A 6His-2Flag dual-tag strategy enables high-yield production of antimicrobial peptide N6 in Pichia pastoris to combat multidrug-resistant Klebsiella pneumonia

Abstract

In response to the global threat posed by multidrug-resistant Klebsiella pneumoniae, this study established an efficient heterologous expression system for the antimicrobial peptide N6 in Pichia pastoris X-33. By employing an innovative 6His-2Flag dual-tag fusion strategy coupled with formic acid-mediated cleavage, we achieved a high-yield production of N6, with a final titer of 1.02 g/L—significantly surpassing previous reports and meeting the threshold for scalable industrial production. The expressed N6 exhibited potent and broad-spectrum antibacterial activity against clinical isolates of K. pneumoniae, with MIC values ranging from 2 to 8 μg/mL. Mechanistic studies revealed that N6 exerted rapid bactericidal and biofilm-eradicating effects through a multimodal action involving membrane disruption, metabolic interference, and induction of oxidative stress. In a murine systemic infection model, the N6 treatment significantly improved survival rates of mice to 70% and reduced bacterial burdens in key organs. This work not only identifies N6 as a promising therapeutic candidate against drug-resistant bacterial infections, but also provides a streamlined and cost-effective synthetic biology platform for the high-level production of antimicrobial peptides, thereby facilitating their clinical translation and industrial application.

Graphical abstract

Highlights

• •An efficient synbio platform for industrial-scale antimicrobial peptide production.
• •Achieves industrial-scale production of peptide N6 with a titer of 1.02 g/L.
• •N6 shows potent bactericidal and biofilm-eradicating activity against K. pneumoniae.
• •In murine infection models N6 increases survival to 70% and mitigates inflammation.
• •N6 kills bacteria via a multimodal mechanism.

1. Introduction

Multidrug-resistant (MDR) bacterial infections have evolved into a pressing global public health crisis. Among these pathogens, Klebsiella pneumoniae, a clinically prevalent Gram-negative opportunistic bacterium, acquires resistance through multiple mechanisms, including biofilm formation, efflux pump upregulation, and horizontal transfer of resistance genes [1,2]. Particularly concerning is the emergence of carbapenem-resistant hypervirulent Klebsiella pneumonia (CR-HVKP), which integrates “refractory resistance” with “high pathogenicity,” posing formidable challenges to clinical treatment [3]. Thus, there is an urgent need to develop novel alternative antimicrobial strategies.

Antimicrobial peptides (AMPs) represent highly promising therapeutic candidates owing to their rapid bactericidal activity, broad antimicrobial spectrum, and low tendency to induce resistance [4]. Marine invertebrate-derived AMPs exhibit remarkable structural diversity and functional specificity, attributed to their long-term adaptation to complex and extreme marine environments [5]. For instance, the arenicin-3 variant NZ17074, derived from the marine invertebrate lugworm Arenicola marina, exhibits potent antibacterial activity against Gram-negative bacteria; however, its inherent cytotoxicity to host cells hinders efficient microbial expression. To address this, N6, a derivative generated by substituting alanine for cysteine residues at positions 3 and 20, was developed. This modification preserves core antimicrobial activity while substantially reducing cytotoxicity, laying a foundation for heterologous expression [6].

Previous studies adopted a SUMO-tag fusion strategy for N6 expression in Pichia pastoris. Although this approach improved fusion protein yield by mitigating toxicity through the SUMO tag, the final yield of active N6 peptide after cleavage and purification was merely 9.7 mg/L [7]. Furthermore, the large molecular weight of the SUMO tag significantly reduces the relative proportion of the small-molecule AMP. Its dependence on SUMO protease for cleavage introduces cumbersome operational steps and increases production costs [8], severely limiting its large-scale industrial application.

Therefore, optimizing heterologous expression systems and tagging strategies is critical for advancing the industrialization of AMPs. As a mature eukaryotic expression platform, P. pastoris combines rapid proliferation with post-translational modification capabilities, effectively addressing folding and stability issues of AMPs [9]. Compared with macromolecular tags such as SUMO, small-molecule tags (e.g., Flag) offer distinct advantages, including low molecular weight, high hydrophilicity, and minimal interference with target protein function [10]. Additionally, commercially available specific antibodies against these tags enable one-step affinity purification and convenient detection, streamlining the production workflow [11]. For tag removal, chemical cleavage methods are more cost-effective, simple, and scalable than expensive enzymatic digestion, making them particularly suitable for large-scale production of small-molecule active peptides such as AMPs [8,12].

Despite the initial establishment of a N6 expression system, critical questions remain regarding its efficacy against clinically K. pneumoniae, underlying mechanism of action, and in vivo therapeutic potential [7]. As a major clinically significant pathogen, K. pneumoniae exhibits a continuously increasing resistance rate to traditional antibiotics, highlighting an urgent need for novel therapeutics [13].

To address the limitations of low yield and complex purification workflows in existing systems, we developed an efficient production strategy for N6 using the P. pastoris platform, with an innovative 6His-2Flag dual-tag design coupled with formic acid-mediated cleavage. We systematically evaluated the antibacterial activities of N6 against clinical K. pneumoniae isolates and its synergistic effects with conventional antibiotics. Furthermore, we validated its in vivo therapeutic efficacy in a murine infection model. This work not only identifies N6 as a promising therapeutic candidate against K. pneumoniae but also establishes a streamlined, cost-effective synthetic biotechnology platform for scalable AMP production—thereby accelerating their clinical translation and industrial application.

2. Materials and methods

2.1. Strains and cell lines

Strains: Nine clinically isolated strains of Klebsiella pneumoniae (76, 77, 111, 113, K11, S469, YB43, YB47, JB22) were provided by Ningbo Municipal Center for Disease Control and Prevention; Staphylococcus aureus ATCC 43300, Staphylococcus epidermidis ATCC 12228, Streptococcus agalactiae ATCC 13813, and Salmonella typhimurium ATCC 14028 were obtained from the American Type Culture Collection (ATCC); Staphylococcus aureus CVCC 546 and Salmonella pullorum CVCC 1802 were sourced from the China Veterinary Culture Collection Center (CVCC); Candida albicans CICC 98001 was acquired from the China Center of Industrial Culture Collection (CICC); Pichia pastoris X-33 was purchased from Invitrogen (Beijing, China). Human normal bronchial epithelial cells (BEAS-2B) were kindly provided by Dr. Xiao Wang from Ningbo University.

2.2. Expression, purification, cleavage, and identification of N6

2.2.1. Construction of the expression vector

The 6His-2Flag-N6 gene sequence was optimized based on codon usage preference in P. pastoris (www.kazusa.or.jp/codon/) using the Reverse Translate Tool (www.bio-informatics.org/sms2/rev_trans.html). The DNA fragment was subsequently ligated into the pPICZαA plasmid with T4 DNA ligase. The recombinant vector was linearised with PmeI and electroporated into competent P. pastoris X-33 cells, with the empty pPICZαA plasmid as a negative control. Positive transformants were screened on YPDS medium (1% yeast extract, 2% glucose, 2% peptone, 2% agar, 1 M sorbitol) containing 100 μg/mL Zeocin and further verified by colony PCR [14].

