Unravelling the role of anaerobic metabolism (pta-ackA) and virulence (misL and ssa) genes in Salmonella Heidelberg shedding using chicken infection model

Abstract

The mechanism of colonisation of the chicken intestine by Salmonella remains poorly understood, while the severity of infections vary enormously depending on the serovar and the age of the bird. Several metabolism and virulence genes have been identified in Salmonella Heidelberg; however, information on their roles in infection, particularly in the chicken infection model, remains scarce. In the present publication, we investigated three Salmonella Heidelberg mutants containing deletions in misL, ssa, and pta-ackA genes by using signature-tagged mutagenesis. We found that mutations in these genes of S. Heidelberg result in an increase in fitness in the chicken model. The exception was perhaps the pta-ackA mutant where colonisation was slightly reduced (2, 7, 14, and 21 days post-infection) although some birds were still excreting at the end of the experiment. Our results suggest that for intestinal colonisation of the chicken caecum, substrate-level phosphorylation is likely to be more important than the MisL outer membrane protein or even the secretion system apparatus. These findings validate previous work that demonstrated the contribution of ackA and pta mutants to virulence in chickens, suggesting that the anaerobic metabolism genes such as pta-ackA could be a promising mitigation strategy to reduce S. Heidelberg virulence.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-023-01241-6.

Introduction

Oral infection of the host by Salmonella enterica induces major cellular changes including rapid drop in pH stomach level and the host temperature increasing. In response, Salmonella is required to respond adaptively to these host environmental stresses largely by transcriptional changes including overexpression of efflux pump-encoding genes, expression of antimicrobial resistance, and chromosomal and plasmid-mediated virulence genes [1]. Type III secretion systems (T3SS) is a major mechanism mediating virulence, secretion of and intracellular injection of effector proteins that manipulate various host cell processes, including the modulation of cell biology and the immune response, enabling the bacteria to avoid detection and clearance [2].

There are several clusters of genes encoding virulence some of which are arranged as T3SS including Salmonella Pathogenicity Island 1 (SPI-1), SPI-2 (ssa), and SPI-3 (misL) encoding genes required for invasion of epithelial cells, survival and multiplication within macrophages, and miscellaneous functions including intestinal colonisation, respectively [3–5].

The misL gene is encoded by SPI-3 and associated with Salmonella invasion locus B (invB) playing a key role in its ability to cause disease in humans and animals [3, 6–8]. This gene is one of a cluster of genes encoding proteins involved in the invasion of host cells. When expressed, the resulting MisL outer membrane protein aids Salmonella to attach to and invade host cells. This protein has been shown to bind to an extracellular matrix protein (fibronectin) found in host tissues, which may facilitate Salmonella invasion [3, 6–8].

Another key component of the Salmonella T3SS located on SPI-2 is the ssa (secretion system apparatus) gene, one of several which is essential for Salmonella to establish and maintain an intracellular niche within host cells, considered critical for its virulence [4, 9]. This pathogen can inject effector proteins encoded by ssa genes through the needle complex of the SPI-2 T3SS into a pore of the host cell. These effector proteins can modulate host cell processes and create a favourable environment for bacterial survival and replication [4, 9].

In Salmonella and other enterobacteria, the pta-ackA genes are key elements in metabolism and energy homeostasis. The genes encode phosphotransacetylase (PTA) and acetate kinase (ACKA) enzymes respectively. Involved in the central metabolic pathway, these enzymes act in the conversion of acetyl-CoA to acetate and ATP production, but also recycling acetyl-CoA [5, 10]. In particular, studies have shown that the expression of these enzymes is upregulated during infection, indicating that they play a role in the survival and growth of the bacterium within the host [5, 10].

Therefore, we performed an in vivo experiment using 1-day-old broiler chicks challenged by three defective Salmonella Heidelberg mutant strains carrying deletions in misL, ssa, and pta-ackA genes.

