The Response of Murine Gut Microbiome in the Presence of Altered rpoS Gene of Klebsiella pneumoniae

Abstract

The murine model is invaluable for studying intricate interactions among gut microbes; hosts; and diseases. However; the impact of genetic variations in the murine microbiome; especially in disease contexts such as Klebsiella pneumoniae (Kp) infection; still needs to be explored. Kp; an opportunistic global pathogen; is becoming increasingly prevalent in regions like Asia; especially China. This study explored the role of the gut microbiota during Kp infection using mouse model; including wild-type and rpoS mutants of Kp138; KpC4; and KpE4 from human; maize; and ditch water; respectively. Under stress conditions; RpoS reconfigures global gene expression in bacteria; shifting the cells from active growth to survival mode. Our study examined notable differences in microbiome composition; finding that Lactobacillus and Klebsiella (particularly in WKp138) were the most abundant genera in mice guts at the genus level in all wild-type treated mice. In contrast; Firmicutes were predominant in the healthy control mice. Furthermore; Clostridium was the dominant genus in all mutants; mainly in ∆KpC4; and was absent in wild-type treated mice. Differential abundance analysis identified that these candidate taxa potentially influence disease progression and pathogen virulence. Functional prediction analysis showed that most bacterial groups were functionally involved in biosynthesis; precursor metabolites; degradation; energy generation; and metabolic cluster formation. These findings challenge the conventional understanding and highlight the need for nuanced interpretations in murine studies. Additionally; this study sheds light on microbiome–immune interactions in K. pneumoniae infection and proposes new potential therapeutic strategies.

1. Introduction

The gut microbiota substantially impacts the gut in sustaining the host health. Infectious interruptions to the balanced gut microbiome have been associated with inflammatory, metabolic, and autoimmune diseases and can even lead to cancer, mental illness, and developmental disorders [1,2]. In recent years, numerous studies have been conducted to investigate which factors impact the gut microbiota, such as antibiotics, diet, health, nutrition, and age [3]. Klebsiella pneumoniae is resistant to nearly all known antibiotics used for treatment and therefore poses the most significant and severe risks to human health. Drug-resistant strains of K. pneumoniae are associated with extremely high lethality rates, where bloodstream and other multiple infections often result in mortality rates of 50% [4]. In this regard, it is a significant challenge to determine an effective clinical treatment. A clinical study found that 22.8% of K. pneumoniae isolates were related to invasive infections [5], potentially indicating the upcoming crisis and challenges in clinical infection therapeutics [6]. As a result, this emphasizes the significance of elucidating the precise relationships and mechanisms of virulence factors and antimicrobial resistance in virulent pathogenic Klebsiella isolates.

Research on gut microbiomes is dominated by murine models due to their low costs, higher rate of reproduction, relativity close resemblance to the human gut system, and ease of experimental manipulation. The disturbance in the microbiome due to pathogen intervention requires investigating the possible cause and relationships between dysbiosis and disease outcomes. Cross-study comparisons can be complicated due to various factors that affect the diversity of the mouse gut microbiome, especially when the alterations in the rpoS gene are infrequent [7,8].

RpoS is a crucial stress response regulator, serving as an alternate sigma factor of RNA polymerase, primarily present in Gammaproteobacteria [9,10]. In various pathogens, RpoS is reported to play a crucial role in cell survival under stressful conditions, such as oxidative stress and exposure to acids [11,12,13]. However, its function in the pathogenesis and impact on virulence may vary, even within the same species [11,12,13,14]. Under harsh conditions, RpoS modifies the comprehensive gene expression of bacteria, causing the cells to shift from an active growth state to the survival stage [15]. Additionally, it has also been demonstrated to promote bacterial colonization in the gut. For instance, RpoS is necessary for various pathogenic bacterial species to effectively colonize the intestine, such as Salmonella enterica serovar Typhimurium, Vibrio cholerae, and the highly pathogenic Escherichia. coli [16,17]. To explore bacterial challenges, we investigated the role of RpoS in gut colonization in mice. The results showed that the absence of RpoS reduces Klebsiella’s capacity to colonize the mouse gut, while this pattern was not followed in wild-type Kp strains (Kp138, KpE4, and KpC4).

