L-Arabinose inhibits Shiga toxin type 2-converting bacteriophage induction in Escherichia coli O157:H7

ABSTRACT

The pathogenicity of Escherichia coli (E. coli) O157:H7 is predominantly associated with Shiga toxin 2 (Stx2) that poses a huge threat to human and animal intestinal health. Production of Stx2 requires expression of stx2 gene, which is located in the genome of lambdoid Stx2 prophage. Growing evidence has implicated that many commonly consumed foods participate in the regulation of prophage induction. In this study, we aimed to explore whether specific dietary functional sugars could inhibit Stx2 prophage induction in E. coli O157:H7, thereby preventing Stx2 production and promoting intestinal health. We demonstrated that Stx2 prophage induction in E. coli O157:H7 was strongly inhibited by L-arabinose both in vitro and in a mouse model. Mechanistically, L-arabinose at doses of 9, 12, or 15 mM diminished RecA protein levels, a master mediator of the SOS response, contributing to reduced Stx2-converting phage induction. L-Arabinose inhibited quorum sensing and oxidative stress response, which are known as positive regulators of the SOS response and subsequent Stx2 phage production. Furthermore, L-arabinose impaired E. coli O157:H7 arginine transport and metabolism that were involved in producing Stx2 phage. Collectively, our results suggest that L-arabinose may be exploited as a novel Stx2 prophage induction inhibitor against E. coli O157:H7 infection.

KEYWORDS:

• L-Arabinose
• Escherichia coli O157:H7
• Shiga toxin type 2 phage
• arginine transport and metabolism
• intestinal health

Introduction

Escherichia coli (E. coli) O157:H7 is an important foodborne pathogen that can cause severe diarrhea, hemorrhagic colitis, and the life-threatening sequelae hemolytic uremic syndrome (HUS).¹ In the past decades, E. coli O157:H7 outbreaks have been frequently reported worldwide, which are associated with a variety of contaminated foods including salad, juice, and processed meat etc.^2–7

The high pathogenicity of E. coli O157:H7 requires Shiga toxin 2 (Stx2) production. Gene coding for Stx2 (stx2 gene) is located in Stx2 prophage that has integrated into bacterial genome.⁸ Induction of Stx2 prophage leads to phage DNA replication as well as subsequent production of Stx2 and formation of active phage.⁹ Of importance, progeny phage particles may infect intestinal commensal bacteria and convert nontoxic strains into virulent pathogens, potentially worsening intestinal disorders.¹⁰ Induction of Stx2 prophage is partly related to activated bacterial SOS response, which is controlled by RecA protein, resulting in the expression of phage antirepressor and self-cleavage of phage repressor cI.¹¹

Since a wide range of antibiotics have been shown to induce Stx2 phage production through the activation of the SOS response, treatment of E. coli O157:H7 infection still remains a considerable public health challenge.¹² An effective therapeutic approach is to develop Stx2 prophage induction inhibitors. For example, cinnamon oil, novel antioxidant compounds (CM092, CM032D, and CM3186B), phenethyl isothiocyanate are currently exploited for limitation of Stx2 phage lytic development in enterotoxigenic E. coli. ^13–15 In particular, some commonly consumed foods such as coffee Arabica, grapefruit seed extract, and cinnamon reduce phage particles in Bacteroides thetaiotaomicron, a common human gut bacteria, indicating the possible role of nutritional strategy in the regulation of prophage induction.¹⁶ L-Arabinose, also known as pectin sugar, has recently gained attention for therapeutic potentials in inflammatory bowel diseases.¹⁷ L-Arabinose was reported to exert anti-inflammatory effects on dextran sodium sulfate-induced colitis in mice.¹⁷ Notably, L-arabinose can improve intestinal health by manipulating the composition of gut microbiota.¹⁷ However, the impact of L-arabinose on Stx2 phage induction and how dietary inhibitors reduce the formation of phage particles is largely unexplored to date.

In this study, we compared abilities of several different dietary sugars on inducing Stx2 phage production and revealed that L-arabinose is a promising dietary prophage induction inhibitor against E. coli O157:H7 Stx2 phage production. The mechanisms underlying the regulation of Stx2 prophage induction by L-arabinose were also investigated.

Results

Screening of functional sugars that inhibit Stx2 prophage induction in E. coli O157:H7

We first assessed to what extent different sugar metabolisms affected bacterial densities and Stx2 phage production of E. coli O157:H7. We found that glucose and xylan promoted the growth of E. coli O157:H7 in comparision with the LB control group (p < 0.05) (Figure 1a; Supplementary Figure S1A), indicating that these sugars should be excluded from further consideration. However, D-xylose, L-arabinose, mannose, and arabinogalactan (AG) significantly decreased bacterial optical densities (p < 0.01) (Figure 1a; Supplementary Figure S1A). We next looked at Stx2 phage production by real-time quantitative PCR. E. coli O157:H7 produced more Stx2 phage (p < 0.05) in medium containing xylan, while glucose, L-arabinose, and arabinogalactan reduced Stx2 phage production (p < 0.01) (Figure 1b). To further confirm whether the inhibition of Stx2-converting phage lytic development contributed to reduced Stx2 phage production, we tested the effects of L-arabinose and arabinogalactan on E. coli O157:H7 growth and Stx2 phage production upon prophage induction provoked by the chemical agent mitomycin C under the minimum inhibitory concentration for typical E. coli (0.5 μg/mL).¹⁴ As expected, mitomycin C treatment significantly inhibited bacterial growth (p < 0.01) (Figure 1c; Supplementary Figure S1B) and increased phage production (p < 0.01) (Figure 1d). However, the addition of L-arabinose maintained E. coli O157:H7 growth (p < 0.01) (Figure 1c; Supplementary Figure S1B) and impaired Stx2 phage lytic development at 180 and 240 min (p < 0.01) (Figure 1d) in the presence of mitomycin C. No similar effects were observed in E. coli O157:H7 after treatment with arabinogalactan (Figure 1c,d; Supplementary Figure S1B). Thus, the above results prove that L-arabinose could reduce Stx2 phage production through inhibition of Stx2 prophage induction.