Materials and methods

Bacterial strains

Salmonella Heidelberg wild-type strain (Lab ID: SH203) was grown routinely in Lysogeny Broth and Lysogeny agar (LB; Becton Dickinson, Maryland, USA) at 37°C. Spontaneous mutants of the S. Heidelberg wild-type strain resistant to nalidixic acid (NAL) (100 µg/mL) or nalidixic acid (100 µg/mL) plus spectinomycin (SPC) (100 µg/mL) were selected to facilitate recovery of the S. Heidelberg mutant strains (SHΔmisL, SHΔssa, and SHΔpta-ackA) and S. Heidelberg wild-type (SH-wt), respectively, during the in vivo experiment.

Construction of Salmonella Heidelberg mutants (SHΔmisL, SHΔssa, and SHΔpta-ackA)

To generate the SHΔmisL, SHΔssa, and SHΔpta-ackA mutants, we used the Lambda-Red technique by replacing the target genes with a kanamycin (Kan) cassette [11]. Briefly, regions upstream and downstream of each target gene (misL, ssaOPQRSTU, and pta-ackA) were PCR amplified from the S. Heidelberg wild-type strain by using primers described in Table S1. Plasmid pKD4 (Kan) was used as the DNA template and the correct length of the PCR products was ensured by agarose gel electrophoresis (1.5%). The S. Heidelberg wild-type strain carrying the λ Red recombinase plasmid (pKD46) was used as the electrocompetent strain to insert by electroporation the PCR products (misL, ssaOPQRSTU, and pta-ackA). After electroporation, transformants were kept at 37°C for 1 h and then plated onto LB+Kan, which were grown at 37°C for 24h. The clones that had integrated the PCR products were selected and confirmed by PCR.

Bacteriophage P22A lysates of the resultant strains were used to transduce each mutation (misL, ssaOPQRSTU, and pta-ackA) to a clean parental genetic background. The kanamycin marker was removed from their chromosomes by transformation using pCP20 plasmid [11]. Next, successive passages in LB broth at 42°C were carried out before streaking onto LB agar supplemented or not with antimicrobials to confirm the completely removal of the resistance cassettes. We also assessed complete lipopolysaccharide (smoothness) by the absence of agglutination using 1:1000 acriflavine with smooth mutants selected and stored at −80°C in Lysogeny broth supplemented with 30% glycerol for future use [12, 13].

In vivo experiment

The experiments were performed according to the Ethical Principles on Animal Experimentation from the Brazilian College of Animal Experimentation and previously approved by the Ethical Committee on the Use of Animals (Process CEUA-006621/18; at May, 10th of 2018). The experiment was carried out in the Laboratory of Avian Pathology at the Faculty of Agricultural and Veterinary Science of São Paulo State University (FCAV/Unesp).

To evaluate the requirement of misL, ssaOPQRSTU, and pta-ackA for S. Heidelberg infection, an in vivo assay was performed using 225 chicks in total. One-day-old broiler chicks were obtained from a commercial hatchery. Upon arrival, sterile swabs were used to sample the bottom of the transport cardboard boxes to confirm the Salmonella-free status of the birds [14]. The chicks were then distributed equally into five groups each of 45 chicks (group 1 (infected with the misL strain), group 2 (ssaOPQRSTU), group 3 (pta-ackA), group 4 (S. Heidelberg wild-type strain), and group 5 (uninfected control)) within metal cages in temperature-controlled rooms. They received sterilised water and antimicrobial-free balanced feed ad libitum throughout the experiment (28 days).

The chicks in each group (1 to 4) were infected orally at 2 days of age with their corresponding mutant strain with group 5 remaining uninfected as a negative control. The birds were inoculated directly by gavage into the crop with 0.2 mL containing 10⁸ CFU. Chicks that developed severe clinical signs were euthanised humanely by cervical dislocation and were not included in the study. Additionally, clinical signs and mortality were daily recorded throughout 28 dpi.