2.2.2. Expression of 6His-2Flag-N6

Positive transformants were selected in YPD liquid medium supplemented with 100 μg/mL Zeocin and cultured at 30°C for 14–18 h. Subsequently, the culture was transferred to 200 mL BMGY medium, with 1% methanol added every 24 h for a total induction duration of 120 h. After induction, the supernatant fermentation solution was collected and subjected to verification by SDS-PAGE. The strain exhibiting the highest expression level was selected for 5 L scale fermentation expression, which was performed as described previously [15].

2.2.3. Purification, cleavage, and characterization of 6His-2Flag-N6

The fermentation broth was centrifuged at 5000 rpm for 30 min, and the supernatant was collected to purify the 6His-2Flag-N6 fusion protein using nickel affinity chromatography. The purified product was desalted, lyophilized, and then cleaved in a 50% formic acid solution at 50°C. Following cleavage, the reaction mixture was subjected to a second round of nickel affinity chromatography. Since the target N6 protein lacks a histidine tag, it did not bind to the resin and was collected in the flow-through fraction, effectively achieving tag removal. The cleavage efficiency was verified by Tricine-SDS-PAGE, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), and zone of inhibition assays.

2.3. Antimicrobial resistance and pathogenicity of K. pneumoniae

2.3.1. Antimicrobial susceptibility of K. pneumoniae

Antimicrobial susceptibility testing of nine clinically isolated K. pneumoniae strains was determined by the Kirby-Bauer disc diffusion method, in accordance with Clinical and Laboratory Standards Institute (CLSI) guidelines [16]. Bacterial suspensions were adjusted to a 0.5 McFarland standard. Sterile cotton swabs were used to uniformly spread the bacterial suspension onto MHA plates. Antibiotic discs were placed on the agar surface with sterile forceps. Three parallel discs were placed for each antibiotic, spaced at least 24 mm apart and at least 15 mm from the plate edge. Following incubation at 37°C for 18 h, the inhibition zone diameters were measured and interpreted based on CLSI criteria. Notably, colistin susceptibility testing exclusively employs the broth microdilution method, with breakpoints limited to Intermediate (I) and Resistant (R) [16].

2.3.2. Detection of virulence genes in K. pneumoniae

The presence of 13 virulence-associated genes (peg344, iutA, mrkD, allS, kfu, entB, aerobactin, iroN, rmpA2, magA, rmpA, ybtS, fimH) was detected by PCR, with primer sequences referenced from Xu [17]. Bacterial genomic DNA was extracted as the PCR template using a DNA extraction kit (Tiangen Biochemical Technology Co., Ltd., China) following the manufacturer's instructions. The total PCR reaction volume was 25 μL, consisting of 2.5 μL 10× PCR Buffer, 2 μL dNTPs, 0.5 μL Taq DNA polymerase, 3 μL DNA template, 1 μL each of forward and reverse primers, and ddH₂O to a final volume of 25 μL. Amplification was performed on an Applied Biosystems PCR instrument (USA) with the following program: 5 min pre-denaturation at 94°C; followed by 30 cycles (denaturation at 94°C for 30 s, annealing at 58°C for 30 s, extension at 72°C for 20 s); concluding with a final extension at 72°C for 5 min. PCR products were separated by 2% agarose gel electrophoresis and visualised using a Geliance 200 gel imaging system (PerkinElmer, USA).

2.4. Antimicrobial and antibiofilm activity of N6

2.4.1. Minimum inhibitory concentration and minimum bactericidal concentration

The minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) of N6 were determined using the microdilution broth method [18]. Briefly, bacterial strains were cultured in MHB medium at 37°C until the logarithmic growth phase. Bacterial suspensions (1 × 10⁵ CFU/mL) were added to a 96-well plate containing twofold serial dilutions of N6 (0.5–128 μg/mL), followed by incubation at 37°C for 18–24 h. PBS was used as the negative control and MHB as the blank control. The MIC value is defined as the lowest concentration of N6 at which bacterial growth is not observed. For MBC determination, samples from wells with no visible growth in the MIC assay were subcultured onto MHA plates. After further incubation, the MBC was recorded as the lowest concentration that resulted in no colony formation. All experiments were performed in triplicate.

2.4.2. In vitro bactericidal kinetics

The time-killing curves for N6 against K. pneumoniae were determined in vitro as previously described [19]. Briefly, K. pneumoniae S469 in the logarithmic growth phase was diluted to 1 × 10⁵ CFU/mL in fresh MHB medium and cultured with shaking at 37°C and 250 rpm. N6 was added to 10 mL bacterial suspensions (in 50 mL conical flasks) at final concentrations of 1×, 2×, and 4× MIC to evaluate bactericidal efficacy. Aliquots of 100 μL were sampled at 0, 0.5, 1, 2, 4, 6, 8, 10, 12, and 24 h post-incubation for colony counting. Colistin was used as the positive control.

2.4.3. Synergy with antibiotics

The checkerboard microtiter assay was employed to determine the combined efficacy of N6 with various antibiotics, including Lincomycin, Vancomycin, Colistin, Erythromycin, Gentamicin, Kanamycin, and Chlortetracycline Hydrochloride. In brief, 180 μL of log-phase K. pneumoniae S469 (1 × 10⁵ CFU/mL) was dispensed into 96-well plates, followed by 10 μL of N6 (1/32× to 4× MIC) and 10 μL of antibiotics. Combination efficacy was assessed using the fractional inhibitory concentration (FIC) index: FIC index (FICI) = MIC (peptide in combination)/MIC (peptide alone) + MIC (antibiotic in combination)/MIC (antibiotic alone). The FICI ≤ 0.5 indicates synergistic effects, 0.5 < FICI ≤ 1 indicates additive effects, 1 < FICI ≤ 4 indicates no interaction, and FICI > 4 indicates antagonistic effects [20].

2.4.4. Biofilm elimination assay

The crystal violet staining method was employed to evaluate the inhibitory effect of peptides on biofilm formation by K. pneumoniae and their ability to disrupt biofilms [21]. Untreated bacteria served as the negative control (A), while fresh MHB medium acted as the blank control (A₀). Biofilm biomass was quantified using the following formula: Biofilms (%) = ((A_peptide−A₀)/(A−A₀)) × 100%.

Effect on nascent biofilms: Log-phase K. pneumoniae S469 (1 × 10⁸ CFU/mL, 180 μL) was added to 96-well plates, followed by peptide solutions (0.5–16× MIC, 20 μL). The plates were incubated at 37°C for 24 h. After incubation, non-adherent bacteria were gently washed away with sterile PBS. Biofilms were stained with 1% crystal violet for 30 min, excess stain was removed by rinsing gently, and the bound crystal violet was dissolved in DMSO. The absorbance of the resulting solution was measured at 600 nm using a SUNRISE microplate reader (TECAN, Austria).