Figure 1. Effects of different sugars on E. coli O157:H7 growth and Stx2 phage production detected by real-time quantitative PCR Cell densities of E. coli O157:H7 (a) and Stx2 phage production (b) upon 24 h growth in LB medium supplemented with 1% glucose, D-xylose, L-arabinose, mannose, xylan, or arabinogalactan (AG). Effects of L-arabinose or AG on the growth of E. coli O157:H7 (c) and Stx2 phage production (d) upon prophage induction forced by 0.5 μg/mL mitomycin C (mitC). All data are presented as mean ± standard deviation and analyzed based on three biological replicates.

Different levels of L-arabinose inhibit Stx2 phage production in E. coli O157:H7

Then, we further tested the role of L-arabinose on inhibition of Stx2 phage production and cultured E. coli O157:H7 with various concentrations of L-arabinose (0–15 mM). While there were no obvious growth inhibition effects (p > 0.05) in 3 and 6 mM treatment groups, L-arabinose treatments at 9, 12, and 15 mM significantly reduced cell densities (p < 0.001) (Figure 2a). Different concentrations of L-arabinose significantly decreased Stx2 phage production (p < 0.01) (Figure 2b). Consistently, Stx2 protein production in culture supernatants was decreased by L-arabinose in a dose-dependent manner (Figure 2c). At 6, 9, 12, and 15 mM, the levels of Stx2 protein were significantly lower than the control group (p < 0.05) (Figure 2c). In contrast, stx2 mRNA expression exhibited noticeable different degrees of upregulation, especially for the 9 mM treatment group (p < 0.01) (Figure 2d). Thus, these data suggest that L-arabinose effectively reduced Stx2 phage production of E. coli O157:H7 in vitro.

Figure 2. L-Arabinose inhibits Stx2 phage production in E. coli O157:H7 Cell densities (a), Stx2 phage production (b), Stx2 protein production (c), and stx2 mRNA expression levels (d) of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose. All data are presented as mean ± standard deviation and analyzed based on three biological replicates.

L-Arabinose reduces Stx2 phage production and mitigates intestinal inflammation in a mouse model

To test whether dietary supplementation with L-arabinose can reduce Stx2 phage production, we used an E. coli O157:H7 infection mice model. The drinking water of mice was supplemented with L-arabinose (1 g/kg Body weight) for eight consecutive days (Figure 3a).¹⁸ At days 1 to 7, mice were gavaged daily with approximately 10⁷ CFU E. coli O157:H7. Results showed that supplementation with L-arabinose decreased mortality (Supplementary Table S1) and body weight loss (p < 0.01) (Figure 3b; Supplementary Figure S2), improved jejunal morphology shown as neatly arranged villi (Figure 3c), reduced colonal inflammatory cell infiltration (Figure 3d), alleviated kidney injury (i.e., mild acute tubular necrosis and lower serum urea concentration) (Supplementary Figure S3), and decreased the levels of proinflammatory cytokines (TNF-α, IL−1β, and IL−6) in the colon tissue (Figure 3e) after E. coli O157:H7 challenge. L-Arabinose treatment had no significant impact on fecal and intestinal E. coli O157:H7 levels (p > 0.05) (Supplementary Figures S4 and S5). However, L-arabinose significantly decreased fecal (Day 1: 7.24 × 10⁴ _Ara versus 8.38 × 10⁴ _{E. coli} copies/100 mg; Day 3: 2.98 × 10⁴ _Ara versus 3.52 × 10⁴ _{E. coli} copies/100 mg; Day 6: 3.28 × 10⁴ _Ara versus 4.21 × 10⁴ _{E. coli} copies/100 mg) (Figure 3f; Supplementary Table S2), jejunum (3.62 × 10⁴ _Ara versus 6.61 × 10⁴ _{E. coli} copies/100 mg) (Figure 3g; Supplementary Table S2), and colon (3.98 × 10⁴ _Ara versus 4.95 × 10⁴ _{E. coli} copies/100 mg) (Figure 3h; Supplementary Table S2) Stx2 phage production (p < 0.05). These findings prove that dietary L-arabinose reduced Stx2 phage production in the gastrointestinal tract and promoted intestinal health.

