High-level biosynthesis and purification of the antimicrobial peptide Kiadin based on non-chromatographic purification and acid cleavage methods

Abstract

Antimicrobial peptides (AMPs) are renowned for their potent bacteriostatic activity and safety, rendering them invaluable in animal husbandry, food safety, and medicine. Despite their potential, the physiological toxicity of AMPs to host cells significantly hampers their biosynthetic production. This study presents a novel approach for the biosynthesis of the antimicrobial peptide Kiadin by engineering a DAMP4–DPS–Kiadin fusion protein to mitigate host cell toxicity and achieve high-level expression. Leveraging the unique properties of the DAMP4 protein, we developed a non-chromatographic purification method to isolate the DAMP4–DPS–Kiadin fusion protein with high purity. The instability of the D–P peptide bond under acidic conditions, combined with the thermal and saline stability of DAMP4, enabled efficient separation of Kiadin through acid cleavage and isoelectric precipitation, yielding Kiadin with 96% purity and a production yield of 29.3 mg/L. Our optimization of acid cleavage temperature, duration, and isoelectric precipitation conditions proved critical for maximizing the purification efficiency and expression levels of Kiadin. The biosynthesized Kiadin exhibited robust bacteriostatic activity against Escherichia coli, Pseudomonas aeruginosa, Acinetobacter baumannii, Bacillus cereus and Staphylococcus aureus. Notably, Kiadin demonstrated significant post-antibiotic effects by disrupting bacterial membrane integrity, inducing cytoplasmic leakage, and inhibiting biofilm formation in E. coli K88 and S. aureus Mu50, without cytotoxicity towards mouse macrophages. In vivo studies further confirmed Kiadin's exceptional therapeutic efficacy against abdominal infections caused by E. coli K88. The acid cleavage and non-chromatographic purification techniques developed in this study offer a cost-effective and efficient strategy for the high-purity production of AMPs.

Introduction

Antimicrobial peptides (AMPs) represent a versatile class of small-molecule proteins renowned for their broad-spectrum antibacterial, antiviral, and antifungal activities. These peptides are ubiquitously found across various life forms, including animals, plants, and microorganisms, and have garnered significant scientific interest due to their unique mechanisms of action and potential applications [1]. Unlike traditional antibiotics, AMPs primarily target bacterial cell membranes, disrupting them through multiple pathways and eliminating bacteria through multiple targets, which significantly reduces the likelihood of bacterial resistance development [2, 3]. AMPs have found extensive applications in medicine, animal husbandry, and the food industry [4,5,6]. For instance, nisin is widely used as a food preservative and can be incorporated into innovative food packaging materials, such as nanofibers and nanoliposomes, offering advanced preservation solutions [7]. In addition, the Bacillus-derived methylene salicylate peptide (BMD) enhances livestock growth performance, while the AMP LL-37 functions as an anti-fibrotic agent by inhibiting excessive collagen synthesis [8, 9]. The biosynthesis of AMPs with superior antibacterial efficacy, safety, and stability remains a pivotal focus of current research. The antimicrobial peptide database (APD3, https://aps.unmc.edu/) catalogs 3,146 natural AMPs from various sources [3]. However, many natural AMPs exhibit vulnerabilities, such as poor resistance to enzymatic degradation and suboptimal safety profiles, necessitating modifications to enhance their practical applications [10, 11]. Kiadin is a modified AMP derived from the natural peptide PGLa found in the skin secretions of the African clawed toad (Xenopus laevis). DiPGLa-H was obtained by tandem duplication of PGLa-H, a truncated form of PGLa, and Kiadin was subsequently created by replacing Val at position 13 in DiPGLa-H with Gly [12]. This modification significantly increased the in vitro bacteriostatic activity of Kiadin and reduced its hemolytic capacity [12, 13]. Researchers performed five Gly mutations on Kiadin to further improve bacteriostatic activity and reduce host toxicity, with Kiadin-2 exhibiting the strongest bacteriostatic activity against Gram-negative bacteria, and Kiadin-5 showing no bacteriostatic activity but the highest safety rating [14]. Overall, these derived peptides did not surpass Kiadin in terms of both enhanced bacteriostatic activity and reduced hemolytic activity, positioning Kiadin as a promising lead compound.

Despite its potential, the chemical synthesis of Kiadin is both complex and costly. Heterologous expression in microorganisms offers a cost-effective and scalable alternative for AMP production. Common expression hosts include Lactobacillus lactis, Pichia pastoris, and Bacillus subtilis. Zheng et al. expressed nisin A with the L. lactis ATCC 11454 mutant, achieving an average yield of 9,960 IU/mL in a 50-ton fermenter [15]. Zhang et al. used P. pastoris to express the antimicrobial peptide AP138L-arg26, achieving a protein concentration of 3.1 mg/mL in a 5-L fermenter, which rapidly killed 99.99% of Staphylococcus aureus in 1.5 h at a concentration of 2 × MIC [16]. Qiu et al. expressed the surfactant Lichenysin with B. licheniformis WX-02, producing 2149 mg/L lichenysin by optimizing the medium conditions, with great potential for application in crude oil extraction [17]. Seo et al. successfully achieved the biosynthesis of human α-defensin 5 (HD5) and β-defensin 2 (HBD2) with the probiotic bacterium Escherichia coli (Nissle 1917) [18]. However, the biosynthesis of Kiadin has not yet been successfully realized due to extensive host toxicity to commonly used expression hosts, limiting its large-scale production and application.

To address this challenge, fusion protein construction is a widely used strategy to mitigate host toxicity. Carrier proteins such as MBP, GST, ELP, SUMO, and TrxA enhance the solubility of fusion proteins and facilitate purification, although these methods often rely on expensive chromatographic techniques [19]. The DAMP4 protein, a novel fusion tag consisting of a four-stranded helical bundle with a hydrophobic core and a hydrophilic exterior, significantly improves fusion protein solubility. The DAMP4 fusion protein can be obtained by facile heating and salting-out precipitation, and the purified AMP can be obtained by removing the DAMP4 protein through isoelectric point precipitation, greatly reducing the production and purification costs of AMPs [20, 21]. Moreover, low-cost cleavage from fusion proteins to produce AMPs is another concern for realizing cost-effective manufacturing. Currently, enzymatic cleavage is mainly used, such as TEV protease, which is highly specific for the sequence ENLYFQG and can effectively cleave the peptide bond between glutamine and glycine at low temperatures [22]. Whereas SUMO proteins act by forming covalent bonds with target proteins (usually lysine residues), SUMO proteases recognize specific x-Gly-Gly-x sequences [23]. However, the production cost of these enzymes and their efficiency in the cleavage system significantly increase the production costs of AMPs. Studies have shown that enzyme reuse reduce the initial activity by 30–40% per cycle [24].

