NhaA facilitates the maintenance of bacterial envelope integrity and the evasion of complement attack contributing to extraintestinal pathogenic Escherichia coli virulence

ABSTRACT

Extraintestinal pathogenic Escherichia coli (ExPEC) is responsible for severe bloodstream infections in humans and animals. However, the mechanisms underlying ExPEC’s serum resistance remain incompletely understood. Through the transposon-directed insertion-site sequencing approach, our previous study identified nhaA, the gene encoding a Na⁺/H⁺ antiporter, as a crucial factor for infection in vivo. In this study, we investigated the role of NhaA in ExPEC virulence utilizing both in vitro models and systemic infection models involving avian and mammalian animals. Genetic mutagenesis analysis revealed that nhaA deletion resulted in filamentous bacterial morphology and rendered the bacteria more susceptible to sodium dodecyl sulfate, suggesting the role of nhaA in maintaining cell envelope integrity. The nhaA mutant also displayed heightened sensitivity to complement-mediated killing compared to the wild-type strain, attributed to augmented deposition of complement components (C3b and C9). Remarkably, NhaA played a more crucial role in virulence compared to several well-known factors, including Iss, Prc, NlpI, and OmpA. Our findings revealed that NhaA significantly enhanced virulence across diverse human ExPEC prototype strains within B2 phylogroups, suggesting widespread involvement in virulence. Given its pivotal role, NhaA could serve as a potential drug target for tackling ExPEC infections.

INTRODUCTION

Extraintestinal pathogenic Escherichia coli (ExPEC) breaches sterile barriers to cause severe extraintestinal diseases in all age groups of humans and animals (1). In humans, ExPEC is responsible for most urinary tract infections and a large proportion of bloodstream infections (2). The ExPEC responsible for poultry infections is avian pathogenic E. coli (APEC), which causes significant economic losses in the poultry industry (3). A common feature of ExPEC is its ability to cause bloodstream infection leading to septicemia (4), and some strains, including those of human and avian origin, exhibit zoonotic potential (5).

To survive and proliferate in the blood, ExPEC must be able to resist the multiple innate defense systems present in the blood, including the effects of phagocytes (6); the complement system (7); lysozymes (8); antimicrobial peptides (9); and coagulation factors VII, IX, and X (10). As part of the innate immune system, the complement system plays a pivotal role in the recognition and direct clearance of invading microbes and aberrant host cells (11, 12). Classical (CP), lectin (LP), and alternative pathways (AP) are triggered via the recognition of antigen-bound IgG (by the C1 complex), conserved sugar residues on bacterial surfaces, and spontaneous hydrolysis of C3, respectively, to activate the complement system (13). Following the formation of C3 convertase, the three complement pathways converge, and C3 convertase cleaves the central complement component C3 into C3a and C3b (14, 15). Moreover, C3b participates in C5 convertase formation, which cleaves C5 and initiates sequential deposition of the remaining complement components to form the C5b-9 membrane attack complex (MAC) (16, 17). These convertase-generated MAC pores in the outer membrane (OM) trigger inner membrane (IM) damage and directly kill microorganisms (11). Complement-mediated lysis of susceptible bacteria is reportedly enhanced by antimicrobial peptides (18), proteases, and lysozyme (19). However, MAC damages bacterial IM and kills bacteria without depending on lysozyme (20, 21).

Multiple factors contributing to resistance to complement-mediated killing have been identified in pathogenic E. coli. Mechanisms promoting resistance include the production of protective polysaccharide layers, interference with the complement cascade, and the expression of factors that contribute to the structural integrity of the cell envelope. Notably, K- and O-antigen capsules, lipopolysaccharides, extracellular polysaccharides, and colanic acid contribute to ExPEC strain survival in the serum (22 – 24). Membrane proteins (OmpA, NlpI, and Prc) contribute to E. coli evasion of complement-mediated serum killing (25 – 27). Additionally, two plasmid-encoded OM lipoproteins, TraT and Iss, have been suggested to interfere with MAC activity (28, 29).

NhaA is an IM protein evolutionarily conserved across all kingdoms of life, and it regulates cellular ion homeostasis (30). The amino acid sequences of NhaA in all ExPEC strains, such as XM, UTI89, RS218, and CFT073, are 100% identical and 98% identical to those of E. coli K12 (30). NhaA consists of 12 transmembrane helices connected by hydrophilic loops (31). It is organized into two functional regions: a pH sensor and catalytic regions containing an ion-binding site, and it is the main electrogenic sodium/hydrogen (Na⁺/H⁺) antiporter. Amino acid residues such as Glu252, Glu241, and Val254 are important in “pH sensing.” Asp65, Pro129, Thr132, and Asp133 are essential for the exchange of sodium/lithium (Na⁺/Li⁺) and H⁺. Gly338, Ala127, and Glu252 are essential for pH sensing and the exchange of Na⁺/Li⁺ and H⁺. In addition, Asp163, Asp164, and Lys300 are essential for Na⁺/H⁺ activity and Li⁺ binding (32, 33). These amino acid residues are conserved in all ExPEC and commensal E. coli K12 strains (30, 33). An NhaA homolog was found to be essential for Yersinia pestis virulence by preventing sodium toxicity in the blood (34). In our previous large-scale screening study, we identified NhaA as an important factor involved in ExPEC virulence (5). The present study aimed to describe in detail the role of NhaA in ExPEC virulence and explore the molecular mechanisms by which NhaA contributes to virulence.