Figure 3. L-Arabinose reduces Stx2 phage production in mice infected with E. coli O157:H7 and promotes intestinal health Experimental timeline examining the effect of dietary L-arabinose (Ara) on Stx2 phage production in mice infected with E. coli O157:H7 (a). See “Materials and methods” for more details. Body weight change (b). Representative hematoxylin and eosin (H&E) staining images of jejunum (c) and colon (d). Expression levels of proinflammatory cytokines (TNF-α, IL − 1β, and IL − 6) in colon tissue (e). Percentage of Stx2 phage production in fecal (f), jejunum (g), and colon (h) samples of mice that were received L-arabinose as compared to E. coli O157:H7 challenge group (defined as 100%). Data in Figures 3b and 3e−3h are shown as mean ± standard deviation and analyzed based on seven biological replicates. nd: Not detected.

L-Arabinose impairs the SOS response of E. coli O157:H7

The induction of Stx2 prophage in E. coli O157:H7 requires RecA-dependent SOS response.⁸ Activated RecA protein induces the cleavage of the SOS regulon repressor LexA and phage repressor cI.¹⁹ To investigate whether the SOS response is involved in the reduction of Stx2 phage by L-arabinose, we measured the concentrations of RecA protein in E. coli O157:H7 cells following 24 h stationary incubation at 37°C. While L-arabinose promoted RecA protein production at 3 and 6 mM (p < 0.001) (Figure 4a), L-arabinose markedly reduced RecA protein production at 9, 12, and 15 mM and dose dependently reduced the expression levels of gene recA (p < 0.05) (Figure 4b; Supplementary Table S3), indicating down-regulated SOS response. Consistent with this observation, the transcription levels of the main SOS regulon genes, including lexA, sulA, umuC, umuD, ruvA, ruvB, and polB were down-regulated in E. coli O157:H7 cells treated with 9, 12, and 15 mM L-arabinose (Figure 4c-i; Supplementary Table S3). Thus, these data suggest that L-arabinose at doses of 9, 12, and 15 mM could reduce Stx2 phage production by impairing the SOS response of E. coli O157:H7.

Figure 4. L-Arabinose impairs the SOS response in E. coli O157:H7 after a 24-h interaction in LB broth RecA protein production of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose and representative Western blot image (a). recA (b), lexA (c), sulA (d) umuC (e), umuD (f), ruvA (g), ruvB (h), and polB (i) mRNA expression levels of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose. All data are presented as mean ± standard deviation and analyzed based on three biological replicates.

L-Arabinose inhibits quorum sensing (QS) and oxidative stress response known to be important positive regulators to control the SOS response of E. coli O157:H7

Activation of QS is conducive to trigger bacterial SOS response as well as subsequent Stx2 prophage induction via QseBC and LuxS.^20,21 The addition of different concentrations of L-arabinose greatly decreased the expression levels of qseB and qseC (p < 0.01) (Figure 5a,b). The high levels of L-arabinose treatments at 12 and 15 mM reduced luxS expression (p < 0.01), whereas neither 3, 6, nor 9 mM treatments affected luxS expression (p > 0.05) (Figure 5c). Given that QS is well known as a positive regulator of bacterial oxidative stress response, which in turn can enhance the SOS response and Stx2 phage production,^11,22 we further evaluated the effects of L-arabinose on several major oxidative stress regulons of E. coli, including oxyR, soxR, and sod.²³ L-Arabinose treatments reduced the mRNA expression of oxyR, soxR, and sod in E. coli O157:H7 (p < 0.05) (Figure 5d-F). Especially, treatments with L-arabinose at 12 and 15 mM almost completely inhibited the expression of oxyR and soxR (Figure 5d,e). Therefore, reduced QS and oxidative stress response may contribute to the inhibition of the SOS response by L-arabinose.

Figure 5. L-Arabinose inhibits quorum sensing and oxidative stress response in E. coli O157:H7 after a 24-h interaction in LB broth qseB (a), qseC (b), and luxS (c) mRNA expression levels of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose. oxyR (d), soxR (e), and sod (f) mRNA expression levels of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose. All data are presented as mean ± standard deviation and analyzed based on three biological replicates.

Genes responsible for arginine transport and metabolism were downregulated after L-arabinose treatments

L-Arabinose treatments at 3 and 6 mM enhanced RecA protein expression by approximately 2.6-fold and 1.6-fold, respectively (Figure 4a), but all exhibited decreased Stx2 phage production (Figure 2b), suggesting additional mechanisms were potentially involved in inhibiting Stx2 prophage induction. To explore this possibility, RNA-Seq was conducted to demonstrate transcriptional signature of L-arabinose metabolism by E. coli O157:H7. Since there was no significant difference on cell densities between the control group and 6 mM of L-arabinose treatment group (p > 0.05), and 15 mM of L-arabinose treatment effectively inhibited both E. coli O157:H7 growth and Stx2 phage production (Figure 2a,b), we tested the transcriptome profiles of the control, 6 mM, and 15 mM treatment groups at 24 h. PCA analysis showed that the separation trend was obvious among three groups (Figure 6a). A total of 1559 genes were found to be differentially expressed under 6 mM L-arabinose treatment, among which 843 were downregulated and 716 were upregulated (Figure 6b). However, 15 mM of L-arabinose treatment resulted in 2574 differentially expressed genes, among which 1310 were downregulated and 1264 were upregulated (Figure 6c).

