Mutation in Wzz(fepE) Linked to Altered O-Antigen Biosynthesis and Attenuated Virulence in Rough Salmonella Infantis Variant

Simple Summary

This study investigates a rough variant of Salmonella Infantis, a common foodborne pathogen in poultry that can cause illness in humans. Researchers compared a smooth and rough version of the bacteria collected from poultry farms in the USA. Using genetic analysis, they found that both versions were highly similar, but the rough variant had a mutation in a gene called Wzz(fepE). This mutation affected the bacteria’s ability to produce O-antigen, a structure important for virulence. We tested how harmful each version was using a chicken embryo model. The smooth strain was more lethal to embryos, while the rough strain showed reduced virulence. This suggests that the mutation in the rough strain may weaken its ability to cause disease. These findings enhance our understanding of Salmonella Infantis and could help in developing new vaccines or treatments to control infections.

Abstract

Salmonella enterica serovar Infantis has emerged as a prevalent foodborne pathogen in poultry with significant global health implications. This study investigates the molecular characteristics influencing virulence in a S. Infantis rough variant collected from a poultry farm in the USA. In this study, whole genome sequencing and comparative genomics were performed on smooth and rough poultry S. Infantis isolates, while chicken embryo lethality assay was conducted to assess their virulence. Comparative genomics between isolates was analyzed using Mauve pairwise Locally Collinear Blocks to measure the genetic conservation. Embryo survival rates between the isolates were compared using the Kaplan–Meier curves. High genomic conservation was observed between the two isolates, but a frameshift mutation was detected in the Wzz(fepE) gene of the rough variant, resulting in early protein truncation. The chicken embryo lethality assay showed that the lethality rate of the smooth strain was higher than that of the rough strain (p < 0.05). This study identifies a frameshift mutation in the Wzz(fepE) gene, leading to protein truncation, which may reduce bacterial virulence by impacting O-antigen biosynthesis in the rough Salmonella Infantis variant. These findings deepen our understanding of S. Infantis pathogenesis and suggest that targeting the Wzz(fepE) gene or related pathways could be a promising strategy for developing effective vaccines and therapeutic interventions.

1. Introduction

Salmonella enterica subsp. enterica serovar Infantis (antigenic formula 6,7,14:r:1,5) has emerged as a prevalent serovar in poultry, contributing to human salmonellosis [1]. Salmonella Infantis typically causes non-life-threatening episodes of gastroenteritis. However, certain strains may exhibit enhanced virulence, enabling them to cause invasive infections that can progress beyond gastroenteritis [2]. Recent years have seen a concerning rise in its prevalence, marked by increased virulence, conferring enhanced environmental fitness [2,3].

Understanding the virulence factors of S. Infantis is essential for controlling its transmission and impact on human health. Like other Salmonella serovars, the persistence and virulence of S. Infantis are significantly influenced by its surface structures, particularly lipopolysaccharide (LPS), which accounts for about 70% of its outer membrane [4,5]. LPS consists of three components: Lipid A, a core oligosaccharide, and the O-antigen. The O-antigen, a polysaccharide of variable chain lengths and composition, forms the outermost part of LPS and plays a crucial role in host–pathogen interaction and immune evasion [4].

The genes involved in O-antigen biosynthesis include those responsible for synthesizing nucleotide sugar precursors specific to the O-antigen, sugar transferase genes for O-unit assembly, and O-unit processing genes that regulate translocation and chain polymerization [6]. These genes are typically organized in a gene cluster located between galF and gnd in Salmonella, although some genes responsible for structural modifications, such as the adding side-chain residues, may be located outside this cluster. [6]. In Salmonella Enteritidis and Salmonella Typhimurium, the O-antigen chain length follows a modal distribution, regulated by the Wzz genes: WzzST (also known as WzzB) regulates the synthesis of long O-antigen chains of 16–35 repeat units, while Wzz(fepE) regulates the synthesis of very long O-antigen chain >100 repeat units [4]. Studies indicate that the distribution of O-antigen chain lengths on the outer membrane enables Salmonella to escape host immune responses, avoid complement-mediated killing, and promote bacterial persistence in the host [4]. Consequently, O-antigen chain length regulators play an important role in S. Typhimurium virulence, highlighting the O-antigen as a critical determinant of pathogenesis [7,8,9].