This study was designed to compare the impact of the RpoS-mutant and wild-type (WT) effects on microbiome variability within the framework of a disease progression in a mouse model. We investigated how comparative effects influence the variability in the gut microbiome of albino laboratory mice, and we found notable differences in the microbiome composition based on the mutation. The microbiome variations in the wild-type (WT) and Mutant-treated mice were not identical. However, differential abundance analysis identified several operational taxonomic units and conserved candidate taxa that could be involved in developing progression or reducing illness.

Our research aimed to explore the microbial diversity of healthy and diseased mice, examining the differences in the relative abundance, phylogenetic diversity, and species richness of the microbial communities in symptomatic mice inoculated with the Klebsiella pathogen. Our findings indicated that infecting mice with K. pneumoniae significantly altered the healthy gut microbiota and that induced gut illness resulted in evident changes in certain bacterial species. Understanding bacterial colonization can help scientists study how bacteria interact with the gut and develop new strategies to prevent harmful bacteria from colonizing or causing infection.

2. Results

2.1. Most Dominant Taxa in Mice in the Presence of Wild-Type and Mutant of K. pneumoniae

The bacterial microbiome from the gut of healthy, mutant, and wild-type K. pneumoniae-treated mice was sequenced. The QIIME cutadapt tool was used for pair-end reads. The excision sequence was selected through the primer fragment, and the mismatched primers were removed. The QIIME DADA2 denoised function for paired-end data was used for quality control by the DADA2 algorithm, and the following steps of de-dimerization, denoising splicing, and chimerization removal were performed. The box plot results are shown in Figure 1. The one-way analysis of variance (ANOVA) results showed significantly different (p < 0.001) alpha diversity indices across all strains of K. pneumoniae, i.e., healthy, wild, and mutant. Alpha diversity analysis was carried out to examine the evenness and richness within a single microbial ecosystem of species.

Seven samples of healthy and diseased mice with three replications were used to check alpha diversity. The results showed that alpha diversity indexes of OTU level with Chao, Shannon, Simpson, observed species, Faith, Pielou, and Goods coverage or gut samples of healthy mice showed higher values compared to diseased mice. In brief, the alpha diversity of healthy mice samples significantly differs from that of diseased mice. The Shannon (4.68), Simpson (0.05), Faith (704.77), and Chao (717.33) indices of healthy mice show significant differences from those of diseased mice. The rarefaction curves of all samples nearly hit the saturation peak (Figure 2A), showing that the sequenced findings adequately reflect the variety present in the samples. The quantity of new unknown amplicon sequence variations (ASVs) did not rise as the sequence depth increased. The SPN and SPD were equal and homogeneous, but the SNPN and SOPD exhibited a considerable variance in abundances across OTUs and poor homogeneity.

2.2. Dominant Microbiota in the Healthy Mice

The abundance of bacterial community at the phylum level in the gut microbiome was observed in healthy control albino mice. After one week of the experiment, Firmicutes showed the maximum comparative abundance of 97.43%, followed by Proteobacteria (1.52%). However, maximum values were observed at the class level for Bacilli, Gammaproteobacteria, and Clostridia (96.97%, 0.36%, and 0.46%). Furthermore, at the order level, maximum comparative abundance was observed for Lactobacillales (96.59%), while Enterobacteriales were absent in abundance. Lactobacillaceae (70.04%) and the absence of Enterobacteriaceae were observed, whereas Streptococcaceae was only found (26.23%) at the family level. Lactobacillus (70.02%) and Klebsiella were absent, while Streptococcus was only 26.14% in the control at the genus level. However, at the species level, the maximum relative abundance of Lactobacillus reuteri in the control (17.33%) was observed (Figure 3A–F).