RESULTS

nhaA was involved in ExPEC virulence in chick and mouse models of systemic infection

In our previous study, we demonstrated the significance of NhaA in ExPEC’s virulence within both mammalian and avian models. However, those virulence tests were conducted using transposon insertion mutant strains and employing a competition assay (mixing mutant and wild-type strains in equal proportions for animal infection) (5). In the present study, we created a non-polar mutant strain by deleting the nhaA gene and assessed both the mutant and wild-type (WT) strains separately in mouse and chicken models to affirm its role in virulence. Furthermore, we reintroduced the nhaA gene into the mutant strain to confirm its contribution to virulence. The wild-type strain killed all eight chicks in 2 days; however, no death was observed in the ∆nhaA mutant group after 7 days (P < 0.001) (Fig. 1A). Additionally, nhaA deletion significantly decreased the bacterial load in the blood (P < 0.001) (Fig. 1B). An average of 1 × 10⁸ CFU/mL bacteria was identified in the blood from wild-type (WT) group chicks; however, no bacteria could be isolated from the blood samples of mutant group chicks (limit of detection, 50 CFU/mL blood), and nhaA gene reintroduction into the mutants almost restored chick mortality and bacterial loads in serum to the WT levels. Similar results were observed in the mouse model (Fig. 1C and D). When their growth was compared in M9 minimal medium, ∆nhaA mutant and WT showed no significant differences in growth in vitro (Fig. 1E). These results demonstrate that nhaA is required for ExPEC virulence in chick and mouse models of systemic infection.

Fig 1.

nhaA deletion significantly attenuated ExPEC XM’s virulence in chick and mouse models. (A) WT, ∆nhaA mutant, and ∆nhaA-complemented strains (5 × 10⁷ CFU) were inoculated into eight 7-day-old chicks per group, and the mortality was monitored for 1 week. The virulence of the ∆nhaA mutant was significantly attenuated compared to that of the WT strain [P < 0.001 by log-rank (Mantel-Cox) test]. (B) nhaA deletion caused a significant decrease of the bacterial load in the blood of infected chicks at 12 h post-infection (P < 0.001 by Student’s t-test). The dotted line represents the blood’s lower limit of detection, 50 CFU/mL blood. (C) WT, ∆nhaA mutant, and ∆nhaA-complemented strains (5 × 10⁷ CFU) were inoculated into eight 6-week-old mice per group, and the mortality was monitored. The virulence of the ∆nhaA mutant was significantly attenuated compared to that of the WT strain [P < 0.001 by log-rank (Mantel-Cox) test]. (D) nhaA deletion significantly decreased the bacterial load in the blood of infected mice at 12 h post-infection (P < 0.01 by Student’s t-test). The dotted line represents the blood’s lower limit of detection, 50 CFU/mL blood. (E) Growth curves for the WT, ΔnhaA, and ΔnhaA-complemented strains administered in M9 minimal medium supplemented with glucose as energy substrates were examined by monitoring the OD₆₀₀ of each sample every 1 h.

Identification of protein loops and amino acid residues of NhaA contributing to ExPEC virulence

NhaA protein is predicted to have a putative secondary structure comprising 12 transmembrane segments connected by 11 hydrophilic loops. We investigated the loops associated with ExPEC virulence by constructing 11 loop-mutant strains (Fig. 2A). We confirmed that the deletion of loops L1, L3, L5, L6, L7, or L10 did not affect the expression of the remaining portion of NhaA protein. On the other hand, the deletion of five other loops (L2, L4, L8, L9, or L11) significantly reduced the expression of remaining protein (Fig. S1A and B), leading to the exclusion of their mutant strains from further studies. The pathogenicity of six loop mutant strains was then evaluated using a chick model. The WT and loop1-mutant strains (pnhaAL1) killed all eight chicks in 4 days, the loop7-mutant strain (pnhaAL7) killed half of the infected chicks in 7 days, and no death was observed in other loop-mutant strains (P < 0.001) (Fig. 2B). Similar results were observed in a mouse model (Fig. 2D). These results indicate the involvement of loops L3, L5, L6, L7, and L10 in ExPEC XM virulence.

Fig 2.

Identification of protein loops and amino acid residues of NhaA contributing to ExPEC virulence. (A) The two-dimensional model of E. coli NhaA. Triangles, mutation sites that inactivate the antiporter; squares, sites where mutations shift the pH profile; circles, sites where mutations affect the K _m for Na⁺ and Li⁺. (B and C) Survival curve of chicks infected with loop- and site-mutation strains (n = 8). (D and E) Survival rates of mice infected with loop- and site-mutation strains at 24 h post-infection (n = 8). Significant differences were evaluated by log-rank (Mantel-Cox) test, and ***P < 0.001, **P < 0.01, and *P < 0.05.

Previous studies have shown that several amino acid residues of NhaA are responsible for the functions of the Na⁺/H⁺ antiporter and pH sensing and response (Fig. 2A). We constructed 11 site-mutation strains and experimentally confirmed that site-mutation deletions did not affect the expression of NhaA protein (Fig. S1C and D). We also validated the role of several key amino acids in pH sensing and regulation as well iron homeostasis functions, as described in previous studies (33, 35, 36) (Fig. S2 and S3). Site mutations of P129L, T132C, and D133C, previously shown to affect the transportation of Na⁺ and Li⁺ (32), did not result in E. coli virulence attenuation, and all chicks died within 3 days. Additionally, two amino acid residues (H225R and G338S), previously identified as essential for E. coli’s pH sensing and response, were investigated. The site mutation of H225R significantly attenuated virulence (P < 0.001); however, the G338S site mutation strain did not exhibit a significant difference from the WT strain. Similarly, among the two-site mutations of A127V and E252C, vital for Na⁺ and Li⁺ transportation and pH sensing, only A127V resulted in significant virulence attenuation (P < 0.01). Moreover, the site mutations of D163C, D164C, and K300C, leading to complete NhaA inactivation as previously demonstrated, significantly reduced virulence (P < 0.001) (Fig. 2C and E).

nhaA deletion decreased E. coli’s resistance to stresses and increased RNase leakage from the periplasm