However, the specific genetic determinants governing O-antigen biosynthesis and their functional impact on S. Infantis virulence remain largely unexplored. O-antigen diversity is essential for bacterial serotyping, which helps in identifying Salmonella strains that vary in host range and disease spectrum [8]. Salmonella serotyping is based on the antigenic formula, which includes both the O-antigen and flagellar (H) antigens [8]. Recent studies have highlighted the emergence of rough variants of S. Infantis with incomplete antigenic formulas. These variants carried flagellar antigens typical of S. Infantis (-:r:1,5) but lacked the O-antigen (6,7) [10,11].

Here, we characterize the genomic and phenotypic attributes of a rough S. Infantis variant to decipher the genetic alterations influencing its virulence by utilizing sequencing, bioinformatics analysis, and chicken embryo lethality assay. This study advances our understanding of S. Infantis pathogenicity and opens new avenues for targeted interventions and vaccine development against this emerging pathogen.

2. Materials and Methods

2.1. Bacterial Strains

Two Salmonella enterica isolates were used in this work. One isolate, designated as Sal_smooth, was a Salmonella enterica serovar Infantis strain from a poultry production facility in the state of Alabama. Another isolate, designated as Sal_rough, was a Salmonella Infantis -:r:1,5 from a boot swab at a pullet farm located in the state of Alabama.

2.2. Whole Genome Sequencing (WGS) and Bioinformatics Analysis

Genomic DNA of Sal_rough and Sal_smooth was extracted from pellet obtained from 1.0 mL of overnight culture grown in Luria–Bertani (LB) broth using the Dneasy Blood and Tissue Kit (Qiagen, Germantown, MD, USA) following the manufacturer’s protocol. Long-read whole genome sequencing was performed by Plasmidsaurus (Eugene, OR, USA) using Oxford Nanopore Technologies (ONT) PromethION. ONT long reads were quality-filtered using Filtlong v0.2.1 (https://github.com/rrwick/Filtlong, accessed date: 10 September 2024), and de novo assembly was performed using Flye v2.9.4 (https://github.com/mikolmogorov/Flye, [12], accessed date: 10 September 2024). The assemblies were further polished with Illumina reads (generated on Illumina NextSeq2000) using Polypolish v0.6.0 (https://github.com/rrwick/Polypolish/releases, accessed date: 10 September 2024) to generate hybrid assemblies. Genome completeness and contamination were assessed using CheckM v1.2.2 (https://github.com/Ecogenomics/CheckM [13], accessed date: 10 September 2024). The genome annotations were performed with NCBI Prokaryotic Genome Annotation Pipeline (PGAP) (https://github.com/ncbi/pgap, accessed date: 10 September 2024).

Serotyping of both S. Infantis isolates was confirmed using in silico serotyping methods on the NCBI pathogen detection webpage (https://www.ncbi.nlm.nih.gov/pathogens/isolates, accessed date: 10 September 2024). Sequence type was determined with in silico multi-locus sequence typing analysis (MLST) (https://github.com/tseemann/mlst, [14]). Multiple genome alignment was carried out with Mauve software (https://darlinglab.org/mauve/mauve.html, accessed date: 10 September 2024). Plasmids were identified using the PlasmidFinder database on ABRicate (https://github.com/tseemann/abricate [15], accessed date: 10 September 2024).