Figure 6. Transcriptome analysis of L-arabinose metabolism by E. coli O157:H7 Transcriptome PCA analysis of the control group, 6 mM, and 15 mM of L-arabinose (Ara) treatment groups (a). Statistics of different expressed genes in E. coli O157:H7 after treatments with 6 mM (b) or 15 mM (c) of L-arabinose. COG function classification of 6 mM (d) and 15 mM (e) of L-arabinose treatment transcriptome. C: Energy production and conversion; D: Cell cycle control, cell division, chromosome partitioning; E: Amino acid transport and metabolism; F: Nucleotide transport and metabolism; G: Carbohydrate transport and metabolism; H: Coenzyme transport and metabolism; I: Lipid transport and metabolism; J: Translation, ribosomal structure and biogenesis; K: Transcription; L: Replication, recombination and repair; M: Cell wall/membrane/envelope biogenesis; N: Cell motility; O: Posttranslational modification, protein turnover, chaperones; P: Inorganic ion transport and metabolism; Q: Secondary metabolites biosynthesis, transport and catabolism; R: General function prediction only; S: Function unknown; T: Signal transduction mechanisms; U: Intracellular trafficking, secretion, and vesicular transport; V: Defense mechanisms; W: Extracellular structures; X: Mobilome. Volcano plots of downregulated genes involved in arginine transport and metabolism after 6 mM (f) and 15 mM (g) of L-arabinose treatments. Differentially expressed genes were selected based on the standard of |log₂(Fold change)| ≥ 1 and p value ˂ 0.05. All data are analyzed based on three biological replicates.

Previous studies have shown that reduced Stx2 phage production was universally accompanied with the inhibition of RNA synthesis.^13–15 We therefore focused on the genes downregulated in E. coli O157:H7. The Clusters of Orthologous Groups of proteins (COG) analysis showed that downregulated genes in 6 mM of L-arabinose treatment group were sequentially mainly related to energy production and conversion, amino acid transport and metabolism, carbohydrate transport and metabolism, and translation, ribosomal structure and biogenesis (Figure 6d), while 15 mM of L-arabinose treatment group mainly downregulated the expression of genes involved in amino acid transport and metabolism, carbohydrate transport and metabolism, transcription, and mobilome (Figure 6e), suggesting that amino acid transport and metabolism may play a potential role in regulating Stx2 prophage induction during L-arabinose treatments. We therefore further analyzed the transcription of genes involved in amino acid transport and metabolism in E. coli O157:H7. Intriguingly, we found that most of downregulated genes were responsible for arginine transport and metabolism, including arginine import (genes artJ1, artJ2, artP, artQ, and artM), arginine biosynthesis (genes argA, argB, argC, argH, and argF), and arginine catabolism (genes speA, adiA, adiC, astB, and astE) (Figure 6f,g, Red plot; Supplementary Tables S4 and S5).²⁴ Taken together, these results indicate that L-arabinose strongly decreased the expression of arginine transport and metabolism genes of E. coli O157:H7.

L-Arabinose inhibits Stx2 phage production by regulating arginine transport and metabolism in E. coli O157:H7

To validate the RNA-Seq transcriptome results, we first tested the intracellular arginine concentrations of E. coli O157:H7. We found that apart from 3 mM of L-arabinose treatment group, intracellular arginine levels in 6, 9, 12, and 15 mM of L-arabinose treatment groups were significantly reduced when compared to the control group at 24 h (p < 0.05) (Figure 7a). Partial least squares discriminant analysis (PLS-DA) further showed distinct bacterial metabolomic profiles among L-arabinose treatment groups and the control group (Figure 7b). The concentrations of spermidine, the main end-product of arginine metabolism, and N2-acetyl-L-ornithine, a critical arginine biosynthetic enzyme, were significantly decreased (p < 0.05) in 6, 9, 12, and 15 mM L-arabinose treatment groups (Figure 7c,d; Supplementary Figure S6). Thus, these data confirm that L-arabinose treatments impaired arginine uptake and metabolism of E. coli O157:H7.

Figure 7. L-Arabinose reduces Stx2 phage production in E. coli O157:H7 by inhibiting arginine transport and metabolism Intracellular arginine levels of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose (a). Partial least squares discriminant analysis (PLS-DA) plot of metabolites in 0, 3, 6, 9, 12, and 15 mM L-arabinose treatment groups (b). Spermidine concentrations in the supernatants of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose (c). N2-Acetyl-L-ornithine concentrations in the supernatants of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM L-arabinose (d). Arginine (Arg) transport and metabolism gene expression levels of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 6 mM or 15 mM L-arabinose (e). Arginine concentrations in the supernatants of culture medium upon 0 and 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM arginine (f). Metabolized arginine concentrations by E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM arginine (g). Cell densities (h) and Stx2 phage production (i) of E. coli O157:H7 upon 24 h growth in LB medium supplemented with 0, 3, 6, 9, 12, and 15 mM arginine. Cell densities (j) and Stx2 phage production (k) of E. coli O157:H7 upon 24 h growth in arginine-deficient (ΔArg) culture medium or arginine-complemented (COMP) culture medium supplemented with 6 mM L-arabinose as the sole carbon source. Data in Figures 7a and 7c−7k are presented as mean ± standard deviation and analyzed based on three biological replicates.