We examined whether nhaA deletion affects bacterial envelope integrity through a series of tests. Initially, we assessed the resistance of the ∆nhaA strain to EDTA, SDS, and a high concentration of NaCl. Notably, the growth of ∆nhaA was not markedly different from that of the WT and ∆nhaA-complemented strains on Luria–Bertani (LB) plates. Furthermore, nhaA deletion did not diminish E. coli’s resistance to 1 or 2 mM EDTA; all strains displayed similar sensitivities to EDTA (Fig. 3A). However, a significant reduction in growth rate was observed for the ∆nhaA strain compared to the WT and ∆nhaA-complemented strains in LB medium supplemented with 2% SDS (Fig. 3A). This growth phenotype was similarly apparent in LB medium containing 4% NaCl, nhaA mutant growth was significantly impaired in LB medium containing 4% NaCl compared to that of the WT and ∆nhaA-complemented strains (Fig. 3B and C). We further performed a periplasmic ribonuclease (RNase) leakage assay (37, 38) to compare the capacity of ∆nhaA, WT, and ∆nhaA-complemented strains to retain periplasmic proteins. When gram-negative bacteria with compromised OM integrity are grown on RNase test agar plates, RNase leakage from the periplasm forms pink halos around the bacterial colonies (39, 40). ∆nhaA showed significantly more leakage of periplasmic RNase into the medium than the WT and ∆nhaA-complemented strains (Fig. 3D). Altogether, these results suggest that nhaA deletion compromised bacterial envelope integrity.

Fig 3.

The effects of nhaA deletion on the envelope integrity of ExPEC XM. (A) EDTA and SDS sensitivity assays of WT, ΔnhaA, and ∆nhaA-complemented strains. The bacterial cultures were adjusted to OD₆₀₀ ≈ 0.4 with double distilled water, and 10-fold serial dilution was performed on LB, LB with 1 mM EDTA, 2 mM EDTA, or 2% SDS agar plates. (B and C) Growth curves for the WT, ΔnhaA, and ΔnhaA-complemented strains inoculated in different culture media were examined by monitoring the OD₆₀₀ of each sample every 1 h. (D) RNase leakage assays. The bacteria were grown on an RNase test agar plate. The pink halos around the colonies indicate the leakage of periplasmic RNase into the agar.

nhaA gene deletion reduced ExPEC resistance to serum bactericidal effects

We explored the molecular mechanism by which NhaA contributes to virulence by comparing resistance to phagocytosis and persistence in RAW264.7 macrophages. No significant difference was observed between the WT, ΔnhaA mutant, and ∆nhaA-complemented strains (data not shown). Next, the ability of these strains to survive in avian and human serum was assessed. The WT, ΔnhaA mutant, and ∆nhaA-complemented strains were incubated in 90% normal avian serum (NAS) at an initial inoculation dose of 5.0 × 10⁶ CFU/mL. The WT strain could still grow in the avian serum, and after 4-h incubation, the cell numbers of the WT strain were increased to 8.5 × 10⁸ CFU/mL. In contrast, nhaA deletion significantly reduced the resistance of ExPEC to serum-bactericidal effects, and the surviving cell numbers dramatically decreased to 4.6 × 10³ CFU/mL, with approximately 184,000-fold changes between the WT and ∆nhaA strains (P < 0.001). nhaA reintroduction into the mutant strain completely recovered ExPEC’s resistance to the WT level (Fig. 4A). An even more obvious difference was observed when the WT, ΔnhaA, and ∆nha-complemented strains were tested in 90% human serum. Although the WT strain could still propagate in human serum, nhaA deletion completely abolished ExPEC’s survival capability (Fig. 4B), and no live bacteria were detected after 4 h of incubation.

Fig 4.

nhaA gene deletion reduced the resistance of ExPEC to serum bactericidal effects. (A) Survival capacity of WT, ΔnhaA, and ΔnhaA-complemented strains incubated in 90% avian serum at an initial inoculation dose of 5 × 10⁶ CFU/mL for 4 h. (B) Survival capacity of WT, ΔnhaA, and ΔnhaA-complemented strains incubated in 90% human serum at an initial inoculation dose of 5 × 10⁶ CFU/mL for 4 h. The dotted line represents the blood lower limit of detection, 50 CFU/mL blood. Data are presented as the mean ± standard deviation. Significant differences were evaluated using the Student’s t-test (A and B), and ***P < 0.001. ns, not significant. (C) Left panel: the morphology of WT, ΔnhaA, and ΔnhaA-complemented strains cultured in LB medium. Right panel: the morphology of WT, ΔnhaA, and ΔnhaA-complemented strains cultured in 90% avian serum. Cells were prepared as described in Materials and Methods for visualization by laser confocal microscopy (scale bars: 5 µm). (D) The morphology of WT, ΔnhaA, and ΔnhaA-complemented strains cultured in LB medium (left panel) and 90% avian serum for 2 h (right panel) and visualized by transmission electron microscopy (scale bars: 1 µm).

Interestingly, only the nhaA mutant cells, but not the WT cells, exhibited morphological aberrations under serum stress. Approximately 100% of the mutant cells displayed filaments and curved shapes when cultured in avian serum; they maintained their rod shape when cultured under normal growth conditions (LB) (Fig. 4C). Under a transmission electron microscope, approximately 56% ∆nhaA cells had increased intermembrane space, invagination of the IM, and formation of intracellular vacuoles, compared to the WT and ∆nhaA-complementation strains (Fig. 4D).

NhaA contributed to resistance to complement-mediated bacterial killing

The WT, ΔnhaA, and ∆nhaA-complemented strains were tested to determine their resistance to complement-mediated bacterial killing by incubation in avian serum with complement or inactivated complement (heat-inactivated avian serum, HIAS). After 4-h incubation, WT and ∆nhaA-complemented strains proliferated from 5.0 × 10⁶ CFU/mL to approximately 1.0 × 10⁹ CFU/mL in either complement-active or inactivated avian sera. Moreover, the ΔnhaA mutant strain decreased from 5.0 × 10⁶ CFU/mL to <1.0 × 10⁴ CFU/mL in the complement-active avian serum. However, complement inactivation recovered the mutant strain’s growth nearly to the WT level, suggesting that NhaA contributed to E. coli’s resistance against complement-mediated bacterial death in avian serum (Fig. 5A). When tested in human complement-active or inactivated sera, the mutant strain could not survive in the complement-active serum. Similarly, complement inactivation significantly enhanced the mutant’s survival in human serum; however, its growth capability was not restored to the WT level (Fig. 5B).