2.3. Presence of Dominant Microbiota in WT Kp-Treated Mice

After one week of the application, maximum comparative abundance was observed for Firmicutes (95.46%), followed by Proteobacteria (4.13%) with WT K. pneumoniae KpC4 application. Similarly, WT KpE4 accounted for 59.81% of Firmicutes and 35.39% of Proteobacteria, and for Kp138, it was 32.41% and 66.83%, respectively. It was also observed that Bacteroidetes and Actinobacteria were found in WT KpE4, accounting for 2.20% and 2.17%, respectively. At the class level, Bacilli, Gammaproteobacteria, and Clostridia were observed in WT KpC4 (91.98%, 3.61%, and 3.48%, respectively), WT KpE4 (56.46%, 30.13%, and 1.51%, respectively), WT Kp138 (30.22%, 66.10%, and 1.67%, respectively). Gammaprotobacteria abundance was found to be the highest in WT Kp138 (66.10%), whereas the lowest abundance was recorded for actinobacteria and Betaprotobacteria in WT KpE4 (2.13% and 2.69%, respectively). At the order level, a maximum abundance of Lactobacillales and Enterobacteriales order was observed in WT KpC4 (91.90% and 3.34%, respectively), WT KpE4 (53.31% and 27.31%, respectively), and WT Kp138 (29.95% and 65.55%, respectively). At the family level, maximum comparative abundance was recorded for Lactobacillaceae in WT KpC4 (91.47%), whereas the minimum was observed in WT KpE4 (22.11%). Similarly, the Enterobacteriaceae family was observed in WT Kp138 in the range of 65.55%. Furthermore, Streptococcaceae was exclusively detected in WT KpE4 (29.84%) and the control group. At the genus level, Lactobacillus abundance was 91.45% in WT KpC4, while a minimum of 22.11% was found in WT KpE4. The maximum relative abundance of Klebsiella was observed in the WT138 (55.25%) group. However, the species level showed a maximum relative abundance of Lactobacillus reuteri in WT KpC4 (22.89%) and a minimum in WT KpE4 (2.01%). Meanwhile, the relative abundance of Lactococcus garvieae was found in a range of 28.53% only in WT KpE4, as illustrated in Figure 3A–F.

2.4. Role of Microbiota in the Mutant Treated Mice

The relative abundance of the microbial community composition at the phyla level in the gut microbiome was observed in albino mice infected by mutant strains of K. pneumoniae (∆Kp138, ∆KpE4, and ∆KpC4). The maximum comparative abundance of Firmicutes and Proteobacteria phylum was observed with the application of mutant-type ∆KpC4 (85.91% and 13.02), ∆KpE4 (66.72% and 30.81%), and ∆Kp138 (88.23% and 8.28%), respectively. It was also observed that the maximum Bacteroidetes phylum was present in ∆Kp138 (2.44%). At the class level, the gut microbiome showed a maximum comparative abundance of Bacilli, Gammaproteobacteria, and Clostridia in ∆KpC4 (62.52%, 11.00%, and 23.20%, respectively). Similarly, Bacilli, Gammaproteobacteria, and Clostridia were present in ∆KpE4- (42.28%, 28.67%, and 23.95%) and ∆Kp138 (74.25%, 2.39%, and 13.69%)-treated mice. It was also observed that the maximum Alphaproteobacteria class was present in ∆Kp138 (5.19%). At the order level, the relative abundance of Lactobacillales and Enterobacteriales was observed in ∆KpC4 (62.06% and 10.15%, respectively), ∆KpE4 (35.26% and 27.63%, respectively), and ∆Kp138 (73.17% and 1.88%, respectively). Clostridiales abundance was the maximum in ∆KpE4 (23.95%), whereas Turicibacterales abundance was 6.72% exclusively in ∆KpE4. Similarly, Rickettsiales was only present in ∆Kp138 (4.75%). At the family level, the application of ∆Kp138 resulted in the lowest taxa of Enterobacteriaceae (1.88%). Additionally, Turicibacteraceae was mainly found in ∆KpE4, comprising 6.72% of the total observed taxa. At the genus level, a maximum abundance of Klebsiella (1.26%) was found in ∆Kp138. However, the species level showed a maximum relative abundance of Clostridium perfringens in ∆KpC4 (21.46%), as shown in Figure 3A–F. The Spearman’s rank correlation between Lactobacillus and Klebsiella was positive, suggesting a strong association between these genera. Nodes represent the ASV/OTU in the sample, and the node size is proportional to its abundance (in Log2 (CPM/n)), while the module with the most nodes in the top 10 nodes is identified by different colors (Figure 4A). The 20 most predominant with the highest correlations bacterial networks are shown in (Figure 4B). The first column is the name of the taxon, and the second to the last column is the relative abundance of the corresponding taxa in each sample of the grouping scheme. The samples are grouped according to the Euclidean distance between their species compositions.