The peptide bond between aspartic acid and proline residues (DP) is acid-sensitive, allowing cleavage by simply adjusting the pH and temperature of the fusion protein [25]. AMPs, which function with α-helical structures, are highly resistant to temperature and pH variations, facilitating the acid cleavage of fusion proteins and the subsequent purification of AMPs. Therefore, the introduction of DP dipeptides (or DPS tripeptides) into fusion proteins will enable low-cost cleavage and large-scale production of high-purity AMPs. Sun et al. explored the effect of connecting sequence lengths between DAMP4 and the AMPs on antimicrobial activity, showing that a long, flexible junction (PGGGGSGGGGGSLVPRGS) retains AMP activity, and low-temperature expression favors correct folding [26]. Furthermore, Sun et al. successfully purified the AMP pexiganan by combining DAMP4 and acid lysis methods, yielding 2 mg/L [21]. Overall, the combination of DAMP4 sequences and acid cleavage methods for AMP expression and purification is still scarce and deserves further exploration.

In this study, we engineered a fusion protein, DAMP4–DPS–Kiadin, and optimized its expression in E. coli. High-purity fusion protein was obtained through high-temperature, high-salt precipitation and centrifugation. To enhance the efficiency of protein cleavage, we systematically evaluated various parameters including the effect of different DPS cleavage sites, temperatures, durations, and types of acid on the cleavage efficiency. By comparing the yield and antimicrobial properties of Kiadin between isoelectric point precipitation and freeze-drying, it was found that the isoelectric point precipitation method was more favorable for the purification of Kiadin. Ultimately, Kiadin with a purity of 96% and a yield of 29.3 mg/L was obtained. Subsequent cellular and animal assays demonstrated that the biosynthesized antimicrobial peptide Kiadin had a good safety profile in vitro and in vivo and was able to effectively inhibit the infection of pathogenic bacteria. These findings underscore the potential for scalable biosynthesis and practical application of Kiadin, marking a significant advancement in AMP research.

Results

Expression, purification, and acid cleavage of the fusion protein DAMP4–DPS–Kiadin

The biosynthesis of the antimicrobial peptide Kiadin is significantly hampered by its inherent toxicity towards common expression hosts, including E. coli, B. subtilis, L. lactis, and P. pastoris, even at low concentrations (Fig. 1A). To overcome these challenges, we engineered a novel fusion protein, DAMP4–DPS–Kiadin, designed to enable cost-effective synthesis and efficient purification of Kiadin. We constructed the fusion protein by linking DAMP4 to Kiadin via Asp-Pro-Ser (DPS) sequences, resulting in the protein M-(EPS-MKQLADSLHQLARQVSRLEHA)₄D-PS-KIAKVALKALKIAKGALKAL-HH, the molecular weight of this fusion protein is 13.7 kDa. AlphaFold3 software predictions show that DAMP4 consists of four AM1 motifs (MKQLADSLHQLARQVSRLEHA) connected by EPS units, with each AM1 motif forming an α-helix (Fig. 1B). The terminal AM1 motif is linked to Kiadin, also an α-helix, via the DPS unit, with the Asp98-Pro99 peptide bond serving as the acid cleavage site. In addition, Kiadin, a modified version of the naturally occurring antimicrobial peptide PGLa, was equipped with two histidine residues at the C-terminus to mitigate hydrolysis risks (Fig. 1B).

Fig. 1

Characteristics, expression, and purification of fusion protein DAMP4–DPS–Kiadin. A Inhibitory activities of the antimicrobial peptide PS–Kiadin–HH against the host-expressing bacteria Escherichia coli BL21(DE3), Bacillus subtilis WB800N, Lactococcus lactis NZ9000, and Pichia pastoris GS115. B Cartoon of DAMP4–DPS–Kiadin generated by PyMOL software. Kiadin was shown by yellow color with starting (K101) and ending (L119) amino acids. AM1-1 to AM1-4 represent identical amino acids (MKQLADSLHQLARQVSRLEHA), which was linked, respectively, by EPS motif. The acid cleavage site is shown as D98–P99–S100 with red color. C Tricine–SDS–PAGE showing the expression and acid cleavage of DAMP4–DPS–Kiadin. Total bacterial proteins before and after IPTG induction and the results of non-chromatographic purification of DAMP4–DPS–Kiadin by hydrochloric acid cleavage at 60 °C for 0 h, 4 h, 24 h and 48 h are shown

The gene encoding the designed fusion protein was cloned into the pET-28a vector and expressed in E. coli BL21 (DE3). Expression results (Fig. 1C) revealed a distinct 13.7 kDa protein band following induction with 1 mM IPTG, confirming successful expression of the DAMP4–DPS–Kiadin fusion protein. Utilizing DAMP4’s high solubility and structural stability under high temperature and salt conditions, the heteroprotein was selectively precipitated and removed, facilitating the purification of the fusion protein (Fig. 1C). Subsequent acid cleavage at the aspartic acid–proline (D–P) bond was performed to release the target antimicrobial peptide (Fig. 1C). The fusion protein was incubated with 60 mM HCl at 60 °C for various durations (0 h, 4 h, 24 h, and 48 h) to monitor cleavage efficiency. Over time, the fusion protein bands diminished, and a new band of approximately 11 kDa emerged, corresponding to the DAMP4 tag (11.2 kDa) (Fig. 1C). After 48 h, most of the DAMP4–DPS–Kiadin had been cleaved into DAMP4-D and PS-Kiadin, confirming successful cleavage. These results demonstrate that the fusion expression of DAMP4 and Kiadin effectively mitigates the antimicrobial peptide's toxicity towards host bacteria, significantly enhancing Kiadin production. The DAMP4 tag and DPS motif facilitated the efficient purification of both the fusion protein and the antimicrobial peptide, promoting cost-effective industrial manufacturing of Kiadin.