Fig 5.

NhaA contributed to the resistance to complement-mediated bacterial killing. (A) Survival capacity of WT, ΔnhaA, and ΔnhaA-complemented strains incubated in 90% NAS and 90% HIAS at an initial inoculation dose of 5 × 10⁶ CFU/mL for 4 h. (B) Survival capacity of WT, ΔnhaA, and ΔnhaA-complemented strains incubated in 90% NHS and 90% HIHS at an initial inoculation dose of 5 × 10⁶ CFU/mL for 4 h. (C) Survival capacity of loop deletion mutants incubated in 90% NAS and 90% HIAS at an initial inoculation dose of 5 × 10⁶ CFU/mL for 4 h. (D) Survival capacity of critical amino acid site mutations incubated in 90% NAS and 90% HIAS at an initial inoculation dose of 5 × 10⁶ CFU/mL for 4 h. The dotted line represents the blood lower limit of detection, 50 CFU/mL blood. Data are presented as the mean ± standard deviation. Significant differences were evaluated using the Student’s t-test, and ***P < 0.001.

The loop deletion mutants and strains with critical amino acid site-mutations were also assessed for their resistance to serum complement. No significant differences were observed in loop1 and loop7 mutant strains when compared to the WT strain. Conversely, other loop mutant strains (L3, L5, L6, and L10) exhibited a notable reduction in survival in the presence of NAS. In the context of complement system inactivation, loop3, loop5, and loop6 mutant strains survival was notably enhanced, while loop10 mutant strain showed marginal effects (Fig. 5C). Turning to the site-mutation strains, mutagenesis involving A127V, D163C, D164C, H225R, and K300C significantly diminished the survival of these mutants in NAS. Notably, inactivation of the complement system markedly increased the survival of the K300C mutant strain by a factor of 20,000 (P < 0.001), whereas its impact was negligible on A127V and H225R site-mutation strains. Importantly, both the D163C and D164C site-mutation strains failed to survive in either NAS or HIAS (Fig. 5D). These findings underscore the significant role of NhaA in ExPEC’s resistance to complement-mediated killing, with specific loops and critical amino acids contributing to this resistance.

NhaA impaired C3b and C9 deposition through complement classical and alternative pathways

The complement system can be activated by three distinct pathways: CP, AP, and LP (7). The roles of these three complement pathways in killing WT and ∆nhaA strains were investigated. The strains were independently incubated in 50% normal human serum (NHS) and 50% NHS with blocked CP (C1q-depleted serum), AP (factor B-depleted serum), or LP (mannose-treated serum). No significant difference was observed in the survival of WT and ∆nhaA strains when incubated in CP- or AP-blocked sera; however, ∆nhaA survival in NHS (P < 0.05) and LP-blocked serum was significantly less than that of the WT strain (P < 0.01) (Fig. 6A). These results suggest that NhaA is involved in the resistance to CP and AP of complement-mediated bactericidal activity.

Fig 6.

NhaA impaired C3b and MAC depositions through classical and alternative pathways. (A) The serum survival of ΔnhaA and WT strains after 3-h incubation in 50% NHS and NHS with the classical (CP^--HS), alternative (AP^--HS), or lectin pathway inhibited (LP^--HS) sera (Left panel). Flow cytometry histogram of C3b deposition on the bacteria after 2-h incubation in different sera (right panel). Data are presented as the mean ± standard deviation. Significant differences were evaluated using the Student’s t-test, and **P < 0.01 and *P < 0.05. (B) C3b deposition levels on ΔnhaA and WT after incubation with 50% normal human serum for the indicated periods (left panel). Flow cytometry histogram of C3b deposition on the bacteria after 2-h incubation in 50% NHS (right panel). (C) C9 deposition levels on ΔnhaA and WT after incubation with 50% normal human serum for the indicated periods (left panel). Flow cytometry histogram of C9 deposition on the bacteria after 2-h incubation in 50% NHS (right panel). The data are presented with mean fluorescence intensity (MFI). The WT strains incubated with fluorescein isothiocyanate (FITC)-conjugated secondary antibodies were used as the control group. The results are expressed as the means ± standard deviations, and the data represent three independent experiments performed in triplicate. Significant differences were evaluated using the Student’s t-test, and ***P < 0.001 and **P < 0.01.

Activation of these three pathways leads to the production and deposition of the complement proteins C3b and C9 on the pathogen surface (15, 17). Flow cytometry analysis revealed that after incubation in 50% NHS for 2 h, both C3b and C9 were deposited on the WT and ∆nhaA cells (Fig. 6B and C), confirming the involvement of the complement system in the demise of both strains. It is worth mentioning that C3b and C9 deposition on ∆nhaA cells significantly increased over time, while it gradually decreased on WT cells. After a 2-h incubation, the deposition of C3b and C9 on ∆nhaA cells was 7.8 (P < 0.001) and 16.8 times (P < 0.01) higher than that on the WT strain, respectively. These results indicate that NhaA impaired the deposition of C3b and C9 on WT cells.