2.5. Microbial Communities’ Comparisons in Healthy, Wild and Mutant Mice

The WT-treated mice guts exhibited a greater microbial abundance of Klebsiella than those of the mutant and healthy mice guts. Very few variations among the bacterial communities of all treated mice were observed. The top abundant genera were Lactobacillus, Klebsiella, and other bacteria (Figure 3). The gut microbiome of albino mice affected by K. pneumoniae displayed notable changes in microbial community composition at the phyla level. Firmicutes and Proteobacteria were most abundant, particularly in WT KpC4 (95.46% and 4.13%) compared to ∆KpC4 (85.91% and 13.02%). Bacteroidetes predominated in ∆Kp138 (2.44%), while in WT KpE4, Bacteroidetes and Actinobacteria were present at 2.20% and 2.17%, respectively. At the class level, Bacilli, Gammaproteobacteria, and Clostridia were prominent. WT KpC4 exhibited the highest levels (91.98%, 3.61%, and 3.48%) compared to mutants.

Alphaproteobacteria was notably elevated in ∆Kp138 (5.19%), while Actinobacteria and Betaproteobacteria were prominent in WT KpE4 at 2.13% and 2.69%, respectively. Analysis at the order level revealed a high abundance of Lactobacillales and Enterobacteriales, particularly in WT KpC4 (91.90% and 3.34%). WT KpE4 displayed dominance of Clostridiales, especially in mutants (23.95%), while Turicibacterales were exclusive to ∆KpE4 (6.72%), and Rickettsiales only present in ∆Kp138 (4.75%). Family-level analysis highlighted Lactobacillaceae predominance in WTC4 (91.47%) and minimal presence in WT KpE4 (22.11%).

Enterobacteriaceae were notably abundant in WT Kp138 (65.55%) compared to mutant treatments. Streptococcaceae were found in control and WT KpE4 (26.23% and 29.84%, respectively), while Turicibacteraceae were exclusive to ∆KpE4 (6.72%). At the genus level, Lactobacillus was highly abundant in WT KpC4 (91.45%) and the least in WT KpE4 (22.11%). Klebsiella was predominant in WT Kp138 (55.25%) compared to mutants. Streptococcus was only present in the control group. Further analysis at the species level revealed Lactobacillus reuteri dominance in WT KpC4 (22.89%), and Lactococcus garvieae was exclusively present in WT KpE4 (28.53%) (Figure 3A–F).

A community composition study was conducted across the samples to identify more species groups and their comparisons using heat maps, with the 20 most abundant genera chosen (Figure 4B). The categorization hierarchy tree depicts the hierarchical connections between all taxa in the sample population, ranging from phylum to genus. The node size is proportional to the average relative abundance of the taxa. The top 20 taxa by relative abundance are also designated by letters in the picture, with the shadow color matching the color of the relevant node. Samples were first grouped based on their components’ similarity and then sorted horizontally and vertically by the clustering findings. Similarly, the classification unit was grouped according to the degree of similarity spread across different samples and placed vertically based on the cluster findings.

We used principal coordinate analysis (PCoA) based on the Bray–Curtis distance and weighted UniFrac distance to analyze the community similarity among samples. The PCoA scatter plot highlights the two characteristics of sample coordinates that accounted for the most significant differences, with contribution impacts of 67.4 and 11%, respectively, depending on the Bray–Curtis distance. This indicated that the gut bacterial community is different in WT compared to mutant and healthy mice (Figure 5B).

A Venn diagram was utilized to identify common and distinct OTUs across diverse variables, such as sample type. This graphic illustration depicted the overlap of operational taxonomic units (ASVs) between the WT and rpoS mutant bacteria in mice (Figure 5A). In the figure, each color block represents a group, the overlapping area between the color blocks indicates the ASV/OTU shared by the corresponding groups, and the number of each block indicates the number of ASV/OTU contained in the block. Specifically, 21 shared ASVs were observed, comprising the bacteria in the mice’s gut. According to the Venn diagram, out of 2239 OTUs, total mice samples showed SCK (101), S1 (144), S2 (341), S3 (106), S4 (224), S5 (701), and S6 (622) OTU levels.