2.3. Identification of O-Antigen Biosynthesis Genes and Wzz(fepE) Mutation in S. Infantis

The assembled genomes of Sal_smooth and Sal_rough were compared to identify genetic variants potentially responsible for possible phenotypic differences between the strains. Specific attention was given to genes involved in O-antigen biosynthesis and regulation, as these are known to influence virulence. The genes involved in O-antigen biosynthesis for S. Infantis was identified by performing a nucleotide BLAST between the Seroseq serotype determinant database (rfb cluster database) from SeqSero2 (http://www.denglab.info/SeqSero2 [16], accessed date: 10 September 2024) and S. Infantis reference genomes on NCBI database. Additional O-antigen biosynthesis genes were identified through annotation. The identified genes were compared with those found in Sal_smooth to Sal_rough genomes. Wzz(fepE) mutation was detected by reference mapping Sal_smooth to Sal_rough genome with BWA/MEM v2.2.1 (https://github.com/bwa-mem2/bwa-mem2 [17], accessed date: 10 September 2024), variant calling was performed using snippy v4.6.0 (https://github.com/tseemann/snippy [18], accessed date: 10 September 2024), and variant visualization was conducted with Integrated Genome Viewer (IGV) v2.17.3 (https://igv.org/doc/desktop/#DownloadPage/ [19], accessed date: 10 September 2024). Confirmation of mutation was achieved through NCBI BLAST (http://www.ncbi.nlm.nih.gov/BLAST, accessed date: 10 September 2024) and additional annotation.

2.4. PCR and Sanger Sequencing

To further confirm the frameshift mutation in the Wzz(fepE) gene of Sal_rough, Sanger sequencing was performed. Genomic DNA was extracted from the culture of Sal_rough, its 20th days of serial passages variant, and Sal_smooth strain using Dneasy Blood and Tissue Kit (Qiagen, Germantown, MD, USA) following the manufacturer’s instructions. PCR targeting the mutation in the O-antigen length regulator (Wzz(fepE)) gene of the S. Infantis rough strain was conducted on a Roche light cycler 480 II system (Roche Molecular Biochemicals, Indianapolis, IN, USA). The primers (synthesized by Integrated DNA Technologies, Coralville, IA, USA) are as follows: forward primer: 5′-AAA CAG ATT AAA TAC GCT GGC CCG A 3′; reverse primer: 5′-GGC GCG TAA AGA TTG TTT CGG ATA A-3′. The products of Wzz(fepE) PCR were sent to ELIM Biopharmaceuticals (Hayward, CA, USA) for Bidirectional Sanger sequencing. Chromatograms were analyzed and visualized using Unipro UGENE v50.0 (https://ugene.net/, accessed date: 10 September 2024).

2.5. Chicken Embryo Lethality Assay

The virulence of S. Infantis isolates was assessed using a modified chicken embryo lethality assay [20]. SPF-embryonated eggs (AVSBio, Norwich, CT, USA) were incubated at 37 °C with 65% humidity until day 11. On day 12, fertile eggs were marked and divided into three groups: two experimental (12 eggs each) and one negative control (10 eggs) groups. Experimental groups included Sal_smooth, Sal_rough, and its 20th days of serial passages variant. The S. Infantis isolates were cultured overnight in LB broth, and the inoculum concentration was adjusted to 10³ CFU/mL. After surface sterilization with 70% ethanol, eggs were inoculated with 100 µL of bacterial suspension or sterile Dulbecco’s phosphate-buffered saline (PBS) as negative control into the allantoic cavity. Inoculation sites were sealed with paraffin wax. Eggs were incubated at 37 °C and candled daily to monitor embryo mortality until day 16 of development (5 days post-inoculation). Embryo death occurring within 24 h of inoculation was considered non-specific and attributed to lethal trauma caused by the inoculation. The experiment was carried out in two biological replicates to ensure the reproducibility of the results.

2.6. Statistical Analysis

Comparative genomics between isolates was analyzed using Mauve pairwise Locally Collinear Blocks (LCBs) to measure the genetic conservation. Embryo survival rates were compared between the isolates using Kaplan–Meier curves generated on GraphPad Prism v10.3.0 (461) (GraphPad Software Inc., Boston, MA, USA). Statistical analyses to compare groups were performed using SAS Studio v.3.81 (SAS Institute Inc., Cary, NC, USA), and differences were evaluated using the log-rank test. p values below 0.05 were considered statistically significant.

3. Results

3.1. Serotyping Confirms the Antigenic Formula of Sal_Smooth and Sal_Rough Strains

In silico serotyping confirmed the Sal_smooth strain as a typical antigenic formula, 7:r:1,5, and the Sal_rough maintained an antigenic formula of -:r:1,5. Both strains were identified as sequence type 32 (ST32), providing a foundation for comparative genomic analysis.