NhaA played a more important role than other membrane proteins in resistance to serum bactericidal activity and in APEC’s virulence

Several membrane proteins, OmpA, Prc, Iss, and NlpI, contribute to ExPEC resistance to human serum-mediated bacterial killing (25 – 27, 29). To evaluate their individual capacities to withstand avian serum-mediated bacterial killing, we generated mutant strains that lacked these proteins and tested them using an ex vivo serum resistance assay. All the WT and mutant strains were incubated in 90% avian serum at an initial inoculation dose of 5.0 × 10⁶ CFU/mL. After 4-h incubation, the WT, ∆iss, ∆ompA, and ∆nlpI mutant strains rapidly increased to approximately 1.0 × 10⁹ CFU/mL, suggesting that these genes (iss, ompA, or nlpI) did not contribute to ExPEC’s resistance to the serum’s bactericidal effects. In contrast, the ∆nhaA mutant strain exhibited a notable low bacterial count, measuring less than 4.9 × 10³ CFU/mL, 350,000-fold (P < 0.01) less than that of the WT (Fig. 7A), suggesting that the gene nhaA was essential in resistance to serum’s bactericidal effects. The mutant strain ∆prc proliferated in avian serum; albeit at a significantly lower level than that of the WT, whereas the ∆iss, ∆ompA, and ∆nlpI mutant strains exhibited no growth or survival differences compared to the wild type in serum. These outcomes suggest that iss, ompA, and nlpI do not contribute to serum resistance in the ExPEC strain used in this study. Meanwhile, the gene prc was found to play a role in serum resistance, albeit relatively less significant than the pivotal role of nhaA. Next, their pathogenicity was assessed using a chick model. As expected, all chicks in the WT, ∆ompA, and ∆iss groups died within 3 days, and six chicks from the ∆nlpI mutant group (75%) and two from the ∆prc mutant group (25%) died. In contrast, all chicks from the ∆nhaA mutant group survived, suggesting that nhaA deletion significantly attenuated the virulence (Fig. 7B). Overall, these results demonstrate that the nhaA gene is significantly more important in conferring resistance to the bactericidal effects of host serum and in contributing to APEC’s virulence compared to previously reported membrane protein genes such as ompA, prc, iss, and nlpI.

Fig 7.

NhaA played an important role in ExPEC resistance to serum-killing activity and ExPEC virulence. (A) The serum survival of WT and ΔnhaA, Δiss, Δprc, ΔompA, and ∆nlpI strains initial inoculation dose of 5 × 10⁶ CFU /mL in 90% NAS for 4-h incubation. The virulence of ΔnhaA and Δprc mutant was significantly attenuated compared to that of the WT strain (P < 0.01 by Student’s t-test). Data are presented as the mean ± standard deviation. Significant differences were evaluated using the Student’s t-test, and **P < 0.01. (B) WT, ΔnhaA, Δiss, Δprc, ΔompA, and ∆nlpI strains (5 × 10⁷ CFU) were inoculated into eight 7-day-old chicks per group, and the mortality was monitored for 1 week. Significant differences were evaluated by log-rank (Mantel-Cox) test, and ***P < 0.001 and **P < 0.01. ns, not significant.

nhaA was important for the full virulence in human ExPEC prototype strains

We further assessed the virulence contribution of NhaA in the neonatal meningitis-associated E. coli prototype strain RS218 and the uropathogenic prototype strains UTI89 (causing cystitis) and CFT073 (causing pyelonephritis), which highlight distinct infection niches and potentially differing virulence mechanisms. The resistance of nhaA mutant strains to serum’s bactericidal effect was examined in 50% human serum, and their virulence was examined in a mouse model of bacteremia, given that neonatal meningitis-associated and uropathogenic E. coli could cause bacteremia (41, 42). nhaA gene deletion in CFT073, RS218, and UTI89 resulted in 2,400- (P < 0.01), 412.5- (P < 0.001), and 29.6-fold (P < 0.01) reductions of E. coli levels in human serum, respectively, compared to their WT strains (Fig. 8A). When tested in a mouse model of bacteremia, seven of the eight mice challenged with UTI89 died, and six of eight mice died when challenged with WT strains of RS218 and CFT073. No death was observed in the mice challenged with ∆nhaA mutant strains of RS218 (P < 0.01) or UTI89 (P < 0.001), and only one mouse died in the group challenged with the CFT073 nhaA mutant strain (P < 0.05) (Fig. 8B). These results suggest that NhaA is also important for full virulence of human prototype ExPEC strains.

Fig 8.

nhaA was important for other ExPEC prototype strains’ resistance to serum-killing activity and their virulence. (A) The serum survival of WT and ΔnhaA in different ExPEC strains at an initial incubation dose of 5 × 10⁶ CFU/mL in 50% NHS for 4 h. Data are presented as the mean ± standard deviation. Significant differences were evaluated using the Student’s t-test, and ***P < 0.001 and **P < 0.01. (B) ΔnhaA of CFT073, RS218, UTI89, and their WT strains were inoculated into eight 6-week-old mice per group, and the mortality was monitored for 1 week. Significant differences were evaluated using the log-rank (Mantel-Cox) test, and ***P < 0.001, **P < 0.01, and *P < 0.0.

DISCUSSION

ExPEC strains cause various infections, including human urinary tract and bloodstream infections and avian colibacillosis (43, 44). Notably, antibiotic-resistant E. coli, particularly ExPEC, causes the most deaths among common bacterial pathogens (45, 46). The ability to resist serum bactericidal activity and survive in the bloodstream is an essential virulence trait of the ExPEC strain (47, 48). In the present study, we demonstrated that besides its Na⁺/H⁺ antiporter activities, NhaA maintained intact cell envelope and morphology and contributed greatly to ExPEC’s serum resistance by facilitating the evasion of complement killing. In addition, we demonstrated that nhaA deletion significantly attenuated the virulence of avian and human ExPECs. Therefore, our study further characterized this important virulence-associated factor and explored its underlying virulence mechanisms.

