at 37°C. Six pigs and 3 calves were inoculated with Salmonella Typhimurium CYP, and 6 pigs and 3 calves were inoculated with the control strain (Salmonella Typhimurium CYP-SlyA', which is not capable of cytopathic effects¹¹).

When pronounced clinical signs, such as diarrhea, dehydration, and pyrexia (typically at 2 to 4 days after inoculation), were observed, affected animals were immediately euthanatized. Euthanasia was performed with xylazine and pentobarbital as described previously. Various samples were collected, including samples of the intestines, blood, spleen, lungs, liver, kidneys, and gonads; the kidneys and gonads were collected because these 2 tissues rarely harbor Salmonella spp, except for the novel cytopathic strains.^1,10,11

In vivo infection-stress experiments involving neuropathogenic strains—Animals were challenge-exposed via oral administration of 4.5 × 10⁶ CFUs of Salmonella Saint-paulNPG/kg; these Salmonella organisms were grown aerobically for 16 hours at 37°C. For each strain, 3 pigs and 3 calves received daily doses of norepinephrine^c (45 μg/kg, IM) starting on the day of inoculation and continuing until the animals were euthanatized. As a control treatment, 3 calves and 3 pigs were challenge-exposed by oral administration of Salmonella Saint-paulNPG and administered a placebo (volume of saline [0.9% NaCl] solution identical to the volume of norepinephrine, IM). Additionally, 3 calves and 3 pigs were challenge-exposed by oral administration of Salmonella Saint-paulNPG and administered daily doses of dexamethasone^j (0.1 mg/kg, IM). The control strain, which was administered to 3 calves and 3 pigs, was Salmonella Saint-paulSARB, which is not capable of causing neurologic disease in cattle.⁷

Animals were monitored for neurologic disease (seizures or clinical signs such as excessive ear fluttering, ataxia, opisthotonus, and proprioceptive placing deficits). When signs of neurologic disease were observed (typically 4 to 9 days after inoculation), affected animals were immediately euthanatized. An animal with neurologic disease was defined as an animal that had convulsions or that had at least 2 of the aforementioned 4 clinical signs of neurologic disease. Animals that did not have signs of neurologic disease were euthanatized 12 to 14 days after inoculation (ie, at least 3 or 4 days after animals with signs of neurologic disease were euthanatized). Animals were euthanatized with xylazine and pentobarbital, as described previously. Samples of the intestines, blood, spleen, and brain were collected from each animal.

Bacteriologic-based detection of Salmonella spp in various tissues—For qualitative assessment of Salmonella spp, microbes were selectively cultured by inoculation of 3 to 5 g of each sample into 100 mL of GN Hajna broth^k and incubation overnight at 37°C in aerobic conditions. Then, 100 μL of the inoculum was transferred into 5 mL of Rappaport-Vassiliadis R10 broth,^k which was also incubated overnight at 37°C in aerobic conditions. Cultures were then transferred to plates containing BGS agar^k; plates were incubated overnight at 37°C. Individual colonies recovered from selective plates were grown overnight in Lennox L broth and identified by use of a Salmonella-speciñc PCR assay targeting the sipB-sipC junction.⁸ All media included antimicrobials specific to the antibiogram of the individual isolate.

For quantitation of Salmonella organisms in blood and spleen tissues as an estimate of systemic pathogen load and as an indirect correlate of clinical signs, samples were directly inoculated on plates containing BGS agar. For blood samples, 100 μL was inoculated onto each of 10 separate plates. For spleen samples, 50 to 75 g of sample was homogenized, filtered with cheesecloth, and then inoculated onto 10 plates containing BGS agar (ie, approx 3 to 5 g of sample/plate). Colonies were enumerated the following day. Because these tissues can harbor similar numbers of bacteria at early time points in the infectious process, CFU data for spleen samples were pooled with counts from blood samples to provide the final data of each time course experiment. The identity of Salmonella strains was confirmed by use of antisera-agglutination-based serogrouping^l in addition to confirmation of the original antibiogram for each isolate.

Statistical analysis—Statistical analysis was performed by use of an ANOVA with the Scheffe F test for multiple comparisons. Analyses were performed by use of a commercially available statistical program.^m Although multiple samples were obtained from each tissue, each tissue-associated value represented a mean for the replicates from that particular tissue.