2.6. Functional Predictions of Bacterial Taxa in Healthy, Wild, and Mutant-Treated Mice

Clustering was used to distinguish between high- and low-abundance taxa, with color gradients representing the similarity in community composition throughout the samples. The results from hierarchical clustering showed that Lactobacillus and Protobacteria were the most abundant taxa at the genus level in the mice gut (Figure 6A). Compared to the respective mutants, wild strains (WT Kp138 and WT KpC4) exhibited a higher abundance of these bacteria when compared to the control group. Mutant ∆KpE4 also exhibited a significant presence of Protobacteria, whereas Clostridium was predominant in mutant ∆KpC4. Hence, Clostridium has emerged as the second most abundant bacteria after Klebsiella.

To examine the significant and notable differences in biomarkers among microbial communities and to distinguish genera across groups, linear discriminant analysis effect size (LEfSe) was used as a standard (LDA) value > 2 (Figure 6B). The control or healthy group exhibited the most remarkable bacterial taxonomic diversity (LDA > 2), featuring three genera primarily from Lactobacillus and Firmicutes (Bacilli). The WT KpC4 group had the second-highest diversity, featuring two genera primarily from Lactobacillaceae and Lactobacillus. The WT KpE4 group showed the lowest diversity (LDA > 2), with only one genus (Lactococcus). LEfSe was also utilized to elucidate significant biomarker differences across the samples.

To shed light on microorganisms roles in the gut of experimental mice groups and to distinguish genera across groups, PPICRUSt2 (system genetic investigation of communities by reconstruction of unobserved states) was utilized for predicting the metabolic activity of the microbial community on the KEGG. (https://www.kegg.jp/) and MetaCyc (https://metac.yc.org/) databases. Differential metabolic pathways across the groups were screened based on logFC and adj p values, and the species makeup of metabolic pathways was examined (Figure 7A). Most bacterial taxa were functionally linked to precursor metabolite, degradation/utilization/assimilation, biosynthesis, energy production, and the establishment of metabolic clusters.

Within the biosynthesis category, nucleoside and nucleotide synthesis were most abundant (35,815.31%). Other biosynthetic categories include amino acid (26,836.14%); lipid and fatty acid (19,413%); carbohydrate (9201.28%); cell structure (10,743.36%); secondary metabolite (4990.92%); aromatic compound (2260.35%); other compound (621.22%); and the biosynthesis of amine and polyamine (1043.42%), aminoacyl-tRNA charging (995.83%), and metabolic regulator biosynthesis (426.17%). Within the degradation/utilization/assimilation category, the most abundant process was found in the degradation of carbohydrates (8605.62%). This was followed by carboxylate degradation (7705.94%), nucleoside and nucleotide (6077.32%), aromatic compounds (5411.78%), secondary metabolite (5334.73%), and amino acid (2436.08%) degradation; C1 compound assimilation and utilization (2116.49%); inorganic nutrient metabolism (1985.95%); amine and polyamine (1360.67%), polymeric compound (1587.09%), alcohol degradation (1199.8%), and fatty acid and lipid degradation (827.98%) (Figure 7B).

3. Discussion

Culturing and traditional isolation techniques have yielded limited insights into Klebsiella infection in the mice gut. In this study, we described the microbial community in healthy, wild, and mutant Kp-treated mice. We used Illumina sequencing to describe the changes in microbial communities during infection under different bacterial strains and host immune response against disease. Firmicutes were the top abundant bacterial phylum in the gut of healthy, wild, and Mutant-treated mice except for clinical wild strain Kp138.