The removal of the nhaA gene had a significant impact on ExPEC’s ability to resist 4% NaCl and 2% SDS, as well as its ability to withstand the lethal effects of human and avian serum. This deletion also resulted in slower growth during the stationary phase when cultured in LB medium. Additionally, the mutants exhibited elongated and curved shapes when cultured in human and avian serum. These findings point toward wide perturbation within bacterial cells, resulting in reduced resilience to both in vitro and in vivo stress conditions. These phenotypes could be attributed to the absence of established functions related to ion transport, pH sensing, and regulation. Such function loss might lead to dysregulation within the bacteria, causing changes in membrane permeability under stress-induced circumstances. Certain results from our research also hint at the possibility of a new, previously unknown function at play. In E. coli, another Na⁺/H⁺ antiporter called NhaB can potentially compensate for the role of NhaA (49 – 51). Consequently, the sole deletion of nhaA gene would not drastically impact Na⁺/H⁺ transportation, pH sensing, or regulation. However, our results showed that exclusive removal of nhaA gene in ExPEC notably reduced its virulence and compromised serum resistance. Furthermore, we analyzed site-specific mutations (P129L, T132C, D133C, and G338S), which were previously recognized to influence Na⁺ and H⁺ transportation or pH sensing and response (32). Surprisingly, these mutations did not diminish ExPEC’s virulence, indicating the potential involvement of an unknown function.

Our RNase leakage assay demonstrated an increased permeability of the OM in the ΔnhaA mutant. This observation was consistent with the results of the SDS tolerance assay, which indicated a disruption in the integrity of the cell envelope in the ΔnhaA mutant. In addition to the downstream consequence of ion level dysregulation discussed earlier, it is worth noting that previous studies have highlighted the potential impact of inner membrane proteins on the biogenesis of the outer membrane, potentially compromising its integrity (52). The damaged OM could promote complement deposition and exacerbate the mortality of ExPEC (27), aligning with our findings. We found that the ΔnhaA mutant strain exhibited a higher complement deposition compared to its WT counterpart (Fig. 6). The complement system is activated at the bacterial OM, initiating bactericidal activity. The disrupted OM in the ΔnhaA mutant might enhance the accessibility of bacterial binding targets for components like C1q in the complement system. Meanwhile, the intact OM in the WT strain is more effective at obstructing or interfering with such interactions. Additionally, the bactericidal effect mediated by the MAC complex primarily involves damaging the OM, which in turn promotes permeability changes and triggers lethal modifications in the IM (20). Furthermore, it is plausible that NhaA contributes to the homeostasis of the IM where it is located. Consequently, the absence of nhaA could lead to instability in the IM, rendering the bacterium more susceptible to MAC attacks. Interestingly, we also observed that the recovery of the ΔnhaA mutant’s growth capability in human sera to the level of the wild-type strain could not be achieved through complement inactivation alone (Fig. 5). This implies the involvement of unknown mechanisms that warrant further investigation in future studies.

We incorporated a selection of previously reported membrane factors (Iss, Prc, OmpA, and NlpI), all of which hold significance for virulence, and then compared their respective roles in serum resistance and virulence with that of NhaA. Our investigation revealed that among these four factors, only NhaA, Prc, and NlpI demonstrated contributions to virulence in our animal tests, and only NhaA and Prc showed involvement in serum resistance (Fig. 7). The failure to replicate the impacts of Iss and OmpA could potentially stem from the presence of redundant alleles (44) or variations in genetic backgrounds across different strains. This discovery emphasizes the need to consider strain-specific variations when evaluating the contributions of virulence factors. Significantly, NhaA stood out as the primary contributor among these factors. The deletion of nhaA leads to complete protection against ExPEC infection, resulting in 0% mortality in animal tests. Notably, this trend held true across three additional human ExPEC prototype strains, namely CFT073 (B2 phylogenetic group, serotype O6: K2: H1), RS218 (B2 phylogenetic group, serotype O18: K1: H7), and UTI89 (B2 phylogenetic group, serotype O18: K1: H7), as depicted in Fig. 8. Given its conservation and essential role in ExPEC, NhaA emerges as a promising drug target. Moreover, a nhaA null mutant presented itself as a strong candidate for the development of live attenuated bacterial vaccines, capitalizing on its demonstrated attenuation of virulence across multiple strains.

In conclusion, our study highlights the significant role played by NhaA in E. coli septicemia and its contribution to resistance against both classical and alternative pathways of host complement defense. Given the ubiquity and high conservation of NhaA in bacteria, these findings have broader implications for other blood-borne bacterial pathogens. The NhaA functions elucidated in this study warrant further investigation of this protein as a potential novel target for both preventing and treating ExPEC infection, as well as for developing live attenuated bacterial vaccines.

MATERIALS AND METHODS

Bacterial strains and culture conditions

ExPEC XM (O2: K1: H7), belonging to the phylogenetic E. coli reference (ECOR) B2, was isolated from the brain of a duck with septicemia and neurological symptoms (53). The strains and plasmids used in this study are listed in Table S1. The bacterial strains (ExPEC XM, RS218, CFT073, UTI89, DH5α, and their derivatives) were routinely cultured at 37℃ in LB medium. For growth studies, the bacteria were grown in M9 minimal salts supplemented with 2 mM MgSO₄, 0.1 mM CaCl₂, and glucose (0.5% vol/vol) added as an energy substrate. The H225R and G338S site-mutant strains were cultivated in LB medium containing 0.6 M NaCl at pH 7.0 and 8.3, or in LB medium supplemented with 0.6 M KCl instead of NaCl at pH 8.3 (36). Selective antibiotics were added when necessary at the following concentrations: ampicillin, 100 µg/mL and chloramphenicol, 25 µg/mL.

Human and chick sera

Normal human serum, C1q-depleted, and factor B-depleted sera were purchased from Complement Technology, Inc. (Tyler, TX, USA). C1q-depleted and factor B-depleted sera supplemented with 5 mM CaCl₂ and 2 mM MgCl₂ served as CP- and AP-blocked sera, respectively. Additionally, to block the LP pathway, NHS was treated with 100 mM mannose (54). For NAS isolation, fresh chick blood purchased from the Harbin Veterinary Research Institute (Harbin, China) was collected in 50 mL tubes and incubated at 37°C for 2 h. Next, the blood was incubated at 4°C for 4 h. Afterward, the tubes were centrifuged at 1,000 × g for 10 min at 4°C. The supernatants were collected, aliquoted, and kept at −80°C until further use. Heat-inactivated human serum (HIHS) and HIAS were prepared by heating the NHS and NAS at 56°C for 30 min.