3.2. High Genomic Conservation Between Sal_Smooth and Sal_Rough Strains

Whole genome comparison revealed a high level of conservation between Sal_smooth and Sal_rough, with 10 homologous Locally Collinear Blocks (LCBs) identified, with a minimum weight of 3099. The genomes of both strains were approximately 4.7 million base pairs in length (Figure 1). However, plasmid analysis revealed that Sal_rough variant harbored a 217,909 bp contig with the IncFIB(pN55391) plasmid replicon and yersiniabactin virulence factors encoded by the fyuA, irp2-irp1-ybtUTE, ybtA, and ybtPQXS gene cluster, while no plasmid was detected in Sal_smooth strain.

3.3. Frameshift Mutation in Wzz(fepE) Gene of Sal_Rough Strain

The analysis of the O-antigen biosynthesis genes in S. Infantis revealed that the region between the galF and gndA genes exhibited 99.6% sequence identity (9352 out of 9389 nucleotides) with the O-antigen gene cluster of S. Choleraesuis (serogroup C1, O-7 antigen) from the Seroseq serotype determinant database. This region in S. Infantis included several genes: EpsG, three glycosyltransferase genes, mannose-1-phosphate guanylyltransferase/mannose-6-phosphate isomerase (manC), phosphomannomutase (manB), and a hypothetical protein. Additional genes involved in O-antigen biosynthesis were identified outside this cluster, including UDP-N-acetylglucosamine–undecaprenyl-phosphate, N-acetylglucosamine phosphotransferase (WecA), O-antigen ligase (rfaL), and O-antigen length regulators (WzzB, and Wzz(fepE)). Comparative analysis between the Sal_rough and Sal_smooth genomes revealed the loss of several O-antigen biosynthesis genes: EpsG and the glycosyltransferases within the cluster in both genomes.

The most notable difference was a frameshift mutation in the Wzz(fepE) gene of Sal_rough, caused by an adenine (A) insertion after nucleotide position 32, extending its length from 1137 bp to 1138 bp (Figure 2). This mutation resulted in an early stop codon at position 26 in the 378 amino acid Wzz(fepE) protein, causing truncation that renders it a pseudogene. Importantly, sequencing analysis further confirmed that the Wzz(fepE) mutation, characterized by the inserted adenine base, was present in Sal_rough but absent in Sal_smooth (Figure 2). Consequently, this mutation was unique to Sal_rough and remained stable over 20 days of serial passages.

3.4. Sal_Rough Demonstrates Attenuated Virulence in Chicken Embryo Model

The chicken embryo lethality assay revealed significantly higher embryo lethality rates in the Sal_smooth group compared to both the negative control (PBS) and Sal_rough (p < 0.05). Notably, no significant difference in embryo lethality was observed between Sal_rough and the negative control (p = 0.721) (Figure 3), confirming the attenuated virulence of the rough variant. Additionally, no difference in virulence was observed between the original Sal_rough and its 20th days of serial passages variant (p = 0.2388), suggesting stable attenuation across repeated passages and multiple bacterial generations.

4. Discussion

Salmonella enterica serovar Infantis is an important zoonotic pathogen with rising prevalence and virulence [21,22,23]. While studies have investigated the role of O-antigen in Salmonella virulence [7,8,9], little is known about its regulatory mechanism in S. Infantis virulence. In this study, genomic and phenotypic comparison between rough (Sal_rough) and smooth (Sal_smooth) S. Infantis strains revealed that though both strains showed loss of the same O-antigen biosynthesis genes, the most significant finding was a frameshift mutation in the Wzz(fepE) gene of the rough variant, resulting in a truncated, non-functional protein that likely disrupted very long O-antigen chain synthesis. This finding aligns with previous studies by Tran and Morona et al. [24] on FepE, a Wzz(fepE) homolog in Escherichia coli, where mutations led to shorter O-antigen chain lengths. The presence of Wzz(fepE) homologs across Salmonella serovars and E. coli suggests a conserved role in the Enterobacteriaceae family [7]. This alteration likely explains the incomplete antigenic profile (-:r:1,5) in Sal_rough and influences its virulence due to impaired resistance to the immune response [25,26].