Effect on mature biofilms: To assess the peptide's disruptive effect on established biofilms, bacterial suspensions were added to 96-well plates and incubated at 37°C for 24 h to form mature biofilms. Following incubation, non-adherent bacteria were gently washed away. Subsequently, wells were supplemented with peptide solutions at varying concentrations and incubated for a further 24 h. Residual biofilms were stained with crystal violet as described above, and absorbance values were measured.

2.5. In vitro and in vivo effects of N6 on K. pneumoniae

2.5.1. Cytotoxicity of N6 towards BEAS-2B cells

The cytotoxicity of N6 was assessed using the CCK-8 assay [22]. The procedure was as follows: BEAS-2B cells were seeded into a 96-well plate at a density of 1 × 10⁵ cells/mL (100 μL per well) and incubated for 24 h. After removing the culture medium, an equal volume (100 μL) of N6 solution was added to each well, with final concentrations ranging from 1 to 128 μg/mL. Cells were further incubated for 24 h. Subsequently, 100 μL of 10% CCK-8 solution was added, and incubation continued for 2 h. Optical density at 450 nm was measured using a microplate reader. The cell viability formula is: Cell viability (%) = (A_s−A_b)/(A_c−A_b) × 100% (where A_s is the absorbance value of the experimental group, A_c is the absorbance value of the PBS group, and A_b is the absorbance value of the blank group).

2.5.2. Treatment of BEAS-2B cell infection with N6

BEAS-2B cells were seeded at a density of 1 × 10⁵ cells per well in a 48-well plate. Following culture to approximately 80% confluence, K. pneumoniae S469 was then added at a multiplicity of infection (MOI) of 5 for 4 h under the same culture conditions. Subsequently, the infected cells were divided into groups for treatment. N6 and colistin treatment groups: Incubated with N6 and colistin respectively at final concentrations of 32–128 μg/mL. Infected group: after infection, add antibiotic-free DMEM medium. Control group: infection without inoculum; all other procedures identical to the infected group. Finally, culture supernatants were collected. Inflammatory cytokine levels IL-10 and IL-1β were measured using ELISA, and cell viability was assessed via the CCK-8 assay.

2.5.3. Treatment of K. pneumoniae-induced infection in mice with N6

This experimental protocol was approved by the Animal Ethics Committee of the Chinese Academy of Agricultural Sciences (Approval No.: IFR-CAAS20251015). The experimental procedures strictly adhered to the relevant provisions of the Guidelines for Ethical Review of Laboratory Animal Welfare issued by the Ministry of Science and Technology of China in 2018. Forty female Balb/c mice (SPF, 19–21 g, six weeks old) were procured from Beijing VTLH Laboratory Animal Technology Co., Ltd. (Beijing, China). Mice were housed under a 12 h light/dark cycle at 24°C ± 2°C and 45 ± 15% humidity with ad libitum access to food and water. Following a 3-day acclimatisation period, all mice were randomly divided into four groups (10 mice per group, housed five per cage). Infection was induced via intranasal inoculation, with each mouse receiving a 20 μL inoculum. The infected and treated groups received a bacterial suspension containing 5 × 10⁷ CFU of K. pneumoniae YB47, while the CK group received an equivalent volume of PBS. Four hours post-challenge, the N6 treatment group and Kanamycin sulphate treatment group received intraperitoneal injections of their respective drugs at a dose of 10 mg/kg. Daily survival status was recorded for survival analysis. On day 7 post-infection, lung and brain tissues were collected for colony-forming unit (CFU) counting and histopathological analysis. Concurrently, blood samples were obtained to measure inflammatory cytokine levels: IL-10, IL-1β, IL-6, and TNF-α. Serum concentrations of inflammatory cytokines were quantified using ELISA kits purchased from Beijing Kangjia Hongyuan Co., Ltd. Harvested lung and brain tissues were rinsed with PBS, fixed in 4% paraformaldehyde at 4°C for 24 h, dehydrated through a graded ethanol series (75–95%), and cleared with xylene. Subsequently, the tissues were embedded in paraffin, sectioned, and stained with hematoxylin-eosin (HE). Finally, the tissue sections were mounted on glass slides and observed under a light microscope.

2.6. Mechanism of bactericidal action

2.6.1. Effect of N6 on the cell membrane of K. pneumoniae

To evaluate the disruption of bacterial membrane integrity by the peptide, K. pneumoniae S469 in the logarithmic growth phase (1 × 10⁹ CFU/mL) was treated with N6 or colistin (1×, 2×, and 4× MIC) for 30 min, 60 min, 90 min, and 120 min. After two washes with PBS, the bacterial cells were stained with propidium iodide (PI, 50 μg/mL) in the dark for 15 min. Detection was performed using a FACS Calibur flow cytometer (BD, USA), with the proportion of PI-positive cells calculated. Untreated bacterial suspension served as the negative control.

2.6.2. Effect of N6 on the morphology of K. pneumoniae cells

The effect of N6 on bacterial cell morphology was assessed by both scanning electron microscopy (SEM) and transmission electron microscopy (TEM). For SEM, bacteria in logarithmic growth phase (1 × 10⁹ CFU/mL) were treated with 4× MIC N6 or colistin at 37°C for 2 h. After treatment, the samples were fixed with 2.5% glutaraldehyde, dehydrated using a gradient ethanol series, and dried at the carbon dioxide critical point. Palladium/gold sputter coating was applied prior to examination and finally observed using a Hitachi SU8000 scanning electron microscope. For TEM, to further examine bacterial morphology and intracellular changes, strains were pretreated as described above, then sequentially processed through fixation, dehydration, embedding, sectioning, and staining steps before observation using a Hitachi H7650B transmission electron microscope.

2.6.3. Effects of N6 on K. pneumoniae metabolism

2.6.3.1. Metabolic activity assay

K. pneumoniae S469 cultured in MHB medium was adjusted to 1 × 10⁶ CFU/mL and incubated with various concentrations (1×, 2×, 4× MIC) of N6 or colistin for 6 h (37°C, 250 rpm). Subsequently, 10 μL of AlamarBlue solution was added, and the mixture was further incubated at 37°C for 1 h. Absorbance was measured at 571 nm. Untreated bacterial cultures were used as the negative control.

2.6.3.2. Intracellular ATP assay

K. pneumoniae S469 was cultured in MHB medium and adjusted to 1 × 10⁸ CFU/mL. The bacterial suspension was then treated with N6 and colistin at different concentrations (1×, 2×, 4× MIC) for 60 min at 37°C. ATP content was subsequently determined using the BacTiter-Glo™ Microbial Cell Viability Assay Kit (Promega) according to the manufacturer's instructions. Fluorescence luminescence values were recorded using a Tecan Infinite M1000 Pro microplate reader.

3. Results

3.1. Expression, purification, cleavage, and identification of N6

3.1.1. Construction of the recombinant plasmids

The 6His-2Flag-N6 gene was subjected to enzymatic digestion using XhoI and XbaI, followed by insertion into plasmid pPICZαA and transformation into Escherichia coli DH5α, resulting in the generation of the recombinant vector pPIC-6His-2Flag-N6 (Fig. 1A). Subsequently, the PmeI was employed for linearizing the recombinant vector and transferring it to P. pastoris X-33. Positive transformants were selected using 100 μg/mL Zeocin.