Then we explored the effect of arginine on Stx2 prophage induction of E. coli O157:H7. Results showed that supplementation with arginine in LB medium promoted gene expression involved in the arginine transport, biosynthesis, and catabolism (p < 0.05) (Figure 7e). Consistent with this observation, arginine concentrations in culture medium were significantly reduced when compared with the initial medium after 24 h of incubation (p < 0.05) (Figure 7f). As the amount of arginine addition increased, metabolized arginine concentrations by E. coli O157:H7 had a rising tendency (Figure 7g). Cell densities were not affected within 0–9 mM, but significantly decreased at doses of 12 mM or 15 mM (p < 0.05) (Figure 7h). Surprisingly, Stx2 phage production was also gradually increased (p < 0.05) (Figure 7i), proving that arginine metabolism promotes Stx2 phage production in E. coli O157:H7.

To further confirm whether arginine participated in Stx2 prophage induction regulation during L-arabinose treatments, we compared Stx2 phage production between arginine-deficient (ΔArg) and arginine-complemented (COMP) culture medium containing L-arabinose as the sole carbon source. As expected, neither cell densities nor Stx2 phage production were affected for the strain, which was grown in arginine-complemented medium compared to that growth in arginine-deficient medium (p > 0.05) (Figure 7j,k), supporting the idea that L-arabinose inhibited arginine transport and metabolism in E. coli O157:H7, thereby impairing Stx2 prophage induction.

Discussion

Infection by E. coli O157:H7 is a global public health problem considering the unavailability of many antibiotics because of their potential to enhance bacterial toxin synthesis and production.¹² Exploration of novel therapeutic strategies for the treatment of E. coli O157:H7 infection is of paramount importance. Stx2 is a major pathogenic factor in E. coli O157:H7, which is encoded by Stx2 lambdoid prophage. Prevention of Stx2 phage lytic development has been shown to be effective against E. coli O157:H7 infectious diseases.¹⁵ For example, Sheng, Rasco, and Zhu¹³ reported that cinnamon oil could be used as a therapeutic agent by inhibiting Stx2 production and Stx2 prophage induction in E. coli O157:H7. However, the compounds that can control the release of Stx2 phage and how these inhibitors modulate prophage induction still remain poorly understood. Especially, few studies explore the potential role of dietary components on regulating Stx2 prophage induction. In this research, we demonstrated for the first time that dietary L-arabinose inhibited Stx2 prophage induction in E. coli O157:H7 and our diet-based antivirulence approach is expected to provide a new paradigm against enteric bacterial infection.

L-Arabinose showed inhibitory effects on Stx2 prophage induction and Stx2 production in E. coli O157:H7. However, 24-h treatments with L-arabinose at different concentrations promoted stx2 gene expression with varied extents. The post-transcriptional processing, translational efficiency, and post-translational modification may alter Stx2 protein levels. Intestinal colonization of E. coli O157:H7 might be associated with similar fecal Stx2 phage production at day 3 and day 6. It is worth mentioning that xylan or glucose promoted E. coli O157:H7 growth and xylan induced more Stx2 phage in E. coli O157:H7, suggesting food taboos for patients suffering from E. coli O157:H7 infection. Similarly to our results, some commonly consumed drinks stimulated Stx2 prophage induction, potentially enhancing the virulence of enterohemorrhagic E. coli (EHEC).²⁵ Prophage induction induced by diet may be a regular occurrence in human gut ecosystem.¹⁶ Thus, it is particularly critical to identify safe food to protect consumers infected with EHEC pathogens and is necessary to fully understand how dietary compounds affect Stx2 or other virulence-encoded prophage induction.

Previous study has proved that the recA-deficient mutant of E. coli O157:H7 produced much less Stx2 phage than the wild-type strain and the virulence of the recA-deficient mutant was strongly reduced.²⁶ This is because activated RecA protein is required to initiate Stx2 prophage induction.¹² L-Arabinose decreased RecA protein production and recA gene expression levels at doses of 9, 12, and 15 mM, which supported the reduction of Stx2 phage. However, at 3 and 6 mM, L-arabinose promoted RecA protein production but still decreased Stx2 protein and Stx2 phage production, indicating that additional regulatory mechanisms may exist. As a global regulator of E. coli O157:H7, QS positively participated in stx2 gene transcription.²¹ E. coli O157:H7 mutants of major genes of QS system, qseB, qseC, or luxS, markedly reduced recA expression and Stx2 phage production.^20,22 In our study, L-arabinose abolished qseB, qseC, and luxS mRNA expression, indicating that QS was likely involved in inhibiting Stx2 prophage induction under L-arabinose treatment. QS can also enhance bacterial oxidative stress,²² which was demonstrated to increase RecA production in EHEC, thereby stimulating the SOS response and promoting Shiga toxin production.¹¹ Our results showed that L-arabinose strongly inhibited oxidative stress response of E. coli O157:H7. Importantly, our study examined that the inhibitory effect of dietary L-arabinose on Stx2 phage production was dependent on the inhibition of arginine metabolism. This may be explained by the mechanism described in a previous study, which showed that impaired arginine utilization by E. coli cells caused the stringent response that was responsible for the limitation of Stx2-converting phage lytic development.¹⁵ By contrast, deprivation of lysine, leucine, methionine, proline, or tryptophan from culture medium effectively promoted lambda prophage induction.²⁷ These findings suggest that the regulatory action of amino acids on prophage induction is type-specific. Nevertheless, effects of other amino acid metabolisms on Stx2 prophage lytic development are rarely investigated. Considering the high consumption of amino acids in human daily diet, a more in-depth understanding of the potential important role of amino acid metabolism on Stx2 prophage induction regulation is crucial. This study also has significant implications for how much L-arabinose should be utilized in clinical practice. Because low doses of L-arabinose could inhibit Stx2 prophage induction only by reducing arginine metabolism, but not by impairing the SOS response, it is quite likely that the high level of L-arabinose intake will achieve better treatment efficacy for patients suffering from E. coli O157:H7. More importantly, different from daily consumed glucose, fructose, or sucrose, L-arabinose is a low calorie sweetener that can even inhibit the activity of intestinal sucrase.^17,18 Thus, the use of L-arabinose in clinical trial is relatively healthy and controllable.