The chicken embryo lethality assay results provided evidence for the impact of a non-functional Wzz(fepE) gene in the virulence characteristics of the rough variant. The significantly lower embryo lethality rates observed in the Sal_rough group compared to Sal_smooth, and the lack of difference between Sal_rough and the negative control, confirm the reduced virulence of the rough variant. These findings are consistent with previous reports by Bravo et al. [27], who suggested that Wzz(fepE) in non-typhoidal Salmonella serovars such as S. Typhimurium primarily confers protection against serum bactericidal activity. These authors also demonstrated that since S. Typhi lacks a functional Wzz(fepE), it was significantly less resistant to serum compared to S. Typhimurium. In S. Paratyphi, very long O-antigen chains also mediate immune evasion, though through a mechanism distinct from non-typhoidal Salmonella [28,29].

In an earlier study, Murray et al. [7] observed functional redundancy between O-antigen chain length regulators Wzz(fepE) and WzzB in S. Typhimurium C5 mutant, where either gene provided complement resistance and virulence in mice. However, although Sal_rough retained the WzzB gene, it showed reduced virulence, suggesting that functional Wzz(fepE) is critical for virulence in this strain. This aligns with another study by Murray et al. [8] showing that serum resistance in S. Typhimurium depends on Wzz(fepE).

Consequently, studies by Pescaretti et al. [30] and da Silva et al. [31] highlighted the essential role of very long O-antigen chains in serum resistance and bacterial survival within macrophages during Salmonella pathogenesis. Crawford et al. [32] found that strains with functional Wzz(fepE) showed higher fitness in mouse colitis models, as evidenced by increased fecal bacterial counts. This advantage was attributed to enhanced bile salt resistance during inflammation, indicating that Wzz(fepE)-mediated very-long-O-antigen chain production contributes to Salmonella virulence.

The rough variant’s attenuation attributed to the Wzz(fepE) mutation makes it a promising candidate for live attenuated vaccine development, aligning with approaches proposed by Galen and Curtiss [33] for using attenuated strains in vaccine production. Modifying the O-antigen length has shown potential for enhancing immunogenicity and addressing T-independent immune responses [34,35]. Han et al. [36] have demonstrated that shortening O-antigen length impacts Salmonella immunogenicity. Thus, targeting Wzz(fepE) could lead to effective vaccines against S. Infantis.

Plasmid analysis revealed the presence of the IncFIB(pN55391) plasmid replicon and associated yersiniabactin virulence factors in the Sal_rough strain but not in Sal_smooth. The IncFIB(pN55391) plasmid replicon, typical of the PESI-like megaplasmid, has been identified in S. Infantis strains including ST32 [37,38], where these strains showed enhanced fitness and bacterial virulence, supporting Salmonella survival in animal hosts [2]. However, its presence in the Sal_rough strain did not exhibit any additional virulence effects, raising questions about the overall role of IncFIB(pN55391)-encoded factors in Salmonella virulence. Further research is needed to understand how IncFIB-encoded factors contribute to Salmonella pathogenesis and fitness across different strains [39].

The chicken embryo lethality assay serves as a well-established model for assessing the pathogenicity and virulence of avian pathogens [40,41]. While valuable for studying virulence, incorporating diverse animal models would provide a more comprehensive understanding of the strain’s attenuated characteristics.

One limitation of this study is that the causal relationship between the mutation in the wzz(fepE) gene and reduced pathogenicity requires further confirmation. Future work will include biochemical assays to quantify O-antigen levels in both rough and smooth Salmonella Infantis strains, imaging studies to observe potential physical differences in bacterial surface structures, and Western blotting to assess variations in immune recognition. Similar studies could be conducted on other Salmonella serotypes, such as S. Typhimurium and S. Enteritidis.

5. Conclusions

This study identifies a frameshift mutation in the Wzz(fepE) gene leading to early protein truncation, which may reduce bacterial virulence by impacting O-antigen biosynthesis. The reduced bacterial virulence, likely due to impaired O-antigen biosynthesis, suggests that targeting Wzz(fepE) could be a promising approach for developing attenuated S. Infantis vaccines. These findings expand our understanding of virulence mechanisms in S. Infantis and may inform strategies for controlling its infections. Future studies will employ CRISPR/Cas9 gene editing to validate the role of Wzz(fepE) in S. Infantis virulence.