Fig. 1.

Expression, purification, and cleavage of 6His-2Flag-N6. (A) Schematic diagram of recombinant expression plasmid pPIC-6His-2Flag-N6. (B) Time course of total secreted protein levels and cell wet weight during high-density fermentation. Samples were taken at the indicated time points (0, 24, 48, 72, 96, 120 h) post-induction. (B) SDS-PAGE analysis of the purification process. Lane M: protein molecular weight marker; Lane 1: unpurified fermentation supernatant; Lane 2: flow-through (unbound material); Lane 3: elution fraction at 25 mM imidazole; Lane 4: elution fraction at 250 mM imidazole.

3.1.2. High-density expression of the 6His-2Flag-N6 fusion protein

Positive transformants were cultivated by high-density fermentation in a 5 L bioreactor. The expression of the target fusion protein increased throughout the induction period, reaching a maximum at 120 h. At this peak expression point, the total secreted protein concentration and biomass reached 1.66 g/L and 0.4 g/mL, respectively (Fig. 1B).

3.1.3. Purification, cleavage, and functional characterization of the target protein

The 6His-2Flag-N6 fusion protein was purified using nickel affinity chromatography. SDS-PAGE analysis of the 250 mM imidazole elution fraction showed a single prominent band, indicating high purity (Fig. 1C, lane 4). To systematically evaluate the process economics, stepwise recovery rates were calculated based on three parallel batches in a 5 L fermenter. The total secreted fusion protein in the fermentation supernatant reached 1.66 g/L, and after the initial Ni-NTA affinity chromatography, 1.31 g/L of purified fusion protein was obtained, corresponding to a recovery rate of 78.9%, with losses primarily attributed to supernatant impurities and unbound fusion proteins.

The purified fusion protein was then subjected to formic acid cleavage, which efficiently separated the target N6 peptide from the 6His-2Flag tag, with complete cleavage achieved at 66 h (Fig. 2A). Analysis of the crude cleavage reaction mixtures revealed antibacterial activity against S. aureus 43300 (Fig. 2B), confirming the successful generation of a bioactive product. The tag-free N6 peptide was subsequently purified (Fig. 2C). Following cleavage, a second Ni-NTA chromatography step was employed to remove the His-tagged residuals, and the flow-through containing the native N6 was collected. The combined recovery rate for the cleavage and secondary purification steps was 77.9%. Thus, the overall recovery rate from the fermentation supernatant to the final purified N6 was calculated as approximately 61.5%, culminating in a final yield of 1.02 g/L with a purity of over 93% (Fig. 2F).

Fig. 2.

Cleavage of 6His-2Flag-N6 and functional characterization of N6. (A) Analysis of the formic acid cleavage time course. Lane M: protein molecular weight marker; Lanes 1-5: cleavage products sampled at 48, 54, 60, 66, and 72 h, respectively. (B) Antibacterial activity of the cleavage reaction mixtures (from panel C) against S. aureus 43300. (C) SDS-PAGE analysis of the purified N6 protein after cleavage. Lane M: Protein molecular weight marker; Lane 1: Fusion protein 6His-2Flag-N6; Lane 2: Flow-through fraction from the nickel column after formic acid cleavage. (D) MALDI-TOF MS analysis of the purified N6 protein. (E) Antibacterial activity of the purified N6 protein against K. pneumoniae S469 at varying concentrations. (F) Identification of N6 protein purity by HPLC. (G) LC-MS/MS fragment coverage map of the purified N6 peptide. (H) LC-MS/MS tandem mass spectrum of the purified N6 peptide.

Its molecular weight was verified as 2475.9 Da by MALDI-TOF MS, matching the theoretical value (Fig. 2D). To further confirm primary structural integrity, we performed LC-MS/MS peptide sequencing. The MS/MS fragment ions provided comprehensive coverage of both the N- and C-terminal regions, demonstrating that no truncation or modification occurred at either terminus, and the full sequence was identical to the theoretical sequence (Fig. 2G and H). Furthermore, the purified N6 peptide exhibited potent, concentration-dependent antibacterial activity against K. pneumoniae S469 (Fig. 2E).

3.2. Antimicrobial resistance and pathogenicity of K. pneumoniae

3.2.1. Drug sensitivity of K. pneumoniae

As shown in Table 1, antibiotic susceptibility testing revealed that all tested strains were resistant to ampicillin and imipenem, while most were susceptible to second and third-generation cephalosporins and meropenem. Notably, strains 113 and 111 showed partial resistance to certain cephalosporins. All tested strains had low MIC values to colistin.

Table 1.

Antibiotic susceptibility of 9 clinically isolated K. pneumoniae strains.

Antibiotic	Klebsiella pneumonia Strains
	76	77	111	113	K11	S469	YB43	YB47	JB22
Ampicillin	R	R	R	R	R	R	R	R	R
Imipenem	R	R	R	R	R	R	R	R	R
Cefotaxime	S	R	R	R	S	S	S	S	S
Cefuroxime	S	R	R	R	S	S	S	S	S
Ceftazidime	S	S	S	S	S	S	S	S	S
Cefoxitin	S	S	S	S	S	S	S	S	S
Meropenem	S	S	S	S	S	S	S	S	S
Gentamicin	S	S	S	S	I	S	S	S	I
Amikacin	S	S	S	S	S	S	S	S	S
Streptomycin	S	S	S	S	S	S	S	S	S
Ciprofloxacin	R	S	S	S	S	S	S	S	S
Nitrofurantoin	S	S	S	S	I	S	S	S	S
Enoxacin	I	S	S	S	S	S	S	S	S
Doxycycline	S	S	R	R	S	S	S	S	S
Colistin	I	I	I	I	I	I	I	I	I

Susceptible, intermediate, and resistant strains are abbreviated as S, I, and R, respectively.

3.2.2. Detection of virulence genes in K. pneumoniae

Virulence genes were detected in nine clinically isolated K. pneumoniae strains, and the results are presented in Fig. 3. All strains carried at least four virulence genes: strains K11 and YB47 harbored the highest number (six genes), followed by strains 77 and JB22 (five genes). The distribution of virulence genes varied significantly: fimH (8/9) and entB (8/9) exhibited the highest carriage rates; ybtS (7/9) and peg344 (6/9) were also relatively common; magA (4/9), iutA (2/9), and rmpA (2/9) showed relatively low carriage rates.

Fig. 3.

Detection of virulence genes in clinical K. pneumoniae isolates by PCR. Lane M represents DNA Marker II, band sizes from top to bottom are 1500, 900, 700, 500, 400, 200, 100 bp; letters A–I correspond to different strains: A: 76, B: 77, C: 111, D: 113, E: K11, F: S469, G: YB43, H: YB47, I: JB22; Lanes 1-14 correspond to detection targets: 1: CK, 2: peg344, 3: iutA, 4: mrkD, 5: alls, 6: Kfu, 7: entB, 8: Aerobactin, 9: IroN, 10: rmpA2, 11: magA, 12: rmpA, 13: ybtS, 14: fimH.