Optimization of acid cleavage conditions for the fusion protein DAMP4–DPS–Kiadin

To enhance the biosynthesis and purification of the antimicrobial peptide Kiadin, we systematically optimized the conditions for acid cleavage of the fusion protein DAMP4–DPS–Kiadin. Our primary focus was the peptide bond between aspartate and proline, which serves as the acid cleavage site. We initially investigated the influence of the number of DP (dipeptides) on acid cleavage efficiency by expressing and purifying the fusion proteins DAMP4–DPS–Kiadin and DAMP4–DPIPDPS–Kiadin. Both were subjected to acid cleavage at 60 °C for 24 and 48 h. The results showed no significant difference in protein bands between the two cleavage durations, indicating that increasing the number of DP dipeptides does not enhance cleavage efficiency (Fig. 2A). Subsequently, we examined the impact of temperature on acid cleavage efficiency. Both fusion proteins were subjected to acid cleavage at 75 °C and 60 °C. At 75 °C, nearly complete cleavage was achieved within 24 h, whereas significant amounts of uncleaved protein remained after 24 h at 60 °C. Extending the cleavage time to 48 h resulted in a decrease in the concentration of the DAMP4 tag, likely due to partial cleavage by the acid at the EPS motif. Furthermore, increasing the number of DP dipeptides did not enhance cleavage efficiency even at the higher temperature (Fig. 2A).

Fig. 2

Optimization of acid cleavage conditions for fusion protein DAMP4–DPS–Kiadin. A Tricine–SDS–PAGE showing protein bands of DAMP4–DPS–Kiadin versus DAMP4–DPIPDPS–Kiadin after cleavage with HCl for different temperatures (60 °C, 75 °C) and different hours (24 h, 48 h), respectively. B Inhibitory activity of acid cleavage products of DAMP4–DPS–Kiadin obtained under different temperatures 75 °C, 78 °C, 81 °C, 84 °C, 87 °C and 90 °C for 1–6 h on E. coli K88. The inhibitory activity of DAMP4–DPS–Kiadin without any treatment was set as control. C Surviving colonies of E. coli K88 treated with acid cleavage products of DAMP4–DPS–Kiadin at 2 h at different temperatures (75 °C, 78°Cand 81 °C), the inhibitory activity of DAMP4–DPS–Kiadin without any treatment was set as control. D Tricine–SDS–PAGE showing protein bands of DAMP4–DPS–Kiadin cleaved in HCl and H₃PO₄ for 4 h and 24 h, respectively

To further elucidate the effect of temperature on lysis efficiency, we evaluated the inhibitory activity of the cleavage products against E. coli K88 at various temperatures (75 °C, 78 °C, 81 °C, 84 °C, 87 °C and 90 °C). The results demonstrated that inhibitory activity increased with higher cleavage temperatures and prolonged cleavage times. Cleavage products at 84 °C, 87 °C, and 90 °C for 1 h achieved 100% killing of E. coli K88. At 75 °C, 78 °C, and 81 °C, the products killed 96.02%, 99.68%, and 99.95% of K88, respectively. After 2 h, products from 75 to 78 °C treatments inhibited 99.73% and 99.99% of K88, while those from 81 to 90 °C treatments achieved 100% inhibition (Fig. 2B, C). These findings underscore that increasing acid lysis temperature or time enhances cleavage efficiency, with 75 °C for 1 h being sufficient to inhibit approximately 96% of E. coli. These data are instrumental in selecting optimal conditions to reduce production costs of AMPs.

Finally, we assessed the effect of different acids on cleavage efficiency using 60 mM phosphoric acid at 60 °C with DAMP4–DPS–Kiadin as the target. Electrophoresis results showed no significant difference between phosphoric acid and hydrochloric acid cleavage after 4 or 24 h, suggesting that the type of acid does not significantly impact cleavage efficiency (Fig. 2D).

Purification and evaluation of antibacterial activity of the antimicrobial peptide Kiadin

To obtain highly purified PS–Kiadin–HH, we addressed the challenge of removing DAMP4-D from the acid cleavage products of the DAMP4–DPS–Kiadin fusion protein. Utilizing isoelectric precipitation as the primary purification method, we adjusted the pH of the cleavage product to the isoelectric point of DAMP4 (pH 6.8). Following a 30-min resting period and subsequent centrifugation, the supernatant revealed a single protein band at approximately 2.6 kDa (Fig. 3A), corresponding to PS–Kiadin–HH, with a concentration of 0.61 mg/mL, yielding 15.25 mg from a 25 mL sample. Given the initial 500 mL volume of the fermentation broth, this method achieved a yield of 30.5 mg/L of the antimicrobial peptide. In addition to a small number of PS–Kiadin–HH bands in the precipitation, a distinct DAMP4-D protein band appeared at approximately 11.2 kDa (Fig. 3A). Comparative analysis with the freeze-drying method revealed that post-freeze-drying and subsequent solubilization in Mueller–Hinton Broth (MHB) medium, centrifugation of the dried powder still showed bands of both PS–Kiadin–HH (2.6 kDa) and DAMP4-D (11.2 kDa), indicating that isoelectric precipitation is superior in removing DAMP4-D and achieving higher purity of PS–Kiadin–HH (Fig. 3A). Further validation of purity using reverse-phase high-performance liquid chromatography (RP–HPLC) demonstrated a main peak for PS–Kiadin–HH at 8.4 min, confirming an approximate protein purity of 96% (Fig. 3B). Mass spectrometry analysis corroborated these findings, showing the actual molecular weight of PS–Kiadin–HH (2620.40 Da) closely aligned with the theoretical value (2607.11 Da) (Fig. 3C). The above results indicate that, based on acid cleavage and isoelectric point precipitation techniques, we successfully obtained a high purity antimicrobial peptide PS–Kiadin–HH with a yield of 29.3 mg/L (30.5 mg/L × 0.96). Since 500 mL of fermentation broth yielded the DAMP4–DPS–Kiadin fusion protein at a concentration of 3.5 mg/mL, the theoretical acid-cleavage Kiadin yield was 31.9 mg/L, so this purification system actually lost about 8.2% of Kiadin.