Faecal excretion

Sixty chicks were used to evaluate the faecal excretion, 15 chicks each from groups 1 to 4. The chicks used to assess faecal excretion were identified individually by leg-bands, and cloaca swabs were taken twice a week throughout 28 days post-infection (dpi) as per our previous studies [13, 15]. Swabs were transferred into 2 mL of Selenite Broth supplemented with 0.04% of novobiocin (SN; Becton Dickinson, Maryland, USA) and incubated for 24 h at 37°C. Subsequently, they were streaked onto Brilliant Green Agar (BGA; Oxoid, UK) supplemented with 100 µg/mL of nalidixic acid for mutant strains (100 µg/mL) with or without spectinomycin (100 µg/mL), incubated for 24 h at 37°C [13]. Presumptive Salmonella colonies were confirmed using Triple Sugar Iron Agar (TSI; Oxoid, UK), Lysine Iron Agar (LI; Oxoid, UK), and Sulphide Indole Motility (SIM; Oxoid, UK) tests followed by slide agglutination test using Salmonella O and H Polyvalent antisera (anti-O; Bio-Rad, USA). We further checked colonies for roughness using acriflavine and the genotype of the strains recovered was confirmed by PCR [13].

Evaluation of cecal colonisation

In order to enumerate bacteria, five chicks from each group were euthanised at two, five, seven, 14, 21, and 28 days post-infection (dpi). Cecal contents were homogenised directly into phosphate-buffered saline pH 7.4 (PBS). Decimal dilutions (v/v) of the PBS homogenates were prepared and a 0.1-mL aliquot of each dilution was inoculated onto BGA^Nal or BGA^Nal+Spc agar plates. Concurrently, 2× concentration of SN broth was added to samples (1:1), whereas plates and enriched samples were incubated at 37°C for 24 h. The CFU/g values were normalised by log₁₀ for statistical analysis purpose. When negative plates were obtained, enriched samples were re-streaked onto BGA^Nal or BGA^Nal+Spc plates and incubated as previously described. The value of 10² CFU/mL was assigned for positives samples for the purpose of further analysis.

Statistical analysis

Data from faecal excretion were compared by Fisher’s exact test, and the values logarithmically transformed for bacterial numbers were submitted to two-way ANOVA followed by Bonferroni multiple comparison test (P ≤ 0.05). All statistical analyses were performed using the software GraphPad Prism, version 9.5.0.

Results

Contribution of misL, ssaOPQRSTU, and pta-ackA to Salmonella Heidelberg faecal excretion

We evaluated faecal excretion for 4 weeks to determine whether S. Heidelberg strains with deletions in misL, ssaOPQRSTU, and pta-ackA genes affect the continuous S. Heidelberg shedding. In this regard, the mutant strains carrying deletions in misL or pta-ackA genes were excreted in similar numbers than the parental wild-type strain during the experimental period (P ≥ 0.05; Fig. 1A; C). In divergence, compared to the wild-type strain, excretion of SHΔssaOPQRSTU strain was statistically higher (P = 0.0003) (Fig. 1B).

Fig. 1.

Faecal excretion of mutants (SHΔmisL, SHΔssa, and SHΔpta-ackA) and wild-type strains by broiler chicks during 4 weeks. A Comparison between the SHΔmisL and wild-type challenge groups; B comparison between the SHΔssa and wild-type challenge groups; C comparison between the SHΔpta-ackA and wild-type challenge groups. *** indicate statistical difference (P = 0.0003), while the absence of * indicates no statistical difference (P ≥ 0.05) by comparison of Fisher’s exact test

The role of misL, ssaOPQRSTU, and pta-ackA to Salmonella Heidelberg gut colonisation by Salmonella Heidelberg in chicks

In order to provide insights into infection of S. Heidelberg in chicks, we carried out Salmonella enumeration through 28 dpi. Although there is no statistical difference at days 2 and 7 p.i., only the SHΔpta-ackA strain was recovered in lower numbers than the parental wild-type strain (P ≥ 0.05; Fig. 2C). Similarly, at day 14, all mutants were recovered less in comparison with the parental strain, whereas during the 21-day p.i. only SHΔmisL and SHΔpta-ackA strains were isolated in lower numbers compared to the wild-type strain (P ≥ 0.05; Fig. 2A; B). After this time, all mutants became predominant in comparison to the parental wild-type strain (P ≥ 0.05; Fig. 2).