3.3. Antibacterial activity of N6

3.3.1. MIC and MBC results for N6

The antibacterial activity of N6 was evaluated by determining its MIC and MBC against various bacterial and fungal strains. As shown in Table 2, N6 exhibited prominent bacteriostatic activity against Gram-negative bacteria (MIC 1–8 μg/mL). Notably, for nine clinically isolated K. pneumoniae strains, the MIC values fell within the range of 2–8 μg/mL. Against Gram-positive bacteria, it demonstrated only moderate activity, exhibiting weak inhibition against S. agalactiae (MIC ≥ 64 μg/mL). Furthermore, N6 also exerted a certain inhibitory effect against C. albicans (MIC 32 μg/mL).The MBC/MIC ratios for all clinical K. pneumoniae isolates ranged from 1 to 4, which meets the bactericidal criterion (≤4).

Table 2.

MIC and MBC values for N6.

Strains	MICs (μg/mL)	MBCs (μg/mL)
Gram-positive bacteria
S. aureus ATCC 43300	8	16
S. aureus CVCC 546	8	8
S. epidermidis ATCC 12228	16	32
S. agalactiae ACCC 61733	>128	>128
S. agalactiae ATCC 13813	64	>128
S. agalactiae PBSA0903	64	>128
Gram-negative bacteria
E. coli ATCC 25922	1	2
E. coli CMCC 44102	2	2
S. typhimurium ATCC 14028	2	2
S. pullorum CVCC1802	8	8
P. aeruginosa CMCC 10104	4	8
P. aeruginosa O15	2	8
A. baumannii 786	2	4
A. baumannii 754	2	8
K. pneumoniae 76	2	2
K. pneumoniae 77	4	8
K. pneumoniae 111	4	4
K. pneumoniae 113	2	4
K. pneumoniae K11	2	2
K. pneumoniae S469	2	2
K. pneumoniae YB43	8	16
K. pneumoniae YB47	2	8
K. pneumoniae JB22	2	4
Fungus
Candida albicans CICC 98001	32	64

3.3.2. In vitro bactericidal kinetics

As depicted in Fig. 4, the bactericidal curves show that N6 at 1×, 2×, and 4× MIC completely eradicated K. pneumoniae S469 within 1 h post-treatment, with no subsequent bacterial rebound. The bactericidal kinetics exhibited a concentration-dependent pattern. In contrast, while colistin exerted concentration-dependent bacteriostatic effects, it failed to achieve complete bacterial eradication.

Fig. 4.

Time-killing curves of N6 and colistin against K. pneumoniae S469.

3.3.3. Synergy with antibiotics

Table 3 presents the in vitro susceptibility test results of K. pneumoniae S469 treated with antimicrobial peptide N6 in combination with seven antibiotics. N6 demonstrated an additive effect in combination with gentamicin (FICI = 0.5625) and erythromycin (FICI = 1.0). Notably, the combination with gentamicin reduced its MIC value by 16-fold. In contrast, combinations with the other five antibiotics (lincomycin hydrochloride, vancomycin, colistin, kanamycin, and chlortetracycline hydrochloride) yielded indifferent effects (FICI > 1), indicating no synergistic or additive interaction between N6 and these agents.

Table 3.

In vitro susceptibility of K. pneumoniae S469 to combination therapy with antimicrobial peptide N6 and antibiotics.

Combination	Variety	MICp/MICa	MICc	FIC	FICI
N6+Lin	N6	2	2	1	1.5
	Lin	64	32	0.5
N6+Van	N6	2	2	1	1.25
	Van	64	16	0.25
N6+ Ery	N6	2	1	0.5	1
	Ery	32	16	0.5
N6+ Col	N6	2	2	1	2
	Col	1	1	1
N6+ Kan	N6	2	2	1	2
	Kan	2	2	1
N6+CTC HCL	N6	2	1	0.5	1.5
	CTC HCL	1	1	1
N6+ Gen	N6	2	1	0.5	0.5625
	Gen	8	0.5	0.0625

3.3.4. Biofilm elimination assay

As shown in Fig. 5A, N6 exerts a concentration-dependent clearance effect on nascent biofilms. After 24 h of treatment at 16× MIC, the biofilm formation rate decreased to approximately 9.9%. Within the 0.5–2× MIC range, N6 showed superior clearance efficacy to colistin; at concentrations above 2× MIC, the two agents showed comparable efficacy.

Fig. 5.

Effects of N6 on K. pneumoniae biofilms: (A) Prevention of biofilm formation and (B) Eradication of mature biofilms.

N6 also displayed clearance activity against mature biofilms (Fig. 5B). Treatment at 0.5–2× MIC cleared ∼20% of mature biofilms, with efficacy increasing with concentration; at 16× MIC, the clearance rate reached ∼61%, indicating N6 has significant clearance efficacy against mature K. pneumoniae biofilms. In contrast, colistin showed weaker clearance efficacy than N6 against mature biofilms across all tested concentrations.

3.4. Effects of N6 on K. pneumoniae in vitro and in vivo

3.4.1. Cytotoxicity of N6 on BEAS-2B cells

To evaluate the biological safety of N6, we assessed its cytotoxic effects on BEAS-2B. As depicted in Fig. 6A, CCK-8 assay results showed that N6 exerted no significant cytotoxicity at concentrations of 1–128 μg/mL, with cell viability consistently exceeding 90%.

Fig. 6.

N6 protects BEAS-2B cells against K. pneumoniae infection. (A): Cytotoxicity of N6 at various concentrations in BEAS-2B cells. (B): Effect of N6 on cell survival rate in K. pneumoniae-infected BEAS-2B cells. (C): Regulation of inflammatory cytokine expression IL-10 by N6 in infected cells. (D): Regulation of inflammatory cytokine expression IL-1β by N6 in infected cells.

3.4.2. N6 treatment of BEAS-2B cell infection

N6 significantly alleviated K. pneumoniae-induced BEAS-2B cell death. As shown in Fig. 6B–K pneumoniae infection markedly reduced cell viability to 29.79% (P < 0.01), confirming successful establishment of the infection model. Both N6 and colistin intervention increased the viability of infected cells in a concentration-dependent manner. At 128 μg/mL, N6 and colistin exhibited comparable protective effects, restoring cell viability to 72.75% and 73.80%, respectively. When concentrations were reduced to 64 and 32 μg/mL, the protective effects of both agents weakened accordingly, with no statistically significant differences observed between their corresponding concentration groups.

Regarding inflammatory cytokine expression (Fig. 6C–D), K. pneumoniae infection significantly increased levels of proinflammatory cytokines IL-1β in BEAS-2B cells (P < 0.01), while the anti-inflammatory cytokine IL-10 was significantly downregulated (P < 0.01), indicating a robust infection-induced inflammatory response. Drug intervention effectively modulated proinflammatory cytokine expression: Specifically, 128 μg/mL N6 exerted the strongest inhibitory effect on IL-1β, reducing its level to near the blank group and significantly outperforming colistin at the same concentration (P < 0.05). Furthermore, N6 effectively restored IL-10 levels suppressed by infection: at 128 and 64 μg/mL, IL-10 levels approached or slightly exceeded those of the blank group, demonstrating potential for promoting inflammation resolution. In contrast, colistin showed limited ability to enhance IL-10 across all concentrations, with levels remaining below the blank group.