Fig. 3

Purification and identification of antibacterial peptide Kiadin. A Supernatants and precipitates of acid cleavage products of DAMP4–DPS–Kiadin treated with isoelectric precipitation (IP) and freeze-drying (FD), respectively, which were showed by Tricine–SDS–PAGE with silver staining. B Characterization of purified PS–Kiadin–HH determined by RP‑HPLC. C Mass spectra of purified PS–Kiadin–HH determined by electrospray ionization mass spectrometry

To comprehensively evaluate the antibacterial efficacy of purified PS–Kiadin–HH, a series of rigorous assays were conducted. Against E. coli K88, both chemically and biologically synthesized PS–Kiadin–HH exhibited a minimum inhibitory concentration (MIC) of 32 μg/mL, demonstrating that the purification process preserved the peptide’s potent antibacterial activity (Fig. 4A). Notably, the MIC values for the peptides PS–Kiadin, Kiadin–HH, and Kiadin against E. coli K88 were also 32 μg/mL, suggesting that the introduction of amino acids HH or PS did not diminish Kiadin's antibacterial efficacy. In contrast, the DAMP4 label displayed no bacteriostatic activity (Fig. 4A). Our investigation extended to other pathogenic bacteria, where biologically synthesized PS–Kiadin–HH demonstrated optimal inhibitory effects. The MIC values were 8 μg/mL for Acinetobacter baumannii ATCC19606 and 16 μg/mL for Pseudomonas aeruginosa ATCC27853, both Gram-negative bacteria (Fig. 4B). Against Gram-positive pathogens, the MIC values were 64 μg/mL for S. aureus MRSA strain Mu50 and 32 μg/mL for Bacillus cereus ATCC11778 (Fig. 4B). Moreover, PS–Kiadin–HH exhibited MIC values of 32 μg/mL, 16 μg/mL, and 16 μg/mL against clinical MRSA strains USA300-LAC, BA01611, and KM041, respectively. For other clinical isolates, including methicillin-resistant Staphylococcus epidermidis strain BB02622, methicillin-resistant Staphylococcus haemolyticus strain BB02541, and multi-drug resistant E. coli ATCC51659, the MIC values were 16 μg/mL, 4 μg/mL, and 16 μg/mL, respectively (Table S1). These findings indicate that PS–Kiadin–HH exhibits broad-spectrum antibacterial activity, with enhanced efficacy against Gram-negative bacteria compared to Gram-positive strains.

Fig. 4

Evaluation of the antimicrobial activity and mechanism of the antimicrobial peptide PS–Kiadin–HH. A Minimum Inhibitory Concentration (MIC) of PS–Kiadin–HH (Biosynthesis), PS–Kiadin–HH (Chemical synthesis) and DAMP4 against E. coli K88. B MIC values of PS–Kiadin–HH (Biosynthesis) against P. aeruginosa ATCC27853, A. baumannii ATCC19606, B. cereus and S. aureus Mu50. C Determination of the post-antibiotic effect (PAE) of PS–Kiadin–HH against E. coli K88. D TEM micrographs of E. coli K88 and S. aureus Mu50 after treatment with PBS and PS–Kiadin–HH, respectively. E Determination of the permeability of PS–Kiadin–HH to the outer membrane of E. coli K88 using NPN and to the integral membranes of E. coli K88 & S. aureus Mu50 using the PI test. F Ability of PS–Kiadin–HH to inhibit the formation of biofilms by E. coli K88 and S. aureus Mu50 at MIC value concentrations

The post-antibiotic effect (PAE) of PS–Kiadin–HH was further assessed by monitoring the growth kinetics of E. coli K88 following a 1-h co-incubation. The results showed a significant PAE, particularly in the 1 × MIC, 2 × MIC, and 4 × MIC groups, where bacterial growth was notably delayed between 5 and 8 h (Fig. 4C). The calculated PAE values for the 1 × MIC, 2 × MIC, and 4 × MIC groups were 2.51, 3.52, and 4.11 h, respectively, indicating that higher concentrations of PS–Kiadin–HH have a stronger inhibitory effect on bacterial regrowth. Transmission electron microscopy (TEM) provided insights into the mechanism of action of PS–Kiadin–HH. Treatment of E. coli K88 with PS–Kiadin–HH caused a marked separation between the inner and outer membranes, accompanied by the formation of wrinkles and pores, which led to significant cytoplasmic leakage and a sparse cytoplasmic distribution (Fig. 4D). Similarly, in S. aureus Mu50 cells, PS–Kiadin–HH treatment resulted in ruptured cell walls, indistinct boundaries between cell walls and membranes, and substantial cytoplasmic leakage (Fig. 4D). These findings suggest that the antibacterial activity of PS–Kiadin–HH is primarily mediated through disruption of the structural integrity of bacterial cell membranes. Further validation of this mechanism was obtained through membrane permeabilization assays (Fig. 4E). Using the fluorescent probe 1-N-phenylnaphthylamine (NPN) to assess outer membrane damage and propidium iodide (PI) to measure intracellular content leakage, our results revealed that PS–Kiadin–HH induces severe outer membrane damage in E. coli K88, with permeability reaching 99% at the MIC concentration (Fig. 4E). PS–Kiadin–HH also significantly disrupted the inner membrane of E. coli K88, causing a 406% increase in permeability, and the cell membrane of S. aureus Mu50, with a 266% increase (Fig. 4E). In addition, PS–Kiadin–HH demonstrated a potent inhibitory effect on biofilm formation. At MIC concentrations, PS–Kiadin–HH significantly reduced the biofilm formation of both E. coli K88 and S. aureus Mu50 (Fig. 4F).

In assessing in vivo safety and efficacy, we examined the cytotoxicity of PS–Kiadin–HH on RAW264.7 mouse macrophages. The results were promising, showing that macrophage viability remained above 80% across various concentrations of Kiadin, with a remarkable 95% survival rate at 128 μg/mL after 24 h, indicating excellent biocompatibility (Fig. 5A). Furthermore, in a murine model of intraperitoneal infection with E. coli K88, treatment with both biosynthesized and chemically synthesized PS–Kiadin–HH significantly increased the survival rate of mice to 50% over 7 days, compared to 100% mortality in untreated controls within 5 days (Fig. 5B). Bacterial load assays in treated mice revealed substantial reductions in bacterial counts across various organs, with reductions of 98.69% in the heart, 99.45% in the spleen, 99.90% in the liver, and 99.92% in the kidneys (Fig. 5C). These comprehensive results unequivocally demonstrate the potential of the antimicrobial peptide PS–Kiadin–HH as an effective treatment for abdominal infections caused by E. coli in mice, highlighting its promising application in clinical settings.