Results

Isolation and preliminary characterization of SGI1-bearing Salmonella Choleraesuis and Salmonella Dublin—From the pools of Salmonella Choleraesuis and Salmonella Dublin isolates, 3 colonies of Salmonella Choleraesuis and 2 colonies of Salmonella Dublin had ampicillin-chloramphenicol-streptomycin-sulfonamide antibiograms. One colony of each serotype yielded the floR-tetR amplicon (Figure 1); this amplicon is present in SGI1.⁸

View Full Size

Bands for PCR-based characterization of SGI1 in Salmonella enterica serotype Choleraesuis (C) and S enterica serotype Dublin (D). Genes evaluated were SO13^3,6, tnpR (resolvase), aadA2 (spectinomycin-streptomycin resistance), floR-tetR (junction between floR [florfenicol resistance] and tetR [regulator of tetracycline resistance]), and pse-1 (carbenicillinase). Predicted amplicon sizes are 800 bp (SO13), 172 bp (tnpR), 250 bp (aadA2), 280 bp (floR-tetR), and 231 bp (pse-1). M = DNA ladder.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1170

• Download Figure
• Download figure as PowerPoint slide

Profile of SGI1 genes in isolates of Salmonella Choleraesuis and Salmonella Dublin with the floR-tetR amplicon—To evaluate the nature of SGI1 in isolates with the floR-tetR amplicon, 4 other SGI1 genes (SO13, tnpR, aadA2, and pse-1) were evaluated. The function of the SO13 gene is unknown,³ although a recent study⁶ revealed that it participates in upregulating invasion genes (and thus virulence) in Salmonella Typhimurium DT104 and other strains containing SGI1.⁶ The tnpR gene is a transposase-encoding gene present in many integrons,¹⁴ whereas aadA2 and pse-1 confer resistance to streptomycin-spectinomycin and ampicillin-amoxicillin, respectively. All 5 genes were detected in both Salmonella Choleraesuis and Salmonella Dublin.

Protozoa-dependent invasion assays for SGI1-bearing Salmonella Choleraesuis and Salmonella Dublin—To determine whether the SGI1-bearing strains of Salmonella Choleraesuis and Salmonella Dublin are hyperinvasive following exposure to protozoa (as has been reported⁵ for Salmonella Typhimurium DT104), bacteria were incubated with free-living protozoa (A castellanii cells) and invasion assays were performed. As a control sample, bacteria were incubated with HEp-2 cells. Bacteria were recovered from protozoa or HEp-2 cells and then subjected to a standard tissue culture invasion assay by use of HEp-2 cells.¹⁰ Protozoa were capable of augmenting invasion for the SGI1-bearing strains of Salmonella Choleraesuis and Salmonella Dublin (Figure 2). The 6- to 7-fold increase was similar to that observed for Salmonella Typhimurium DT104, which parallels that reported⁵ for rumen protozoa-mediated hyperinvasion and hypervirulence. No such effect was observed for Salmonella Fyphimurium TH11 (a DT104 strain that lacks SGI1⁸) or for the SGI1-free SARB strains of Salmonella Choleraesuis and Salmonella Dublin.¹²

View Full Size

Protozoa-mediated effects on the invasion of HEp-2 cells (black bars) and Acanthamoeba castellanii cells (white bars) by SGI1-bearing isolates of Salmonella Choleraesuis (Sch) and Salmonella Dublin (Sdub). Mean ± SEM values for invasion percentage were determined after recovery from HEp-2 or A castellanii cells. Control isolates included SGI1-bearing S enterica serotype Typhimurium DT104, Salmonella Typhimurium TH11 (ie, SGI1-free Salmonella Typhimurium DT104), and Salmonella CholeraesuisSARB and Salmonella DublinSARB (both of which lack SGI1). Invasion percentage was calculated as follows: 100 × (CFUs recovered from the HEp-2 cells/CFUs added to the HEp-2 cells). *Within an isolate, value differs significantly (P < 0.05) from the value for the HEp-2 cells.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1170

• Download Figure
• Download figure as PowerPoint slide

In vivo experiments to assess protozoa-associated virulence for Salmonella Typhimurium DF104 in pigs and cattle—Salmonella organisms were incubated with protozoa and then inoculated into pigs and calves to determine whether protozoa-associated Salmonella Typhimurium DT104 hypervirulence in calves could extend to pigs. As a control sample, the same Salmonella strains were incubated with HEp-2 cells prior to in vivo inoculation.