3.4.3. Therapeutic effects of N6 on K. pneumoniae-induced mouse infection

To evaluate the pathogenicity of intranasal infection with K. pneumoniae and the therapeutic efficacy of N6 and kanamycin, we monitored mouse survival within 7 days post-infection. As shown in Fig. 7A, mice in the CK group survived the entire experiment. Infected mice started to die from day 3 post-infection, with all succumbing by day 7—validating the high lethality of K. pneumoniae and successful establishment of the infection model. For the treatment groups, N6 treated mice exhibited a 70% survival rate by day 2, which remained stable thereafter. Mice treated with kanamycin survived the first 3 days, followed by a gradual decline in survival rate that ultimately stabilized at 70%. These results demonstrate that both N6 and kanamycin significantly enhanced the final survival rate of infected mice compared to the untreated infected group.

Fig. 7.

Protective effects of N6 against K. pneumoniae infection in mice. (A) Survival probability within 7 days post-infection. (B) Bacterial load (lg CFU/g) in mouse lung tissue at 7 days post-infection. (C) Bacterial load (lg CFU/g) in mouse brain tissue at 7 days post-infection. (D–G) Serum cytokine levels in mice at 7 days post-infection: (D) IL-10, (E) IL-6, (F) TNF-α, and (G) IL-1β (all in pg/mL).

On day 7 post-inoculation, the lung and brain tissues of mice were collected for bacterial colony counting and histopathological analysis, while serum was harvested to measure inflammatory cytokine levels. As shown in Fig. 7B–C, bacterial loads in the brain and lung tissues of infected mice were significantly higher than those in the CK group (P < 0.01). Following treatment with N6 or kanamycin, bacterial loads in all examined tissues were markedly reduced. Regarding inflammatory responses, as shown in Fig. 7D–G, serum levels of IL-1β, IL-6, IL-10, and TNF-α in infected mice were significantly elevated compared to the CK group (P < 0.01). After drug administration, levels of these inflammatory cytokines were effectively modulated, largely returning to the levels observed in the CK group.

Histopathological examination yielded the following observations: In lung tissue, the infected group displayed alveolar wall thickening, granulocytic infiltration, hemorrhage, and vascular congestion (Fig. 8A). The CK group, N6-treated group, and kanamycin-treated group exhibited only focal alveolar wall thickening and minimal granulocytic infiltration, with rare vascular congestion (Fig. 8B–D). In brain tissue, the infected group showed meningeal inflammatory cell infiltration, hemorrhage, neuronal edema, and ventricular dilation (Fig. 8E). The remaining three groups predominantly exhibited neuronal shrinkage, whereas the N6-treated group displayed mild neuronal edema (Fig. 8G) without other significant abnormalities (Fig. 8F–H).

Fig. 8.

Protective effects of N6 on lung and brain tissues in K. pneumoniae-infected mice (Histological Observation, HE Staining).

(A–D): Histopathological sections of lung tissue (A): Infected Group: Prominent neutrophil infiltration in alveolar walls (green arrows); scattered necrotic cell debris in alveolar walls (orange arrows). (B): CK Group: Only minimal neutrophil infiltration in alveolar walls (green arrow). (C): N6 Treatment Group: Minimal neutrophil infiltration in alveolar walls (green arrow). (D): Kanamycin (KAN) Treatment Group: Minimal neutrophil infiltration in alveolar walls (green arrow).

(E–H): Histopathological sections of brain tissue. (E): Infected Group: Minimal lymphocytic and neutrophilic infiltration in the meninges (green arrows); mild hemorrhage (yellow curved polygon). (F): CK Group: No significant abnormalities detected in brain tissue. (G): N6 Treatment Group: No significant abnormalities in the meninges or brain parenchyma. (H): Kanamycin (KAN) Treatment Group: No significant abnormalities in the meninges or brain parenchyma.

Overall, the infected group presented the most pronounced pathological lesions in both lung and brain tissues. The CK group, N6-treated group, and kanamycin-treated group primarily showed mild focal changes in lung tissue, while brain tissue was characterized by neuronal shrinkage. The overall severity of pathological lesions in these groups was significantly less pronounced than that in the infected group.

3.5. Mechanism of action

3.5.1. Effect of N6 on the cell membrane of K. pneumoniae

Flow cytometry analysis was employed to assess the membrane disruption effects of N6 and colistin on K. pneumoniae, as shown in Fig. 9. The results demonstrated that both agents induced concentration-dependent membrane damage, but with distinct temporal dynamics and efficacy. For N6, treatment at 1× MIC resulted in low lysis rates (3.76%–7.41%) without clear time dependence. In contrast, at 2× MIC, the lysis rate increased progressively over time, reaching 69.4% by 120 min, while at 4× MIC, it rose rapidly and continuously to a peak of 89.8% at 120 min, indicating potent and sustained membrane damage at higher concentrations. Under colistin treatment, 1× MIC also showed limited lysis (2.25%–9.53%) without time dependence, and 2× MIC induced fluctuating lysis rates between 15.3% and 21.1%, with a maximum value significantly lower than that of N6 at the same concentration. At 4× MIC, colistin caused an initial peak (82.8% at 60 min), followed by a decline and a partial recovery (79.2% at 120 min). In summary, N6 exhibited stronger and more consistent membrane disruption than colistin, particularly at elevated concentrations, with clear time-dependent activity.

Fig. 9.

FACS analysis of time- and concentration-dependent membrane permeability of K. pneumoniae S469 exposed to N6 and colistin. Rows correspond to groups treated with N6 (1× MIC, 2× MIC, 4× MIC) and colistin (1× MIC, 2× MIC, 4× MIC), respectively. Columns represent exposure times (30 min, 60 min, 90 min, 120 min). Scatter plots of Percp-A vs. SSC-A depict dynamic changes in membrane permeability of K. pneumoniae S469 across different drug concentrations and exposure times.

3.5.2. Effects of N6 on cell morphology of K. pneumoniae

The effects of a 2 h treatment with N6 and colistin on K. pneumoniae morphology were examined by SEM and TEM. SEM images (Fig. 10A) revealed that untreated cells maintained a normal, smooth morphology with intact membranes. In contrast, treatment with 4× MIC N6 significantly reduced bacterial abundance and induced the formation of extensive reticulated structures on the cell surface. Under the same conditions, colistin primarily caused numerous membrane protrusions.

Fig. 10.

Morphological changes of K. pneumoniae S469 treated with N6 or colistin. (A) Scanning Electron Microscopy (SEM) images of the PBS control, N6-treated, and colistin-treated groups (left to right) at magnifications of ×5k, ×20k, and ×50k (top to bottom). (B) Transmission Electron Microscopy (TEM) images of the PBS control, N6-treated, and colistin-treated groups (left to right) at magnifications of ×5k, ×20k, and ×40k (top to bottom).