Fig. 5

Evaluation of the safety and in vivo efficacy of PS–Kiadin–HH. A Cytotoxicity of PS–Kiadin–HH (Biosynthesis) against RAW264.7 cells. B Survival rate of E. coli K88 infected mice after treatment with PBS, biosynthetic PS–Kiadin–HH and chemical synthesis of PS–Kiadin–HH, respectively. E. coli K88 was used to construct an intraperitoneal infection model, 16 mice in each group. C Bacterial loads in hearts, livers, spleens, and kidneys from infected mice on the second day of treatment, with 5 mice in each group

Discussion

AMPs represent a promising frontier in human health, livestock production, and food preservation due to their extensive antibacterial, antiviral, antifungal, and immunomodulatory properties [27,28,29]. Among these, nisin is particularly noteworthy as the first antimicrobial peptide granted FDA approval under the Generally Recognized As Safe (GRAS) designation. Its applications in meat and dairy preservation, wound healing, oral health, and gastrointestinal infections underscore its clinical significance. The market for nisin is projected to grow at a compound annual growth rate (CAGR) of 5.45% from 2022 to 2027, driven by ongoing research and expanding applications [30, 31]. Other notable AMPs include LL-37, a human-derived peptide with demonstrated antibacterial and immunomodulatory effects, currently advancing through early phase clinical trials for cancer treatment [32]. In addition, Piscidin-1, a 22-amino-acid peptide from fish mast cells, exhibits potent antimicrobial activity against various pathogens, including critical viruses, such as the catfish virus (CCV), frog virus 3 (FV3), and HIV-1 [33,34,35].

Despite their promising potential, the broad application of AMPs is hindered by high chemical synthesis costs. Biosynthesis presents a feasible alternative to reduce these costs. For example, Nisin can achieve a yield of 9960 IU/mL in Lactobacillus casei ATCC 11454. Similarly, His6-Intein-KR12AGPWR6 has been successfully purified using Ni–NTA resin in E. coli BL21 (DE3), with AMPs released via intein self-cleavage [36]. However, technical challenges such as host toxicity, low expression levels, and complex purification processes often lead to high manufacturing costs for AMPs. This study focuses on the biosynthesis of Kiadin, an antimicrobial peptide, using E. coli as the expression host. The DAMP4 fusion protein was employed to mitigate host toxicity associated with Kiadin. This approach, combined with acid cleavage techniques targeting the aspartyl-prolyl (DP) bond, enabled the successful production and purification of Kiadin with a remarkable purity of 96% and a yield of 29.3 mg/L. This yield is notably high compared to the typical range of 0.6–35.6 mg/L reported for other AMPs.

The development of Kiadin involved truncating and tandemly repeating PGLa to generate DiPGLa-H, followed by substituting Val at position 13 with Gly to create Kiadin. The modified peptide demonstrated a minimum inhibitory concentration (MIC) of 0.75 μM against E. coli ATCC 25922 and 3 μM against Klebsiella pneumoniae ATCC 13883. Kiadin also exhibited low hemolytic activity and minimal cytotoxicity [12]. In vitro, Kiadin showed broad-spectrum antibacterial activity against A. baumannii ATCC 19606, P. aeruginosa ATCC 27853, S. aureus Mu50, and B. cereus ATCC 11778. Transmission electron microscopy (TEM) further revealed that Kiadin affects the cell membranes and walls of E. coli K88 and S. aureus Mu50. In vivo, Kiadin improved the survival rate of mice in an E. coli K88-induced peritoneal infection model by 50% and reduced bacterial load in various organs by over 98%.

One major challenge in the biosynthesis of Kiadin is its significant host toxicity, which has impeded its production in common expression hosts, such as E. coli, B. subtilis, L. lactis, and P. pastoris. For instance, Plectasin’s yield was found to be three times lower when expressed in B. subtilis WB800N compared to E. coli [37], highlighting the substantial impact of the host system on AMP yield. To address this, this study utilized the fast-growing E. coli BL21(DE3) strain and reduced host toxicity through the construction of a DAMP4 fusion protein. The DAMP4 protein's strong hydrophobic interactions and resistance to high temperatures and salt concentrations facilitated the use of high temperatures, sodium sulfate precipitation, and centrifugation for purification [38]. This non-chromatographic method is simpler and more cost-effective compared to traditional chromatographic techniques, which are operationally complex and expensive. For instance, the purification of AMPs Jg7197.t1, Jg7902.t1 and Jg7904.t using His-tag and SUMO-tag requires nickel-column affinity chromatography and inclusion body purification, leading to higher operational costs and complexity [39]. Furthermore, DAMP4 protein's stability and resistance to protease hydrolysis suggest higher expression yields in heterologous expression systems [26]. We also innovatively compared isoelectric precipitation and freeze-drying methods for removing the DAMP4 tag. The isoelectric precipitation method proved to be more effective, achieving higher purity of Kiadin by minimizing the solubility of DAMP4 at its isoelectric point (pH = 6.8) while maintaining the solubility of Kiadin. This method is more cost-effective compared to freeze-drying, resulting in a final purified product, PS–Kiadin–HH, which retains the same broad-spectrum antibacterial activity as chemically synthesized Kiadin.

However, expressing long peptides, proline-rich peptides, or utilizing fusion protein strategies like DAMP4 presents considerable technical challenges. Peptides exceeding 50 amino acids, such as LL-37, often suffer from reduced translation efficiency due to ribosome stalling [40]. Furthermore, these longer peptides are highly susceptible to misfolding and aggregation, leading to the formation of inclusion bodies and significantly reduced solubility [40]. The inherent instability of long peptides under conditions of low pH and high temperatures during protein purification further complicates the production of active forms. For proline-rich antimicrobial peptides (PR-AMPs), the structural properties of proline introduce unique barriers, including ribosomal stalling, particularly in repetitive proline sequences, thereby impairing translation efficiency [41]. Moreover, PR-AMPs have the potential to disrupt the host's protein synthesis machinery, adversely affecting cellular processes and decreasing overall expression yields [41]. Given that both the DAMP4 and DPS sequences contain proline residues, expressing proline-rich peptides via the DAMP4 strategy may exacerbate host cell toxicity, further complicating the expression process. In a previous attempt, we were unable to successfully express a larger construct, DAMP4-(DPS-kiadin)₄, in E. coli BL21(DE3). This failure was likely due to the increased protein size and positive charge, which altered the hydrophobicity and net charge of the fusion protein. These alterations, particularly under acidic conditions, likely induced charge repulsion, preventing correct folding [42]. This experience underscores that while the DAMP4 fusion strategy holds potential, it faces significant limitations when applied to long, proline-rich peptides, necessitating further optimization for efficient production.