Protozoa were able to enhance the systemic virulence of Salmonella Typhimurium DT104 in pigs (Figure 3). This effect was not as dramatic as that observed in calves, as indicated by the results for the present study as well as results of another study⁵ All Salmonella Typhimurium DT104—inoculated calves were euthanatized by 48 hours after inoculation because of severe clinical manifestations of salmonellosis (eg, pyrexia, diarrhea, lethargy, anorexia, and dehydration). Clinical signs were mild and transient in the Salmonella Typhimurium TH11-inoculated calves and in all pigs; thus, these animals were not euthanatized until 96 hours after inoculation. Protozoa-associated Salmonella Typhimurium DT104 were recovered to a significantly greater extent than were Salmonella Typhimurium DT104 not exposed to protozoa. The SGI1-free strains were not recovered from samples of blood or spleen obtained from inoculated pigs, regardless of protozoa exposure. Salmonella organisms were not recovered from most of the spleen tissues obtained from animals challenge-exposed with SGI1 free strains or strains that were not exposed to protozoa.

View Full Size

Evaluation of the protozoal-mediated effects on time-dependent systemic infectivity of Salmonella Typhimurium DT104 in pigs (SW) or calves. Data represent the mean ± SEM number of CFUs of Salmonella Typhimurium DT104 or Salmonella Typhimurium TH11 (ie, the SGI1-free version of Salmonella Typhimurium DT104) recovered from calves or pigs inoculated with protozoa (Pz) that contained Salmonella spp or with HEp-2 cells (no Pz) that contained Salmonella spp; values were calculated for 3 or 6 samples, each of which was assayed in triplicate. The last data points (48 hours after inoculation for Salmonella Typhimurium DT104-inoculated calves and 96 hours after inoculation for pigs and Salmonella Typhimurium TH11-infected calves) represent cumulative data for samples of blood and spleen. *Within a time point, value differs significantly (P < 0.05) from the relevant control treatment (eg, Salmonella Typhimurium DT104 vs Salmonella Typhimurium TH11 or Pz vs no Pz).

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1170

• Download Figure
• Download figure as PowerPoint slide

In vivo experiments to assess protozoa-associated virulence for SGI1-bearing host-adapted Salmonella spp in pigs and cattle—In vivo experiments were conducted to determine whether protozoa-associated hyper-virulence could be detected in 2 new SGI1-bearing strains of Salmonella spp adapted to pigs (Salmonella Choleraesuis) and cattle (Salmonella Dublin). Salmonella organisms were incubated with protozoa and then inoculated into calves and pigs. As a control sample, the same Salmonella strains were incubated with HEp-2 cells prior to in vivo inoculation. Control strains included the SARB versions of each serotype.

Protozoa were able to enhance the systemic virulence of SGI1-bearing Salmonella Choleraesuis in pigs (Figure 4). This effect was evident at an earlier onset of detectable pathogen load (36 hours for Salmonella CholeraesuisSGI plus protozoa and 48 hours for all others) and an earlier onset of pathogen burden necessitating euthanasia (48 hours for Salmonella CholeraesuisSGI plus protozoa and 60 hours for all others). Protozoa were incapable of augmenting the virulence of SGI1-bearing Salmonella Dublin.

View Full Size

Protozoal-mediated effects on time-dependent systemic infectivity of SGI1-bearing host-adapted Salmonella spp in pigs (A) or calves (B). Data represent mean ± SEM number of CFUs of Salmonella organisms recovered from pigs or calves noculated with protozoa (Pz) that contained Salmonella spp or with HEp-2 cells (no Pz) bearing Salmonella spp; values were calculated for 3 samples, each of which was assayed in triplicate See Figure 3 for remainder of key.