Treatment of E. coli O157:H7 cells with high concentrations of L-arabinose (9, 12, and 15 mM) could be efficacious against bacterial growth in vitro. Few studies have investigated the growth inhibitory role of dietary sugars. Our results showed that L-arabinose treatments downregulated gene expression levels mainly involved in energy production and conversion, carbohydrate transport and metabolism as well as amino acid transport and metabolism. This possibly leads to decreased utilization of certain essential nutrients, which support E. coli growth. Further studies are needed to address the underlying mechanisms. L-Arabinose-induced bacteriostatic effect may be the reason why inhibition of Stx2 phage lytic development did not result in healthier bacterial population. Notably, the OD₆₀₀ value of E. coli O157:H7 was increased after mitomycin C treatment. Previous studies have shown that mitomycin C can induce filamentation, which might explain this observation.^28–30 All in all, L-arabinose is an excellent candidate for improving gut health during E. coli O157:H7 infection.

Conclusion

Dietary L-arabinose could prevent Stx2 prophage induction in E. coli O157:H7. High levels of L-arabinose supplementation diminished bacterial SOS response, a positive regulator of Stx2 prophage, potentially through inhibiting bacterial QS and oxidative stress response. Importantly, we revealed for the first time that L-arabinose decreased arginine transport and metabolism, which contributed to reduced Stx2 phage production. Altogether, L-arabinose has great potential to be considered as a functional anti-Stx2 prophage food.

Materials and methods

Bacterial strain

E. coli O157:H7 strain MB41-1 (GenBank: CP039834.1) was obtained from Prof. Shiyan Qiaoʼs laboratory at China Agricultural University. The stain was grown in lysogeny broth (LB) (Beijing Aoboxing Bio-tech Co., Ltd, Beijing, China) at 37°C.

E. coli culture in LB medium supplemented with different sugars

Approximately 10⁷ CFU E. coli O157:H7 was resuspended in 100 µL PBS. Subsequently, the bacterial suspension was inoculated into 8 mL LB liquid medium containing 1% glucose (Solarbio, Beijing, China), D-xylose (Aladdin, Shanghai, China), L-arabinose (Aladdin, Shanghai, China), mannose (Aladdin, Shanghai, China), xylan (Aladdin, Shanghai, China), or arabinogalactan (YuanyeBiotech, Shanghai, China). Following 24 h stationary incubation at 37°C, bacterial samples were cooled down in an ice bath. Bacterial cells were precipitated by centrifugation at 12,000 rpm for 10 min and then resuspended in 200 μL PBS for OD₆₀₀ value detection using a microplate reader (Synergy, BioTek, USA).

Mitomycin C induction

Mitomycin C induction experiments were conducted in a 96-well plate at 37°C. Each well was seeded with 10⁷ CFU E. coli O157:H7. Mitomycin C (0.5 μg/mL) was added into E. coli culture medium at 60 min of incubation.

Construction of streptomycin-resistance E. coli O157:H7 strain

Competent E. coli O157:H7 cell was prepared via the CaCl₂ method and then transformed with pET28a vector (Miaoling Biotechnology Co., Ltd, Wuhan, China) to confer streptomycin resistance. Streptomycin-resistant E. coli strain was used to infect mice and counted on LB-agar plates supplemented with streptomycin (50 μg/mL).

E. coli O157:H7 intestinal colonization assay

Animal experiments were performed in accordance with the animal experimental ethics committee guidelines of China Agricultural University under protocol number AW80902202-1-1. E. coli O157:H7 intestinal colonization assay was conducted as the method previously described.³¹ Briefly, seven-week-old male C57BL/6J mice (SPF Biotechnology Co., Ltd, Beijing, China) were orally administrated with 10⁷ CFU E. coli O157:H7. The infected mice were anaesthetized and euthanized by cervical dislocation at day 3. The duodenum, jejunum, ileum, cecum, and colon were harvested, and the digesta was squeezed out. Each intestinal tissue sample was ground and homogenized in PBS for colony forming units (CFU) analysis.

Dietary L-arabinose/arabinogalactan intervention studies

Six-week-old male C57BL/6J mice were kept for 7 days as an adaption period before being used in experiment. During formal experimental period, mice were fed ad libitum and given free access to drinking water containing 5.63 mg/mL L-arabinose or arabinogalactan. Daily, mice were received approximately 10⁷ CFU E. coli O157:H7 and observed for body weight loss and death. Sixteen hours after E. coli gavage, fecal samples were collected and resuspended in PBS (100 mg/mL) to extract phage DNA. At the end of the experimental period, mice were sacrificed to collect blood, intestinal and kidney tissues, as well as digesta samples for further analysis.