The cleavage of fusion proteins is a crucial step for purifying AMPs. Current methods include enzyme cleavage, self-cleavage, and chemical cleavage. Enzyme cleavage methods, such as using SUMO enzyme or TEV enzyme, offer high specificity but are expensive and unsuitable for large-scale production [22]. Self-cleavage methods, based on ELP-intein fusion tags, are simple but increase host expression burden and prolong expression time, requiring multiple washes and centrifugation, which can reduce yield and increase costs [43]. Chemical cleavage methods, on the other hand, offer unique advantages. Historically, it has been known that aspartyl-prolyl peptide bonds are unstable and break easily under acidic conditions [44, 45]. For example, cleavage of the Asp-Pro bond in tobacco mosaic virus proteins using 90% formic acid at 37 °C achieved a cleavage efficiency of 38% [44, 45]. Another study found that the X-Pro bond (X = A, D, E, F, I, or S) was the most unstable peptide bond when the X-position was Asp, while the X-position was the most stable when it was the nonpolar amino acid Ile when under acidic conditions [46]. This study compared the effects of single and multiple DP sites on acid cleavage efficiency. We designed sequences with one DP site and DPIPDPS sequences with two DP sites and found no significant difference in cleavage efficiency between them.

Temperature and time are pivotal factors influencing acid cleavage efficiency, with lower temperatures and shorter cleavage durations suggesting reduced production costs. Previous research has demonstrated that complete cleavage of Asp-Pro bonds can be achieved at 85 °C for 3 h or at 55 °C for 48 h, while only partial cleavage occurs at 37 °C after 12 h [47]. Our experiments revealed minimal cleavage efficiency at 60 °C for 4 h, whereas treatment at 75 °C for 24 h significantly outperformed treatment at 60 °C for 48 h. Further analysis indicated that temperatures ranging from 75 to 90 °C for 2 h markedly enhanced cleavage efficiency. The inhibitory activity assay indicated that lysis products treated at 84–90 °C for 1 h or at 81 °C for 2 h completely inhibited E. coli K88 growth, whereas treatments at 75 °C and 78 °C for 2 h achieved 96% inhibition. This suggests that optimal lysis temperatures and durations should be chosen based on cost-efficiency for industrial applications. In addition, previous studies reported that adding 10% acetic acid or a 70–90% formic acid solution to tobacco mosaic virus proteins resulted in DP bond cleavage, while an 84% cleavage of the carboxypeptidase inhibitor from potato was achieved using a 10% aqueous solution of acetic acid (pH 2.5) [45, 46]. This study compared the effects of strong (HCl) and medium–strong (phosphoric acid) acids at identical pH levels on cleavage efficiency and found no significant differences between the acids, irrespective of whether the cleavage duration was 4 h or 48 h. It is noteworthy that Kiadin, the antimicrobial peptide of interest in this study, maintains an α-helical secondary structure and exhibits high tolerance to elevated temperatures, high salt concentrations, and strong acids. Proteins with more intricate tertiary structures may not be suitable for this cleavage approach.

Conclusions

This study represents a significant advancement in the production of the antimicrobial peptide Kiadin, addressing the critical issue of host toxicity. By employing the DAMP4–DPS–Kiadin fusion protein and leveraging E. coli BL21 (DE3) as the expression host, we successfully mitigated the toxic effects typically associated with Kiadin. Utilizing purification techniques, including ammonium sulfate precipitation, acidic cleavage, and isoelectric point precipitation, we achieved an impressive 96% purity and a yield of 29.3 mg/L for Kiadin. The production of Kiadin in large quantities is of great value for its low-cost manufacturing and application.

Materials and methods

Materials

The PS–Kiadin–HH peptide (PSKIAKVALKALKIAKGALKALHH-NH₂) with a molecular weight of 2607.11 Da, was custom synthesized by Sangon Biotech (Shanghai) with 97.71% purity. Unless otherwise specified, all biochemical reagents, buffers and culture media were procured from Sangon Biotech (Shanghai). Restriction endonuclease and T4 DNA ligase were purchased from New England Biolabs (Beijing Ltd.). The strains and plasmids utilized in this study are detailed in supplementary Table S2, while the primers are listed in supplementary Table S3. E. coli, A. baumannii, P. aeruginosa, B. cereus and B. subtilis in this experiment were cultured in LB medium. S. aureus, L. lactis and P. pastoris were cultured in TSB, GM17 and YPD medium, respectively.

Agar disk diffusion test

Bacterial suspensions of E. coli BL21 (DE3), B. subtilis WB800N, L. lactis NZ9000, and P. pastoris GS115 were adjusted to a concentration of 1.5 × 10⁸ cfu/mL. Following a 100-fold dilution, 100 µL of each suspension was spread on respective solid media plates (LBA, GM17, and YPDA). Wells were punched in the agar, and 50 µL of the synthesized PS–Kiadin–HH peptide was added to each well. Plates were incubated overnight at 37 °C, and results were photographed the following day.

The expression of fusion protein

The nucleotide sequence encoding the DAMP4–DPS–Kiadin fusion protein (M (EPS MKQLADSLHQLARQVSRLEHA)₄D-PS-KIAKVALKALKIAKGALKAL-HH) was synthesized by Jiangsu Junzhong Biologicals after codon optimization (http://www.jcat.de/) for E. coli (Table S4). This sequence was cloned into the pET-28a plasmid using T4 ligase to construct the recombinant plasmid pET28a–DAMP4–DPS–Kiadin. Relevant primers are listed in Table S3. The plasmid was sequentially transformed into E. coli DH5α and E. coli BL21 (DE3). The tertiary structure of the DAMP4–DPS–Kiadin fusion protein was predicted using AlphaFold3 and visualized with PyMOL software.

E. coli BL21 (DE3) carrying the pET28a–DAMP4–DPS–Kiadin plasmid was streaked on an LB agar plate and incubated overnight at 37 °C. Single colonies were inoculated into 10 mL of LB medium containing 50 µg/mL kanamycin and incubated overnight at 37 °C with constant shaking. The culture was then diluted 1% into 500 mL of LB medium with kanamycin and incubated at 37 °C until an OD600 of 0.6–0.8 was reached. IPTG was added to a final concentration of 1 mM, and the culture was shaken at 16 °C for 16 h. Cells were harvested by centrifugation at 12,000 g for 10 min, and the supernatant was discarded.