Fig. 2.

Salmonella counts in cecal contents from broiler chicks challenged by SHΔmisL, SHΔssa, and SHΔpta-ackA, and SH wild-type sampled at 2, 5, 7, 14, 21, and 28 days post-infection. A Comparison between the SHΔmisL and wild-type challenge groups; B comparison between the SHΔssa and wild-type challenge groups; C comparison between the SHΔpta-ackA and wild-type challenge groups

Discussion

As a first step towards determining whether S. Heidelberg strains with deletions in misL, ssaOPQRSTU, or pta-ackA genes affect their colonisation of the gut, we observed that mutations in central metabolism (pta-ackA) and virulence (misL and ssa) genes did not appear to be essential for effective colonisation, measured by faecal shedding and evaluation of cecal colonisation by using the chicken infection model.

Previous work has reported the presence of misL in different Salmonella serovars isolated from different sources [16–18]. As an important virulence factor that accomplish varied functions in host cell, misL contributes to Salmonella colonisation in different hosts, including mice, chickens, and pigs [19, 20]. In contrast to our findings, previous surveys have suggested that mutations in misL promote reduction in the invasion of host cells, consequently determined as a potential target for bacterial attenuation [3, 6–8]. However, these studies used S. Typhimurium in HeLa cells rather than S. Heidelberg [3], and S. Typhimurium in colonic carcinoma (T84 cells) [6]. Interestingly, Morgan et al. [8] demonstrated that the S. Typhimurium misL mutant was attenuated in chicks.

SPI-2 T3SS mediates bacterial survival in macrophages and the establishment of systemic disease [2]. While assessing the SPI-2 T3SS mutant (ssaOPQRSTU), we did not find any evidence of attenuation. Gene mutation in Salmonella strains can impair the function of the SPI-2 T3SS, including the ssa gene, which can offer a promising strategy for the development of novel antimicrobial agents against Salmonella infections, leading to reduced virulence [4, 9].

To determine whether a central metabolic pathway that codes for phosphotransacetylase (PTA) and acetate kinase (ACKA) enzymes is essential for S. Heidelberg during infection in chickens, we constructed a strain carrying a deletion in pta-ackA genes to confirm previous results obtained with newly hatched chickens [5] and mice [21] challenged with S. Typhimurium. Here we have demonstrated that in 2, 7, 14, and 21 days p.i., the SHΔpta-ackA was defective for colonisation of chicks, corroborating previous results [5, 21]. Although mutations in the pta-ackA have been shown to reduce the virulence of Salmonella, indicating that these enzymes are necessary for the bacterium to cause disease [5, 21], we cannot yet consider it a complete attenuation.

Besides conversion of acetyl-CoA, phosphotransacetylase is also related to the consumption of propionate excreted during growth on 1,2-propanediol [5]. It is thus entirely possible that strains lacking pta-ackA, which are unable to utilise acetyl-CoA or other electrons acceptors that are more energetically favourable, will be used under anaerobic conditions [13].

Our results suggest that for intestinal colonisation of the chicken caecum, substrate-level phosphorylation is likely to be more important than the MisL outer membrane protein or even the secretion system apparatus. These findings validate previous work that demonstrated the contribution of ackA and pta mutants to virulence in chickens [5], suggesting that the anaerobic metabolism genes such as pta-ackA could be a promising mitigation strategy to reduce S. Heidelberg virulence. Finally, the construction of a mutant strain carrying multiple deletions in virulence and anaerobic metabolism genes deserves to be investigated further.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This work was supported by research grants from The São Paulo Research Foundation (FAPESP 2018/03189-0, 2018/21301-2, and 2020/06076-2).

Declarations

Competing interests

The authors declare no competing interests.