Measurement of serum urea levels

The serum was prepared by centrifugation of the whole blood at 4,500 for 10 min. Serum urea levels were detected using an automatic biochemical analyzer (Hitachi 7600, Tokyo, Japan).

Hematoxylin and eosin (H&E) staining

The intestinal and kidney tissues of mice were fixed in 4% paraformaldehyde solution and then dehydrated using a graded ethanol series (70% to 100%). Subsequently, the samples were cleared with xylene and embedded in paraffin wax. Serial sections (5-μm thickness) were cut using LEICA RM2135 rotary microtome (Leica Microsystems GmbH), and then stained with hematoxylin and eosin. The bright field images of the tissue sections were taken by a Zeiss Axio Imager microscope (Carl Zeiss Microscopy, Germany).

Determination of proinflammatory cytokines in colon tissue

Six hundred microliters of PBS and 1% phenylmethanesulfonyl fluoride (PMSF) were added to colon tissue, followed by vigorous shaking at 4°C for 2 min. The supernatant containing the tissue protein was collected after centrifugation at 5,000 rpm for 5 min. The concentrations of proinflammatory cytokines TNF-α, IL−1β, and IL−6 were then measured using respective commercially available enzyme-linked immunosorbent assay (ELISA) kits (Thermo Fisher Scientific, MA, USA).

Quantitative analysis of Stx2 phage

Stx2 phage DNA was extracted using the rapid phage DNA extraction kit (Mei5 Biotechnology Co., Ltd, Beijing, China). Specifically, the mixture containing Stx2 phage was centrifuged at 12,000 rpm for 10 min. The supernatant was filtered through a 0.22 µm PVDF membrane (Millipore, USA). An equal volume of Hank’s balanced salt solution (HBSS) containing Ca²⁺ and Mg²⁺ was added, followed by incubation with DNase I and RNase A for 30 min at 37°C to eliminate bacterial DNA and RNA. PEG 8,000 was added to a final concentration of 10% (w/v) and left overnight at 4°C. After centrifugation at 12,000 rpm for 10 min at 4°C, the phage pellet was lysed in the lysis buffer (4.5 M guanidinium isothiocyanate, 44 mM sodium citrate pH 7.0, 0.88% sarkosyl, 0.72% 2- mercaptoethanol), followed by incubation with 20% SDS for 10 min at 70°C and then with ice bath for 5 min. One hundred microliters impurity precipitation solution was added to the mixture. After centrifugation at 14,800 rpm, 4°C for 10 min, the supernatant containing phage DNA was collected. DNA was extracted using the centrifugal adsorption column method. Serials dilutions of plasmids containing stx2 gene cloned into a plasmid vector were analyzed to generate a standard curve and calculate absolute copies of Stx2 phage.

ELISA quantification of Stx2 protein

The Stx2 protein concentrations in bacterial culture supernatants were measured by ELISA-based assays (Shanghai Enzyme-linked Biotechnology Co., Ltd, Shanghai, China). Liquid samples diluted 1: 2 and antibody-enzyme conjugate were incubated at 37°C for 30 min in the antigen-coated wells. After washing five times with wash buffer, 100 µL/well of colorimetric reagent was incubated at 37°C for 10 min. Following incubation, 50 µL/well of 2 M H₂SO₄ was used to stop the reaction. Absorbance of all reactions wells was recorded at 450 nm using a microplate reader (Synergy, BioTek, USA).

Bacterial RNA extraction, cDNA synthesis, and quantitative reverse transcription-PCR (RT-PCR)

Bacterial total RNA samples were prepared using a bacterial RNA rapid extraction kit (Aidlab Biotechnologies Co., Ltd, Beijing, China) according to the manufacturer’s instructions. The concentration of RNA was determined using a NanoDrop spectrophotometer (Thermo Fisher Scientific, MA, USA). Each cDNA was synthesized using the SynScript^TM III cDNA Synthesis Mix (Tsingke Biotechnology Co., Ltd, Beijing, China) following the protocols. The 10 μL RT-PCR reaction mixture consisted of 0.4 µL each of the forward and reverse primers, 5 µL of 2 × TSINGKE Master qPCR Mix (SYBR green I) (Tsingke Biotechnology Co., Ltd, Beijing, China), 1.0 µL of cDNA template, and 3.2 μL of nuclease-free H₂O. RT-PCR was performed on a Riche light cycler 96 Real-Time PCR System (Switzerland). The primers used in this study were listed in Supplementary Table S6. The 16S rRNA gene was served as reference gene. The mRNA expression level of specific gene was normalized to the control (2^−ΔΔCt method).

Western blot analysis

E. coli O157:H7 protein samples were extracted with bacterial protein lysate buffer (Shanghai Sangon Biotechnology Co., Ltd, Shanghai, China). Total protein concentrations in prepared samples were determined by Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, MA, USA) and were subsequently adjusted to reach the same final protein concentrations. A capillary-based auto Western blotting system was used to analyze the RecA protein expression (Protein Simple Wes, CA, USA). All procedures were completed based on the manufacturer’s instructions and default settings. Primary RecA protein antibody (Abcam, Cambridge, UK) was used at 1: 300 dilutions. Data were analyzed with the Compass Software associated with the Wes instrument (ProteinSimple, San Jose, CA). Band density was first corrected with the sample total protein and was then normalized to that of the control group.