Non-chromatographic purification of and acid cleavage of fusion protein

The bacterial pellet was resuspended in 50 mL of equilibration buffer (50 mM NaCl, 25 mM NaH₂PO₄, pH 7.5). Na₂SO₄ was added to a final concentration of 0.8 M, and the suspension was incubated in a 90 °C water bath for 30 min, then cooled to room temperature. The mixture was centrifuged at 39,000 g for 10 min, and the supernatant was collected. The pH of the supernatant was adjusted to 3 with hydrochloric acid (HCl), diluted fivefold with ultrapure water, and left to stand at room temperature for 30 min. Following centrifugation at 39,000 g for 10 min, the precipitate was resuspended in 10 mL of ultrapure water. HCl was added to the suspension to a final concentration of 60 mM, and acid cleavage was performed at 60 °C for 48 h with intermittent sampling.

Small molecule protein gels were prepared using the Tricine–SDS–PAGE Gel Preparation Kit (Sangon Biotech). Sample preparation, staining, and destaining were performed according to the manufacturer's protocol, with Protein Stains Q used for silver staining. All products post-acid cleavage was neutralized to pH 7 with sodium hydroxide before running the protein gels.

Optimization of acid cleavage conditions

To evaluate the impact of the cleavage site on fusion protein cleavage efficiency, the DAMP4–DPIPDPS–Kiadin fusion protein was constructed similarly to DAMP4–DPS–Kiadin, with DPIPDPS as the linkage sequence. Plasmid construction, transformation, and induced expression followed the same protocols. The concentration of DAMP4–DPIPDPS–Kiadin was standardized, and acid lysis was performed at 60 °C and 75 °C for 24 and 48 h. Samples were analyzed by Tricine–SDS–PAGE.

To assess the effects of cleavage time and temperature, 10 mL of purified DAMP4–DPS–Kiadin solution was subjected to acid digestion (HCl) for 6 h at temperatures ranging from 75 to 90 °C (75 °C, 78 °C, 81 °C, 84 °C, 87 °C, and 90 °C) in a water bath. Equal amounts of the acid cleavage product and E. coli K88 bacterial solution (10⁶ CFU/mL) were co-cultured at 37 °C. A 50 µL series of dilutions was taken every hour, plated on MHA plates, and incubated overnight at 37 °C. Colony counts were performed the next day, with three replicates for each condition.

To determine the effect of acid type on cleavage efficiency, hydrochloric acid (HCl) and phosphoric acid (H₃PO₄) were added to 10 mL of purified DAMP4–DPS–Kiadin solution to a final concentration of 60 mM. Acid lysis was conducted at 60 °C for 4 and 24 h, and samples were analyzed by Tricine–SDS–PAGE.

Purification of PS–Kiadin–HH

Isoelectric precipitation method: To isolate PS–Kiadin–HH, the sample underwent isoelectric precipitation. A 20 mL sample, post 48-h acid lysis, was adjusted to pH 6.8 and allowed to stand at room temperature for 30 min. Following this, the sample was centrifuged at 38,000 g for 20 min. The precipitate contained DAMP4, while the supernatant contained the target antimicrobial peptide, Kiadin. Both the supernatant and precipitate were separately sampled, labeled, and subjected to Tricine–SDS–PAGE analysis along with freeze-dried samples.

Freeze-drying method: Another method for purifying PS–Kiadin–HH involved freeze-drying. A 20 mL sample, post 48-h acid lysis, was poured into a petri dish and frozen at −80 °C. The frozen sample was then transferred to liquid nitrogen to prevent thawing and subsequently placed in a freeze dryer for approximately 24 h until a flocculent powder formed. This powder was carefully collected into a sterile 50 mL centrifuge tube. After dissolving the powder in 20 mL of MHB medium, the solution was centrifuged at 20,000 g for 20 min. The precipitate contained DAMP4, and the supernatant held the target antimicrobial peptide, Kiadin. Both components were separately sampled and labeled.

Minimal inhibitory concentration (MIC) test

The antibacterial efficacy of the peptides was evaluated using a modified method from the National Committee for Clinical Laboratory Standards (NCCLS). Bacterial cells in the log phase were diluted in Mueller–Hinton Broth (MHB) to achieve a final concentration of 0.5–1 × 10⁵ CFU/mL. Fifty microliters of peptides at various concentrations were mixed with 50 μL of the bacterial solution in sterile 96-well plates. The plates were incubated at 37 °C for 24 h. The MIC was determined by measuring absorbance at 492 nm using a microplate reader, with the MIC defined as the concentration that inhibits 95% of bacterial growth.

Mass spectrometry

Protein quantification was performed using analytical reversed-phase high-performance liquid chromatography (RP-HPLC) on an LC-40 HPLC system (Shimadzu, Kyoto, Japan), equipped with a Jupiter 300 Å C18 column (Phenomenex, Torrance, CA). The mobile phase consisted of Buffer A (0.1% trifluoroacetic acid (TFA) in water) and Buffer B (90% acetonitrile and 0.1% TFA in water). Gradient elution ranged from 30 to 65% Buffer B over 35 min at a flow rate of 1 mL/min, with detection at 214 nm.

For molecular weight identification of proteins and peptides, liquid chromatography–mass spectrometry (LC–MS) was employed. This setup combined a Waters Alliance HPLC system (Waters, Milford, MA) with a Quattro Micro API Quadrupole system (Waters, Milford, MA). The LC utilized a Kinetex C18 column (100 Å, 2.6 μm, 100 mm × 4.6 mm, Phenomenex, Torrance, CA). The mobile phases were A (0.01% TFA in water) and B (90% acetonitrile, 0.01% TFA in water). The gradient transitioned from 30 to 65% B over 35 min, with a column flow rate of 0.6 mL/min and detection at 214 nm. Positive-ion electrospray ionization (ESI) was used for mass spectrometry, with voltages for the capillary, cone, extractor, and radiofrequency lens set to 3 kV, 24 V, 3 V, and 0 V, respectively.