Citation: American Journal of Veterinary Research 71, 10; 10.2460/ajvr.71.10.1170

• Download Figure
• Download figure as PowerPoint slide

In vivo experiments to evaluate the cytopathicity of strain Salmonella Typhimurium CYP in pigs—In vivo experiments were conducted to determine whether strain Salmonella TyphimuriumCYP could evoke cytopathic effects and polysystemic salmonellosis in pigs, similar to that observed in veal calves.¹⁰ Salmonella TyphimuriumCYP was inoculated into young calves and pigs. As a control sample, animals were inoculated with Salmonella TyphimuriumCYP-SlyA', which is not capable of cytopathic activity¹¹ As an additional control sample, tissue samples were obtained from animals inoculated with the host-adapted serotypes.

Pigs did not harbor Salmonella TyphimuriumCYP in any of the extraintestinal sites (Table 1). In contrast, Salmonella TyphimuriumCYP was recovered from all tissues obtained from the calves, including organs not usually associated with salmonellosis (eg, kidneys and testes). The noncytopathic strain Salmonella TyphimuriumCYP-SlyA' was recovered from all intestinal samples and from the spleen of 1 calf. Both host-adapted serotypes were recovered from extraintestinal sites (ie, blood, spleen, and lungs) usually associated with these pathogens in their preferred hosts. Renal and gonadal tissues did not yield either host-adapted strain from the respective host.

In vivo experiments to evaluate the neuropathogenic effects of Salmonella Saint-paulNPG in pigs—In vivo experiments were conducted to determine whether strain Salmonella Saint-paulNPG could lead to neurologic disease in pigs, similar to that observed in stressed calves.⁷ Salmonella Saint-paulNPG was inoculated into young calves and pigs, and these animals were also administered consecutive daily doses of norepinephrine or dexamethasone to mimic stress. As a control sample, animals were inoculated with Salmonella Saint-paulSARB, which is not capable of eliciting signs of neurologic disease in calves.⁷

Pigs did not have signs of clinical disease, and Salmonella organisms could only be recovered from the intestines (Table 2). In contrast, calves had signs of neurologic disease when inoculated with Salmonella Saint-paulNPG and administered norepinephrine or dexamethasone. Salmonella organisms were isolated from the intestines, blood, spleen, and brain of calves with signs of neurologic disease. Nonneuropathogenic Salmonella Saint-paulSARB was only recovered from the intestines of calves, similar to results in another study.⁷

Discussion

Several unique strains of Salmonella spp have unusual virulence and pathogenicity properties in calves. However, these effects have not been confirmed clinically in pigs. Hypervirulence in cattle is attributable to the presence of both SGI1 and survival within protozoa, whereas cytopathogenicity and neuropathogenicity have been associated with specific strains of Salmonella spp under certain conditions. Stress has been implicated in the etiology of the encephalopathies related to the neuropathogenic Salmonella spp.

Analysis of results for the study reported here indicated that protozoa-associated hypervirulence of Salmonella Typhimurium DT104 can be observed in pigs, although the effect is not as pronounced as that found in cattle. In pigs, Salmonella Typhimurium DT104 infection peaked at 60 hours after inoculation but decreased with time, whereas the pathogen increased in a time-dependent logarithmic manner in cattle until clinical signs warranted euthanasia at 48 hours. The SGI1-bearing Salmonella Choleraesuis had enhanced virulence following exposure to protozoa, but this effect was only related to the onset of disease. Salmonella Dublin associated with protozoa did not have significant enhancement of virulence in cattle, and this may have been related to its inherent putative maximal degree of virulence in calves.

Further comparisons revealed that Salmonella TyphimuriumCYP could not evoke cytopathic effects and polysystemic salmonellosis in pigs similar to those observed in cattle. Salmonella Saint-PaulNPG caused neurologic disease in cattle, but similar effects were not observed in pigs. These 2 findings are not altogether surprising because the Salmonella-associated cytopathogenicity and neuropathogenicity appear to be dependent on neurohormonal factors in cattle that may be divergent from those in swine.

Overall, these Salmonella strains typically are more virulent in cattle. For SGIl-bearing Salmonella spp, protozoa-associated hyperinvasion led to augmented virulence in pigs, and this was evident in the onset of disease for Salmonella Choleraesuis. From a clinical perspective, the hypervirulence phenomenon has not been reported in pigs because these animals have a lower exposure to protozoa than do ruminants (ie, protozoamediated hypervirulence of Salmonella spp can occur in pigs, but the incidence is probably not as high as that in cattle). In contrast, cytopathic and neuropathogenic strains were not capable of expressing their respective phenotypes in pigs.