The results align with the earlier research on the microbiota in the liver [18]. The rpoS mutation in the present study significantly impacted the bacterial diversity index in mice. According to Chao1, Simpson, observed-species, and Shannon values, the virulence of distinct Klebsiella strains was accompanied by variations in the microbial community richness. The gut microbiota is crucial for preserving human health, and new research indicates a link between intestinal microbial dysbiosis and long-term illnesses, including chronic hepatitis and diabetes [19,20,21]. Furthermore, the gut microbiota is thought to be an extended host defense mechanism that strengthens the body defenses against pathogen invasion and disease. The mechanism controlling the dynamic interactions between pathogenic bacteria and microbiota is yet fragmented. Nevertheless, in the current investigation, we explained how healthy mice’s natural microbiota was affected by the opportunistic pathogen K. pneumoniae and its variants. Our results suggest that the K. pneumoniae WT Kp138, WT KpE4, and WT KpC4 isolates can significantly alter the microbiota composition in healthy mice, leading to gut abnormalities. Through comparative analysis of the microbiota compositions between healthy and infected mice, we also found evidence that the alterations in particular bacterial species in the gut correlate with infection.

We discovered that inoculation with K. pneumoniae by oral gavage (Kp138, KpE4, KpC4, and their respective mutants) altered the composition of gut microbiota in both mice groups that were infected with wild-type and mutant, respectively, compared to the healthy mouse group. Still, a direct comparison of the gut microbiota of WT and Mutant-treated mice revealed particular changes linked to the gut infection. Most notably, the gut infection was linked to a significant reduction in the Lactobacillus spp., an abundance that is a crucial part of the gut microbiota and a member of the Lactobacillales within the Bacilli class of Firmicutes [22]. Numerous investigations have further shown the vital role of Lactobacillus in host health. Some strains, such as Lactobacillus rhamnosus, have been isolated and used as probiotics [23,24]. We speculate that the significant decline in Lactobacillus abundance may be closely associated with a gut infection caused by Klebsiella due to the critical function that Lactobacillus plays. It has been shown that enteropathogenic bacteria, such as Salmonella typhimurium and enterotoxigenic Escherichia coli, are less virulent when pro-biotic Lactobacilli are present [25,26,27]. In our study, wild-type K. pneumoniae-treated mice exhibited higher disease infection and were accompanied by a decreased abundance of Lactobacillus.

Conversely, mice treated with the mutant K. pneumoniae did not develop severe Klebsiella infection or show a reduction in Lactobacillus. Healthy mice gut possessed a higher abundance of Lactobacillus and a small fraction of Protobacteria, which proved that Klebsiella is present harmlessly in the mice gut and plants as an endophyte https://medicalxpress.com/news/2024-03-klebsiella-pneumoniae-opportunistic-pathogen-harmless.html [28]. When K. pneumoniae isolates produce a gut infection, Lactobacillus may also be a decisive factor in disease outcome. However, aside from Lactobacillus, K. pneumoniae isolates-induced changes in microbiota composition may play a role in the pathogenicity of Kp infection.

The study offers novel insights into the microbial community structure, providing a detailed exploration of bacterial diversity and abundance in the gut during Klebsiella infection. It lays a theoretical foundation for developing innovative disease management strategies. Using Illumina-based sequencing, our research signifies the initial characterization of bacterial diversity, abundance, and functional predictions in the gut microbiota of healthy, wild-type and mutant Kp-treated mice. Significant variations in alpha diversity indices were recorded across all mouse groups under investigation. Further experiments are warranted to elucidate the specific functions of different bacteria in the gut microbiome. In summary, our research shows that a K. pneumoniae gut infection causes alterations in the gut microbiota in healthy mice and, most notably, a decrease in Lactobacillus. This suggests that alteration in microbiota compositions is an essential pathogenic mechanism underlying gut infection.

4. Materials and Methods

4.1. Ethical Statement

The Animal Care and Use Committee of Yunnan Agricultural University (GB14925–2010), Kunming, China, approved all of the animal experiments.

4.2. Mice Collection

Kunming mice (Mus musculus albus), aged 6 to 8 weeks and weighing 20 ± 2 g, were acquired from Kunming Medical University, Kunming, China. The management of the mice was conducted in strict adherence to the guidelines outlined in the “Experimental Animal Environment and Facilities” (National Standard GB 14925-2010) and Yunnan Provincial Regulations on the Administration of Laboratory Animals (No. 59). The experiment was carried out after a 3–4 days acclimation period following the mice purchase. The mice were provided with food and water and were placed individually in aerated cages upon their transfer to the Biosafety Level 2 laboratory, where the infectious experiments were conducted.