Bacterial transcriptome analysis

E. coli O157:H7 transcriptome analysis was performed by Shanghai Majorbio Bio-pharm Technology Co., Ltd (Shanghai, China). Briefly, the total RNA from bacterial samples was extracted, followed by RNA concentration determination and agarose gel electrophoresis. rRNA was removed and mRNA was fragmented into small sizes (approximately 200 bp) using fragmentation buffer. Subsequently, the fragmented mRNA was used as the template for the first-strand cDNA synthesis with random primers in the presence of reverse transcriptase. In the process of second strand cDNA synthesis, DTTP is replaced by dUTP in dNTPs reagent. Sequencing was completed on the Illumina Hiseq platform with 2 × 150/300 bp pair-end reads.

Arginine metabolism experiments

LB culture medium containing different concentrations of arginine (Solarbio, Beijing, China) was sterilized using a 0.22 µm filter membrane (Pall Life Sciences, USA) before use. Arginine-deficient and arginine-complemented culture medium were purchased from Coolaber (Beijing, China). E. coli suspensions (approximately 10⁷ CFU) were inoculated in 8 mL of medium. Following incubation at 37°C for 24 h, bacterial cultures were centrifuged at 12,000 rpm, 4°C for 10 min. The E. coli precipitation and supernatant were respectively collected for subsequent analysis.

Determination of arginine concentrations

Arginine concentrations were determined using the automatic amino acid analyzer (Hitachi L8900, Japan). Samples were hydrolyzed with 6 M HCl at 110°C for 22–24 h and freeze-dried to remove water, HCl, and volatile organics. The residue was then dissolved in 2 mL of pH 2.2 citrate buffer. To analyze arginine, the supernatant was filtered through a 0.45 µm nylon membrane syringe filter (Pall Life Sciences, USA).

LC-MS/ms-based non-target metabolomic analysis

LC-MS/MS (Thermo Fisher Scientific, MA, USA) was used to perform non-target metabolomics analysis and analyze spermidine and N2-Acetyl-L-ornithine relative contents in the supernatant. In brief, the same volume of liquid samples were extracted using a 400 µL methanol: acetonitrile (1: 1, v/v) solution. The mixtures were then sonicated at 40 kHz for 30 min at 5°C. The samples were placed at −20°C for 30 min to precipitate proteins. After centrifugation at 13,000 g at 4°C for 15 min, the supernatant was carefully transferred to new microtubes and evaporated to dryness under a gentle stream of nitrogen. The samples were reconstituted in 100 µL loading solution of acetonitrile: water (1: 1, v/v) by brief sonication in a 5°C water bath. Extracted metabolites were spun for 15 min at 13,000 g at 4°C on a bench-top centrifuge and cleared supernatant was transferred to sample vials for LC-MS/MS analysis. The metabolites were identified by searching database including HMDB (http://www.hmdb.ca/), Metlin (https://metlin.scripps.edu/), and Majorbio Database. The data were analyzed on the free online platform of majorbio cloud platform (cloud.majorbio.com).³²

Quantification of spermidine and N2-acetyl-L-ornithine concentrations by ELISA

Quantitative analysis of spermidine and N2-acetyl-L-ornithine was respectively performed on the supernatants derived from E. coli O157:H7 cultures by ELISA according to the manufacturer’s instructions (Shanghai Enzyme-linked Biotechnology Co., Ltd, Shanghai, China).

Statistical analysis

Graphs were generated with GraphPad Prism 8.0 software (GraphPad Software, La Jolla, CA, USA). Statistical analysis was performed using IBM SPSS Statistics 22.0 software (Corporation, Armonk, NY). Studentʼs t-test was used to detect statistical significance between two treatment groups. If more than two treatment groups were involved, one-way ANOVA and post hoc Tukey’s tests were performed for multiple comparisons. Results represent mean ± standard deviation (SD). The same small letters indicate that there was no significant difference, while different letters indicate that there was a significant difference. Statistical significance was considered as p < 0.05.

Author contributions

Conceptualization, J.H., J.W. and D.H. Visualization, J.H.,Y.P., X.L., J.W. and D.H. Methodology, J.H., X.Z., L.K., Y.L., J.W. and D.H. Investigation, J.H., X.Z., Y.W., J.W. and D.H. Funding acquisition, J.W., D.H. and J.H. Writing-original draft, J.H. Writing-review and editing, S.Z., D.H. and J.W.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The raw RNA-Seq transcriptome data of E. coli O157:H7 analyzed in this study is available in the NCBI Sequence Read Archive (SRA) under BioProject PRJNA884078 https://www.ncbi.nlm.nih.gov/bioproject/PRJNA884078.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19490976.2023.2221778.

Additional information

Funding

This work was financially supported by the National Key Research and Development Program of China (No. 2022YFA1304201 and 2021YFD1300201), the National Natural Science Foundation of China (No. 32172750, 32125036, 31972596, 31902170, and 31630074), the Open Project Program of Key Laboratory of Feed Biotechnology (KLFB-IFR-202201), the China Agricultural Research System (CARS-35), the 111 Project (B16044), and Postgraduate Innovation Funding Project of China Agricultural University (No. 2022TC173).