Determination of the PAE of PS–Kiadin–HH

3 mL of PS–Kiadin–HH at 0 × MIC, 1 × MIC, 2 × MIC, and 4 × MIC were each mixed with 3 mL of log-phase bacterial suspension (1.35 × 10⁶ cfu/mL) in separate 15 mL tubes (test group, T). A bacterial control tube (C) was also prepared by adding 6 mL of log-phase bacterial suspension (6.75 × 10⁵ cfu/mL). The resulting drug concentrations in each tube were 0 × MIC, 1 × MIC, 2 × MIC and 4 × MIC, respectively. All tubes were incubated in a 37 °C water bath for 1 h. Then, 0.1 mL from each tube was transferred to 9.9 mL of pre-warmed medium (37 °C) to reconstitute the bacteria. This timepoint was recorded as time zero. At 0, 1, 2, 3, 4, 5, 6, and 8 h after reconstitution, serial dilutions were performed from each tube, and 0.1 mL was plated on agar for colony counting. The procedure was repeated three times, and the average value was taken.

The post-antibiotic effect (PAE) was calculated using the formula PAE = T – C, where T and C represent the time (in hours) required for the bacterial count (cfu/mL) in the test and control groups to increase by tenfold after drug removal.

TEM

For TEM analysis, S. aureus Mu50 and E. coli K88 were initially cultured overnight in a shaker. This was followed by a secondary activation, where 500 μL of the culture was transferred into 50 mL of fresh liquid medium and incubated at 37 °C with shaking at 250 rpm for 2 h until reaching the log phase. After incubation, the cultures were centrifuged at 4 °C and 3000 rpm for 10 min and subsequently washed twice with pre-cooled PBS under the same conditions. The bacterial pellets were resuspended in PBS, thoroughly mixed, and the optical density measured at 600 nm. The suspension was then diluted with PBS to achieve a final concentration of 10⁸ CFU/mL. For the antibacterial assay, the bacterial liquid was mixed with Kiadin at its MIC value and incubated at 37 °C in a shaker at 250 rpm for 1 h. The samples were then centrifuged at 4 °C and 3000 rpm for 5 min, washed twice with PBS, and fixed overnight at 4 °C with 500 μL of 2.5% glutaraldehyde. The next day, the samples were washed three times with PBS for 15 min each, post-fixed with osmium tetroxide for 2–4 h, and washed again three times with PBS for 10 min each. Dehydration was performed using a series of ethanol gradients (30%, 50%, 70%, 80%, 90%, and 100%), twice at each concentration. The samples were then embedded in a 1:1 mixture of anhydrous acetone and epoxy resin for 30 min, followed by pure epoxy resin, and cured at a constant temperature overnight. Finally, the specimens were sectioned using an ultrathin sectioning machine, stained with uranium dioxide acetate and lead citrate, and examined under a TECNAI G2 SPIRIT BIO transmission electron microscope (FEI, America).

Membrane permeabilization

NPN (Sigma-Aldrich, catalog no. 104043, ≥ 98%) was used to evaluate the outer membrane integrity of PS–Kiadin–HH-treated E. coli K88. Log-phase E. coli K88 cells (OD600 = 0.2) were incubated with NPN (10 μM) in 5 mM HEPES buffer (pH 7.4, containing 5 mM glucose) for 30 min. Fluorescence (excitation λ = 350 nm, emission λ = 420 nm) was measured using a microplate reader (Varioskan Flash; Thermo). Cell suspensions (100 μL) were mixed with peptide solutions in a 96-well black plate, and fluorescence was recorded until it plateaued. Each experiment was performed in duplicate with three biological replicates. NPN uptake was calculated as NPN uptake (%) = (F_obs – F_buffer)/(F_positive – F_buffer) × 100%, where polymyxin B (10 μg/mL) was the positive control, and untreated cells served as the negative control.

Inner membrane integrity assay. E. coli K88 and S. aureus Mu50 cells were washed and resuspended in 0.01 mol l − 1 of PBS (pH 7.4) to obtain an OD600 of 0.5, followed by the addition of 10 nmol l − 1 of PI (Thermo Fisher Scientific, catalogue no. P1304MP) in the presence of PS–Kiadin–HH. After incubation for 30 min, fluorescence was measured with the excitation wavelength at 535 nm and emission wavelength at 615 nm.

Antibiofilm detection

The anti-biofilm activity of PS–Kiadin–HH was evaluated using a crystal violet staining method. Briefly, 100 μL of bacterial suspension at 1 × 10⁶ CFU/mL was incubated with MIC concentration of PS–Kiadin–HH for 48 h. The biofilm formed was stained with 1% crystal violet, and the absorbance was measured at 595 nm.

Cytotoxicity assays

The cytotoxicity of peptides was evaluated using the murine macrophage cell line RAW264.7 via the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Cells were seeded at a density of 1.0–2.0 × 10⁴ cells per well in 96-well plates and exposed to varying concentrations of PS–Kiadin–HH (2–128 μg/mL) for 18–24 h at 37 °C under a 5% CO₂ atmosphere. After incubation, cells were treated with 0.5 mg/mL MTT for 4 h at 37 °C. The supernatant was removed, and the formazan crystals formed were dissolved in 150 μL of dimethyl sulfoxide (DMSO). Absorbance was measured at 570 nm using a Varioskan Flash microplate reader (Thermo Fisher Scientific, USA).

Mouse model of abdominal infection

An abdominal infection model in mice was established as previously described [48]. Male C57BL/6 mice, aged 6–8 weeks, were used as experimental subjects and divided into six groups: survival and dissection groups injected with PBS, and survival and dissection groups injected with either chemically synthesized or biosynthesized PS–Kiadin–HH. Each survival group consisted of 16 mice, and each dissection group consisted of 5 mice. The control group was injected with PBS. All six groups were injected intraperitoneally with a lethal dose of E. coli K88 at the start of the experiment. Four groups were administered 64 μg/mL of either biosynthesized or chemically synthesized PS–Kiadin–HH 30 min post-infection, followed by additional injections at 6-h intervals for a total of three injections. The remaining two groups received PBS injections as controls. The mortality rate of the 16 mice in each survival group was observed over a 7-day period. Mice in the dissection groups were euthanized 1 ~ 2 day post-infection, and their hearts, livers, spleens, and kidneys were harvested, homogenized, and serially diluted up to 10⁷ times with PBS. Aliquots of 100 μL from each dilution were plated on Luria–Bertani agar (LBA) plates and incubated overnight at 37 °C. The next day, bacterial counts and survival rates were analyzed using GraphPad Prism software. All experimental protocols and the use of animals were approved by the Animal Protection and Use Committee of Northwest A&F University.