4.3. Experimental Setup

This study was conducted at the State Key Laboratory for Conservation and Utilization of Bioresources in Yunnan, Kunming, China. Three wild-type (WT) strains of K. pneumoniae (Kp138 from humans, KpC4 from maize, and KpE4 from ditch water environment) were isolated and stored at −80 °C in the “State Key Laboratory for Conservation & Utilization of Bioresources in Yunnan Agricultural University, Kunming, China.

Mutants of all three strains were constructed using the allelic-exchange method, where homologous recombination integrated the suicide plasmid psr47S containing a 1.5-kbp fragment of the rpoS gene into the chromosomal rpoS gene of K. pneumoniae through conjugation. This results in a single-crossover mutant with antibiotic resistance. A counter-selection process using sucrose eliminates the vector’s selective markers, forming unmarked double-crossover mutants. This two-step process produces seamless mutations with precision to a single DNA base pair [29]. Before the experiment, 100 μL samples from every six bacterial strains were taken from the preserved sample at −80 °C and cultivated in LB broth at 37 °C with 180 rpm for 24 h. A spectrophotometer was used to monitor the cell growth, and suspension of six bacteria strains was adjusted to a concentration of 1 × 10⁸ CFU/mL, which was then cultured in 5 mL of Luria Bertani (LB) medium. A total of 21 mice were used in this experiment, and seven treatments, including a control, were applied. Three mice were used for each treatment (n = 3). Mice were dissected on day 13 post-infection [30].

4.4. Gut-Colonization Model

In the first step of the experiment, the mice were administered a 100 μL dose of (K. pneumoniae) wild-type and mutant inoculum via oral gavage slowly to avoid choking. To detect the signs of distress, animals were routinely checked for their weight and other visible signs. The mice infected with the pathogen exhibited different moderate to severe symptoms of distress, such as hunched body posture, separation from the group, inactivity, fur piloerection, asthma, and weight loss, which were recorded during the experiment. The mice guts were carefully incised and examined visually to assess any damage.

4.5. Microbiome Analysis in the Presence of Wild-Type and Mutant Strain

The incised gut was carefully collected, washed with ddH₂O twice, kept in sterile zipper bags, and stored at −80 °C. The OMEGA Soil DNA Kit (M5635-02) from Omega Bio-Tek (Norcross, GA, USA) was used to extract genomic DNA from all samples according to the manufacturer’s instructions. The extracted DNA samples were kept at −20 °C until further procedure. The concentration and purity of the extracted DNA were determined using a NanoDrop NC2000 spectrophotometer from Thermo Fisher Scientific (Waltham, MA, USA), and the DNA’s quality was evaluated using agarose gel electrophoresis. The libraries were produced in batches, measured, and sequenced on an Illumina Platform using a 500 bp read length (2 × 250 bp). Amplicon identification and quantification evaluations were analyzed using R packages (v3.2.0). Initially, sequence quality was assessed using FASTQC software (v.0.11.9) (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ accessed on 19 July 2023). Subsequently, DADA2 [31] was employed to generate amplicon sequence variants (ASVs). Briefly, sequences underwent visual inspection before applying filtering and trimming functions, followed by error modeling for denoising and merging the reads using default parameters.

4.6. 16S rRNA Gene Amplicon Sequencing

The V3-V4 region of the bacterial 16S rRNA genes was amplified by PCR using the forward primer 338F (5′-ACTCCTACGGGAGGCAGCA-3′) and reverse primer 806R (5′-GGACTACHVGGGTWTCTAAT-3′). The primers were modified with sample-specific 7 bp barcodes to facilitate multiplex sequencing. The PCR mixture consisted of 5 μL of buffer, 2 μL of dNTPs (2.5 mM), 0.25 μL of Fast pfu DNA Polymerase (5 U/μL), 1 μL each of forward and reverse primers (10 μM), 14.75 μL of ddH₂O, and 1 μL of DNA template. The thermal cycling technique was performed with a 5 min denaturation at 98 °C, followed by 25 cycles of denaturation for 30 s, annealing at 53 °C for 30 s, and extension at 72 °C for 45 s. The amplification was determined by a final extension step of 5 min at 72 °C. The PCR amplicons were purified using Vazyme VAHTSTM DNA Clean Beads (Vazyme, Nanjing, China) and quantified with the Quant-iT PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, CA, USA). Following individual measurements, the amplicons were pooled in equimolar amounts. Pair-end sequencing with a read length of 2 × 250 bp was performed using the Illumina NovaSeq technology and the NovaSeq 6000 SP Reagent Kit (500 cycles) at Shanghai Personal Biotechnology Co., Ltd, Shanghai, China.