Recombinant DNA techniques

PCR, DNA ligation, electroporation, and DNA gel electrophoresis were performed as described by Sambrook and Russell (55). The DNA-modifying and restriction enzymes used in this study were purchased from New England Biolabs (Ipswich, MA, USA) or Thermo Fisher Scientific (Waltham, MA, USA). Restriction fragments, PCR products, and recombinant plasmids were purified using the MiniBEST DNA Fragment Purification or MinElute Gel Extraction Kit (Takara, Dalian, China) as recommended by the supplier. Deletion mutants of ExPEC XM, RS218, CFT073, and UTI89 were constructed using the bacteriophage lambda-red homologous recombination system described by Datsenko and Wanner (56). For complementation, the coding sequence of nhaA and its putative promoter regions were amplified from the XM strain and independently cloned into pGEN-MCS (57, 58) using EcoRI and BamHI restriction sites. Next, site-directed and loop mutations were constructed using the complementation plasmid. To generate the NhaAD65C mutant, specific primers were meticulously designed with incorporated BamHI and EcoRI cleavage sites at the 5′ end, inclusive of the promoter sequence. ΔnhaA-D65C-F and ΔnhaA-D65C-R primers were crafted to introduce a site-specific mutation within the nhaA gene, using the homologous arms of the mutation sequence. The upstream and downstream fragments were separately amplified with ExPEC XM DNA as the template and pGEN-MCS-nhaA-F/ΔnhaA-D65C-R and pGEN-MCS-nhaA-R/ΔnhaA-D65C-F as the primers. Following gel electrophoresis and purification of the fragments, they were linked through overlap PCR, using pGEN-MCS-nhaA-F and pGEN-MCS-nhaA-R primers after ligation. The resultant product and the plasmid pGEN-MCS underwent digestion using BamHI and EcoRI, after which the target fragment was extracted and ligated to pGEN-MCS. The ligated product was then introduced into DH5α competent cells. Upon selection of individual colonies and subsequent identification, the recombinant plasmids were extracted and electrophoretically transferred to ΔnhaA competent cells for further investigation. The list of primers used for mutagenesis and complementation can be found in Table S2 in the supplemental material.

Animal experiments

Healthy 7-day-old specific pathogen-free white leghorn chicks and 6-week-old BALB/c mice were used in this study. The mice and chicks were purchased from WeiTong LiHua Experimental Animal Technology (Beijing, China) and the Harbin Veterinary Research Institute, respectively. For virulence experiments in vivo, the chicks (n = 8) were injected with 5 × 10⁷ CFU of ExPEC XM and its derivatives intratracheally. Next, the challenged chicks were observed for 7 days, and death or survival was recorded. To determine the bacterial load, chicks were euthanized after 12 h, and blood was serially diluted in sterile phosphate buffered saline (PBS) and plated on LB agar. For loop-mutant and site-mutant strains, the mice (n = 8) were injected with 5 × 10⁷ CFU of ExPEC XM and its derivatives intraperitoneally, and the bacterial load in blood and mortality were determined after 12 and 24 h, respectively. For other ExPEC strains, challenge doses in RS218, CFT073 and their mutants were 5 × 10⁷ CFU, and UTI89 and its mutants were 5 × 10⁶ CFU. The challenged mice were observed for 7 days, and death or survival was recorded.

Envelope stress assays

The XM, ∆nhaA mutant, and ∆nhaA-complemented strains were initially cultured in LB medium until their OD₆₀₀ values reached 0.4. For the high EDTA and SDS resistance assays, LB-grown bacteria (OD₆₀₀ ≈ 0.4) were washed thrice with double distilled water, and a 10-fold serial dilution was performed. Next, 10 µL of each dilution was spotted onto LB plates and LB plates supplemented with 1 mM EDTA, 2 mM EDTA, or 2% SDS. Next, the plates were incubated for 36 h at 37°C. For the high osmotic pressure resistance assay, the bacteria were grown in LB medium or LB medium containing 4% NaCl and incubated at 37°C with shaking at 180 rpm. Afterward, the optical density was measured every hour at 600 nm to assess bacterial growth. For the RNase leakage assay (39, 40), the bacterial suspension from a single fresh colony was adjusted to OD₆₀₀ ≈ 0.4 with double distilled water. Next, 10 µL of the culture was spotted onto LB agar plates containing 0.2% (wt/wt) yeast RNA and 2.5% (wt/wt) toluidine blue O and incubated for 24 h at 37°C.

Cell culture and infection

RAW264.7 (ATCC) murine macrophage cells were cultured in Dulbecco’s minimal essential medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum and incubated at 37°C in a 5% CO₂ atmosphere. For quantitative mass proteomics analysis, RAW264.7 cells (2.5 × 10⁶ cells/well) were seeded onto 12-well tissue culture plates and inoculated with bacterial culture diluted in DMEM (5% FBS) at a multiplicity of infection of 100:1. After centrifugation at 500 × g for 5 min at 25°C, the plates were incubated for 1 h at 37°C in 5% CO₂. The infected cells were washed once in PBS and incubated in DMEM containing 200 µg/mL gentamicin at 37°C in 5% CO₂ for the indicated times. Subsequently, three washes with PBS were performed, and the cells were lysed with 1 mL of water. Serial dilutions of the antibiotic on LB were used to determine the intracellular CFU.

Serum survival assay

Overnight bacterial cultures in LB were diluted 1:100 and grown to an OD₆₀₀ of 0.4–0.6 at 37℃ with shaking at 180 rpm. Next, the strains were washed twice with PBS, and the OD₆₀₀ values were normalized to 1.0 (5 × 10⁸ CFU/mL). The cultures were inoculated at 1:100 in 90% NAS (90% NAS + 10% LB) or 50% NHS (50% NHS +50% LB). After 4-h incubation, live bacterial counts were determined by plating the appropriate dilutions on LB agar plates.