Consistent findings were observed by TEM (Fig. 10B). Untreated cells showed intact membranes and uniformly distributed cytoplasmic electron density. After N6 treatment, cells were surrounded by reticulated structures and exhibited electron-lucent cytoplasm with substantial content leakage. The colistin-treated cells, however, displayed membrane protrusions along with uneven electron density and cytoplasmic leakage.

3.5.3. Effects of N6 on metabolic activity of K. pneumoniae

3.5.3.1. Metabolic activity assay

As shown in Fig. 11A, both N6 and colistin inhibited the metabolic activity of K. pneumoniae S469, with distinct concentration-response patterns. N6 exerted a concentration-dependent inhibitory effect, with metabolic activity decreasing progressively from 1× MIC to 4× MIC. Relative to CK group, metabolic activity was significantly reduced at 2× MIC (P < 0.05) and 4× MIC (P < 0.01). In contrast, colistin showed weak inhibition at 1× MIC, significantly enhanced inhibition at 2× MIC (P = 0.01), but no additional enhancement at 4× MIC.

Fig. 11.

Effects of N6 on metabolic activity of K. pneumoniae.

In all panels: CK, blank control; N6, N6 treatment group; Col, colistin treatment group. 1×, 2×, and 4× MIC denote 1, 2, and 4 fold the minimum inhibitory concentration, respectively. Different letters (a-d) mark significant differences. (A) Bacterial metabolic activity measured by the Alamar Blue assay. (B) Intracellular ATP levels, shown as luminescence intensity. (C) Intracellular ROS levels detected by DCFH-DA fluorescence.

3.5.3.2. Intracellular ATP detection

Following 60-min drug treatment, intracellular ATP levels of K. pneumoniae S469 were measured and shown in Fig. 11B. Both drugs exhibited concentration-dependent effects, but their impacts on ATP levels differed markedly. At 1× MIC, N6 induced a slight reduction in ATP levels, while colistin caused a minor elevation—neither difference reached statistical significance compared to the CK group. In contrast, at higher concentrations (2× and 4× MIC), the two drugs exerted diametrically opposed effects: N6 elicited a drastic reduction in ATP levels (P < 0.01), whereas colistin induced a significant elevation (P < 0.01) relative to the control.

3.5.3.3. Intracellular Reactive Oxygen Species (ROS) detection

Intracellular ROS detection results (Fig. 11C) revealed distinct concentration-response relationships between N6 and colistin in inducing oxidative stress in K. pneumoniae. At 1× MIC, N6 exhibited a significantly elevated ROS-related fluorescence intensity compared to the CK group (P < 0.01), with a further increase in fluorescence intensity observed at higher concentrations. In contrast, colistin did not cause a significant increase in intracellular ROS levels at 1× MIC; a significant elevation in ROS levels was only detected at 2× MIC (P < 0.01). Notably, only at 4× MIC did colistin induce ROS levels comparable to those in the high-concentration N6 group.

4. Discussion

Large-scale heterologous expression of AMPs poses a critical bottleneck for their clinical translation. Host system selection, fusion tag design, and fermentation process optimization are key determinants of the target product's yield, bioactivity, and industrialization potential [12]. To address N6's inherent characteristics, such as its strong positive charge which may induce host toxicity, a fusion expression strategy incorporating a “6×His-2×Flag dual tag with a formic acid cleavage site” was used in this work. The 2×Flag tag effectively neutralizes N6's positive charge to ensure host compatibility, and the 6×His tag enables efficient downstream purification via nickel affinity chromatography. Compared to previous strategies relying on large tags (e.g., SUMO) [7], the short peptide tags significantly enhanced the yield of cleaved target peptide, highlighting the utility of short peptide tags for high-yield AMP expression.

For purification, the integration of a formic acid cleavage site allowed efficient, low-cost separation of the tags from N6. Unlike protease-based cleavage methods that incur high costs, complete cleavage was achieved via formic acid treatment, ensuring both high efficiency and cost-effectiveness. Following cleavage, a secondary nickel column purification step removed free tags, uncleaved fusion proteins, and other impurities in one step, significantly streamlining the downstream workflow.

Using this integrated strategy, we achieved a marked increase in N6 yield, with the final purified product exceeding 1.02 g/L, significantly outperforming recent comparable reports [23] and meeting the threshold for industrial-scale production [12]. The core achievement is the simultaneous realization of multiple goals: mitigated host toxicity, enhanced expression efficiency, streamlined purification, and a balanced yield-cost trade-off. This study provides a practical solution for the industrial-scale production of N6, and the established “short tag-chemical cleavage” strategy serves as a technical paradigm for the large-scale heterologous expression of other difficult-to-express AMPs.

Currently, AMPs in clinical development exhibit MIC ranges from <0.03 μg/mL to 32 μg/mL, whereas conventional antibiotics typically have MIC values between 0.5 and 4 μg/mL [24]. In the present study, N6 exhibited MIC values of 2–8 μg/mL against clinical K. pneumoniae isolates, which is comparable to the MIC range of conventional antibiotics and underscores its potential as a viable therapeutic candidate.

The increasing prevalence of multidrug-resistant (MDR) bacteria has severely limited clinical therapeutic options. As a “last-line” defense against MDR Gram-negative infections, colistin faces growing clinical limitations [25]. Our in vitro time-kill kinetics demonstrated that while colistin transiently inhibits K. pneumoniae, this is consistently followed by bacterial regrowth—a phenomenon that aligns with clinical reports of colistin treatment failure in various isolates [26,27]. In contrast, the antimicrobial peptide N6 exhibited potent and rapid bactericidal activity. At 1×, 2×, and 4× MIC, N6 completely eradicated K. pneumoniae S469 within 1 h in a concentration-dependent manner, with no regrowth observed. Consistent with this, other AMPs, such as AA139, SET-M33 [28], and NNS5-6 [29], have demonstrated comparable efficacy, underscoring the therapeutic potential of this class of agents.

In this study, combinations of N6 with gentamicin or erythromycin yielded additive effects, attributable to their complementary mechanisms. Gentamicin targets the 30S ribosomal subunit to disrupt protein synthesis initiation [30], while erythromycin binds to the 50S subunit, blocking transpeptidation and/or translocation [31]. Both require passage across the cell membrane. N6, a cationic peptide (+5 charge) with demonstrated membrane-disrupting activity [32], likely enhances the intracellular delivery of these antibiotics, improving access to their targets and boosting efficacy. This “membrane disruptor + intracellular inhibitor” rationale is supported by prior studies: Sushmasri observed additive effects between gentamicin and colistin against Enterobacteriaceae, and She reported marked synergy (FICI < 0.266) between the colistin derivative SPR741 and erythromycin against XDR K. pneumoniae [33,34]. Although N6 combinations in our study yielded additive (FICI 0.5–1.0) rather than synergistic effects, this outcome remains clinically valuable, as it may reduce the required dosage of individual drugs, potentially lowering the risk of aminoglycoside-induced nephrotoxicity, and establishes a promising therapeutic paradigm for combating MDR K. pneumoniae.