4.7. Sequence Analysis

Microbiome bioinformatics analyses were carried out using QIIME2 2022.11 [32], with minor alterations based on the official tutorials at https://docs.qiime2.org/2022.11/tutorials/. In brief, the raw sequence data underwent demultiplexing using the Demux plugin, followed by primer trimming using the Cutadapt plugin [33]. Sequences were then quality filtered, combined, denoised, and chimeras eliminated using the DADA2 plugin [31]. Amplicon sequence variations (ASVs) that were nonsingletons were aligned using MAFFT [34] and used to create a phylogenetic tree using FastTree2 [35].

4.8. Taxonomic Composition Analysis

The classify-sklearn naive Bayes taxonomy classifier within the feature-classifier plugin [36] was used to classify ASVs, employing the database provided by [37]. We analyzed the 10–20 most abundant taxa from each group using comprehensive data.

4.9. Alpha and Beta Diversity Index Analysis

Alpha-diversity metrics Chao1 [38], Shannon [39,40], Observed species, Faith’s PD [41], Simpson [42], Good’s coverage [43] and Pielou’s evenness [44], and beta diversity metrics (weighted UniFrac [45], Jaccard distance, Bray–Curtis dissimilarity, and unweighted UniFrac[46]), were calculated using the diversity plugin. Samples were rarefied to an equal number of sequences per sample.

QIIME2 and the R package (v3.2.0) were mainly used to analyze sequence information. QIIME2 generated ASV-level alpha diversity metrics such as the Chao1 richness estimator, Shannon diversity index, observed species, Faith’s PD, Pielou’s evenness, Simpson index, and Good’s coverage. These metrics were represented using box plots. In addition, sorted abundance curves at the ASV level were used to cR pRompare the evenness and richness of the samples. Beta diversity analysis investigated fundamental differences in bacterial communities within samples using Jaccard distance metrics [47], UniFrac distance metrics [45,46], and Bray–Curtis dissimilarity metrics [48]. These metrics were shown using principal coordinate analysis (PCoA) and hierarchical clustering with the unweighted pair-group approach with arithmetic means (UPGMA) [49]. In addition, principal component analysis (PCA) was performed using genus-level compositional profiles.

4.10. Analysis of Differentiation among Groupings at Each Taxonomic Level

PERMANOVA (permutational multivariate analysis of variance) [50], Permdisp [51], and ANOSIM (analysis of similarities) [52,53] were used with QIIME2 to determine the importance of variations in microbial community framework between groups. A Venn diagram was created utilizing the R package “VennDiagram” to depict shared and unique ASVs among samples or groups based on ASV occurrence, regardless of relative abundance [54]. In addition, LEfSe (linear discriminant analysis effect size) was used with default parameters to find differentially abundant taxa across groups [55].

4.11. Construction of Association Network and Prediction of Microbial Metabolic Functions

A co-occurrence association network analysis was conducted using SparCC analysis, with a pseudo-count value set to 10⁶. Utilizing random matrix theory-based approaches implemented in the R package RMThreshold, a correlation coefficient threshold of 70 was established. Employing the correlation coefficients, we built a co-occurrence network with nodes representing ASVs and edges representing correlations between these ASVs. The network was displayed using the R tools graph. Microbial function predictions were performed using PICRUSt2 (Phylogenetic Investigation of Communities by Re-construction of Unobserved States) [56], which used data from the MetaCyc (https://metacyc.org/) and KEGG (https://www.kegg.jp/) databases. Based on the projected results, different metabolic pathways across groups were discovered using adjusted p values, species composition analysis of metabolic pathways, and log fold changes (logFC) to predict the principal coordinate analysis (PCoA) at the operational level of the operational unit.