Western blot analysis

The bacterial strain was initially cultured overnight and then transferred at a 1:100 dilution into LBK medium, a medium equivalent to LB but with an equimolar concentration of KCl in the place of NaCl. The culture was incubated at 37°C with agitation at 180 rpm for 4 h. Following this incubation, bacterial cells were harvested and suspended in PBS. The resuspended cells were subjected to sonication and then centrifuged at 5,000 × g for 10 min to collect the supernatant. After subsequent centrifugation at 100,000 × g for 1 h, the supernatant was discarded, and the resultant pellet was washed with PBS. The pellet underwent another centrifugation at 100,000 × g for 1 h to obtain the final pellet, which was then resuspended in a solubilization buffer (1.5% n-dodecyl-β-D-maltopyranoside (DDM), 150 mM 3-(N-morpholino) propane sulfonic acid (MOPS), 30% glycerol, 3 mM dithiothreitol (DTT)). This procedure yielded membrane proteins for further analysis.

For protein analysis, either membrane proteins or whole bacterial proteins were separated by 12.5% SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane (Merck). The PVDF membranes were probed with a polyclonal antibody against Flag (Abcam, Cambridge, MA, USA; ab18230), and detection was performed using an anti-mouse IgG antibody labeled with KPL DyLight 800 (Thermo Fisher, SA5-10176). Antibodies against GroEL (Abcam, ab90522) or RpoB (Abcam, ab191598) served as loading controls for the western blot analysis. Membrane images were captured using an Azure500 imager (Azure Biosystems, CA, USA).

Isolation of membrane vesicles and Na⁺/H⁺ antiporter activity determination

Na⁺/H⁺ antiporter activity assays were carried out using everted membrane vesicles (59), which were obtained as follows: the bacteria were cultured in LBK medium at 180 rpm and 37°C for 5 h, then centrifuged at 5,000 × g for 10 min. The collected bacteria were suspended in Buffer A (consisting of 10 mM Tris-HCl, 0.14 M choline chloride, 0.5 mM dithiothreitol, and 0.25 M sucrose at pH 7.4). After sonication, supernatants were obtained through centrifugation. The resulting membrane vesicles were once more pelleted by centrifugation at 100,000 × g for 1 h and washed twice with Buffer A. The pellet was then suspended in Buffer A and stored in small amounts at −80°C.

The assay for antiport activity was based on the measurement of Na⁺-induced changes at the ΔpH as described (36, 59). The reaction mixture comprised Buffer B (10 mM Tris-HC1, 0.14 M choline chloride, and 0.5 mM dithiothreitol, pH as indicated) and 1 µM acridine orange in a 0.2 mL volume in 96-well plates. Each assay contained 100 µg of everted membrane vesicles. Quenching was initiated with the addition of 10 mM dl-lactate. After stabilization of the fluorescence signal, 100 mM NaCl was added. Fluorescence was measured using excitation and emission pairs of 490 and 530 nm.

Confocal microscopy

For microscopic analysis, the strains were grown in the medium for the indicated time. Next, the medium was removed, and the cell pellet was washed twice with PBS. The pellet was then resuspended in 20 µg/mL Hoechst (Thermo Fisher) and incubated at room temperature for 20 min. The pellet was then incubated with 20 µg/mL FM4-64 (Biorigin, Beijing, China) at 4°C for 1 min. A 5 µL aliquot of the culture was then immobilized on adhesion microscope slides. After air-drying, the slides were cover-slipped and imaged using the confocal laser scanning microscopy platform Zeiss LSM880 and LSM800 (Oberkochen, Germany).

Transmission electron microscopy analysis

The strains were cultured in LB or 90% NAS for 2 h at 37°C with continuous shaking at 180 rpm. Next, the cultures were harvested by centrifugation at 5,000 × g for 5 min before being washed twice with double distilled water and fixed in 2.5% glutaraldehyde. The specimens were dehydrated in propylene oxide for 10 min and embedded in an epoxy resin. All samples were sent to the Harbin Veterinary Research Institute for electron microscopy analysis (Hitachi H-7650, Tokyo, Japan).

Flow cytometry analysis

The strains (5 × 10⁶ CFU) were incubated at 37°C in 50% NHS for different periods. Next, the cultures were washed thrice with PBS. For the deposition of C3b, bacteria were incubated with FITC-conjugated anti-C3b antibody (Abcam) for 30 min at room temperature. ExPEC XM incubated in HIHS with FITC-conjugated anti-C3b antibody was used as a control. The bacteria were incubated with antibodies against C9 (Abcam) at room temperature for 30 min. Following three washes with PBS, the primary antibody-labeled bacteria were incubated with FITC-conjugated secondary antibodies (Abcam) for 30 min at room temperature. The bacteria labeled with FITC-conjugated secondary antibodies, without incubation with primary antibodies, were used as controls. The surface deposition of molecules was analyzed using a multiparameter flow cytometer (Apogee, Hertfordshire, UK).

Statistical analysis

Survival was analyzed using the Kaplan–Meier survival method with a log-rank (Mantel-Cox) test, and all other binary comparisons were analyzed using the Student’s t-test. P < 0.05 was considered statistically significant.

ETHICS APPROVAL

All animal experimental procedures were conducted following the Beijing Administration Guidelines for the Use of Laboratory Animals and were approved by the Review Board of the Harbin Veterinary Research Institute and the Animal Care and Use Committee of Heilongjiang Province (220929-01-GR).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/iai.00039-23.

Fig.S1 to S3, Tables S1 to S2. iai.00039-23-s0001.docx.

Fig. S1 (Validation of NhaA protein expression in the membranes for site and loop mutations), Fig. S2 (Validation of pH sensing and regulation functions of nhaA mutant strains with site mutations H225R and G338S), Fig S3 (Na+/H+ antiporter activity of mutant strains with the G338S and A127V site mutations), Table S1 (Strains and plasmids), and Table S2 (Oligonucleotides).

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.