This study identified a high prevalence of the virulence genes fimH and entB among nine clinical K. pneumoniae isolates. The fimH gene encodes an adhesin that mediates bacterial adhesion to host cells and biofilm formation, representing a conserved virulence factor in Enterobacteriaceae [35,36]. The entB gene is involved in the biosynthesis of enterobactin, a high-affinity siderophore that facilitates iron acquisition under host iron restriction. Its high prevalence (88.9%) in these isolates highlights the adaptive fitness of clinical strains within the host environment [37]. Other genes, including iucA, rmpA, and peg-344, are associated with hypervirulent phenotypes, enhancing pathogenicity by regulating capsular synthesis, iron uptake, and mucoid morphology [[38], [39], [40]]. Notably, even hypervirulent strains (e.g., K11, YB47) in this study displayed N6 MIC values ranging from 2 to 8 μg/mL, demonstrating that N6 retains robust antimicrobial activity against hypervirulent K. pneumoniae.

Biofilm formation is a key mechanism underlying antibiotic resistance in K. pneumoniae, as it physically shields bacteria from antibiotics and host immune clearance while also promoting the horizontal transfer of resistance genes [41,42]. In this study, the N6 demonstrated potent anti-biofilm activity: at 2× MIC, it suppressed nascent biofilm formation by 90.1%, while at 16× MIC, it achieved approximately 61% eradication of mature biofilms. Such anti-biofilm effects are also observed in other AMPs. For example, AMP Kla acts synergistically by inhibiting both biofilm formation and efflux pump activity against gentamicin-resistant strains [43].

Bronchial epithelial cells (BEAS-2B) constitute a critical mucosal barrier, providing physical protection, maintaining fluid homeostasis, and secreting mucus and surfactant [44]. Infection with K. pneumoniae significantly impaired BEAS-2B viability and triggered a marked pro-inflammatory response, characterized by the upregulation of IL-1β and the concurrent downregulation of IL-10 (P < 0.01). This imbalance in cytokine production is consistent with the known pathogenesis of K. pneumoniae, where bacterial adhesins and LPS activate host inflammatory signaling, leading to cytotoxicity [45,46]. Specifically, elevated IL-1β amplify inflammation, activate caspase-3, and drive excessive epithelial cell apoptosis [47], while the concurrent downregulation of IL-10 weakens a critical brake on the NF-κB pathway, creating a feed-forward loop that amplifies inflammation and tissue injury [2]. Given N6's cationic nature, we propose that its anti-inflammatory effect may, at least in part, stem from direct LPS neutralization, although the precise molecular mechanism warrants further investigation.

In our murine pneumonia model, infection with K. pneumoniae YB47 triggered progressive clinical deterioration. By day 2 post-infection, mice exhibited ruffled fur, lethargy, and diminished activity. By day 3, severe neurological manifestations—including trunk ataxia, abnormal posture, and rotatory ataxia—emerged, suggestive of central nervous system (CNS) invasion and impaired cerebral function [48,49]. All untreated mice succumbed to the infection within 7 days. In contrast, N6 treatment (10 mg/kg, i.p.) significantly improved survival, with 70% of mice surviving—a level of protection comparable to the kanamycin-positive control.

Histopathological and molecular analyses confirmed that N6 significantly reduced bacterial burdens in the lungs and brains of infected mice (P < 0.01). The treatment alleviated pulmonary lesions, including alveolar wall thickening, granulocyte infiltration, and hemorrhage. It also mitigated brain pathology, such as meningeal inflammatory cell infiltration, neuronal edema, and ventricular dilatation. Serum cytokine profiling revealed that N6 significantly mitigated infection-induced elevations in pro-inflammatory cytokines (IL-1β, IL-6, TNF-α) and restored anti-inflammatory IL-10 to baseline levels, effectively rebalancing the host inflammatory response. These in vivo findings collectively demonstrate that N6's therapeutic efficacy stems from its integrated antibacterial, anti-inflammatory, and anti-biofilm activities observed in vitro, highlighting its potential for treating severe K pneumoniae infections—including those with life-threatening CNS dissemination.

Flow cytometry revealed that both N6 and colistin disrupted the cell membrane of K. pneumoniae in a concentration- and time-dependent manner. Membrane disruption was significantly greater at 4× MIC than at 1× MIC for both agents, consistent with previously reported concentration-dependent effects [32,50]. Notably, the membrane disruption rate for colistin decreased at 90 min (from 82.8% to 64.4%), likely due to bacterial membrane repair or adaptive remodeling—a finding consistent with the stress-induced responses previously described [51]. In contrast, N6 maintained a higher disruption rate (89.8%) at 120 min than colistin (79.2%), demonstrating its superior and sustained membrane-damaging activity at high concentrations.

Electron microscopy further revealed distinct morphological changes induced by the two agents. N6 treatment resulted in extensive reticulated structures on the cell surface, consistent with membrane leakage and abnormal folding as reported previously [32,52]. In contrast, colistin primarily caused membrane blebbing, presumably due to its interference with membrane lipid biosynthesis and subsequent membrane destabilization [53]. Although both agents ultimately led to cell lysis and content leakage, they induced clearly distinct paths of ultrastructural damage.

Beyond membrane disruption, N6 exerts bactericidal effects by disrupting core bacterial metabolic processes. Alamar Blue assays revealed a significant suppression of metabolic activity in N6-treated K. pneumoniae, which was accompanied by a sharp decline in intracellular ATP levels and a marked accumulation of ROS. Collectively, our data support a model in which N6 exerts its bactericidal activity through a multimodal mechanism that integrates “membrane disruption,” “metabolic inhibition,” and “oxidative stress”: initial membrane damage triggers the leakage of vital cellular components, concurrent metabolic suppression depletes energy reserves, and ROS accumulation inflicts further oxidative damage to macromolecules.

Conclusion: This study establishes a cost-effective Pichia pastoris platform for high-yield recombinant N6 production via a 6His-2Flag dual-tag strategy with formic acid cleavage. N6 exhibits potent activity against multidrug-resistant Klebsiella pneumoniae, mediated by membrane disruption, metabolic inhibition, and oxidative stress. It inhibited preformed biofilms, demonstrated minimal cytotoxicity, and improved mouse survival to 70% in vivo. These results position N6 as a promising clinical candidate while validating the dual-tag system as a scalable antimicrobial peptide production platform.

CRediT authorship contribution statement

Yexuan Wang: Writing – review & editing, Writing – original draft, Methodology, Investigation. Manli Cao: Investigation. Mengyin Deng: Investigation. Na Yang: Writing – review & editing, Project administration, Methodology, Data curation. Da Teng: Validation, Project administration, Funding acquisition. Ya Hao: Project administration. Keye Xu: Resources. Jianhua Wang: Writing – review & editing, Project administration. Ruoyu Mao: Writing – review & editing, Project administration, Methodology, Funding acquisition, Conceptualization.

Data availability statement

No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Funding

This study was supported by the National Key Research and Development Program of China (No. 2023YFD1301103, China), and the National Innovation Program of Agricultural Science and Technology at the Chinese Academy of Agricultural Sciences in the Feed Research Institute (No. CAAS-ASTIP-2023-IFR-01, Beijing, China).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.